Evidence Report/Technology Assessment

Number 76

Ephedra and Ephedrine for Weight Loss and

Athletic Performance Enhancement: Clinical
Efficacy and Side Effects

U.S. Department of Health and Human Services

Agency for Healthcare Research and Quality

This report may be used, in whole or in part, as the basis for development of clinical practice
guidelines and other quality enhancement tools, or a basis for reimbursement and coverage
policies. AHRQ or U.S. Department of Health and Human Services endorsement of such
derivative products may not be stated or implied.
AHRQ is the lead Federal agency charged with supporting research designed to improve the
quality of health care, reduce its cost, address patient safety and medical errors, and broaden access
to essential services. AHRQ sponsors and conducts research that provides evidence-based
information on health care outcomes; quality; and cost, use, and access. The information helps
health care decisionmakerspatients and clinicians, health system leaders, and policymakers
make more informed decisions and improve the quality of health care services.

This document is in the public domain and may be used and reprinted without permission except
those copyrighted materials noted for which further reproduction is prohibited without the
specific permission of copyright holders.

Suggested Citation:
Shekelle P, Morton, S., Maglione M, et al. Ephedra and Ephedrine for Weight Loss and Athletic
Performance Enhancement: Clinical Efficacy and Side Effects. Evidence Report/Technology
Assessment No. 76 (Prepared by Southern California Evidence-based Practice Center, RAND,
under Contract No 290-97-0001, Task Order No. 9). AHRQ Publication No. 03-E022. Rockville,
MD: Agency for Healthcare Research and Quality. February 2003.

ii

Preface
The Agency for Healthcare Research and Quality (AHRQ), through its Evidence-Based
Practice Centers (EPCs), sponsors the development of evidence reports and technology
assessments to assist public- and private-sector organizations in their efforts to improve the
quality of health care in the United States. The reports and assessments provide organizations
with comprehensive, science-based information on common, costly medical conditions and new
health care technologies. The EPCs systematically review the relevant scientific literature on
topics assigned to them by AHRQ and conduct additional analyses when appropriate prior to
developing their reports and assessments.
To bring the broadest range of experts into the development of evidence reports and health
technology assessments, AHRQ encourages the EPCs to form partnerships and enter into
collaborations with other medical and research organizations. The EPCs work with these partner
organizations to ensure that the evidence reports and technology assessments they produce will
become building blocks for health care quality improvement projects throughout the Nation. The
reports undergo peer review prior to their release.
AHRQ expects that the EPC evidence reports and technology assessments will inform
individual health plans, providers, and purchasers as well as the health care system as a whole by
providing important information to help improve health care quality.
We welcome written comments on this evidence report. They may be sent to: Director,
Center for Practice and Technology Assessment, Agency for Healthcare Research and Quality,
6010 Executive Blvd., Suite 300, Rockville, MD 20852.
Carolyn M. Clancy, M.D.
Director
Agency for Healthcare Research and Quality
Jean Slutsky, P.A., M.S.P.H.
Acting Director, Center for Practice and
Technology Assessment
Agency for Healthcare Research and Quality

The authors of this report are responsible for its content. Statements in the report
should not be construed as endorsement by the Agency for Healthcare Research and
Quality or the U.S. Department of Health and Human Services of a particular drug,
device, test, treatment, or other clinical service.

iii

Acknowledgment
We would like to thank Dr. Lori Love and Ms. Constance J. Hardy of the Food and Drug
Administration for providing us with adverse event reports related to ephedra. We thank Dr.
Catherine MacLean, a rheumatologist, and Dr. Barbara Vickrey, a neurologist, for providing
advice on cases related to their fields. We thank Drs. Karen Miotto, a psychiatrist, and Martin
Iguchi, a psychologist, for review of psychiatric case reports. We thank Ms. Birgit Danila, MA,
for translation of Danish studies.

iv

Structured Abstract
Objectives. To assess the efficacy of herbal ephedra-containing dietary supplements and
ephedrine on weight loss and athletic performance, through comprehensive literature review and
synthesis of evidence. We also assessed safety of these products through review of adverse
events reported in clinical trials, published case reports of adverse events, reports on file with the
U.S. Food and Drug Administration (FDA), and a file of reports kept by a manufacturer of
ephedra products, Metabolife.
Search Strategy. We searched for studies of herbal ephedra and ephedrine using the following
electronic databases: Medline, EmBase, BIOSIS, Allied & Complementary Medicine Database
(AMED), MANTIS, the Cochrane Controlled Clinical Trials Register Database, International
Pharmaceutical Abstracts, Pascal, and SciSearch. We were able to obtain unpublished studies by
posting notices in relevant journals and through contacts on our Technical Expert Panel. The
FDA provided us with copies of over 1,000 adverse event reports (AERs) related to herbal
ephedra and 125 adverse event reports related to ephedrine. The Metabolife files contained
18,502 cases.
Selection Criteria. Only studies of weight loss that were controlled trials of human subjects with
treatment of at least eight weeks duration were accepted to assess efficacy. For assessment of
athletic performance, only controlled trials of human subjects were accepted, but no minimum
follow-up was specified. Reports of adverse events from controlled trials were included
regardless of treatment duration. We reviewed all available reports of death, myocardial
infarction (heart attack), cerebral vascular accident (stroke), seizure, and serious psychiatric
illness reported to the FDA prior to September 30, 2001 and contained in their ephedra or
ephedrine files, and all case reports identified in our literature search .
Data Collection and Analysis. We found 59 articles that corresponded to 52 controlled clinical
trials of ephedrine or herbal ephedra for weight loss or athletic performance. Forty-four were
controlled trials assessing ephedra or ephedrine for weight loss. Of these, 18 were excluded from
pooled analysis because they had treatment durations of less than eight weeks. Thirteen articles
corresponding to six trials were excluded for a variety of reasons. For the outcome of weight loss
the effects of ephedra/ephedrine were examined in six different types of comparisons:
(1) ephedrine versus placebo; (2) ephedrine plus caffeine versus placebo; (3) ephedrine plus
caffeine versus ephedrine; (4) ephedrine versus other active treatment; (5) ephedra versus
placebo; and (6) ephedra plus herbs containing caffeine versus placebo. Only four placebocontrolled trials assessed the combination of ephedra plus herbs containing caffeine, and only
one trial assessed ephedra without herbs containing caffeine. Because of their small number and
heterogeneity, eight athletic performance trials were compared and contrasted using only a
narrative review and were not synthesized statistically. We also conducted a pooled metaanalysis on those adverse event symptoms that occurred frequently in the controlled trials.
In reviewing the individual adverse event reports, we searched for documentation that an
adverse event had occurred, documentation that the subject had consumed ephedra within 24
hours prior to the adverse event, or a toxicological examination revealing ephedrine or one of its
associated products in the blood or urine. We also sought evidence that an adequate investigation
v

had assessed and excluded other potential causes. Cases that met all these criteria were labeled
sentinel events. Cases that met the first two criteria but had other possible causes of the event
were labeled possible sentinel events. Classification as a sentinel event does not imply a
proven cause and effect relationship. We used clinical judgment of expert clinicians to assess
whether other causes had been adequately evaluated and excluded.
Main Results. Weight Loss. Short-term use of ephedrine, ephedrine plus caffeine, or dietary
supplements containing ephedra with or without herbs containing caffeine is associated with a
statistically significant increase in short-term weight loss (compared to placebo). The addition of
caffeine to ephedrine is associated with a statistically significant modest increase in short-term
weight loss. The observed effects on weight loss of ephedrine plus caffeine and ephedracontaining dietary supplements with or without herbs containing caffeine are approximately
equivalent: a weight loss approximately two pounds per month greater than that with placebo, for
up to four to six months. No studies have assessed the long-term effects of ephedrine or ephedracontaining dietary supplements on weight loss; the longest published treatment duration was six
months.
Athletic Performance. The effect of herbal ephedracontaining dietary supplements on
athletic performance has not been assessed. The few studies that assess the effect of ephedrine on
athletic performance have included only small samples of fit individuals (young male military
recruits) and have assessed its effect only on very short-term immediate performance. These data
support a modest effect of ephedrine plus caffeine on very short-term athletic performance. One
study reported the addition of caffeine to ephedrine is necessary to produce an effect on athletic
performance. No studies have assessed the sustained use of ephedrine on performance over time.
Safety Issues. There is sufficient evidence from controlled trials to conclude that the use of
ephedrine and/ or the use of ephedra-containing herbal supplements or ephedrine plus caffeine is
associated with two to three times the risk of nausea, vomiting, psychiatric symptoms such as
anxiety and change in mood, autonomic hyperactivity, and palpitations. The controlled trials
studied relatively few people and in aggregate were insufficient to evaluate events with a risk of
less than 1.0 per one thousand.
The majority of case reports are insufficiently documented to make an informed judgment
about a relationship between the use of ephedrine or ephedra-containing dietary supplements and
the adverse event in question. Prior ephedra consumption was associated with two deaths, three
myocardial infarctions, nine cerebrovascular accidents, three seizures, and five psychiatric cases
as sentinel events. Prior consumption of ephedrine was associated with three deaths, two
myocardial infarctions, two cerebrovascular accidents, one seizure, and three psychiatric cases as
sentinel events. We identified 43 additional cases as possible sentinel events with prior ephedra
consumption and seven additional cases as possible sentinel events with prior ephedrine
consumption. About half the sentinel events occurred in persons aged 30 years or younger.
Conclusions. Ephedrine, ephedrine plus caffeine, and ephedra-containing dietary supplements
with or without herbs containing caffeine all promote modest amounts of weight loss over the
short term. There are no data regarding long-term effects on weight loss. Single-dose ephedrine
plus caffeine has a modest effect on athletic performance. The available trials do not provide any
evidence about ephedrine or ephedra-containing dietary supplements, as they are used by the
general population, to enhance athletic performance. Use of ephedra or ephedrine plus caffeine is
vi

associated with an increased risk of gastrointestinal, psychiatric, and autonomic symptoms. The
adverse event reports contain a sufficient number of cases of death, myocardial infarction,
cerebrovascular accident, seizure, or serious psychiatric illness in young adults to warrant a
hypothesis-testing study, such as a case-control study, to support or refute the hypothesis that
consumption of ephedra or ephedrine may be causally related to these serious adverse events.

vii

Contents
Evidence Report............................................................................................................................. 1
Chapter 1. Introduction .................................................................................................................. 3
Purpose..................................................................................................................................... 3
Scope of Work ......................................................................................................................... 3
Background .............................................................................................................................. 3
The Problem of Obesity ..................................................................................................... 5
Enhancing Physical Performance ...................................................................................... 6
History and Pharmacology....................................................................................................... 7
Chapter 2. Methodology .............................................................................................................. 13
Original Proposed Key Questions.......................................................................................... 13
Weight Loss ..................................................................................................................... 13
Athletic Performance ....................................................................................................... 13
Safety Assessment ........................................................................................................... 13
Technical Expert Panel .......................................................................................................... 14
Assessment of Adverse Events ........................................................................................ 14
Literature Search.................................................................................................................... 15
Additional Sources of Evidence....................................................................................... 16
Article Review ....................................................................................................................... 17
Extraction of Study-Level Variables and Results............................................................ 17
Meta-Analysis ........................................................................................................................ 18
Selection of Trials for Meta-Analysis.............................................................................. 18
Stratification of Interventions .......................................................................................... 19
Weight Loss Effect Size................................................................................................... 20
Performance of Meta-Analysis ........................................................................................ 21
Analysis of Dose .............................................................................................................. 23
Publication Bias ............................................................................................................... 23
Meta-Regression .............................................................................................................. 23
Safety Assessment ................................................................................................................. 24
Controlled Trial Adverse Events ..................................................................................... 24
Case Report Adverse Events............................................................................................ 26
Peer Review ........................................................................................................................... 32
Chapter 3. Results ........................................................................................................................ 73
Results of Literature Search................................................................................................... 73
Efficacy Analysis ............................................................................................................. 73
Adverse Events Analysis ................................................................................................. 73
Efficacy .................................................................................................................................. 73
Weight Loss ..................................................................................................................... 73
Athletic Performance ....................................................................................................... 77
Safety Assessment ................................................................................................................. 78
Controlled Trials .............................................................................................................. 78
Case Reports .................................................................................................................... 80
ix

Chapter 4. Limitations ............................................................................................................... 197

Chapter 5. Conclusions .............................................................................................................. 201
Efficacy ................................................................................................................................ 201
Weight Loss ................................................................................................................... 201
Athletic Performance ..................................................................................................... 202
Adverse Consequences ........................................................................................................ 202
Chapter 6. Future Research........................................................................................................ 205
References.................................................................................................................................. 207
Evidence Tables ......................................................................................................................... 217
Acronyms................................................................................................................................... 228
Appendix 1. Bibliography
Appendix 2. Metabolife Serious Adverse Events
Appendix 3. Reviewer Comments

Tables
Table 1. Herbs containing caffeine commonly combined with ephedra in products
marketed for weight loss or improved physical performance....................................... 11
Table 2. Technical expert panel members ................................................................................... 34
Table 3. Technical expert panel suggestions about data collection ............................................. 35
Table 4. Measures used in assessing causality............................................................................. 36
Table 5 Ephedra/ ephedrine search methodology........................................................................ 37
Table 6 Ephedra/ephedrine search methodology additional databases..................................... 38
Table 7. Categories of adverse events.......................................................................................... 39
Table 8. Report reviewers ............................................................................................................ 40
Table 9. Weight loss trial inclusion results................................................................................ 104
Table 10. Ephedrine versus placebo .......................................................................................... 105
Table 11. Publication bias tests.................................................................................................. 106
Table 12. Ephedrine + caffeine versus placebo ......................................................................... 107
Table 13. Ephedrine + caffeine versus ephedrine...................................................................... 108
Table 14. Ephedrine versus another active weight loss therapy ................................................ 109
Table 15. Ephedra versus placebo ............................................................................................. 110
Table 16. Ephedra + herbs containing caffeine versus placebo................................................. 111
Table 17. Meta-regression results .............................................................................................. 112
Table 18. Exercise trials by Bell and colleagues ....................................................................... 113
Table 19. Summary table of meta-analysis of adverse events reported controlled trials .......... 114
Table 20. Summary table of other of adverse events reported in controlled trials .................... 115
Table 21. Distribution of adverse events in the FDA file according to the Excel
Spreadsheet ................................................................................................................. 116
Table 22. Evidence table of case reports ................................................................................... 117
Table 23. Summary of adverse events with ephedra consumption............................................ 169
Table 24. Summary of adverse events with ephedrine consumption......................................... 170
Table 25. Summary of adverse events not reviewed in detail ................................................... 171
Table 26. Summary data of key variables from Metabolife file analysis .................................. 172
Table 27. Comparison of serious cases identified by RAND and by Metabolife...................... 174
Table 28. Summary of Metabolife medical records................................................................... 177

xi

Figures
Figure 1. Screening form for literature ........................................................................................ 41
Figure 2. Quality review form for literature ................................................................................ 42
Figure 3a. Adverse events analysis form for death, MI, stroke cases.......................................... 47
Figure 3b. Adverse events analysis form for seizure cases ......................................................... 52
Figure 3c. Adverse events analysis form for psychiatric cases ................................................... 57
Figure 4. Brief data collection form for case reports ................................................................... 63
Figure 5. Examples of MIPER files............................................................................................. 64
Figure 6. Example of duplicate case ............................................................................................ 69
Figure 7. Metabolife record screener form .................................................................................. 71
Figure 8. Literature flow ............................................................................................................ 186
Figure 9. Ephedrine versus placebo forest plot....................................................................... 187
Figure 10. Ephedrine versus placebo funnel plot.................................................................... 187
Figure 11. Ephedrine + caffeine versus ephedrine alone forest plot....................................... 188
Figure 12. Ephedrine + caffeine versus placebo funnel plot .................................................. 188
Figure 13. Ephedrine + caffeine versus ephedrine alone forest plot....................................... 189
Figure 14. Ephedra + herbs containing caffeine versus placebo forest plot ........................... 190
Figure 15. Ephedra + herbs containing caffeine versus placebo funnel plot .......................... 190
Figure 16. Effect sizes by comparison group............................................................................. 191
Figure 17a. Flow of evidence for adverse events analysis, part 1 ............................................. 192
Figure 17b. Flow of evidence for adverse events analysis, part 2 ............................................. 193
Figure 17c. Flow of evidence for adverse events analysis, part 3 ............................................. 194
Figure 18. Flow of MIPER ID Numbers ................................................................................... 195

xii

Agency for Healthcare Research and Quality

Evidence Report/Technology Assessment

Number 76

Ephedra and Ephedrine for Weight Loss and Athletic

Performance Enhancement: Clinical Efficacy
and Side Effects
Summary

Overview
At the direction of the funding agencies (the
National Institutes of Health Office of Dietary
Supplements (ODS), the National Center for
Complementary and Alternative Medicine
(NCCAM), and the Agency for Healthcare
Research and Quality (AHRQ)), and in
consultation with our Technical Expert Panel, we
addressed research questions regarding the efficacy
of herbal ephedra and ephedrine for weight loss
and athletic performance through a
comprehensive literature review and synthesis of
evidence. We assessed the safety of these products
through review of clinical trials. Meta-analysis was
performed where appropriate. In addition, we
reviewed herbal ephedra- and ephedrine-related
adverse events reports on file with the U.S. Food
and Drug Administration (FDA), published case
reports, and reports to a manufacturer of ephedracontaining products. It is expected that the results
of this review will be used to direct further
research.

Reporting the Evidence

The following questions were provided to us by
the funding agencies and guided this evidence
report.
Weight Loss
1. What is the evidence for efficacy of ephedracontaining dietary supplement products in
weight loss, over a sustained period of time?
2. Can efficacy for weight loss be attributed to
ephedra alone, or ephedra in combination
with other ingredients (e.g., caffeine)?
3. Does ephedra have additive effects with other
agents?
4. What dosage levels of ephedra are necessary
to achieve weight loss?

Athletic Performance
1. What is the evidence for efficacy of ephedracontaining dietary supplement products in
terms of energy enhancement and
enhancement of athletic performance, over a
sustained period of time?
2. Can efficacy for energy enhancement and
enhancement of athletic performance be
attributed to ephedra alone, or ephedra in
combination with other ingredients (e.g.,
caffeine) that produce energy enhancement
and/or enhancement of athletic performance?
3. Does ephedra have additive effects with other
agents?
4. What dosage levels of ephedra are necessary
to achieve energy enhancement and
enhancement of athletic performance?
Safety Assessment
1. Does use of ephedra-containing dietary
supplement products over a sustained period
of time increase the risk of cardiovascular
disease (CVD) or other serious and lifethreatening events in specific populations?
2. What populations are at risk of CVD and
other life-threatening events through use of
ephedra over a sustained period of time?
3. Can the risk for adverse events in these
populations be attributed to ephedra alone,
or in combination with other ingredients
(e.g., caffeine)?
4. Does ephedra have additive effects with other
agents?
5. What dosage levels of ephedra produce risk of
CVD or other life-threatening events?

U . S . D E PA R T M E N T O F H E A LT H A N D H U M A N S E R V I C E S P u b l i c H e a l t h S e r v i c e

6. Do ephedra-containing dietary supplement products alter

physiologic markers of cardiovascular function?
7. What are the metabolic actions of ephedra, so as to
explain its beneficial and adverse effects?
In addition to answering these 15 questions about ephedracontaining dietary supplement products, we were also asked to
synthesize the available information on the same questions for
the purified alkaloid, ephedrine.
After searching published reports, journal articles, conference
presentations, and various sources of unpublished studies, we
identified 52 controlled clinical trials of ephedrine or herbal
ephedra for weight loss or athletic performance in humans. The
FDA provided us with copies of over 1,000 adverse event
reports (AERs) related to herbal ephedra and 125 AERs related
to ephedrine. These reports often included interviews with
patients and/or family members, extensive medical records, and
copies of product labels. We identified 65 case reports in the
literature and received a disk of 15,951 reports containing
18,502 cases from Metabolife, a manufacturer of ephedra
products.

Methodology
Efficacy. Data for the efficacy analysis were abstracted from
reports of controlled trials onto a specially designed form
containing questions about the study design, the number of
patients and comorbidities, dosage, adverse events, the types of
outcome measures, and the time from intervention until
outcome measurement. We selected the variables for abstraction
with input from the projects technical experts. Two physicians,
working independently, each extracted data from the same
reports and resolved disagreements by consensus.
In selecting studies for the meta-analysis of weight loss
efficacy, we considered only those trials of at least 8 weeks
treatment duration. Our technical expert panel judged that
shorter treatment durations were insufficient to assess weight
loss. In selecting studies on athletic performance, we found that
these studies varied widely with respect to intervention. Because
of this heterogeneity, we compared and contrasted these studies
in a narrative review, rather than performing a statistical
synthesis.
The effects of ephedra/ephedrine on weight loss were
examined in six different types of comparisons: (1)ephedrine
versus placebo; (2) ephedrine plus caffeine versus placebo; (3)
ephedrine plus caffeine versus ephedrine; (4) ephedrine versus
other active treatment; (5)ephedra versus placebo; and (6)
ephedra plus herbs containing caffeine versus placebo. The last
comparison subgroup contained only a single trial; thus, effect
sizes were estimated only for the first five. The effect size was
calculated by dividing the outcome of a study (e.g., difference
in weight loss per month between the two groups) by its
standard deviation, which produces a unitless measure that is
useful when comparing studies that assess outcomes (such as
weight) that are similar but are measured differently (e.g.,

weight loss in pounds versus change in body mass index). Effect

sizes were pooled separately for each of the five comparison
subgroups. In addition, we used meta-regression to conduct a
cross-subgroup synthesis on the effect sizes of the subgroups
with a placebo comparison: ephedrine versus placebo;
ephedrine plus caffeine versus placebo; and ephedra plus herbs
containing caffeine versus placebo.
Safety. We reviewed each report of a controlled trial
(regardless of treatment duration) for data on adverse events.
Adverse events were recorded onto a spreadsheet that identified
each study arm, the description of the adverse event as listed in
the original article, and the numbers of subjects and adverse
events in each arm. We then compared event rates in the
ephedra or ephedrine groups to those in the placebo groups.
We conducted a meta-analysis on those adverse event
symptoms for which appreciable numbers of events were noted
in the controlled trials.
Adverse event reports compiled by the FDA concerning
ephedra or ephedrine were also reviewed by our physician
reviewers. Within the time and resource constraints of this
report, we reviewed all available reports of death, myocardial
infarction (heart attack), cerebral vascular accident (stroke),
seizure, and serious psychiatric illness filed prior to September
30, 2001. We also reviewed published case reports as well as
event reports filed with Metabolife, a manufacturer of ephedracontaining products. After screening, all case reports were
subjected to a review.
Based on input from our technical expert panel and the
literature on methods to assess adverse event reports, we
identified three important criteria for inclusion of such reports:
1. Documentation of an adverse event that met our selection
criteria.
2. Documentation that the person having the adverse event
took an ephedra-containing supplement or ephedrine
within 24 hours prior to the event (for cases of death,
myocardial infarction, stroke, or seizure).
3. Documentation that alternative explanations for the
adverse event were investigated and were excluded with
reasonable certainty.
We classified cases that met all three of these criteria as
sentinel events. Cases in which the event might have had
other possible causes but the pharmacology of ephedrine could
have contributed were classified as possible sentinel events.
Cases of death, myocardial infarction, cerebral vascular
accident, and seizure were reviewed by internists, with
additional review (as indicated) by a cardiologist, neurologist,
or rheumatologist. Psychiatric cases were reviewed by a
psychiatrist specializing in addictions and a psychologist with
expertise in substance abuse. The criterion for use within 24
hours was not required for psychiatric cases.

Findings
Efficacy for Weight Loss. We identified 44 controlled trials
that assessed use of ephedra or ephedrine used for weight loss.
Of these, 18 were excluded from pooled analysis because they
had a treatment duration of less than 8 weeks. Six additional
trials were excluded for a variety of other reasons. Of the
remaining 20 trials included in the meta-analysis, only five
tested herbal ephedra-containing products. Together, these 20
trials assessed 678 persons who consumed either ephedra or
ephedrine. The majority of studies of both ephedra and
ephedrine are plagued by methodological problems
(particularly, high attrition rates) that might contribute to bias.
These methodological limitations must be considered when
interpreting any conclusions regarding the efficacy of these
products. Nevertheless, the evidence we identified and assessed
supports an association between short-term use of ephedrine,
ephedrine plus caffeine, or dietary supplements that contain
ephedra with or without herbs containing caffeine and a
statistically significant increase in short-term weight loss
(compared to placebo). Adding caffeine to ephedrine modestly
increases the amount of weight loss. There is no evidence that
the effect of ephedra-containing dietary supplements with herbs
containing caffeine differs from that of ephedrine plus caffeine:
Both result in weight loss that is approximately 2 pounds per
month greater than that with placebo, for up to 4 to 6 months.
No studies have assessed the long-term effects of ephedracontaining dietary supplements or ephedrine on weight loss;
the longest duration of treatment in a published study was 6
months.
Efficacy for Physical Performance Enhancement. The
effect of ephedrine on athletic performance was assessed in
seven studies. No studies have assessed the effect of herbal
ephedra-containing dietary supplements on athletic
performance. The few studies that assessed the effect of
ephedrine on athletic performance have, in general, included
only small samples of fit individuals (young male military
recruits) and have assessed the effects only on very short-term
immediate performance. Thus, these studies did not assess
ephedrine as it is used in the general population. The data
support a modest effect of ephedrine plus caffeine on very
short-term athletic performance. No studies have assessed the
sustained use of ephedrine on performance over time. The only
study that assessed the additive effects of these agents reported
that ephedrine must be supplemented with caffeine to affect
athletic performance.
Safety Issues. The data on adverse events were drawn from
clinical trials and case reports published in the literature,
submitted to the FDA, and reported to Metabolife, a
manufacturer of ephedra-containing supplement products. The
strongest evidence for causality should come from clinical trials;
however, in most circumstances, such trials do not enroll
sufficient numbers of patients to adequately assess the
possibility of rare outcomes. Such was the case with our review

of ephedrine and ephedra-containing dietary supplements.

Even in aggregate, the clinical trials enrolled only enough
patients to detect a serious adverse event rate of at least 1.0 per
1,000. For rare outcomes, we reviewed case reports, but a causal
relationship between ephedra or ephedrine use and these events
cannot be assumed or proven.
Evidence from controlled trials was sufficient to conclude
that the use of ephedrine and/or the use of ephedra-containing
dietary supplements or ephedrine plus caffeine is associated
with two to three times the risk of nausea, vomiting, psychiatric
symptoms such as anxiety and change in mood, autonomic
hyperactivity, and palpitations.
The majority of case reports are insufficiently documented to
make an informed judgment about a relationship between the
use of ephedrine or ephedra-containing dietary supplements
and the adverse event in question. For prior consumption of
ephedra-containing products, we identified two deaths, three
myocardial infarctions, nine cerebrovascular accidents, three
seizures, and five psychiatric cases as sentinel events; for prior
consumption of ephedrine, we identified three deaths, two
myocardial infarctions, two cerebrovascular accidents, one
seizure, and three psychiatric cases as sentinel events. We
identified 43 additional cases as possible sentinel events with
prior ephedra consumption and seven additional cases as
possible sentinel events for prior ephedrine consumption.
About half the sentinel events occurred in persons aged 30
years or younger. Classification as a sentinel event does not
imply a proven cause and effect relationship.
We did not assess the plethora of additional symptoms that
have been reported in the published literature and the FDA
Medwatch file for ephedra-containing dietary supplements and
ephedrine products.

Future Research
Our analysis of the evidence reveals numerous gaps in the
literature regarding the efficacy and safety of ephedracontaining dietary supplements. First, long-term assessments of
the effectiveness of herbal ephedra or ephedrine for promoting
weight loss are lacking. We identified no study with a treatment
duration longer than 6 months. To improve health outcomes
and reduce the risk of morbidities associated with being
overweight, sufficient weight loss (5 to 10 percent of body
weight) and long-term weight maintenance are necessary.
Therefore, the benefit of ephedrine or herbal ephedracontaining dietary supplements for health outcomes is
unknown.
Evidence regarding the effect of herbal ephedra or ephedrine
on physical performance that reflects its use in the general
population (repeated or long-term use by a representative
sample) is also needed.
In order to assess a causal relationship between ephedra or
ephedrine consumption and serious adverse events, a

hypothesis-testing study is needed. Continued analysis of case

reports cannot substitute for a properly designed study to assess
causality. A case-control study would probably be the study
design of choice.

Availability of the Full Report

The full evidence report from which this summary was taken
was prepared for the Agency for Healthcare Research and
Quality (AHRQ) by the Southern California-RAND Evidencebased Practice Center, under Contract No. 290-97-0001. It is
expected to be available in March 2003. At that time, printed
copies may be obtained free of charge from the AHRQ
Publications Clearinghouse by calling 800-358-9295.
Requesters should ask for Evidence Report/Technology
Assessment No. 76, Ephedra and Ephedrine for Weight Loss and
Athletic Performance Enhancement: Clinical Efficacy and Side
Effects. . In addition, Internet users will be able to access the
report and this summary online through AHRQs Web site at
www.ahrq.gov.

www.ahrq.gov
AHRQ Pub. No. 03-E021
March 2003
ISSN 1530-440X

Evidence Report

Chapter 1. Introduction
Purpose
This evidence report details the methodology, results, and conclusions of a comprehensive
literature review and synthesis of evidence on the efficacy and safety of ephedra and ephedrine,
either alone or in combination with other substances, to promote weight loss or to enhance
athletic performance. Meta-analysis was performed where appropriate.

Scope of Work
At the direction of the funding agencies (National Institutes of Health Office of Dietary
Supplements (ODS), National Centers for Complementary and Alternative Medicine (NCCAM),
and Agency for Healthcare Research and Quality (AHRQ) and in consultation with our
Technical Expert Panel (see Table 2, Chapter 2), we addressed research questions regarding the
efficacy of herbal ephedra and synthetic ephedrine for weight loss and athletic performance. We
assessed the safety of these products through review of clinical trials. In addition, we reviewed
herbal ephedra-related adverse events reports on file with the U.S. Food and Drug
Administration (FDA), published case reports, and reports to a manufacturer of ephedra
products. It is expected that the results of this review will be used to direct further research.
In searching for evidence of efficacy and safety, we were directed to assess studies using
both the isolated alkaloid, ephedrine, and whole herb or extracts of the herb ephedra.

Background
A 2000 survey by manufacturers of ephedra-containing supplement products estimated that
three billion servings of these products were consumed in the prior year; these findings were
revealed during testimony at a Public Meeting on the Safety of Dietary Supplements Containing
Ephedrine Alkaloids held August 8, 2000. According to Michael McGuffin, an industry
spokesperson, this figure represented a 65 percent increase in sales volume over the previous five
years and would correspond to approximately $6.8 billion in total sales.1 Use of ephedrine
alkaloidcontaining products to promote weight loss or enhance athletic performance has
garnered a great deal of media attention over the last year. This attention is due in part to a
number of well-publicized adverse events reportedly associated with the use of ephedra or
ephedrine alkaloidcontaining products.2-7
Herbal ephedra has been used in China to treat respiratory conditions for over 5,000 years;8
however, the herb is not used for weight loss or physical performance enhancement in eastern
medicine. Its active alkaloid, ephedrine, was first used in western medicine as an asthma
treatment in the 1930s. Since then, ephedrine and other sympathomimetic alkaloids have been
used in many over-the-counter (OTC) decongestants and cold medicines. It was not until the
3

early 1990s that herbal ephedra and other products containing ephedrine began to be promoted as
weight loss aids in the United States.
Federal regulation of dietary supplement products differs considerably from that of products
that are deemed drugs. Dietary supplement products, including those that contain herbal ephedra
(as distinct from the purified alkaloid ephedrine), are regulated by the Dietary Supplement
Health and Education Act (DSHEA) of 1994. Under DSHEA, new products that contain only
supplement ingredients that were sold in the United States before October 15, 1994 do not
require FDA review before they are marketed, because they are presumed to be safe based on
their history of use by humans. Manufacturers of a dietary supplement that contains a new
ingredient not sold as a dietary supplement before 1994 must notify FDA of their intent to
market that product and must demonstrate reasonable evidence for the safety of the product to
humans. In turn, FDA can bar the new ingredient from the marketplace for safety reasons.
However, manufacturers are not required to perform clinical or other studies to establish the
safety of their products before marketing. Once a dietary supplement is marketed, FDA can
restrict its use or order its removal from the marketplace only if it can prove that the product is
not safe. In contrast to the rules for dietary supplements, before a drug product can be marketed,
the manufacturer must obtain FDA approval by providing convincing evidence that it is both safe
and effective.
On October 11, 1995, in response to a growing number of adverse event reports submitted to
the FDA about ephedra-containing products (more than 300 at the time), the FDA convened an
open meeting of the Special Working Group on Food Products Containing Ephedrine Alkaloids
(a working group of the Food Advisory Committee) to assess the potential public health
problems associated with dietary supplements and other food products that contained botanical
sources of ephedrine alkaloids (that is, ephedra). The reported adverse events involved primarily
the cardiovascular and central nervous systems. Most events occurred in young to middle-aged
women, often those using the products for weight loss or to increase energy. Based on the reports
and the evidence they heard, the working group found sufficient evidence to suggest that adverse
effects were associated with the use of ephedrine alkaloids, that safe levels should be established
and that warning labels should appear on products containing the ephedrine alkaloids, regardless
of their source.
In August 1996, the FDA convened a meeting of its Food Advisory Committee to continue
the discussion of the safety of ephedrine alkaloidcontaining foods and supplements. By that
time, the number of adverse events reported to the FDA had doubled from the year before to over
600. As a result of that meeting, some members recommended removal of dietary supplements
containing ephedra from the market.9 Other members suggested that the FDA develop rules on
use that would help reduce the risk of adverse events. In 1997, the FDA published a proposed
rule on use of dietary supplements containing ephedrine alkaloids. It proposed a dose limit of 8
mg ephedrine alkaloid per serving, a daily limit of 24 mg, a duration limit of 7 days, and various
label warnings. After the rule was published in the Federal Register, the FDA received a large
number of comments from consumers, physicians, scientists, and supplement manufacturers. In
response, the General Accounting Office (GAO) audited the methods used by the FDA to
develop the proposed rules. In July 1999, the GAO reported that the FDA had insufficient
4

evidence to support dosage and duration limits. As a result, in early 2000, the FDA withdrew a
large part of the 1997 proposal.
However, the controversy over ephedra has continued. From 2000 to 2002, more than 100
people sued makers of ephedra products,7 and from 1992 through 2002, more than 1000 health
problems were reported to FDA. These reports led a nonprofit consumer group, Public Citizen,
to file a lengthy petition in 2001, asking the FDA to ban the production and sale of ephedra
products. In the fall of 2001, the National Football League banned the substance following the
deaths of several high school and college athletes after alleged use of ephedrine-containing
products, and in January 2002, the Canadian government issued a warning against use of
ephedra.
On June 14, 2002, the U.S. Department of Health and Human Services proposed an expanded
scientific evaluation of ephedra. The agenda for that research will be based on the findings of the
current report.

The Problem of Obesity

From 1999 through 2002, the prevalence of obesity in the United States increased by 1
percent per year, reaching a level of 19.8 percent among the adult population. This increase
represents a 65 percent rise in the prevalence of obesity from 1991 to 2002 (from 12 percent to
19.8 percent), although a precise comparison is difficult because of changing definitions of
obesity.10-12 Obesity is currently defined as a body mass index of 30 or greater: BMI is obtained
by dividing body weight (in kilograms) by the height (in meters) squared. Overweight
individuals are those whose BMI falls between 25 and 29.9. According to that definition, by the
year 2000, the majority of Americans (56 percent) were overweight.10 Moreover, according to
the 1999 National Health and Nutrition Examination Survey (NHANES), 13 percent of children
and adolescents are currently seriously overweight and are displaying increasing rates of obesityrelated chronic diseases such as Type II diabetes, not previously seen in children.13 Attempts to
meet the body weight goal of the Healthy People 2000 Initiative14 (reducing the prevalence of
overweight among adults to less than 20 percent of the population) have failed.
The United States is not alone is facing rising rates of obesity. In Canada, between 1985 and
1998, the overall prevalence of obesity increased in adults from 5.6 percent to 14.8 percent and
from 1981 to 1996, it tripled in children.15, 16 The World Health Organization reports that there
are more than 300 million obese people in the world, and the rising rate of obesity is no longer
solely a problem of industrialized countries, but one that is rapidly appearing in developing
countries.17, 18
In addition to Type 2 diabetes, other serious health risks are associated with obesity. Rates
and severity of hypertension, dyslipidemia, insulin resistance (Syndrome X), coronary artery
disease, stroke, sleep apnea, osteoarthritis, certain cancers, and other conditions increase with
increasing weight.19, 20 Further, obesity increases the rate of mortality as well as morbidity,
especially mortality associated with heart disease and diabetes.21 Using data from five large
prospective cohorts, Allison and colleagues estimated that in 1991, 280,000 deaths were
attributable to excess weight.22 Patients with a BMI greater than 30 accounted for more than 80
percent of the obesity-attributable deaths.
5

The costs of obesity to the health care system are large and growing with the increasing rates
of obesity. In 1986, when only 34 million Americans were clinically obese, a conservative
estimate of the economic costs related to obesity was $39.3 billion.23 By 1995, one study24
estimated direct cost for obesity at $70 billion, although another study estimated these costs at 25
percent lower.25 The costs of obesity are estimated to be higher than those for either smoking or
excessive drinking.26
Intentional weight loss by obese persons leads to reductions in risk factors for disease. A
minimum loss of 5 percent to 10 percent of body weight followed by long-term weight
maintenance can improve health outcomes.27 Despite this finding, only 42 percent of obese
people surveyed by Galuska and colleagues reported that their doctor recommended weight
loss.28, 29 Still, much of the population reports that they are actively trying to lose weight: a 2000
survey showed that one third (38 percent) of subjects were actively trying to lose weight and
another third (36 percent) were trying to maintain their weight.11 Furthermore, among those who
were overweight, 45 percent of subjects were actively trying to lose weight, and 35 percent were
trying to maintain their weight. Among those who were obese, 66 percent of subjects were
actively trying to lose weight, and 21 percent were trying to maintain their weight.11 In a
population-based study of 14,679 U.S. adults in 5 states using the 1998 BRFSS data, seven
percent reported using nonprescription weight loss products; 2 percent reported using
phenylpropanolamine and one percent reported using ephedra products from 1996 to 1998.
More women used ephedra products than men; 1.6 percent of women and 0.4 percent of men
reported using weight loss products containing ephedra. Extrapolated nationally, this study
estimated that during 19961998, 2.5 million Americans used weight loss products containing
ephedra. This study also has data to suggest that many individuals are not aware they are taking
weight loss products that contain ephedra. Of the 183 respondents in Michigan who responded to
the questions about using ephedra and reported that they took other nonprescription weight
loss products, 33 percent reported using name-brand products that claim to contain both ephedra
products and chromium picolinate.30
This estimate of use of ephedra-containing products may be low. Heber and Greenway state
that the most widely used herbal products for weight loss contain ephedrine alkaloids.31 Among
230 (61 percent) of 376 adults in the St. Paul/Minneapolis area who reported using an herbal
product during the past 12 months, 44 (19 percent) used ephedra. Of these 44, 20 (45 percent)
used ephedra for weight loss. Therefore, 5.3 percent of these adults (20 of 376) reported using
ephedra for weight loss.32

Enhancing Physical Performance

Stimulants have a long history of use in athletic performance, dating back to the early
1900s.33 Several serious accidents in the late 1960s, including the death of a cyclist using
amphetamines, spurred the International Olympic Committee (IOC) to ban stimulants from use
during competition. However, this ban was not fully enforceable until a reliable screening test
became available in 1972.34 Use of OTC stimulants is regulated somewhat differently than that
of stimulants available only by prescription, because OTC stimulants are widely available in
products used to treat common conditions such as colds or congestion. Therefore, these

compounds are not banned outright, but athletes whose use of these substances exceeds some
reporting threshold are subject to censure.34
Use of dietary supplements by athletes is common and somewhat more frequent than that of
the general public. In a review of 51 studies, Sobal and Marquart35 found that 56 percent of
athletes used one or more supplements. Another study of college athletes reports a 42 percent
prevalence of supplement use.36 Supplement use was higher among men than among women,
and higher among elite athletes and in particular sports such as body building, weight lifting, and
ultramarathon running.35 A survey among elite Australian swimmers supports this finding: 94
percent of them reported using dietary supplements. All participants reported using vitamins
and/or minerals, and 61 percent reported using herbal preparations.37 Supplement use is prevalent
even among younger athletes: 20 to 25 percent of adolescents are reportedly using supplements.
For all athletes, performance enhancement is cited as a common reason for use, and multivitamins are the most frequently used dietary supplements.35
OTC stimulants, particularly ephedrine or its related alkaloids, are among the substances
most frequently detected on drug screens or reported in surveys. In a series of 1,256 positive
drug screens identified by IOC laboratories in 1989, 40 percent involved stimulants. Ephedrine
alkaloids accounted for 75 percent of the stimulants reported.38 Evaluation of drug use by student
athletes in the most recent National Collegiate Athletic Association (NCAA) survey (2001)
showed that ephedrine and amphetamine use increased from 1997 to 2001, at a time when use of
many other substances was declining.39 Ephedrine was used by 3.5 percent of responders in
1997, a figure that increased to 3.9 percent by 2001. In a survey of 511 subjects attending a
gymnasium, self-reported use of ephedrine exceeded that of anabolic steroids (25 percent versus
18 percent for men; 13 percent versus 3 percent for women).40 The authors asserted that
extrapolating these figures to the general public would suggest that 2.8 million people have used
ephedrine-containing products to improve athletic performance within the last three years.
Further, there may be a subset of committed users of ephedrine products who take high doses for
extended periods of time. Gruber and Pope41 reported on a cohort of female weightlifters, 56
percent of whom were using doses of 120 mg ephedrine daily for over one year. Some
individuals had been using such doses continuously for over five years, and the majority of these
women continued to use ephedrine despite the presence of adverse symptoms.

History and Pharmacology

Ephedra species have a long history of medicinal use, documented in medical treatises from
China and India. Some experts have called it the oldest medicinal plant in continuous use:42 A
species of ephedra was found in a Neanderthal grave and was presumably used medicinally.43
Use by Dioscorides, the famous Hellenistic herbalist, has been documented, as has use in Europe
from the 15th to the 19th centuries.44 Use of ephedrine, the principal alkaloid in ephedra, gained
notoriety during modern times when it was learned that the drug was given parenterally to
Japanese kamikaze pilots during World War II.45 Over 40 species of ephedra are found
throughout Asia, Europe, and the Mediterranean area, as well in North and South America.

Botany
Ephedra, or ma huang, is the common name for any one of three species grown medicinally
in China and recognized in the Chinese Materia Medica: Ephedra sinica, Ephedra equisentina
and Ephedra intermedia.46
The branches of this small twiggy shrub have been used in the practice of traditional Chinese
medicine to treat colds, fevers, and wheezing and as a diaphoretic and diuretic.47 Botanically
related species have also been used in traditional Indian and Tibetan medicine for similar
indications.48 In the modern discipline of phytomedicine, ephedra has been approved by the
German Commission E to treat diseases of the respiratory tract with mild bronchospasm in
patients over 6 years of age.49 In addition to the three species mentioned above, others, such as
Ephedra distachya or Ephedra gerardiana may be used for preparation of commercial
products.49 North American ephedra species, such as Ephedra nevadensis, commonly known as
mormon tea, reportedly contain little or no ephedrine.50, 51
Phytochemistry
The active components of ephedra (about 1.32 percent by weight) are the phenylalaninederived alkaloids such as (-)-ephedrine, (+)-pseudoephedrine, (-)-norephedrine, and
(+)-norpseudoephedrine, which is also called cathine.45, 52 Alkaloid content and composition may
vary based on species and growing conditions such as geographic location, altitude, and soil
pH.53-56 Ephedra is harvested in the fall, when the alkaloid content is highest.57 Even though the
total alkaloid content can vary from 0.5 percent to 2.3 percent, ephedrine accounts for the
majority of the alkaloids (up to 90 percent of total), followed by pseudoephedrine (generally up
to 27 percent of total).45, 58, 59 One species of ephedra, Ephedra intermedia, has been reported to
have reversed ratios of ephedrine to pseudoephedrine, with approximately 30 percent ephedrine
and up to 75 percent pseudoephedrine.57, 60 Norephedrine content is generally very low in
commercial ephedra species.60 Ratios of the most common alkaloids vary in commercial
preparations, but ephedrine and pseudoephedrine account for 90 to 100 percent of the alkaloids
measured.61 The relative potency of the alkaloids is discussed below.
Pharmacology of Ephedrine/ Ephedra
Although ephedrine was first isolated from ma huang in 1887, it was not until early in the
twentieth century that the pharmacology of ephedrine and its related alkaloids was considered by
Western medicine.44, 62-64 Ephedrine is defined as a mixed sympathomimetic agent, which acts
indirectly by enhancing the release of norepinephrine from sympathetic neurons and directly by
stimulating alpha and beta adrenergic receptors.65 The other, related, alkaloids have similar
activities, although they are less potent than ephedrine.62 Thus, the pharmacologic activity of a
given ephedra sample depends on its alkaloid composition.
In the cardiovascular system, ephedrine increases heart rate and therefore cardiac output.65, 66
Because of its peripheral vasoconstriction activity, ephedrine increases peripheral resistance and
can lead to a sustained rise in blood pressure. As a result, parenteral ephedrine has been used to
treat shock and hypotension associated with cesarean section. Elevations in blood pressure
appear to be dose dependent in humans.67 However, there appears to be a threshold effect: doses
under 50 mg do not necessarily result in increased blood pressure.
8

In the lung, ephedrine acts via the beta (2) adrenergic receptors to relax bronchial smooth
muscle.65, 66 However, ephedrines use as a bronchodilator (which began in 1924),63 has largely
been supplanted by more selective agents for chronic use. Currently, ephedrine is used as a
decongestant and for the temporary relief of shortness of breath due to bronchial asthma.
Because of its lipid solubility, ephedrine crosses the blood-brain barrier where it acts as a
central nervous system stimulant.65, 66 Immediate effects are attributable to stimulation of
dopamine release, but ephedrine also acts on central adrenergic receptors, which increases
release of central norepinephrine. This combination of adrenergic and dopaminergic effects
leads, in the short term, to improved mood and heightened alertness with decreased fatigue and
desire for sleep. Physical activity also increases. Concern exists that because of its chemical
similarity to amphetamines, ephedrine may have potential for abuse. Ephedrine has demonstrated
reinforcing effects in humans that are similar to those of amphetamines but not as strong.68 At
higher doses, the release of norepinephrine causes anxiety, restlessness, and insomnia.
Ephedrine and its alkaloids may promote weight loss via several mechanisms. First,
ephedrine may exert an anorexic effect via central effects of norepinephrine on satiety centers in
the hypothalamus.69 Second, stimulation of beta (3) receptors in brown fat, via release of
catecholamines, leads to increased lipogenesis.70 Third, of the three principal alkaloids of
ephedra (ephedrine, pseudoephedrine, and phenylpropanolamine), ephedrine is the most potent
thermogenic agent (a substance that increases the portion of ingested calories that are dissipated
as heat, at the expense of energy storage).
Pharmacokinetics
Ephedrine is readily absorbed from the gastrointestinal tract, with peak concentrations of an
oral, immediate-release dose achieved at approximately two to three hours.71-73 It is distributed
widely throughout the body, crossing the blood-brain barrier, as mentioned above, as well as the
placenta. The half-life of ephedrine in the blood (the time required to reach half the peak
concentration) is six hours.73 Metabolism does occur in the liver, but the majority of ephedrine
(6097 percent) is excreted unchanged via the urine.74, 75 The pharmacokinetics of
pseudoephedrine and phenylpropanolamine (norephedrine) are similar.73
The disposition of a pharmaceutical preparation of ephedrine (25 mg) in ten healthy
volunteers was compared with that of three botanical preparations that contained a roughly
equivalent alkaloid dose.76 Among the four products tested, the time to achieve peak
concentration ranged from 2.61 to 3.05 hours, and the elimination half-life (time that was
required for half of the ingested product to be eliminated) varied from 4.85 to 6.47 hours. None
of the botanical preparations tested was found to have statistically different pharmacokinetics
from the purified ephedrine. These results are not completely confirmed by a second study,
which compared purified ephedrine with a botanical preparation.77 This study found that the
absorption of the botanical preparation was slower and took almost twice as long as the
pharmaceutical preparation to reach maximal concentration (3.90 versus 1.69 hours). However,
maximal concentration of ephedrine was actually higher for the botanical preparation.
Elimination half-life was between five and six hours for both preparations.

Combination Formulas Used for Weight Loss

Pharmaceutical preparations of ephedrine for weight loss often include caffeine and/or
aspirin. Caffeine alone has been shown to stimulate thermogenesis and weight loss, both as an
isolated alkaloid and as a botanical tea.67, 78, 79 Further, caffeine potentiates the thermogenic
effects of ephedrine by acting as an adenosine receptor antagonist and inhibiting cellular
phosphodiesterase activity.80, 81 Botanical preparations often mimic these combined formulations
by including caffeine- or salicylic acidcontaining herbs or those that contain sympathomimetic
amines such as Sida cordifolia (country mallow) or Citrus aurantium (bitter orange) (see
Table 1). Other herbs frequently included in botanical weight loss formulas include those with
diuretic or laxative actions.

10

Chapter 2. Methods
Original Proposed Key Questions
The topic of this report was nominated by the National Institutes of Health (NIH) Office of
Dietary Supplements (ODS). The following questions were originally proposed:

Weight Loss
1. What is the evidence for efficacy of ephedra-containing dietary supplement products for
weight loss, over a sustained period of time?
2. Can efficacy for weight loss be attributed to ephedra alone, or ephedra in combination
with other ingredients (e.g., caffeine)?
3. Does ephedra have additive effects with other agents?
4. What dosage levels of ephedra are necessary to achieve weight loss?

Athletic Performance
1. What is the evidence for efficacy of ephedra-containing dietary supplement products in
terms of energy enhancement and enhancement of athletic performance, over a sustained
period of time?
2. Can efficacy for energy enhancement and enhancement of athletic performance be
attributed to ephedra alone, or ephedra in combination with other ingredients (e.g.,
caffeine) that produces energy enhancement and/or enhancement of athletic
performance?
3. Does ephedra have additive effects with other agents?
4. What dosage levels of ephedra are necessary to achieve energy enhancement and
enhancement of athletic performance?

Safety Assessment
1. Does use of ephedra-containing dietary supplement products over a sustained period of
time increase the risk of cardiovascular disease (CVD) or other serious and lifethreatening events in specific populations?
2. What populations are at risk of CVD and other life-threatening events through use of
ephedra over a sustained period of time?
3. Can the risk for adverse events in these populations be attributed to ephedra alone, or in
combination with other ingredients (e.g., caffeine)?
4. Does ephedra have additive effects with other agents?
5. What dosage levels of ephedra produce risk of CVD or other life-threatening events?
6. Do ephedra-containing dietary supplement products alter physiologic markers of
cardiovascular function?
7. What are the metabolic actions of ephedra, so as to explain its beneficial and adverse
effects?

13

In addition to the questions related to ephedra-containing dietary supplement products, the

sponsor also requested a review of the scientific literature on ephedrine (the purified alkaloid)
regarding its efficacy and safety. A brief review of the mechanism of action of ephedra was also
requested.
We were also asked about the gaps in knowledge about the effects of ephedra, alone or in
combination with other agents, on weight loss, energy enhancement, and enhancement of athletic
performance. We were asked to focus on the following categories of potential consumers:
children, adolescents, young athletes (male and female), and adults (male and female).

Technical Expert Panel

Each AHRQ evidence report is guided by a Technical Expert Panel (TEP). We invited a
distinguished group of basic scientists and clinicians, including individuals with expertise in
cardiac electrophysiology, exercise, herbs, obesity and human nutrition, pharmacognosy (the
study of developing drugs from plant and animal sources), pharmacology, and toxicology. Panel
members are listed in Table 2.
Our expert panel meeting was held at RANDs Arlington, Virginia, office on Wednesday,
November 28, 2001. Margaret Coopey, the Task Order Officer, represented AHRQ. Dr. Paul
Coates, head of the NIH ODS, also attended. At the meeting, we discussed the focus of the
report. The TEP agreed that we should review articles that discuss either ephedra or ephedrine.
Studies or case reports on pseudoephedrine were not to be reviewed, except in the context of
ephedra/ephedrine. We agreed to include a brief description of the other alkaloids
(pseudoephedrine, norephedrine, etc.) in the introduction to our report.
The TEP also provided a number of suggestions regarding data collection. These suggestions
are shown in Table 3.

Assessment of Adverse Events

With regard to adverse events, EPC staff and the TEP recognized that, even in aggregate, the
number of patients included in randomized trials was likely to be too few to allow adequate
statistical power to assess the rate of serious adverse events (such as death, myocardial
infarction, stroke, or seizure) due to ephedra. Because of this likelihood, the EPC staff
recognized the necessity of relying on case reports to help inform the sponsor regarding the key
questions concerning serious adverse events. A long discussion occurred at the TEP meeting
about criteria for assessing causality based on case reports. The framework for this discussion
was based on an unpublished article by Cynthia Mulrow, MD (C. Mulrow, personal
communication). This paper summarized the criteria used in all of the major published
algorithms for establishing different levels of causality in case reports of adverse events from
drugs (see Table 4). Our TEP judged that, to establish definite causality from case reports, a dechallenge/re-challenge test needed to be performed (that is, it had to be documented that the
adverse event in question went away when the offending drug was withdrawn and reoccurred
when the offending drug was reinstated). Clearly, such a de-challenge/re-challenge was not
possible or feasible in the case of serious adverse events such as death or myocardial infarction.
14

Consequently, our TEP judged that case reports alone would be insufficient to establish definite
causality between ephedra use and serious adverse events. The TEP discussed the key
characteristics of a case report that would signal the need for additional study. Such
characteristics would include the following:

Documentation (preferably medical) that the adverse event occurred.

Documentation that the patient took ephedra and that the dose and timing were consistent
with the known pharmacology of ephedrine (for cases of death, myocardial infarction,
stroke, or seizure). (The TEP later quantified this characteristic for acute events such as
stroke or myocardial infarction to mean a dose preferably within six hours of the adverse
event and in no cases greater than 24 hours before the adverse event.)

Performance of an evaluation sufficient to rule out other potential causes for the adverse
event.

The TEP and EPC staff discussed extensively the types of information necessary to satisfy
this last criterion. The TEP agreed that the absence of data could not be construed as a negative
result. For example, the absence of information about prior cardiac disease could not be
construed as an absence of cardiac disease. Furthermore, the TEP emphasized that verbal
histories indicating no prior history of serious conditions were not sufficient to rule out
alternative explanations for the most serious adverse events, since unrecognized preexisting
cardiac disease, congenital abnormalities, berry aneurysms in the cerebral circulation, and other
such conditions occur with some frequency and are known to cause death, myocardial infarction,
or stroke without warning in otherwise healthy individuals. Realizing that it would be very
difficult to attempt to define all of the possible evaluations and interpretation of results in the
abstract, the TEP left it to EPC staff to resolve these issues, guided by the three characteristics
listed above.

Literature Search
Our search for controlled human studies of the effects of ephedra and ephedrine began with
an electronic search of library databases in April 2001. Tables 5 and 6 show our specific search
strategies. We started with Medline, which is maintained by the U.S. National Library of
Medicine and is widely recognized as the premier source for bibliographic coverage of
biomedical literature. It encompasses information from Index Medicus, the Index to Dental
Literature, and the Cumulative Index to Nursing and Allied Health Literature (allied health
includes occupational therapy, speech therapy, and rehabilitation), as well as other sources of
coverage in the areas of health care organization, biological and physical sciences, humanities,
and information science as they relate to medicine and health care. We also searched EMBASE,
the Excerpta Medica database produced by Elsevier Science, which is a major biomedical and
pharmaceutical database indexing over 3,800 international journals. EMBASE currently contains
over six million records, with more than 400,000 citations and abstracts added annually. We also
searched BIOSIS, the most complete database for the life sciences; the Allied & Complementary
Medicine Database (AMED); the Manual Alternative and Natural Therapy Index System
15

(MANTIS), which is the largest index of peer-reviewed articles in the area of complementary
and alternative forms of therapy; and the Cochrane Controlled Clinical Trials Register Database.
AMED is produced by the Health Care Information Service library in the United Kingdom. It
covers journals in allied health professions as well as complementary and alternative medicine.
Similarly, MANTIS covers manual, alternative, and natural therapy. The Cochrane Collaboration
is an international organization that helps people make well-informed decisions about health care
by preparing, maintaining, and promoting the accessibility of systematic reviews on the effects
of heath care interventions. The Cochrane Register of Controlled Trials is available on CD-ROM
by subscription.
Our TEP then suggested that we search three additional databases: the International
Pharmaceutical Abstracts; Pascal (produced by the Institut de lInformation Scientifique et
Technique (INIST) of the French National Research Council (CNRS), whose subject areas
include physics, chemistry, biology, medicine, psychology, applied sciences, technology, earth
sciences, and information sciences); and SciSearch. SciSearch contains all records published in
Science Citation Index and additional records from about 1,000 journals whose table of contents
pages are listed and indexed in the weekly Current Contents publications. Every subject area
within the broad fields of science, technology, and biomedicine is included. Mary Hardy, MD,
and Margaret Maglione, MPP, reviewed a total of 1,780 retrieved titles. Of those, 452 articles
were deemed relevant to our undertaking and were ordered. Thirty-four additional articles were
found through mining reference lists, and 64 were contributed by the TEP or AHRQ. We
reviewed the reference list of every retrieved article for additional literature we might have
missed and ordered any we found. Literature was tracked using ProCite and Access software.

Additional Sources of Evidence

We obtained the report Safety Assessment and Determination of a Tolerable Upper Limit
for Ephedra, published in December 2000 by CANTOX Health Sciences and funded by the
Council for Responsible Nutrition, an association of dietary supplement manufacturers. We
ordered copies of all literature cited in this report. We also obtained transcripts of a public
meeting, held in Washington, DC on August 8 and 9, 2000 and sponsored by the HHS Office on
Womens Health, on the safety of dietary supplements containing ephedrine alkaloids. We
contracted with physicians proficient in Japanese and Chinese to search for scientific literature in
their native languages. These searches identified little on the use of ephedra for weight loss and
exercise enhancement because ephedra is not used in that manner in Eastern cultures. In addition,
we found nothing about ephedra on Phytonet, a European database. We also contacted Baptist
University, Hong Kong, which has a database on herbal medicine, as well as the Taiwan Poison
Control Center, but did not receive any data from either.
On January 31, 2002, we spoke to Dr. Phillip Waddington, Director of the Natural Health
Products Directorate for Health Canada. He agreed to send us 60 adverse event reports regarding
ephedra/ephedrine products. However, at the time of this report, we had not received anything.
In January 2002, we created an announcement regarding our projects need for any
unpublished studies on the use of ephedra/ephedrine for weight loss or exercise enhancement.
The announcement was submitted to both the journal Phytomedicine and the Herbalgram
newsletter. The intent was to reach individuals who might know of small studies being done on
16

ephedra or ephedrine of which the TEP were not aware. We receive no responses to this
announcement.
In March 2002, we obtained a recent monograph on ephedra, written by Dennis McKenna,
from the Institute for Natural Products Research, a nonprofit research and education foundation.
Finally, Wes Seigner, an attorney for the Ephedra Education Council in Washington, D.C.,
agreed to send us unpublished industry studies. We developed a confidentiality agreement, and
Mr. Seigner sent us several reports on then-unpublished controlled trials conducted by members
of the council.

Article Review
We reviewed the articles retrieved from the various sources against our exclusion criteria to
determine whether to include them in the evidence synthesis. A one-page screening review form
(checklist) that contains a series of yes/no questions was created to track the articles (Figure 1).
After being evaluated against this checklist, each article was either accepted for further review or
rejected. Two physicians and a policy analyst, each trained in the critical analysis of scientific
literature, independently reviewed each study, abstracted data, and resolved disagreements by
consensus. The principal investigator resolved any disagreements that remained unresolved after
discussions among the reviewers. Project staff entered data from the checklists into an electronic
database that was used to track all studies through the screening process.
To be accepted for analysis, studies had to be controlled clinical trials according to the
following definitions:
Randomized controlled trial (RCT). A trial in which the participants (or other units) are
definitely assigned prospectively to one of two (or more) alternative forms of health care, using a
process of random allocation (e.g., random number generation, coin flips).
Controlled clinical trial (CCT). A trial in which participants (or other units) are either:
(a) Definitely assigned prospectively to one of two (or more) alternative forms of health
care using a quasi-random allocation method (e.g., alternation, date of birth, patient
identifier)
OR
(b) Possibly assigned prospectively to one of two (or more) alternative forms of health
care using a process of random or quasi-random allocation.

Extraction of Study-Level Variables and Results

We abstracted data from the articles that passed our screening criteria onto a specialized
Quality Review Form (QRFsee Figure 2). The form contains questions about the study design,
the number of patients and comorbidities, dosage, adverse events, the types of outcome
measures, and the time from intervention until outcome measurement. We selected the variables
for abstraction with input from the projects TEP. Two physicians, working independently, each
17

extracted data from the same articles and resolved disagreements by consensus. A senior
physician resolved any disagreements not resolved by consensus.
To evaluate the quality of the studies, we collected information on the study design,
withdrawal/dropout rate, method of random assignment (and blinding), and method for
concealment of allocation (the attempt to prevent selection bias by concealing the assignment
sequence prior to allocation). We also calculated the percentage of attrition by dividing the
number of persons who dropped out of the trial (i.e., the number of people who entered the trial
minus the number who completed the trial) by the number of persons entering the trial. The
elements of design and execution (randomization, blinding, and withdrawals) have been
aggregated into a summary score developed by Jadad.82 The Jadad score rates studies on a 0 to 5
scale, based on the answer to three questions:

Was the study randomized?

Was the study described as double-blind?
Was there a description of withdrawals and dropouts?

One point is awarded for each yes answer, and no points are given for a no answer.
Additional points are awarded if the randomization method and method of blinding were
described and were appropriate. A point is deducted if the method is described but is not
appropriate. Empirical evidence has shown that studies scoring 2 or less show larger apparent
differences between treatment groups than do studies scoring 3 or more.83

Meta-Analysis
Selection of Trials for Meta-Analysis
In selecting trials for the meta-analysis of weight loss, we considered all weight loss trials
that included a treatment duration of at least eight weeks. Our TEP suggested that shorter
treatment durations were insufficient to assess long-term weight loss. Trials on athletic
performance encompassed a wide variety of interventions. Because of this heterogeneity, we
compared and contrasted athletic performance studies in a narrative review and did not perform a
meta-analysis. This section focuses on methods used for the meta-analysis of the weight-loss
trials.
Trial Inclusion
The available weight loss trials were judged to be sufficiently clinically homogeneous to
support a pooled analysis. For some trials, several publications presented the same outcome data.
In these cases, we picked the most informative of the duplicates; for example, if one publication
was a conference abstract with preliminary data and the second was a full journal article, we
chose the latter. The publications dropped for duplicate data do not appear in the evidence table
but are noted in the text of Chapter 3, Results. We note that multiple citations of the same article
were removed at the title screening stage of the project.
Based on input from our TEP, we chose weight loss as the most clinically relevant outcome
for the included trials. In order for a trial to be included in the analysis, the associated publication
18

had to report on weight loss as an outcome, provide data prior to the crossover point if the trial
was a crossover design, and contain sufficient statistical information for the calculation of an
effect size. We calculated an effect size for every comparison of interest, e.g., ephedra versus
placebo, at each relevant follow-up time-point, as described below. The effect size is calculated
by dividing the difference between the weight loss in the treatment group and the weight loss in
the placebo group by its standard deviation. The effect size is a unitless measure that is useful
when comparing trials assessing outcomes that are similar (such as weight loss) but are measured
in different ways (pounds versus body mass index). We synthesized effect sizes within
comparison and follow-up subgroups. The percentage of weight lost, compared to pretreatment
weight, is another clinically relevant outcome. However, we did not choose this outcome for our
primary analysis for two reasons. First, pooling percentage of weight loss within a treatment
group (e.g., an ephedra group) eliminates the placebo comparison from the trial and therefore
does not make use of the strength of the randomized controlled design. Comparison of the
treatment group to the placebo group within a trial utilizes the full strength of a randomized
controlled trial, as patients who are similar in all aspects except treatment assignment are
compared to each other. Thus, if one wanted to perform an analysis of weight loss percentage,
we would advise pooling the difference in weight loss percentage between the treatment and
placebo groups. The second, and more important, reason for not performing an analysis of
weight loss percentage, regardless of whether the internal placebo comparison is made, is lack of
data. The vast majority of trials did not report percentage of weight loss as an outcome. As a
result, we would have had to make two assumptions in our calculations. First, to estimate mean
percentage weight loss for a group in a trial, we took the ratio of mean weight loss between
baseline and follow-up divided by mean baseline weight. The mean of a set of ratios does not
equal the ratio of the means, but this would have been the best estimate we could obtain. Second,
to estimate the standard deviation of our ratio, we would have had to use the delta method to
approximate the standard deviation and furthermore would have had to estimate the correlation
between the baseline and follow-up weights to be 0.5. We are unable to check either of these
assumptions. In contrast, the vast majority of trials did report weight loss as an outcome, and also
presented the standard deviation of this statistic. Hence, weight loss became our primary
outcome for analysis.

Stratification of Interventions
The literature included 6 different types of comparisons: (1) ephedrine versus placebo;
(2) ephedrine plus caffeine versus placebo; (3) ephedrine plus caffeine versus ephedrine;
(4) ephedrine versus other active treatment; (5) ephedra versus placebo; and (6) ephedra plus
herbs containing caffeine versus placebo. Only one trial compared the effect of ephedra alone
versus placebo. If a trial had other treatment arms such as caffeine only, we dropped those arms
from our analysis. Effect sizes were pooled separately within each comparison subgroup. In
addition, a cross-subgroup synthesis using meta-regression was conducted on the ephedrine
versus placebo; ephedrine plus caffeine versus placebo; and ephedra plus herbs containing
caffeine versus placebo effect sizes as well as a direct within-study comparison for those few
studies that presented data for more than one comparison, as described below.

19

Weight Loss Effect Size

For each trial, we calculated effect sizes for any of the six comparisons of interest for which
the study provided data. The majority of trials included only one comparisonbetween a single
treatment (e.g., ephedrine) arm and placebo. One trial84 included both an ephedrine plus caffeine
plus aspirin arm and an ephedrine plus caffeine arm. However, we combined these arms into a
single ephedrine plus caffeine arm, based on the clinical reasoning that aspirin has relatively
little effect on weight loss.
Nevertheless, a small number of trials contained more than one relevant comparison between
arms and thus contributed more than one effect size to be considered for analysis. Doublecounting patients is a concern if a trial contributed more than one effect size to an analysis, and
patients were included more than once in calculating those effect sizes. For example, if a trial
had one placebo arm, an ephedrine arm, and an ephedrine plus caffeine arm, it contributed two
effect sizes, both based on the same placebo patients. Fortunately we encountered relatively few
instances of double-counting of patients within the analyses. One trial85 included two ephedrine
doses and a placebo arm and thus contributed two ephedrine versus placebo effect sizes, that is,
two effect sizes within a single comparison group.
Four trials84, 86-88 contributed effect sizes in more than one of the six comparison groups.
Since we conducted the comparison group analyses separately, the four latter trials do not
double-count patients within comparison group analysis. We discuss the possible influence of
multiple effect sizes per study on the meta-regression analysis below.
For each trial, we extracted the means and standard deviations of weight loss between
baseline and the relevant follow-up times for each arm, if available. For example, if a trial with
placebo and ephedra arms reported follow-up data at two months, we extracted the means and
standard deviations of weight loss at two months for the ephedra and placebo arms. If trials did
not report a weight loss mean for any arm, or this mean could not be calculated from the given
data, the trial was excluded from the meta-analysis.
We initially considered four separate treatment duration measurement times: two months,
three months, four months, and six months. However, only one ephedra trial89 and two ephedrine
plus caffeine trials89, 90 reported an outcome measure for a treatment duration of six months.
These numbers are too small to perform a separate pooled analysis on six-month outcomes.
Thus, we considered three treatment durations: The two-month duration of treatment included
only outcomes for 8 weeks of treatment. However, for the analysis of three-month treatment
durations, we included data collected anytime between 12 and 15 weeks, and for the four-month
analysis, we included data collected between 18 and 24 weeks. We also analyzed the rate of
monthly weight loss, as described below.
The large majority of included trials reported weight loss in kilograms; some trials reported
weight loss in pounds. Since an effect size is unitless, data expressed in either unit of measure
could be extracted for analysis. One trial91 reported weight loss only in terms of body mass index
(BMI). Because this measure involves both height and weight, we first transformed the study
data to kilograms by assuming an average height of 68 inches (within a range of reasonable
values, the height that was chosen made little difference in the results).
20

As mentioned above, for each arm in each included trial, we also calculated the mean
monthly weight loss by dividing by the number of months of treatment. Thus, using our previous
example, we calculated the mean monthly weight loss for the placebo and ephedra arms
respectively by dividing the associated two-month mean weight loss by two. For those trials that
had more than one treatment duration time, we used the longest treatment duration time data to
calculate the monthly weight loss. We extracted both weight loss at specific time points (e.g.,
two, three, and four months) and monthly weight loss to compare the results for both types of
outcomes. This comparison allows us to check trends in weight loss, for example, whether
weight loss is linear or dampens over time. Using meta-regression, we verified that weight loss
was linear over the range of time for which data were available by comparing pooled monthly
weight loss rates based on the two-month, three-month, and four-month data separately in each
comparison group. Thus, our primary analysis focuses on monthly weight loss. We note that the
included trials had relatively short-term follow-up; thus, our results address only short-term
weight loss and should not be extrapolated beyond four months.
If a trial reported a standard deviation of weight loss at a relevant follow-up time, we
extracted those data and used them to calculate the standard deviation of the monthly weight
loss. Eight trials84, 87, 88, 92-96 failed to report a standard deviation for weight loss at a given
follow-up time, or a standard deviation could not be calculated from the given data. For these
trials, we imputed the standard deviation of the monthly weight loss by using those trials and
arms that did report a standard deviation. We averaged the monthly weight loss standard
deviations by weighting all arms equally in the imputed value calculation. For those trials
missing standard deviations, we then used the imputed monthly weight loss standard deviation to
calculate the standard deviation for weight loss at the relevant follow-up time.
For each pair of arms, an unbiased estimate97 of Hedges g effect size98 and a 95 percent
confidence interval were calculated. A negative effect size indicates that the treatment arm
(ephedrine or ephedrine plus caffeine, or ephedra plus herbs containing caffeine) is associated
with a larger weight loss at follow-up (or a larger monthly weight loss) than is the comparison
arm, e.g., the placebo.

Performance of Meta-Analysis

We estimated a pooled random-effects estimate99 by combining effect sizes for comparison

subgroups that contained three or more effect sizes. We also report the chi-squared test of
heterogeneity p-value.97
Forest plots were constructed for each comparison subgroup. Each individual trial effect size
is shown with confidence intervals as a box whose area is inversely proportional to the estimated
variance of the effect in that trial. The pooled estimate and its confidence interval are shown as a
diamond at the bottom of the plot with a dotted vertical line indicating the pooled estimate value.
A vertical solid line at zero indicates no treatment effect.
For each trial, we calculated the monthly weight loss percentage for each treatment group
and the placebo group. Monthly weight loss percentage is defined as the mean monthly weight
loss divided by the mean baseline weight in that group. Unfortunately, we were not able to
calculate monthly weight loss percentage on an individual level. To determine the standard
21

deviation of the monthly weight loss percentage, we used the delta method100 and assumed a
correlation of 0.5 between the baseline and follow-up weights.101 For each comparison subgroup,
we pooled monthly weight loss percentages in the treatment groups and placebo groups
separately using a random effects model99 and produced associated 95% confidence intervals.
We acknowledge that combining estimates within treatment groups only, or placebo groups only,
does not take advantage of the randomization and pairing of treatment and control within a trial.
This lack of pairing, and the fact that the monthly weight loss percentage in the treatment group
must be compared to the associated monthly weight loss percentage in the placebo group, should
be kept in mind when interpreting the results of this analysis.
Sensitivity Analyses
When relevant, we conducted sensitivity analyses on subgroups of trials to determine the
robustness of our conclusions. In order to assess the possible impact of attrition, we divided the
trials into two groups: (1) those with less than 20 percent attrition in all arms and (2) all others.
Twenty percent attrition is a commonly accepted threshold above which concerns about bias
increase, due to loss to follow-up. For trials in which attrition was unknown, we assumed it was
not less than 20 percent. We conducted the main analyses for the two attrition strata separately.
We also conducted further analyses on the attrition rates. To determine whether the attrition
rate varied between treatment and placebo groups within a trial, we first collapsed all the
treatment groups together within a trial and estimated a single attrition rate for treatment. We
then conducted a paired t-test that assessed whether the difference between the treatment and
placebo attrition rates within a trial was significantly different from zero. All studies were
weighted equally in this analysis. We also categorized each trial as significant or not significant
based on its effect size. Trials that had more than one effect size agreed in terms of significance
(in other words, the trial reported consistent result with respect to significance at multiple time
points). We then categorized each trial as to whether the attrition rate for the treatment group was
higher than, lower than, or the same as that of the placebo group. We examined the bivariate
distribution of studies into these six categories, (three relationships between group attrition rates
categories, and whether each of these relationships was significant or nonsignificant), and
conducted a chi-squared test of the association between significance and the relationship between
group attrition rates.
When relevant, we also performed our calculations a second time, excluding the trial by
Moheb and colleagues.84 This trial was presented only in abstract form and provided only the
total sample size, not the sample sizes for each arm; thus, we had to assumed equal sample sizes
across arms.
For the ephedrine plus caffeine versus placebo trials, we performed two sensitivity analyses.
In the first, we dropped one trial102 that had synephrine in the ephedrine plus caffeine arm. In the
second, we dropped one arm of one trial103 in which aspirin was combined with ephedrine plus
caffeine; the sensitivity analysis was performed with the ephedrine plus caffeine arm alone.
A final sensitivity analysis concerned the choice of summary statistics to pool. Instead of
pooling effect sizes or standardized mean differences, we applied a weighted mean
difference approach. In the latter, we pooled the absolute differences in weight loss between the
22

treatment and placebo groups, inversely weighted by the trial variances of the differences. That
is, we did not first divide the differences by their standard deviations to produce effect sizes and
then weight by the inverse variances of the effect sizes. If the variances are not homogeneous
and/or the variances are not well estimated, these two methods may not produce the same results.
The weighted mean difference approach has the appeal of being conducted entirely in the clinical
units of interestin this case, pounds.

Analysis of Dose

We tested for a dose effect using a random-effects meta-regression model.104 A separate

model was fitted within each comparison subgroup. We defined a low dose of ephedrine as 10
20 mg; a medium dose of ephedrine as 4090 mg; and a high dose of ephedrine as 100150 mg.
We characterized each dose level as an indicator variable in a main-effects model and chose the
medium-dose group as the level to exclude. The meta-regression approach allowed us to test
directly the efficacy of low and high doses versus the excluded medium dose group, as well as to
estimate the effect size for each dose level.

Publication Bias
We assessed the possibility of publication bias by evaluating a funnel plot of effect sizes for
asymmetry, which can result from the nonpublication of small trials with negative results. These
funnel plots include a horizontal line at the fixed-effects pooled estimate and pseudo95%
confidence limits.105 If bias due to nonpublication exists, the distribution is asymmetric or
skewed. Because graphical evaluation can be subjective, we also conducted an adjusted rank
correlation test105 and a regression asymmetry test106 as formal statistical tests for publication
bias. The correlation approach tests whether the correlation between the effect sizes and their
variances is significant, and the regression approach tests whether the intercept of a regression of
the effects sizes on their precision differs from zero; that is, both formally test for asymmetry in
the funnel plot. We acknowledge that other factors, such as differences in trial quality or true
study heterogeneity, could produce asymmetry in funnel plots.

Meta-Regression
As described above, in order to compare monthly weight loss effect sizes across
comparisons, we conducted a random-effects meta-regression.104 The observations in this metaregression were all monthly weight loss effect sizes across the ephedrine, ephedrine plus
caffeine, and ephedra plus caffeine-containing herbs comparisons. The variables are indicator
flags, one for each comparison. Only one trial85 had multiple effect sizes in the regression, and
we did not account for the correlation between these two effect sizes in our model.
Three trials84, 86, 88 contained both ephedrine and ephedrine plus caffeine arms. For these
trials, we were able to conduct a direct, or head-to-head, comparison of these treatments by
pooling the effect sizes for each trial together. In the estimation of an effect size in this situation,
the comparison group is that group of individuals who received ephedrine alone. Thus, a
negative effect size means that ephedrine plus caffeine is associated with a larger monthly weight
loss than is ephedrine alone. This direct comparison is more robust than the cross-group metaregression described above, because the former compares groups only within a trial. However,
due to the small number of trials that provided more than one treatment arm and the lack of any
23

direct comparisons of ephedrine alone or ephedrine plus caffeine versus ephedra, we conducted
both analyses.
Interpretation of the Results
To aid in interpreting our results, we back-transformed all pooled estimates to weight loss in
pounds. In order to do this, we multiplied each pooled estimate by the average standard deviation
across trials, and then further multiplied by 2.2 to transform kilograms to pounds. In this way, we
were able to equate our unitless pooled effect size with weight loss in pounds. However, we note
this back-transformation requires assuming a particular underlying standard deviation. Readers
may wish to apply their own standard deviation, based on the particular patient population to
which they wish to apply the results.
We conducted all analyses and drew all graphs using the statistical package Stata.107

Safety Assessment
Controlled Trial Adverse Events
Data Collection
Each trial that we identified was examined to determine whether it reported data on adverse
events. Adverse events were recorded onto a spreadsheet that identified each study arm, the
description of the adverse event as listed in the original article, the number of adverse events in
each category, and the number of subjects in each arm.
Meta-Analysis
The strongest level of evidence for attributing an adverse event to an exposure comes from
placebo-controlled randomized trials of the exposure in question. In this evidence report, such
evidence would come from placebo-controlled trials of ephedra or ephedrine. We therefore
searched all such trials that we identified and extracted from each trial the adverse events that
were reported associated with it, as described above. Because each event was counted as if it
represented a unique individual, and because a single individual might have experienced more
than one adverse event, this method may have overestimated the number of people having an
adverse event. We then compared event rates in the people who received ephedra or ephedrine
with those in people who received placebo. We performed a meta-analysis on those adverse
events for which there was an appreciable number of reports in the randomized trials.
We collected data on adverse events for the randomized controlled trials. For each adverse
event, e.g., vomiting, and for each treatment group and for the placebo group, we abstracted
either the number of events or the number of people, depending on how the trial chose to report
events. The majority of trials recorded the number of events, rather than the number of unique
people who experienced the event. We treated all events as if they occurred in unique
individuals, which, as we stated, may overestimate the number of people apparently affected in a
particular event category.
We note that some trials recorded zero events in a particular event category, and these data
were thus recorded. However, some trials recorded no data for a certain event category or
24

recorded no adverse events at all. These trials did not enter the adverse event meta-analysis, in
that we did not assume zero observed events if a trial did not mention a particular type of event.
By excluding these trials, we may have underestimated the number of patients for whom a
particular adverse event was not observed. We note that, for the power calculation (described
below) for serious adverse events (deaths, myocardial infarctions, strokes, seizures, and serious
psychiatric symptoms), the sample sizes of all trials were taken into account, regardless of
whether they mentioned these serious events. We assumed that such serious events would have
been recorded had they been observed, so that a record of zero or no mention of a serious event
could both be taken to mean that no such events were observed.
After abstracting the data, we identified mutually exclusive subgroups of similar events,
based on clinical expertise. When we subgrouped events, we again treated all observed events as
having occurred in unique individuals. For example, we considered nausea and vomiting as a
single subgroup. For a trial that reported nausea events and vomiting events separately, we
assumed the events that occurred in each category were unique and occurred in different
individuals. The number of individuals who were at risk of being affected is the total number of
patients in the trials relevant group (placebo or treatment).
For each event subgroup, we report the number of trials that provided data for any event in
the subgroup. We also report the total number of individuals in the placebo groups in the relevant
trials who were observed to have experienced the event (calculated as described above) and the
total number of patients in the placebo groups in those trials. We then report the analogous
counts for all applicable treatment groups (ephedrine, ephedrine plus caffeine, ephedra) in the
relevant trials. We specifically do not provide crude placebo and treatment rates (total number of
affected patients divided by total number of patients at risk). Such crude rates do not weight
trials appropriately.
Based on clinical importance and the availability of data, we chose a limited number of event
subgroups for meta-analysis. For each chosen event subgroup, we estimated the pooled odds
ratio across the trials that reported on any events in the subgroup, as well as a 95% confidence
interval for the pooled odds ratio. Given that many of the events were rare, we utilized exact
conditional inference to perform the pooling rather than applying the usual asymptotic methods
that assume normality. Asymptotic methods require corrections if zero events are observed:
Generally, half an event is added to all cells in the outcome by treatment two-by-two table in
order to allow estimation, since these methods are based on assuming underlying continuity.
Such corrections can have a major impact on the results when the outcome event is rare. Exact
methods do not require such corrections. We conducted the meta-analysis using the statistical
software package StatXact.108
We also conducted a power calculation to determine the lowest adverse-event rate that the
clinical trials we identified had at least 80 percent power to detect. That is, we assumed a sample
size equal to all the trials combined, and assuming a two-sided test of level 0.05, we determined
the lowest detectable adverse-event rate. This calculation was performed to assess the statistical
power we actually had available to detect adverse events if few or none were observed. Even if
no adverse events are observed, we cannot necessarily conclude that the rate is zero, because the
sample size available may have been too small to detect a rare event.
25

Case Report Adverse Events

Data Collection
Because the clinical trial data had low statistical power to detect a rate of serious adverse
events, we therefore assessed case reports of adverse events associated with ephedrine or
ephedra-containing dietary supplement use in order to inform the sponsors regarding the Safety
Assessment key questions concerning serious adverse events. We reviewed case reports from
three sources: the FDA MedWatch file, published case reports, and a file kept by the ephedra
supplement manufacturer, Metabolife. Published case reports were identified through our
literature search process previously described.
FDA Medwatch Data
In September 2001, the FDAs Office of Nutritional Products, Labeling, and Dietary
Supplements produced an Excel spreadsheet with a master list of adverse-event report case
numbers and summary information, in response to our request for all herbal ephedrarelated
adverse-event reports from their database. After several discussions and several months of work,
the dataset construction algorithm was reproduced and limited to only herbal ephedrarelated
adverse-events reports, because some ephedrine adverse events had mistakenly been entered into
the initial Excel file. We also received several sets of compact disks containing portable
document format (PDF) files of events reported in the FDA Adverse Reaction Monitoring
System (ARMS) for the dates specified. Documents retrieved included MedWatch Reports (FDA
form 3500); Consumer Complaint Injury Reports (FDA Form 2516); Complaint/Injury Followup Forms (FDA Form 2516a); Adverse Reaction Questionnaires (Form A); letters from family
members, health care professionals, or lawyers; affidavits collected from witnesses during FDAheld investigations; police reports; medical records, including physician notes (both inpatient and
outpatient), emergency department reports, nurse notes, and laboratory reports; product labeling
and related information; and product analysis results.
The second master list of only ephedra-related adverse-event reports was created at the
product level, so that adverse-event report identification numbers (IDs) were repeated if multiple
products appeared in one report. Because our analysis was at the adverse-event report level,
where there were multiple products per single ID, we joined those into one record. We
established a cutoff date of September 30, 2001, the production date for the CDs that contained
actual reports. Our analysis does not include case reports filed after that cutoff date, since these
files had not been redacted of identifying information.
The data were analyzed in a series of steps. First, we coded each unique report according to
type of adverse event listed in the summary information on the Excel spreadsheet. The categories
into which we grouped the reports are listed in Table 7. Then, we separated those reports with
events coded as most serious (death, stroke, myocardial infarction (MI), seizure, and certain
psychiatric symptoms) from those considered moderately serious. Reports that contained events
considered most serious were analyzed using specialized data-collection instruments called
Adverse Events Analysis Forms (AEA Formssee Figures 3a3c). We developed these
instruments to collect information from the corresponding PDF file or published case report on
whether the report was actually on ephedra or ephedrine, whether the data were adequate to
analyze the report, and whether or not the adverse event qualified as a sentinel event (see
26

below). When a case report dealt with more than one individual, an AEA form was completed
for each individual.
To understand the other potentially serious adverse events, we reviewed all case reports that
had been grouped into the categories of other serious cardiovascular, other serious
neurological, and psychiatric in our initial review of the master Excel file. For this review, we
used a brief data abstraction form (Figure 4). This brief form was developed to assess the
evidence supporting the prior use of ephedra and to define the adverse event more precisely.
Again, when a case report contained more than one subject, a brief form was completed for each
subject. Then, the data collected in the brief review were used to justify including certain moreserious events into the more-detailed review described above. These more-serious adverse events
included ventricular tachycardia/fibrillation, cardiac arrest, pulmonary arrest, transient ischemic
attack, and brain hemorrhage. Select adverse-event reports were then reviewed a second or third
time by project staff physicians to reach an implicit judgment about whether an adequate
investigation of other potential causes had been performed. Internists performed the initial
reviews of cases of death, myocardial infarction, stroke, and seizure, and were assisted (as
appropriate) by a cardiologist, rheumatologist, or neurologist. Psychiatric events were reviewed
by two experienced professionals: a psychiatrist specializing in addictions and a psychologist
who leads RANDs Drug Policy Research Center. All cases were reviewed by two individuals,
with differences resolved by consensus.
As part of additional work we were requested to perform, we received hard copies of
MedWatch data on ephedrine, organized in the same manner as the data on ephedra. We first
reviewed all these events with our short form to identify the serious adverse events. These events
were then reviewed using the same methods we developed for the ephedra database. Two types
of adverse events associated with ephedrine were not associated with ephedra. The first involved
the intravenous use of ephedrine given during surgery; several such reports were filed by medical
personnel. The second involved attempted suicide. We note these two types of case reports in our
analysis.
Literature Cases
During our literature search, we identified published case reports of adverse events
associated with ephedra use. These published case reports were then reviewed using the same
criteria used for the MedWatch events.
Metabolife File
We received the following materials from the Food and Drug Administration (FDA), which
had, in turn, received them from Metabolife:

A CD-ROM labeled MIPER (described in more detail below).

Photocopies of medical information pertaining to 43 cases (also described in more detail

below).

27

A two-page Listing of Key Complaint for the Metabolife Medical Records Submitted,
which is a listing of the key complaints for 46 cases, with photocopied medical
information. (Note: we received medical information for only 43 cases.)

A two-page sheet entitled Index of Redacted Consumer Medical Records with

Corresponding MIPER Numbers, which contained a listing of the 46 cases with
additional medical record information and the file numbers for related information on the
MIPER CD-ROM.

Three reviews of the Metabolife adverse-event file, which Metabolife commissioned.

Note that to prevent their assessment from biasing our own, we did not read any of these
reviews prior to our assessment, but did review them briefly when our assessment was
completed.

A file entitled 77 serious AEs as identified by Metabolife, which contains photocopies

of reports of events that were selected by Metabolife as being the most serious in nature.
Most, but not all, of these reports were contained on the MIPER CD-ROM. Again, in
order to avoid bias, we did not examine this file until after our initial assessment was
performed.

Several journal articles, all of which were already in our possession.

Later, we also received a report entitled Adverse Event Reports from Metabolife that had
been prepared for Sen. Richard J. Durbin, Rep. Henry A. Waxman, and Rep. Susan A. Davis by
the Minority Staff Report Special Investigations Division, Committee on Government Reform,
U.S. House of Representatives, and which consisted of an analysis of the MIPER CD-ROM.
The MIPER CD-ROM contains several thousand files of adverse event reports organized in
20 folders. The adverse-event files are numbered from 15111 to 35069, and are continuous,
except for three gapsbetween 21121 and 22035; 25535 and 27472; and 30627 and 35047.
Each file is a TIF picture file, generally of a single sheet of paper, on which is recorded
information regarding the potential adverse event or events. This information was recorded in
many different ways, including an email record of a telephone conversation between a company
representative and the consumer; typed or handwritten letters from the consumer to the company;
handwritten notes of telephone conversations with consumers, written on either a rudimentary
form or on whatever piece of paper seems to have been handy at the moment; and a form
developed by Metabolife for systematically collecting information about possible adverse events.
Examples of all of these types of files are presented in Figure 5. Personal identifiers had been
redacted from the files we received.
Each consumer could experience one or more adverse events. We referred to a particular
adverse event for a person as a case, and our analysis was conducted at the case level, rather
than at the person level. Thus, a person could contribute more than one case to our analysis. We
use this terminology throughout the remainder of the report. Practically speaking, in most
instances of serious adverse events such as death, heart attack, or stroke, a person contributed
only a single case in this manner.
28

In general, each file on the MIPER CD-ROM contained only a single sheet of paper. We did
identify some files that were exactly the same as other files, and we excluded these files from our
analysis. The information on a case might reside in a single file or in more than one file. For
example, if a letter from a consumer concerning one of the adverse events experienced by that
consumer was three pages long, each page resided in a separate file. If possible, we tried to
identify all files that pertained to the same case (which we called duplicate files), so that we
would not count a case more than once. Whether we identified all such instances is unknown,
since information in each file was insufficient to allow us to check for duplicate files by
matching on key variables such as age, gender, and the type of adverse event. No other
mechanism for checking was possible within the time and resources available to the project.
Therefore, while we did our best to identify and exclude or in some other way resolve duplicate
files, we cannot be certain that all such files were identified. An example of the difficulty in
identifying duplicates is given in Figure 6. In this instance, Metabolife had identified in the 77
serious AEs document that these two files belonged to the same case. We would not have been
able to make this determination, because the files are separated by more than 7000 numbers on
the MIPER CD-ROM (file 16897 and file 24209), and the notes in one file specify seizure,
whereas the other file states no history of seizure.
In contrast to a duplicate file, a file might contain information on more than one case, either a
set of adverse events all experienced by the same consumer or one or more adverse events
experienced by several different consumers (see Figure 5, Example 5c). For this reason and
because of duplicate files, the number of cases of possible adverse events does not equal the
number of files.
In order to review this large CD-ROM dataset within the given time frame, we chose to have
the initial data collected by a team of abstractors, each working on a portion of the MIPER
CD-ROM. We retained six nurses, each with many years of experience in medical record
abstraction. We developed a one-page data collection form to collect key variables related to age,
gender, nature of the reported adverse event, and need for hospitalization, which is reproduced in
Figure 7. After undergoing training by the principal investigator, each nurse abstractor completed
a sample of 135 records, each of which was reviewed in a group meeting with the principal
investigator to identify areas of possible misinterpretation and vague language. Based on this
experience, we revised the form and developed a codesheet to define how certain complaints
were to be coded. Formal inter-rater reliability testing was performed on a 1 percent systematic
sample of the MIPER files. This sample was stratified into two parts, the larger (N = 114)
portion containing only a single adverse event in each file and the smaller portion (N = 16)
containing more than one case per file. Inter-rater reliability was assessed using both absolute
percentage agreement among abstractors and the kappa statistic, which adjusts for agreement due
to chance. Kappa varies between 0 and 1.0, with values of 0.4 to 0.6 usually indicating moderate
agreement beyond chance, 0.6 to 0.8 indicating substantial agreement beyond chance, and
greater than 0.8 indicating almost perfect agreement.109 Inter-rater reliability testing
demonstrated a kappa statistic of greater than 0.8 or absolute agreement of 95 percent or greater
for all variables, indicating almost perfect agreement, for the one case, one file (N = 114)
records. For the files with multiple cases, two produced disagreement over the number of
multiple cases contained in the file. For the remaining 14 multiple-case files, this analysis
29

showed levels of reliability similar to the one case, one file analysis. Based on these results, we
concluded that the inter-rater reliability for the six nurse abstracters trained in this manner was
acceptable for this project. Each nurse was then assigned approximately one-sixth of the MIPER
file. Questions that arose during abstraction were posted by email or telephone to our EPCs lead
physician abstracter (WAM), who answered their questions, reviewed files himself, and
consulted with the principal investigator on decisions requiring nuanced judgment. He
maintained the codesheet, keeping it upto-date and redistributing it to the abstractors whenever
changes or additions were made.
We reviewed the forty-three cases that included photocopies of medical information.
Personal identifiers had been redacted. Some of these cases were related to cases contained on
the MIPER CD-ROM. However, matching these cases was a challenge. As previously noted, we
were sent a two-page Index of Redacted Consumer Medical Records with Corresponding MIPER
Numbers, which indicated a number and the associated files on the MIPER CD-ROM.
Unfortunately, the medical records we received were not numbered. Furthermore, a second table
that we received entitled Listing of Key Complaint for the Metabolife Medical Records Submitted
contained a list of main complaints, also numbered. However, the two numbering systems did
not agree. We numbered the cases in the order in which we received them in the shipping box.
Our numbering system and the two numbering systems we received start out in agreement, but
discrepancies occur as we progress through and compare the three systems. We did our best to
resolve them.
Analysis of Case Reports
In our draft report, we assigned a likelihood of causality to selected cases, based on our
modification of published methods. Many of the peer review comments received for this report
pertained to our attempts to assign causality. These comments varied widely, ranging from
critiques of our method for being too conservative (meaning, in the opinion of some reviewers,
we had excluded or assigned too low a level of causality to certain cases) to critiques for being
too liberal (meaning, in the opinion of some reviewers, we had assigned too high a level of
causality to certain cases). Often, these conflicting comments concerned the same cases. We
believe these peer review comments demonstrate that case report reviews involve considerably
more subjective interpretation than do reviews of randomized trials. Because our goal in this
evidence report is to report the evidence as objectively as possible, we ceased to assign
assessments of causality to the case reports. Rather, we tried to identify those cases that would be
classified medically as idiopathic in etiology, meaning the cause is not known. For such cases,
given the known pharmacology of ephedrine, if use of ephedra or ephedrine was documented, a
potential role for ephedra or ephedrine in causing the event must be considered. We classified
such cases as sentinel events.
In order to be classified as a sentinel event, three criteria had to be met:
1. Documentation existed that an adverse event meeting our selection criteria occurred.
2. Documentation existed that the person having the adverse event took an ephedracontaining supplement within 24 hours prior to the event (for cases of death, myocardial
infarction, stroke, or seizure).
3. Alternative explanations were investigated and excluded with reasonable certainty.
30

Within the time and resources available for this evidence report, we were able to do an indepth review of FDA case reports only for those events classified as death, myocardial infarction
(which included acute coronary syndromes), stroke (which included intracerebral hemorrhage),
seizures, and severe psychiatric symptoms (see below). Cases that met all three criteria were
classified as sentinel events. Cases where another condition by itself could have caused the
adverse event, but for which the known pharmacology of ephedrine made it possible that ephedra
or ephedrine may have helped precipitate the event, were classified as possible sentinel events.
Probably not related was used for events that had other clear causes discovered on detailed
investigation and to which the pharmacology of ephedrine was unlikely to have potentially
contributed. We also classified many cases as having insufficient information because crucial
information was missing, such as the presence of ephedrine or a metabolite in the blood or
documentation that the patient took ephedra within 24 hours prior to the event (for cases of
death, myocardial infarction, stroke, or seizure); or other possible causes were insufficiently
investigated. (We also classified as sentinel events a few cases that, on detailed review, led us
to question whether an event meeting our inclusion criteria had actually occurred.)
We translated the criteria for identifying sentinel events into the following set of procedures:

We required medical record documentation that an adverse event had occurred.

For adverse events described as seizure, cases described as generalized tonic-clonic

seizures underwent further review.

For psychiatric symptoms, we reviewed cases described as psychosis, mania or severe

agitation, severe depression, hallucinations, confusion or delusion, suicide attempt,
paranoia, or violence.

We required (for all but psychiatric events) that there be documentation that the subject
had consumed ephedra or ephedrine within 24 hours prior to the adverse event, or that a
toxicological examination revealed ephedrine or one of its associated products in the
blood or urine. Cases with no such documentation were not reviewed further. For the
Metabolife cases, we assumed ephedra use to have been within the prior 24 hours for all
but psychiatric events.

For psychiatric cases, we did not require documentation that the product was taken within
24 hours prior to the event. Ephedrine psychosis (as with amphetamine psychosis in
general) is associated with prolonged use, which may lead to neurotoxicity, resulting in
depletion of dopamine and other brain monoamines.110

To be eligible for detailed review to investigate other potential causes of death, a file
required evidence that an autopsy had been performed, and the results had to be available.

To be eligible for detailed review to investigate other potential causes for cases of
myocardial infarction, coronary angiography had to have been performed and the results
had to be available.
31

All cases of stroke that met the criterion of having consumed ephedra or ephedrine within
24 hours were reviewed in more detail. To be classified as a sentinel event, reports of
thrombotic stroke needed to have an assessment for a hypercoagulable state and
vasculitis, reports of embolic stroke needed to have an embolic evaluation performed,
whereas reports of hemorrhagic stroke required an examination to assess structural
problems with the circulatory system of the brain.

Other potential causes of seizure were assessed by searching cases for the results of vital
signs, brain imaging (CT or MRI), serum glucose and electrolytes, blood calcium and
magnesium, an EEG, and prior history of a seizure disorder or substance abuse.

For cases with psychiatric symptoms, cases in which patients had a history of psychiatric
or severe psychological problems were excluded from further review as reports of
possible sentinel events. Cases where the patient reported use of or tested positive for
other substances known to cause psychiatric symptoms were also excluded as possible
sentinel events. For patients with a prior psychiatric history or use of other substances,
these cases were classified as inconclusive.

One of the key questions we were asked to answer by the sponsoring agencies concerned the
relationship between dose and the likelihood of serious adverse events. We do not believe such
an analysis is justifiable based on the case report evidence presented here, for the following
reasons. First, such an analysis assumes a cause-and-effect relationship that has not been proven
by conventional standards of medical science. Second, it would rely to a great extent on patients
recall of dose after having suffered an adverse event, which increases the likelihood of recall
bias. Third, and most important, for more than half the adverse-event cases, no dose data were
available.

Peer Review
This report was subjected to a lengthy peer review process. An initial draft report was
prepared in July 2002. We received comments from 37 reviewers, including representatives from
the American Herbal Products Association; Centers for Disease Control and Prevention;
Consumer Healthcare Products Association; Council for Responsible Nutrition; Food and Drug
Administration; National Center for Complementary and Alternative Medicine; National
Institute of Diabetes and Digestive and Kidney Diseases; National Institute of Neurological
Disorders and Stroke; National Heart, Lung and Blood Institute; National Institute of Health
Office of Dietary Supplements; National Institute of Health Office of Research on Womens
Health; National Nutritional Foods Association; Public Citizen Health Research Group; Center
for Science in the Public Interest; Utah Natural Products Alliance; and members of the U.S.
House of Representatives and U.S. Senate. Additional work requested, involving case report
assessments, was performed during Autumn 2002. The safety section of the revised report,
which contains the new material, was reviewed by additional experts in December 2002. A
complete list of Reviewers is in Table 8.

32

We considered each peer review comment (more than 100 pages in total) and detail our
responses in Appendix 3. Service as a reviewer of this report should not in any way be construed
as agreeing with or endorsing the content of the report.

33

Chapter 3. Results
Results of Literature Search
Efficacy Analysis
Figure 8 displays the flow of the literature review. As a result of computerized library
searches, reference mining, talking to experts, and searching government files (see Methods), we
ordered 553 articles. Of those 553, we were unable to obtain 20 articles, mostly foreign or very
old background articles, none of which appeared (from their titles and keywords) to be clinical
trials of ephedrine or ephedra.
Of the 533 articles collected, 57 reported results from randomized clinical trials or controlled
clinical trials that assessed the effects of either ephedrine or herbal ephedra on weight loss or
athletic performance. The 57 articles, which corresponded to 52 unique controlled clinical trials,
went on to further review and data abstraction. Articles that did not go on to initial data
abstraction included 66 case reports or case series articles that reported adverse events. One
hundred fifty-eight were rejected because they did not discuss ephedra or ephedrine, although
they may have discussed phenylpropanolamine, pseudoephedrine, or caffeine, or provided
general background on herbal medicine, weight loss, or athletic performance enhancement. One
hundred twenty-four articles were rejected because they were not RCT/CCTs and did not report
adverse events. Forty-eight articles were rejected because they did not study human populations.
Another seven articles were duplicates of articles already on file. Fifty-four additional articles
were rejected for design, including previous reviews of ephedra or ephedrine, descriptions of its
chemical properties, editorials, commentaries, letters to journal editors that did not report new
cases, and newspaper or trade journal stories. Eighteen RCT/CCTs were rejected because they
did not concern weight loss or athletic performance.

Adverse Events Analysis

The adverse events analysis includes 52 controlled trials and 46 of the 66 case reports/case
series articles (six articles were rejected because they were duplicates of articles or reports
already included in the analysis).

Efficacy
Weight Loss
Of the 52 unique controlled trials that assessed the effects of either synthetic ephedrine or
herbal ephedra on weight loss or athletic performance, 44 of those assessed the effects of
ephedra, ephedrine, or ephedrine and other compounds on weight loss. Of these 44 trials, 18
were excluded from pooled analysis because they had treatment duration of less than eight weeks
(the longest published weight loss intervention was six months, and no studies assessed postintervention weight maintenance). Six more trials were excluded for a variety of reasons (See
Table 9). We classified the comparisons made in the remaining 20 trials into six categories:
73

1.
2.
3.
4.
5.
6.

Ephedrine versus placebo

Ephedrine plus caffeine versus placebo
Ephedrine plus caffeine versus ephedrine alone
Ephedrine versus another active pharmaceutical for weight loss
Ephedra versus placebo
Ephedra plus herbs containing caffeine versus placebo

For the 16 trials that reported baseline sample sizes, the attrition rate in the treatment arms
averaged 27 percent, whereas the attrition rate in the placebo arms averaged 29 percent. This
difference was not statistically significant. Five trials reported more dropouts from the treatment
than from the placebo group: Four of these trials reported a statistically significant benefit for the
treatment, and one did not. Eight trials reported more dropouts from the placebo group than the
treatment group. Five of these trials reported a statistically significant benefit for treatment,
whereas three did not. Three trials reported an equivalent number of dropouts from the treatment
and placebo groups. No significant association was found between the frequency of favorable
results and the relative proportion of dropouts in the treatment and placebo groups.
Ephedrine Versus Placebo
We identified five trials (which contained six comparisons) that assessed the effect of
ephedrine versus placebo.84-86, 88, 94 A study by Pasquali had two comparison arms that assessed
different doses.85 The scores on Jadads scale (0-5) for these trials were 1, 2, 2, 3, and 3,
respectively. All five were described as randomized, placebo-controlled trials. Three of the trials
(with four comparisons) reported results at three months, and three of the trials reported results at
four months. The random effects pooled estimate of the rate of weight loss per month was an
effect size of -0.50 (95% CI: -0.85, -0.15), which translates to a monthly weight loss of 1.3
pounds more than weight lost on placebo (Table 10 and Figure 9). The pooled average percent
weight loss in the ephedrine-treated patients, compared to pretreatment weight, was 11 percent at
4 months.
A sensitivity analysis on only those trials that scored three or higher on the Jadad scale
yielded a pooled estimate of effect substantially lower than the main analysis (effect
size = -0.20); this difference was statistically significant (p = 0.049). All of these trials had an
attrition rate greater than 20 percent; therefore, no sensitivity analysis on attrition could be
performed. A final sensitivity analysis, in which the trial by Moheb84 was dropped, did not
materially change these results.
In our dose analysis, only high doses of ephedrine resulted in a weight loss that was
significantly greater than zero, and the difference in weight loss between medium dose trials and
high dose trials approached statistical significance (p = 0.052). Neither graphical nor statistical
tests yielded evidence of publication bias (See Table 11 and Figure 10).
We interpret these data to indicate that the use of ephedrine is associated with a statistically
significant increase in weight loss (1.3 pounds of weight loss per month) compared with that of
placebo for up to four months of use.

74

Ephedrine plus caffeine versus placebo

We identified 12 trials that assessed the effect of ephedrine plus caffeine versus placebo for
weight loss.84, 86-88, 90-92, 95, 96, 103, 111, 112 Six trials 86, 87, 92, 95, 96, 112 had scores of three or greater on
the Jadad scale, a threshold that in other settings has been associated with less bias.83 Seven were
described as randomized, double blind, placebo-controlled trials. Four of the trials measured
weight loss at two months, four trials measured it at three months, and five trials measured it at
four months. The random effects pooled estimate of the rate of weight loss per month was an
effect size of -0.85 (95% CI: -1.1, -0.61), which translates to a weight loss of 2.2 pounds per
month above that with placebo (Table 11 and Figure 11). The pooled average percent weight loss
in the ephedrine plus caffeine treated patients, compared to pretreatment weight, was 11 percent
at 4 months.
A sensitivity analysis on only those trials that scored three or greater on the Jadad score
yielded a result similar to the main analysis. Another sensitivity analysis on only those trials that
had less than 20 percent attrition yielded a similar pooled estimate effect size of -0.74
(95% CI: -1.2, -0.3). Two trials96, 112 in this category were randomized, double-blind, placebocontrolled trials with an attrition rate of less than 20 percent. The first112 reported an effect
somewhat greater than the main analysis (effect size = -1.35; 95% CI: -2.2, 0.54). The second96
had a smaller effect (effect size = -0.46; 95% CI: -1.3, 0.34). A fourth sensitivity analysis in
which the trial by Moheb84 was dropped did not change the result compared with the primary
analysis, nor did sensitivity analyses that dropped trials that also included synephrine or aspirin.
In our dose analysis, there was a trend toward increased weight loss with higher doses
(weight loss greater than placebo of 2.0, 2.2, and 2.6 pounds per month for low, medium, and
high doses, respectively) but these differences were not statistically significant. Neither visual
nor graphical tests revealed any evidence of publication bias (See Table 11 and Figure 12).
We interpret these data to indicate that the use of the combination of ephedrine and caffeine
is associated with a significantly greater (2.2 pound) weight loss per month than is associated
with placebo, for up to four months duration.
Ephedrine plus caffeine versus ephedrine
We identified three trials that included arms that compared a combination of ephedrine and
caffeine to ephedrine alone.84, 86, 88 The Jadad scores for these trials were 1, 2, and 3 respectively,
and all three had attrition rates of greater than 20 percent. The random effects pooled estimate of
the rate of weight loss per month was -0.31 (95% CI: -0.60, -0.02), which equates to a weight
loss of 0.8 pounds per month more than with ephedrine alone (Table 13 and Figure 13). There
were too few trials to perform any sensitivity analysis.
We interpret these data to indicate that addition of caffeine to ephedrine is associated with a
statistically significant increase in weight loss per month of about 0.8 pounds, over that
attributable to ephedrine alone.
Ephedrine versus another active weight loss therapy
We identified two trials that compared ephedrine with another active weight loss therapy
(Table 14). The trials, both Danish, are briefly described here. In 1994, Breum and colleagues113
75

published the results of a randomized controlled trial comparing the effect of dexfenfluramine to
a combination of ephedrine and caffeine. At 15 weeks, the dexfenfluramine group had lost an
average of 6.9 kg (15.2 lb.), whereas the ephedrine and caffeine group had lost 8.3 kg (18.3 lb.),
a difference that was not statistically significant. The other Danish study,87 published in 1981,
compared the effects of the Elsinore pill (a prescription that contained ephedrine and caffeine)
with those of diethylpropion. At 12 weeks, the diethylpropion patients had a median weight loss
of 8.4 kg (18.5 lb.), while those taking Elsinore pills lost a median of 8.1 kg (17.8 lb.), a
difference that was not significant. Each of these two trials87 included approximately 40
ephedrine and 40 other treatment patients. The approximate weight loss in the ephedrine groups
was 8.4 kg ( approximately 4.0 kg SD) (18.4 8.8 lb.) over three months. Based on a two-sided
test of significance level 0.05 and assuming the same variance in both groups, trials of this size
have only 59 percent power to distinguish between an 8.4 kg weight loss in the ephedrine group
and a 6.4 kg (14.1 lb.) weight loss in the active treatment group, i.e. a difference of 30% between
the groups. In order to attain 80 percent power, a study would need 67 ephedrine patients and 67
comparison treatment patients.
Ephedra versus placebo
We identified a single trial that assessed the effect of herbal ephedra versus placebo.114 This
trial was described as a randomized, double-blind, parallel group assessment of Metab-O-Lite, a
dietary supplement that contains ephedra and other compounds but does not contain caffeine or
herbs that contain caffeine. The duration of the trial was three months. Those in the ephedra arm
lost 1.8 pounds more per month than did those in the placebo arm (95% CI: -2.7, -1.0)
(Table 15). This trial scored four on Jadads scale and had 17 percent attrition.
Ephedra plus herbs containing caffeine versus placebo
We identified four trials that assessed the effect of herbal ephedra plus herbs containing
caffeine versus placebo.89, 93, 115, 116 The Jadad scores for these four trials were 5, 5, 2, and 2
respectively; and all four were described as randomized placebo-controlled trials. Two of the
trials reported outcomes at two months, one trial reported three-month outcomes, and one trial
reported four-month results. The pooled random effects estimate of the rate of weight loss per
month of these four trials was -0.81 (95% CI, -1.12, -0.51), which equates to a weight loss of 2.1
pounds per month more than that for placebo, for up to four months (Table 16 and Figure 14).
The pooled average percent weight loss in the ephedra-treated patients, compared to pretreatment
weight, was 5.2 percent at four months. A sensitivity analysis for only those trials that scored
three or more on the Jadad scale yielded a result similar to the main analysis. All studies assessed
medium doses of ephedra; therefore no analysis of dose effect was possible. Neither visual nor
graphical tests revealed any evidence of publication bias (Table 11 and Figure 15).
We interpret these data to indicate that the use of a combination of ephedra plus herbs
containing caffeine is associated with a statistically significant increase in weight loss per month
of 2.1 pounds compared with that of placebo, for up to four months duration.
Meta-regression analysis
In order to assess the effects of ephedrine, ephedrine plus caffeine, and ephedra plus herbs
containing caffeine on weight loss, we conducted a meta-regression analysis. The results are
displayed in Table 17 and Figure 16. The table shows the pooled monthly weight loss in pounds
76

and its confidence interval for each comparison group versus placebo. All are significantly
different from zero, indicating that all treatments are associated with an increased weight loss as
compared to placebo. The last column, which compares ephedrine alone, ephedrine plus caffeine,
and ephedra plus herbs containing caffeine, shows no significant differences among these
treatments. Figure 16 lists the comparison groups in order of least effective versus placebo (left)
to most effective (right). The individual effect sizes converted to pounds within each comparison
are plotted vertically as circles, with the circle area inversely proportional to trial variance. We
connect the pooled-effect sizes in pounds within each comparison group by line segments,
showing the visible downward trend from left to right.
These data indicate that both ephedrine plus caffeine and ephedra plus herbs containing
caffeine are somewhat more effective than ephedrine alone in promoting weight loss and that
there is no difference in effect between ephedrine plus caffeine and ephedra plus herbs
containing caffeine. To help put these data in context, we note that placebo-controlled trials of
some FDA-approved weight loss pharmacotherapies have shown losses of 6-10 pounds more
than placebo, over 6-12 months, for patients taking sibutramine117-120 or orlistat;121-125 or 16
pounds more than placebo, at 9 months, for patients taking phentermine.126

Athletic Performance
We found eight published controlled trials of the effects of synthetic ephedrine on athletic
performance; most were crossover designs, and all but one also included caffeine. One trial,127
which assessed the effect of ephedrine and exercise training on basal metabolic rate but did not
report athletic performance outcomes, is not described below. The remaining seven trials were
not appropriate for pooled analysis because they included various types of exercise and outcome
measures. Thus, they are discussed here individually. We found no trials of the effect of herbal
ephedra on athletic performance.
Six trials by Bell and colleagues assessed the exercise capacity of small groups of healthy
males (all trials included 24 subjects or fewer). In their first trial,128 healthy subjects who were
not athletically trained were divided into four treatment groups: caffeine, ephedrine, ephedrine
plus caffeine, and placebo. Outcome measures included oxygen consumption (VO2), carbon
dioxide production (VCO2), and peak time to exhaustion. Ephedrine plus caffeine was reported
to improve parameters of exercise performance such as oxygen consumption, time to exhaustion
or carbon dioxide production by 20 to 30 percent, but neither caffeine nor ephedrine alone had
significant effects. A follow up trial,129 using a similar population and outcome measures but a
lower dose of caffeine and ephedrine, showed similar effects on exercise performance and fewer
side effects. Nausea and vomiting were reported in a third of subjects given 1 mg/kg ephedrine
with 5 mg/kg caffeine, but none of the subjects given a lower dose (0.8 mg/kg ephedrine and 4
mg/kg caffeine) experienced symptoms. A third trial130 assessed the effects, in a field trial, of
ephedrine plus caffeine on VO2 and time to complete a standardized exercise test, and again
reported improvements in the group treated with ephedrine plus caffeine. A fourth trial131 tested
the effects of ephedrine plus caffeine on body temperature regulation and oxygen consumption
during sub-maximal exercise in a hot environment and found that the combination did not
increase body temperature significantly. A fifth trial132 compared the effects of placebo, caffeine,
ephedrine, and a combination of ephedrine plus caffeine on muscle endurance in men performing
77

weight circuit training. This trial showed an improvement in muscle endurance, but only on the
first of three repetitions. The most recent trial133 reported that, compared to placebo, ephedrine
plus caffeine consistently improved exercise performance during stationary biking. The Bell
trials are summarized in more detail in Table 18.
A trial by Sidney134 assessed the effects of ephedrine versus placebo or no treatment (for
baseline measures) on performance on a variety of physical function tests among 21 healthy
young men. No statistical differences were seen among the groups on performance of any of the
tests, including VO2, measures of endurance and power, reaction time, hand-eye coordination,
speed, and self-perceived exertion. These results agree with the finding by Bell and colleagues
that ephedrine alone did not demonstrate significant effects on athletic performance.
In conclusion, the effects of ephedrine on athletic performance have not been well studied.
The populations studied have been small and exclusively male, and the method of administration
of ephedrine does not replicate the patterns of use reported for the general public. Effects of
ephedrine on exercise performance are most often studied acutely (e.g., one to two hours after a
single dose) in contrast to assessing the effects of chronic use on conditioning and performance.
The one trial that did assess the effect on strength training did not find a sustained benefit of
ephedrine supplementation. In addition, to show even a short-term effect of ephedrine,
combination with caffeine was required. We identified no trials that assessed the sustained effect
of ephedrine on aerobic conditioning or strength training and no trials that tested the effects of
herbal ephedra on athletic performance.

Safety Assessment
Controlled Trials
We initially considered all 52 clinical trials of ephedra and ephedrine for the safety assessment.
Two trials were excluded from the odds ratio meta-analysis because they did not contain a placebo
group.113, 138 Numerous symptoms were reported as adverse events. We grouped clinically similar
symptoms as follows:

Psychiatric symptoms: those symptoms described in the original clinical trials as

euphoria, neurotic behavior, agitation, neuropsychiatric, depressed mood,
giddiness, irritability, and anxiety;

Autonomic hyperactivity: those symptoms described in the original clinical trials as

tremor, twitching, jitteriness, insomnia, difficulty sleeping, sweating,
increased sweating, and increased perspiration;

Nausea/ vomiting: those symptoms described in the original clinical trials as nausea,
vomiting, abdominal pain, upset stomach, heartburn, and gastroesophageal
reflux;

78

Palpitations: those symptoms described in the original clinical trials as palpitations,

irregular heartbeat, loud heartbeat, heart pounding, and increased or stronger
heartbeat;

Tachycardia: those symptoms described in the original clinical trials as tachycardia and
slightly elevated heart rate;

Hypertension: those symptoms described in the original clinical trials as hypertension,

increased systolic blood pressure, and increased diastolic blood pressure; and

Headache.

Table 19 presents the pooled estimate of the odds ratio for those adverse events for which
data were sufficient to justify meta-analysis. The odds ratio will slightly overestimate the risk
ratio for these events, as they occurred in 10 to 20 percent of subjects. This analysis reports a
statistically significant increase of between 2.15 and 3.64 percent in the odds for the adverse
events of psychiatric symptoms, autonomic hyperactivity, nausea/vomiting, and palpitations.
There is a trend toward an increase of similar magnitude in the report of hypertension, but this
increase was not statistically significant. There was also a non-statistically significant trend
towards an increase in headaches. There were too few trials of ephedra or ephedrine alone to
support analyses specific to these products; the subgroup analysis of adverse events involving
ephedrine plus caffeine was similar to the main analysis. In our dose analysis, there was a trend
toward higher risk of adverse events with higher doses of ephedrine, but data were sparse, and
these differences were not statistically significant (for example, adjusted odds ratios of
autonomic hyperactivity were 3.0 and 12.5 for medium- and high-dose ephedrine respectively,
but the 95% confidence intervals overlapped; adjusted odds ratios for the three cardiovascular
outcomes combined were 2.7 and 7.9 for medium- and high-dose ephedrine, a difference that
was not statistically significant). The pattern of symptoms with statistically significant increases
in occurrence is consistent with the pharmacology of ephedrine.
Table 20 presents frequency data concerning the other adverse events reported in the clinical
trials. Meta-analysis was not performed on these data, primarily due to small numbers of events.
No serious adverse events (e.g., death, myocardial infarction, stroke, etc.) were reported in
the 52 clinical trials that reported sample sizes. Therefore, the rate for these adverse events is
zero. Even in aggregate, these trials had sufficient statistical power only to detect a serious
adverse event rate of 1.0 in 1000, given the small numbers of patients studied in these trials. For
trials of ephedra, statistical power in aggregate was sufficient only to detect a rate of serious
adverse events of 4.0 in 1000. A conventional definition of a rare adverse event is about 1 in
1000. We also note that these data come from patients enrolled in clinical trials: Data from the
pharmaceutical literature support the contention that patients taking pharmaceuticals outside of
clinical trials may have a greater risk of particular adverse events than do patients selected to
participate in clinical trials.139 Therefore, in community practice, the rate for serious adverse
events may be higher than that seen in clinical trials.
79

Case Reports
Because, even in aggregate, the numbers of subjects enrolled in clinical trials have been too
small to assess the possibility of rare but serious side effects, we assessed case reports of serious
events allegedly associated with ephedra use.
FDA Medwatch Data and Literature Cases
Figure 17 is a graphical representation of the case report evidence used in the safety
assessment. The first master list produced by the FDAs Office of Nutritional Products, Labeling,
and Dietary Supplements contained 1,848 adverse event reports. The second master list, from
which events associated with ephedrine were removed, contained 1,783 reports. When we
combined event reports of identical ID number but different products, we reduced the
observations by 88 to 1,695 total observations but did not lose any data. In addition, we removed
137 reports listed in the master Excel file as having been filed after our September 30, 2001 cutoff date. The master Excel list also contained 214 reports (dated before September 30, 2001) for
which no PDF data files existed on the CDs. Because documentation was essential for review of
each report, these reports were removed from our analysis, which left 1,344 reports in our final
dataset. In Table 21, we show the result of a chi-squared test of independence, which tests the
association between the type of event distribution (death; stroke; myocardial infarction; other)
and the type of data (available; after September 30, 2001; not available). This test rejected the
null hypothesis of no association (p < 0.001), indicating that the distribution of events was
different for the different data types. Thus, bias may exist, because the events we included were
different in type than those we had to exclude. Since more cases of death were reported, as a
percentage of total cases, in the data subsequent to September 2001, it is possible that our results
would be different had we had the opportunity to include the cases filed after September 2001.
To the 1,344 unique and available reports that met the cut-off date (Batch 1), the FDA added
another 125 reports (Batch 2) that consisted predominately of adverse events related to
ephedrine. Together, 1,469 reports from the FDA MedWatch files were reviewed. Within the
1,344 reports from Batch 1, 158 cases reported on the most serious adverse events (death, stroke,
and myocardial infarction), and 1,186 reported on other adverse events according to the master
Excel spreadsheet. Of the 1,186 case reports, we found 935 reports that fit the categories of
other serious cardiovascular, other serious neurological, and psychiatric. (We did not
examine the remaining 251 adverse event reports because the descriptors in the master excel
spreadsheet appeared to fall outside our focus of serious adverse events.) The 935 reports
contained data on 965 subjects, of which 922 reported taking ephedra. From the brief review, we
determined that 158 of these subjects reported events serious enough (ventricular
tachycardia/fibrillation, cardiac arrest, pulmonary arrest, transient ischemic attack, brain
hemorrhage, seizure, or psychiatric symptoms) to warrant including their file in the more
detailed review. Within the 125 reports of Batch 2 (reporting on 130 subjects), we found 106
subjects reporting ephedra or ephedrine use. Thirty-three of those subjects reported events
serious enough (ventricular tachycardia/fibrillation, cardiac arrest, pulmonary arrest, transient
ischemic attack, brain hemorrhage, seizure, or psychiatric symptoms) to warrant including their
file in the more detailed review.

80

Of the 533 articles retrieved from the medical literature for this report, sixty-six were case
reports or case series of adverse events. Six reports were rejected as duplicates, leaving 60 case
reports or case series, reporting on 99 subjects. Of these, further review identified four as not
reporting on ephedra or ephedrine. Of the remaining 95 subjects, 46 had adverse events that went
on to detailed review, and 49 were not reviewed further.
From all sources, 84 deaths, 26 myocardial infarctions, 56 cerebral vascular accidents
(strokes or cerebral hemorrhage), 30 other cardiac events, eight other neurological events,
40 cases of seizure, and 91 cases of psychiatric events. We identified two deaths, three
myocardial infarctions, nine cerebrovascular accidents, three seizures, and five psychiatric cases
as sentinel events with prior ephedra consumption. Three deaths, two myocardial infarctions, two
cerebrovascular accidents, one seizure, and three psychiatric cases were identified as sentinel
events with prior ephedrine consumption. We identified an additional 43 cases as possible
sentinel events with prior ephedra consumption and an additional seven cases as possible sentinel
events with prior ephedrine consumption. About half of the sentinel events occurred in persons
aged 30 years or younger. Classification as a sentinel event does not imply a proven cause and
effect relationship.
What follows are short descriptions of the adverse event cases that we reviewed further for
other potential causes for the adverse event. Table 22 presents data abstracted from each
reviewed case. Tables 23 and 24 summarize the adverse events by gender, age, and type of event.
Cases classified as something other than sentinel events include, where feasible, our reasons for
so classifying them. Clinical detail regarding outcome is recorded to the extent it was available in
the source material.
Deaths
Sentinel Events
FDA CasesEphedra
A 21-year-old male collapsed and died during a physical agility run at school after taking
Hydroxycut and Ripped Fuel one time to increase energy. At autopsy, ephedrine and caffeine
were found in the blood at concentrations of 0.02 mg/l and 0.31 mg/l, respectively. Although the
autopsy report itself was not included in the FDA documents we received, a detailed description
of the autopsy was found in the police notes that were included in the FDA file. According to
these notes, the autopsy report stated that the coronary arteries were normal and that the
diagnosis was acute arrhythmia due to ephedrine. (13914)
A 22-year-old female who weighed 183 pounds collapsed and died while standing in line to
purchase ice cream. She had a history of asthma that was characterized as well-controlled. She
had congenital hydrocephalus with a shunt placed. She was taking Slacker II. Ephedrine was
found in the blood. The autopsy report stated that the coronary arteries were free of
atherosclerosis. There was no myocarditis. The brain was normal, except for the presence of the
shunt. There was no other cause of death. The death certificate listed cardiac arrhythmia due to
ephedrine-containing diet medication. (14390)

81

FDA CasesEphedrine
A 30-year-old female died suddenly. Her husband stated that she had been taking MiniTabs
in order to lose weight. She was found unresponsive by her husband and was brought to the
emergency department in full cardiac arrest. Blood toxicology screen showed an ephedrine level
of 24 micrograms (g) per milliliter (ml). Examination of the heart and brain did not reveal any
evidence of cause of death. Final pathological diagnosis included acute drug toxicity
ephedrine with the opinion that this autopsy illustrates an instance of death due to acute drug
toxicity. (3275432)
A 33-year-old male was found dead. He had been taking an over-the-counter preparation
named Max Brand Two-Way ephedrine tablets. Autopsy did not reveal any obvious cause of
death, particularly with respect to the brain and heart. However, the blood ephedrine level was
13.4 g per ml. The final pathological diagnosis was drug intoxication, with the opinion that the
person died as the result of drug intoxication with ephedrine. (3289590)
Literature CaseEphedrine
A 28-year-old male truck driver was, according to his family, taking up to 600 mg of
ephedrine per day for six years. After having consumed 250 mg of ephedrine one day, he
collapsed while baling hay. Autopsy did not reveal any pathologic process to account for his
death. Specifically, the coronary arteries, valves, and myocardium, including the conduction
system, were normal. Only ephedrine and guaifenesin were identified on toxicology screens.
The conclusion was that death was most likely due to a cardiac arrhythmia triggered by the
combination of an excessive use of ephedrine and strenuous labor on a hot day. (348)
Possible Sentinel Events
FDA CasesEphedra
A 36-year-old female began taking Natures Nutrition-Formula One in August 1993. In
December 1993, she was taken to the emergency department and found to have low potassium
but signed out Against Medical Advice. In May 1994, she had severe stomach cramps and later
that day was found unconscious by her daughter. She was taken to the hospital, where she died
five days later without having recovered consciousness. According to her husband, she had no
history of heart, thyroid, or blood pressure problems. The medical record documents a past
history of bulimia and anorexia. There is a brief autopsy form consisting of a series of handwritten notes and check marks. Next to the word cardiovascular is a handwritten check mark,
suggesting the heart had been examined and found normal, but no dictation describes the heart.
There is an extensive description of the brain, which was normal. Elsewhere in the medical chart
is a statement that the patient was thought to have had an acute myocardial infarction with adult
respiratory distress syndrome and heart failure. Her creatine phosphokinase (CPK) isoenzymes
and myoglobin (MB) fractions were both elevated. The emergency department record reported a
toxicology screen positive for ephedrine, phentermine, and chlorpheniramine. A cardiology note
and an echocardiogram reported severe global cardiac dysfunction and a left ventricular ejection
fraction of 25 percent. We classified this case as no higher than a possible sentinel event because
there was no evidence available to us of an adequate examination of the heart. It is possible that
this woman could have had acute myocarditis, which led to global cardiac dysfunction. (9508)

82

A 32-year-old male truck driver was found dead in his truck by the police. In the truck were
found Natures Nutrition-Formula One, Natures Nutrition-Formula Three, a bottle of Tylenol
with codeine, Vicks Formula 44, Nyquil, Ibuprofen, and Rexall cold tablets. The FDA files do
not contain a copy of the actual autopsy report, but notes state that the autopsy report said the
heart was enlarged, the coronary arteries were normal, and there was a slight case of pneumonia.
The cause of death was listed as myocarditis, bronchiolitis, and pneumonia. Toxicology screen
was negative for ephedrine but did show pseudoephedrine and doxylamine. We classified this
case as a possible sentinel event because the myocarditis could have contributed to his death and
the etiology of the myocarditis is unknown. (10276)
A 38-year-old male collapsed and died after jogging. Prior to jogging, he had had a cup of
coffee and Ripped Fuel supplements. At autopsy, he was found to have triple vessel coronary
artery disease and cardiomegaly. Because of this preexisting condition, we classified this as a
possible sentinel event. (12485)
A 21-year-old male on his college wrestling team was trying to lose weight, perhaps as much
as 17 pounds in a few days, to achieve a weight limit for an upcoming meet. He had been taking
Thermogenics Plus for an unknown length of time. He began to feel weak while sitting in the
sauna. He left the sauna, went to get a drink, and collapsed. On autopsy, the cause of death was
listed as sudden cardiovascular collapse with rhabdomyolysis and dehydration. The heart exam
revealed normal coronary arteries. We classified this as a possible sentinel event since the
intense effort to lose weight likely led to dehydration, which then led to cardiac collapse. (12722)
A 15-year-old female collapsed and died while playing soccer. She had been taking Ripped
Fuel for an unknown length of time. At autopsy, she was found to have a congenital abnormality:
an anomalous origin of the left coronary artery from the pulmonary circulation (the BlandWhite-Garland Syndrome). This condition is not usually associated with life beyond infancy, if
left uncorrected. Due to this preexisting condition, the case was classified as a possible sentinel
event. (12843)
A 26-year-old male had been taking Ripped Fuel for two weeks prior to his death. There is
no autopsy report in the FDA files, but detailed notes state that the autopsy revealed he died of
acute aortic dissection. According to the file, he had been having back pain two weeks prior to
his fatal event and went to the emergency department complaining of severe chest pains. He was
diagnosed with esophagitis and sent home. Four days later, he returned to the emergency room
again with severe chest pains and was told the source of his pain was the chest wall. The next
day, he was found dead by his girlfriend. According to his family history, several relatives had
also had an aortic dissection, including a niece who had had an aortic dissection at age 18. The
notes state, This appears to have been some form of genetic connective tissue abnormality.
There is also evidence of prior borderline hypertension. Toxicology screen for cocaine,
ephedrine, and amphetamines was negative. Only caffeine and acetaminophen were found in the
blood. This case was classified as a possible sentinel event, because it appears the patient was
genetically predisposed to aortic dissection. (13906)
A 26-year-old female was found dead by her father. She had been taking a product called
Diet Fuel for six months. The adverse event report stated, Coroner felt this was a massive heart
83

attack. Autopsy concluded she suffered from tachycardia, high blood pressure, and restriction of
the coronary artery. The death certificate stated that death was due to dissection of the left
anterior coronary artery. Ephedrine and pseudoephedrine were found in the blood. This case
was classified as a possible sentinel event. (14019)
A 35-year-old male, who had been taking Hydroxycut for seven days to increase energy,
came home from work early because he was not feeling well, went to the bathroom, and was
later found unresponsive. At autopsy, an 80 percent stenosis of the proximal left anterior
descending coronary artery was found, along with moderate stenosis of the right and left
circumflex coronary arteries. The cause of death was listed as atherosclerotic coronary vascular
disease. Because of this preexisting condition, we classified this case as a possible sentinel event.
(14638)
Literature CaseEphedra
A 23-year-old man was found dead in his apartment by his sister. He had been using Ripped
Fuel. Autopsy showed no gross evidence of a pathologic process. Microscopic examination of
the heart reported multifocal and confluent myocyte necrosis with healing of approximately 1 to
2 weeks; mild perivascular, focal endocardial, and focal epicardial fibrosis; and moderate
myocyte hypertrophy and vascular congestion. There was no evidence of myocarditis. Blood
toxicology screen was negative, but urine toxicology screen showed ephedrine at 1.6 g per ml.
We classified this case as a possible sentinel event. (258)
Literature CasesEphedrine
A 42-year-old male was found dead at home. On autopsy, he was found to have had an
intracranial hemorrhage and was also found to have hypertensive cerebral vasculopathy. He had
been taking a street drug (speed) that contained ephedrine, and toxicology screen was positive
for ephedrine at 2.7 g per ml. We classified this case as a possible sentinel event due to the
preexisting vasculopathy. (44)
An 84-year-old female was found in a coma and subsequently died. On clinical investigation,
she was found on computerized tomography (CT) scan to have a subarachnoid hemorrhage and a
right subdural hemorrhage. At autopsy, she was found to have a ruptured berry aneurysm along
with cerebral atherosclerosis. She had been taking an unknown drug containing ephedrine, and
her blood toxicology screen was positive for ephedrine. We classified this case as a possible
sentinel event due to the preexisting berry aneurysm. (44)
A 44-year-old male was taking ephedrine as a replacement for daily coffee and cocoa. He
had a sudden cardiorespiratory arrest and died. Autopsy revealed an acute thrombus in the left
anterior descending coronary artery. The report states, All other coronary lumina were patent,
although calcified with focal narrowing to approximately 50 percent. Due to the preexisting
coronary artery disease, we classified this case as a possible sentinel event. (224)
Probably Not Related
FDA CaseEphedra
A 24-year-old male collapsed and died during a training run. He had reportedly taken Ripped
Fuel, although none was found in his personal possessions, nor were traces of amphetamines
84

found on toxicology screen (which looked for ephedrine as well). At autopsy, he was found to
have died of massive sickle-cell crisis. As a result, we classified this case to be probably not
related to ephedra use. (13672)
FDA CasesEphedrine
A 40-year-old male was taking Max Alert to stay awake on the job. He had odd symptoms
that were not a recognizable illness, described as nausea, dizziness, sweats, irritability,
dehydration, respiratory problems, etc. The report then states that he was killed in a car accident
while driving. We classified this case as probably not related to ephedrine use and note that it
may be the same case as Case 1902493, which has a virtually identical description, both cases
having been filed within one month of each other. (1859087)
A young male was killed in an automobile accident while driving home in the early morning
from his night shift job at a hotel. The patient had been using Max Alert to stay awake. The
cause of death at autopsy was laceration of the aorta. We classified this case as probably not
related. (1902493)
A 30-year-old male was found dead. He had had chronic low back pain and a chronic pain
disorder and was on numerous medications. Autopsy did not reveal any cause of death; however,
the toxicology screen was positive for alcohol, fentanyl, phenylpropanolamine, ephedrine,
pseudoephedrine, bupropion, nordiazepam, alprazolam, chlordiazepoxide, and nortriptyline.
Cause of death was listed as mixed drug and alcohol intoxication. We classified this case as
probably not related. (3491515)
A 29-year-old male died. Toxicology screens of blood and urine revealed that he was
positive for morphine, hydrocodone, acetaminophen, diphenhydramine, hydromophone,
promethazine, dihydrocodeine, codeine, pseudoephedrine, ephedrine, brompheniramine,
phenothiazine, cannabinoids, and nicotine. We judged this case as probably not related.
(3772362)
Insufficient Information
FDA CasesEphedra
A 37-year-old male collapsed and died after taking Metabolife for weight loss and energy. At
autopsy, ephedrine was found in the blood. In addition, one coronary artery was found to be 70
to 80 percent stenosed. These data were recorded on a single page of telephone conversation
notes in the file. Because there was no other documentary evidence regarding this case, we
classified it as having insufficient information to judge. (13806)
A 56-year-old male who had recently started taking Thermogen Plus Liquid fat complexor
tablets was found slumped over in his bathtub after a barbecue. He had a history of hypertension
and was on a calcium channel blocker as well as daily baby aspirin. Within the year prior to his
death, he had had a normal treadmill test. He also had elevated cholesterol for at least four years
prior to his death. Four days before his death he was noted to have heartburn. No autopsy report
was in the FDA records. A report of a phone conversation stated that the death certificate listed
cardiac arrhythmia of unknown etiology as the cause of death. We counted this case as having
insufficient information, because no autopsy report was available to us, and he had preexisting
85

hypertension and elevated cholesterol. Without the finding of the autopsy report that detailed the
examination of the cardiovascular system, we can come to no other conclusion. (14465)
Myocardial Infarctions/ Acute Coronary Syndromes
Sentinel Events
FDA CaseEphedra
A 45-year-old male took two tablets of Natures Nutrition-Formula One prior to suffering a
myocardial infarction. The patient had also smoked for 30 years and had been a practicing
alcoholic until one year prior to the event. At angiography, the coronary arteries were found to be
normal. (10024)
FDA CaseEphedrine
A 23-year-old female took four Midnight Ecstasy tablets as a sexual stimulant. Shortly
thereafter, she began developing symptoms of autonomic hyperactivity followed by palpitations,
shortness of breath, and pink frothy sputum. She was taken to the emergency room, where she
was found to be in pulmonary edema. Cardiac enzymes indicated acute myocardial injury. Urine
toxicology screen was positive for marijuana and amphetamines. Ephedrine was not mentioned
in the toxicology report. Coronary angiography did not reveal any sign of coronary artery
disease. Her recovery was complicated by a presumed infection, but she was ultimately
discharged from the hospital in good condition. (3446357)
Literature CasesEphedra
A 30-year-old male body builder was taking ma huang as instructed by the product label.
He presented to the emergency department complaining of chest pain. Vital signs revealed
tachycardia and no hypertension. Electrocardiogram was consistent with acute inferior cardiac
ischemia. Cardiac enzymes confirmed a myocardial infarction. Urine toxicology screen was
negative for cocaine and amphetamines. Emergency cardiac catheterization demonstrated normal
coronary arteries with mild global left ventricular hypokinesis and mild left ventricle
hypertrophy. He recovered well. (244)
A 19-year-old male experienced chests pains 30 minutes after taking Dymetadrine Xtreme at
the recommended dose. Vital signs revealed tachycardia and elevated respiratory rate of 22. The
physical examination was described as unremarkable except for diaphoresis. Electrocardiogram
was consistent with an inferolateral myocardial infarction, and myocardial necrosis was
confirmed by cardiac enzymes. Toxicology test was negative for cocaine. Cardiac catheterization
was reported as showing only minimal intimal disease of the distal left anterior descending
artery. The patient was reported to have recovered well. (516)
Literature CaseEphedrine
A 35-year old woman was taking a dietary supplement containing ephedrine for weight loss.
She had the acute onset of chest pain, diaphoresis (perspiration), and shortness of breath. She
was admitted to the hospital, where an electrocardiogram and cardiac enzymes were consistent
with acute myocardial infarction. Results of a cardiac catheterization were reported as normal
cardiac function and normal coronary arteries. She was discharged with a diagnosis of acute
myocardial infarction secondary to cardiac spasm. (224)
86

Possible Sentinel Events

FDA CasesEphedra
A 37-year-old male took Eola Amp II Pro Drops for twelve days prior to suffering an
inferior myocardial infarction. At coronary angiography, his mid-right coronary artery was found
to be 95 percent stenosed. The patient received percutaneous transluminal coronary angioplasty.
Due to the preexisting condition, we classified this case as a possible sentinel event and note the
product was later reported to include illegal doses of ephedrine. (9372)
A 54-year old-male with a history of hypertension had been taking Natures NutritionFormula One for approximately three to four months prior to suffering an inferior myocardial
infarction. He also had a cardiac arrest from which he was successfully resuscitated. On
angiography, he was found to have an 80 percent stenosis of the right coronary artery along with
total occlusion of the obtuse marginal artery. He was treated with angioplasty. Because of the
preexisting condition, we classified this case as a possible sentinel event. (9504)
A 35-year-old male took five capsules of Metabolift prior to a vigorous workout. He then had
an acute inferior myocardial infarction, for which he received thrombolytic therapy. At
angiography, his left anterior descending coronary artery had ectasia, which is indicative of
coronary artery disease. Therefore, we classified this case as a possible sentinel event. (10009)
A 38-year-old female took Herbalife Original Green for one day. The next day, she suffered
an inferior myocardial infarction, for which she received thrombolytic therapy. At
catheterization, the posterior descending artery was found to be totally obstructed. The left
coronary artery was found to be normal except for one area with a 70 percent lesion. The
toxicology screen was negative for cocaine but positive for amphetamines. Ephedrine was not
mentioned in the toxicology report. The diagnosis was presumed right coronary spasm with
Prinzmetals Angina. In the setting of coronary artery disease, we classified this case as a
possible sentinel event. (13009)
A 37-year-old female with a family history of coronary artery disease was both overweight
and a cigarette smoker. She was taking Metabolife 356 to lose weight. She suffered a myocardial
infarction and was found on angiography to have total occlusion of the right coronary artery with
diffuse disease in the left anterior descending coronary artery and the left circumflex artery. She
received percutaneous transluminal coronary angioplasty with placement of a stent. Because of
the preexisting condition, we classified this case as a possible sentinel event. (14114)
A 43-year-old female had a heart attack. Earlier that day, she had taken six tablets of MetabO-Lite. She was a cigarette smoker and had a lipid disorder. At coronary catheterization, the left
main coronary artery was found to be normal, the left anterior descending artery had 2030
percent stenosis, the left circumflex artery had no disease, and the right coronary artery had 30
percent stenosis. Because of the preexisting coronary artery disease, we classified this case as a
possible sentinel event. (14530)

87

Cerebrovascular Accident/Stroke
Sentinel Events
FDA CasesEphedra
A 26-year-old female began taking Thermo Slim and The Accelerator daily for weight loss.
Three days later, her legs became weak, and she reported feeling like she was going to pass out.
She was taken to the emergency room where she was found to have a probable basal ganglia
hemorrhage on CT scan. She also had paranoid psychosis. She was a long-time intravenous drug
abuser and alcohol abuser. She had also smoked cigarettes for ten years. She was not taking oral
contraceptive pills. Blood pressure in the emergency room was 129/71. Toxicology screen was
positive for acetone and for benzodiazapines. It was negative for cocaine, amphetamines, and a
host of other substances, but ephedrine was not specifically examined. Ephedrine was not
mentioned in the toxicology report. She signed out Against Medical Advice from that hospital
and ended up in another hospital later that day in restraints. Another toxicology screen was
positive for phenylpropanolamine and benzodiazepine. At this hospital, rheumatologic and
embolic evaluation was negative. (10874)
A 42-year-old female who was taking Power Trim began having headaches. Approximately
one week later, when she was scheduled to undergo a root canal procedure to treat a dental
abscess, her daughter heard a thump on the floor and found her mother shaking, lying on the
floor, unresponsive. She was taken by ambulance to the emergency room, where she was
observed to have a focal seizure, which then generalized. She had no prior history of seizure.
Toxicology screen was negative. Glucose was 93. CT scan revealed a possible small area of
hemorrhage in the upper right parietal region. An MRI with contrast revealed a 1 by 1-1/2
centimeter area of acute hemorrhage without mass effect and without abnormal vascularity.
Digital subtraction angiography did not reveal any evidence of arteriovenous malformation. The
patient remained seizure-free on anticonvulsant therapy. (11062)
A 31-year-old female had been taking Trim Easy for weight loss for nine months. She
occasionally took up to 6 caplets at a time and smoked 34 cigarettes a day. She was found in the
bathroom unconscious and brought to an emergency room. Her blood pressure was 143/86. CT
scan showed a large intracerebral hematoma. Cerebral angiography did not show any source of
the hemorrhage, and MRI also was remarkable only for the bleed. Medical records documented
improvement over a period of time, but she was left with substantial physical limitations.
(11105)
A 28-year-old male who was described as a weight trainer took Ripped Fuel. He also smoked
two packs of cigarettes per day. During sexual relations with his wife, he had a headache and
became dizzy. He was taken to the emergency department where he was found to have a right
middle cerebral artery infarction that was suggestive of vasospastic phenomenon. At the time
of his admission, his blood pressure was 132/78. His toxicological examination was positive for
benzodiazapines. Ephedrine was not mentioned in the toxicology report. Cerebral angiogram was
negative. Rheumatologic and hypercoagulability evaluations were negative. There was a
negative transesophageal echocardiogram. Urine screen showed ephedrine, pseudoephedrine,
and caffeine, according to a discharge summary, which also described the illness as cerebral
infarction associated with ephedrine and caffeine use. (11675)
88

A 39-year-old male Navy diver, who regularly took Ultimate Orange for energy, presented to
the ship doctor with right hand and leg numbness. The symptoms had occurred 1 hours after
using the product, during a workout. He also reported taking omeprazole, creatine, and vitamins.
CT scan and angiogram were negative except for an intracerebral hemorrhage. Blood pressure
initially was 140/72. The patient made a good recovery and was able to go back to work but was
prohibited from further diving. (12980)
A 29-year-old male in the Army, who was taking Ultimate Orange and had a history of
migraine headaches, reported that he was running on a treadmill, when he experienced the worst
headache of his life. He was taken to the emergency department, where he had some right-sided
weakness, which improved over the next 12 hours. The CT scan and lumbar puncture were
normal. The following morning, most of his symptoms had resolved, but then he suddenly
developed total hemiplegia. He underwent an emergency MRI and angiogram, which showed a
complete occlusion of the right middle cerebral artery. The patient had a very stormy course,
ultimately resulting in a right hemispherectomy to control swelling. A full hypercoagulability
workup was done and was normal. Echocardiogram did not reveal an embolic source.
Microscopic examination of the brain tissue did not show any sign of vasculitis. (13418)
A 53-year-old female was taking Slim Caps. In June 2000, she presented with the clumsy
hand syndrome on the right. At the time, her blood pressure was 204/128. She stated that in the
past, her systolic blood pressure had been 135. CT scan at the time of the event showed a lacunar
infarct in the right frontal parietal region. She made a good recovery. A hypercoagulability
workup was negative. Echocardiogram was done and showed no evidence of clot. (14372)
A 46-year-old female took Xenadrine for 4 days. While at work, she stood up, had a seizure,
and became unresponsive. She was taken to the emergency department. A CT scan showed a
left-sided frontal cortex stroke. MRI showed occlusion of the left middle cerebral artery. Blood
pressure in the emergency room was 102/69. MR angiography was negative. Transesophageal
echocardiogram was negative. Hypercoagulability workup was negative. (14473)
Literature CaseEphedra
A 33-year-old male presented to the emergency department with the sudden onset of left
hemiparesis and slurred speech. He smoked one pack of cigarettes a day and took bupropion for
smoking cessation. He consumed Thermadrene 8 hours prior to the onset of neurologic
symptoms. There was no hypertension. A noncontrast CT scan did not show any abnormalities.
There was no evidence of intracerebral hemorrhage. He was treated with tissue plasminogen
activator. Normal tests included a sedimentation rate, clotting studies, urine toxicology screen,
homocysteine, antinuclear antibodies, Factor V level, complete hypercoagulability evaluation,
echocardiogram, VDRL, and carotid and vertebral duplex studies. He was given a diagnosis of
new cerebrovascular accident in the right middle cerebral artery. He was left with a permanent
disability. (552)
Literature CasesEphedrine
A 19-year-old female with a history of anorexia/bulimia and alcoholism presented after
taking 15 to 18 tablets that contained ephedrine 25 mg (along with guaifenesin). She had
previously been taking this product 3 to 10 tablets at a time to lose weight, and the case report
89

was silent regarding whether or not this increased dose was a suicide attempt. Initial presentation
was unremarkable, but while in the emergency department, she developed a severe headache and
right-sided paralysis. Blood pressure was elevated at 136/98. A CT scan showed left parietofrontal cerebral hemorrhage with extension into the left lateral ventricle. Angiography did not
document a source of the bleed. She was treated with an emergency craniotomy and hematoma
evacuation. No arteriovenous malformation was found. She survived but had a major residual
neurologic deficit. We classified this case as a sentinel event, but also note this may have been an
unrecognized suicide attempt. (184)
A 20-year-old woman took two capsules of a purported amphetamine look-alike. Two
hours after consumption, she developed a severe headache, nausea, hemiparesis, and aphasia.
Ten hours later, she sought admission to a hospital. On initial evaluation, her blood pressure was
not elevated and she had a right homonymous hemianopsia and a dense right spastic hemiparesis.
At that time, sedimentation rate, clotting studies, rheumatologic studies, and other tests
evaluating both vasculitis and the hypercoagulable state were negative. CT scan showed a left
external capsular hemorrhage with shift of the midline structures. Angiography showed
alternating narrowing and dilation of several branches of the middle cerebral artery. She made a
slow and incomplete recovery. Analysis of the capsules demonstrated the presence of ephedrine,
phenylpropanolamine, and caffeine. (514)
Possible Sentinel Events
FDA CasesEphedra
A 32-year-old female who was taking Eola Amp II Pro Drops for weight loss suffered a
brain stem stroke. She had been to the emergency room twice earlier in the day prior to her
stroke and both times was thought to be having an allergic reaction to peanut butter. She was
taking oral contraceptive pills. Evaluation of the cause of her stroke included a transesophageal
echocardiogram that was negative except for an atrial septal aneurysm, but this was not thought
to be the cause of the stroke. Lumbar puncture was negative. Cerebral angiography showed the
basal artery was occluded with embolus. Because no hypercoagulability workup was noted, we
could not judge this case as more than a possible sentinel event, and also note this product was
later reported to include illegal doses of ephedrine. (9296)
A 56-year-old female who was taking Eola Amp II Pro Drops for three months suffered a
lacunar infarct. She had preexisting hypertension, total cholesterol of 238, and triglycerides of
529. She also had a 60-pack-per-year history of smoking. CT scan was negative. The MRI
revealed microvascular changes. Because of her preexisting conditions, we classified this case as
a possible sentinel event, and also note this product was later reported to include illegal doses of
ephedrine. (9335)
A 24-year-old female had a right internal capsule stroke after taking one dose of Super Fat
Burners. She was not taking oral contraceptive pills or on any other medications. Carotid
arteriogram was normal, echocardiogram was normal, and drug screen was negative. She had
had two miscarriages in the past. While most of the hypercoagulability evaluation was normal,
one test suggested the presence of the lupus anticoagulant. Because the antiphospholipid
antibody syndrome was not effectively excluded, we classified this case as a possible sentinel
event. (10094)
90

A 64-year-old female with a history of hypertension, paroxysmal atrial fibrillation, and two
transient ischemic attacks was on propranolol, isradipine, and aspirin therapy, and had taken Fit
America Natural Weight Control Aid. She was found unconscious in the bathroom by her
husband and taken to the emergency room, where she was found to have had an embolic stroke.
She was given heparin, which resulted in a small left temporal parietal intracerebral hemorrhage.
She was found to be in atrial fibrillation. Carotid ultrasound was negative. Due to the preexisting
condition of atrial fibrillation, we classified this case as a possible sentinel event. (12713)
A 47-year-old male, who had a long history of hypertension but had not taken medication in
20 years, suffered a right lentiform nucleus bleed that manifested itself as left-sided paralysis.
The notes stated that just prior to the event, he took Purple Blast to lose weight; however, there
was no drug or alcohol use. When he arrived in the emergency department, his blood pressure
was 196/94, and it later fell to 187/107. Chest x-ray revealed cardiomegaly. CT scan showed
large acute intracranial hemorrhage in mid portion of right cerebral hemisphere. His
triglyceride levels were 364. Because of the history of long-standing untreated hypertension, we
classified this case as a possible sentinel event. (12733)
A 41-year-old female with a history of hypertension who had been taking Diet Phen and had
four stroke events over a two-month period. Blood pressure in the emergency department after
one event was 170/108, and at another time, it was 158/99. She was put on coumadin and
discharged home after her first stroke but then was readmitted three times in the next two months
for recurrent stroke, all of which occurred while she was on anticoagulants. She suffered
additional neurologic events consistent with brain stem infarction. She had numerous magnetic
resonance angiograms, the first of which revealed an irregularity of the basilar artery.
Subsequent studies showed the basilar artery totally occluded. The interpretation of these
angiograms was the subject of considerable discussion, with the final interpretation in the notes
being basilar artery vasculitis. Rheumatologic evaluation and hypercoagulability evaluation
were negative. This case was difficult for us to assess since amphetamines have been linked to
vasculitis, but her vasculitis could also have had other etiologies and therefore we classified this
case as a possible sentinel event. (12888)
A 25-year-old female who was taking Natural Trim presented with slurred speech and rightsided weakness. She was not on oral contraceptives and was a nonsmoker for five months. Blood
pressure recorded in the nurses notes was 144/88. MRI revealed a lacunar stroke. Carotid duplex
was negative. Echocardiogram showed a mitral valve prolapse with a patent foramen ovale with
a right-to-left shunt that was described as minimal. No clot was seen. The patient had total
resolution of symptoms. Hypercoagulable workup was normal, but incomplete. No tests of
protein S or C (proteins involved in blood clotting) were reported in the FDA material.
Therefore, we classified this case as a possible sentinel event. (14378)
A 42-year-old male who was taking Slim N Up awoke with paresthesia on the left and
difficulty walking and talking. He was a nonsmoker for five years. The patient had a known
diagnosis of hypertension, was on Diltiazem, and also had hypercholesterolemia. Emergency
department blood pressure was 132/89. Lumbar puncture in the emergency room was traumatic,
with 35 red blood cells in tube 1 and 0 red blood cells in tube 4. A transcranial Doppler study
was negative, carotid duplex study was negative, echocardiogram was negative, and MRI
91

showed left cerebella infarct. Subsequent admissions were for seizure control. We classified this
case as a possible sentinel event due to the prior hypertension and hypercholesterolemia and the
lack of complete hypercoagulability workup. (14434)
A 55-year-old female who had been taking Metabolife 356 for 60 days developed a
headache, had a seizure, and became unconscious. She was taken to the emergency department,
where she was found to have a large subarachnoid hemorrhage. An emergency angiogram
showed a large right posterior communicating artery aneurysm, which was subsequently clipped.
She had a stormy course complicated by meningitis, hydrocephalus, and placement of a
ventricular peritoneal shunt. Because of the preexisting large aneurysm, we classified this case as
a possible sentinel event. (14553)
Literature CaseEphedra
A 33-year-old man who had been taking ma huang (4060 mg per day of ephedra alkaloids)
for energy and weight training awoke with a severe Wernickes aphasia. He had not complained
of prior headache or other symptoms. He had a slight right-sided facial and arm weakness and a
right Babinski sign. His blood pressure was 140/60, and his pulse was 54 per minute. Brain CT
showed signs of extensive left middle cerebral artery infarct. Cervical ultrasound duplex
scanning and cerebral angiography were normal. Cerebral CSF examination was normal. The
report contained no coagulopathy assessment other than D-dimers (cross-linked fibrin molecules
that may be a diagnostic marker for venous thromboembolism), which were within the normal
range. Creatinine was in the normal range. Transesophageal echocardiography and ECG were
also normal except for a patent foramen ovale. We classified this case as a possible sentinel
event, because there was no additional documentation about the details of the coagulopathy
evaluation. (270)
Literature CasesEphedrine
A 37-year-old male who ingested 10 pills of a street drug containing ephedrine (15.3 mg
per tablet, as identified by subsequent analysis), developed sudden right body numbness and, on
evaluation, was found to have pure right body sensory loss with a left thalamic infarct on MRI.
Laboratory studies included normal prothrombin time and partial thromboplastin times,
sedimentation rate, electrocardiogram and transthoracic electrocardiogram. The patient refused
arteriography. As a result, we classified this case as a possible sentinel event. (44)
A 20-year-old man was admitted with nausea, vomiting, and headache that began one hour
after he took an unknown quantity of what he called speed. Urine drug screen on the day of
admission revealed only ephedrine, in particular excluding amphetamine, phenylpropanolamine,
caffeine, and other drugs. CT scan obtained on admission demonstrated blood in the
subarachnoid space, which was confirmed by lumbar puncture. Angiography on the day of
admission was normal, and rheumatologic evaluation was negative. A repeat angiogram seven
days later showed features typical of vasculitis. We classified this case as a possible sentinel
event. (438)

92

Insufficient Information
FDA CasesEphedra
A 36-year-old female who was taking Natures Nutrition-Formula One for weight loss
developed a headache. Based on CT and MRI, she was diagnosed as having had a stroke. She
was a nonsmoker and was not taking oral contraceptive pills. This information was obtained
from notes of a conversation with the patient herself. The notes stated that medical records were
requested; however, none appeared in the FDA file. Therefore, we classified this case as having
insufficient information. (9521)
A 30-year-old female who took Metabolife 356 for one week developed a headache while
eating and had a stroke. A friend drove her to the hospital; en route, they encountered a
paramedic, who obtained a blood pressure reading of 249/131. She stated that she was on no
medications, had no hypertension, didnt smoke, and didnt drink. Unfortunately, no additional
information appeared in this record. Therefore, we classified this case as having insufficient
information. (13829)
A 36-year-old female who took Metabolife 356 had respiratory failure and a possible
stroke. The file contains a note stating that medical records were requested but were never
received. Thus, we classified this case as having insufficient information. (13905)
FDA CaseEphedrine
A 32-year-old female who, according to case notes, was taking over 100 Maxi Thins per
day for five years and was addicted to product, had three cerebral hemorrhages and two strokes
and was hospitalized for two months. No additional information is available. Thus, we classified
this case as having insufficient information and also note the extraordinary dose of ephedrine
being consumed. (1823550)
Literature CaseEphedrine
A 68-year-old man who had been taking an over-the-counter anti-asthma pill containing
4060 mg of ephedrine per day for 10 years had a left temporal-parietal hematoma with rupture
into the lateral ventricle. Angiography showed changes consistent with vasculitis, and
pathological examination from material obtained during surgical evacuation of his hematoma
showed necrotizing angiitis of the small vessels. The patient improved with prednisone. We
classified this case as having insufficient information. (515)
Other Cardiac and Other Neurological Cases
Near Sudden Death
A 22-year-old male who regularly used Ripped Force along with a variety of other
supplements collapsed while lifting weights and had a ventricular fibrillation cardiac arrest
complicated by hypoxic encephalopathy. Although he had a history of asthma, this was not felt
to be an asthmatic attack. Toxicological examination revealed ephedrine, pseudoephedrine,
methyl ephedrine, and phenylpropanolamine. An echocardiogram ruled out asymmetric septal
hypertrophy but did reveal a reduced left ventricular ejection fraction (3540 percent) and an
increased left ventricular end diastolic dimension. CT scan showed no brain tumor or bleed. A
pulmonary consultant who saw him in the hospital stated that he doubts that this incident was
related to asthma. He recovered to the point where he could feed himself but he does not
93

remember his friends and has substantial mental disability. We classified this as a possible
sentinel event since the echocardiogram results raise the possibility that this patient could have
had a cardiomyopathy, which then could be the cause of the cardiac arrest. (12851)
A 28-year-old female who took Herbalife Original Green had a cardiac arrest later that same
day while playing softball. According to the affidavit, she needed to be defibrillated four times
and now has a permanently implanted defibrillator. Unfortunately, other than this affidavit, no
information was available. Therefore, we classified this case as having insufficient information.
(13031)
A 32-year-old female had been taking Natural Trim for two weeks. On one morning, after
taking Natural Trim, she ate lunch, went outside her office building to smoke a cigarette, and had
a witnessed cardiac arrest. A physician bystander initiated CPR, and she was taken to the
Emergency Department with decorticate posturing. Although successfully resuscitated, she was
left with permanent heart and brain damage. She also had a permanent intracardiac defibrillator
implanted. Notes from the FDA investigator said that her hospital records showed she had
chronic obstructive pulmonary disease, ventricular fibrillation, and cardiomyopathy versus
acute myocarditis with possible contributing factor of cardiotoxic diet pill. She had a left
ventricular ejection fraction of 35 percent with global left ventricular hypokinesia, and
endomyocardial biopsy found mild focal hypertrophy and mild focal interstitial fibrosis. No
medical records were available in the FDA files. Therefore, we classified this case as having
insufficient information. (13643)
Cardiomyopathy
A 28-year-old female who had been taking ephedrine tablets (2000 mg per day) for eight
years to lose weight presented with dilated cardiomyopathy. She denied any other chronic
alcohol or drug use except tobacco. She reported that after abruptly reducing the dose of
ephedrine to only 75 mg per day, she rapidly developed symptoms of dyspnea, fatigue, and
orthopnea (difficulty breathing while lying flat). Exhaustive diagnostic evaluation, including
cardiac catheterization and endomyocardial biopsy, revealed no specific diagnosis, and the
patients dilated cardiomyopathy was characterized as idiopathic. We classified this case as a
possible sentinel event, but note the extraordinary level of ephedrine use. (110)
A 39-year-old male with a history of hypertension presented with dyspnea on exertion,
orthopnea, and edema. He had been taking numerous Herbalife supplements (including Original
Green) for three months, which provided a total of between 7 and 21 mg of ephedrine alkaloid
daily. Exhaustive diagnostic evaluation, including endomyocardial biopsy, yielded a diagnosis of
hypersensitivity or eosinophilic myocarditis. He was treated with azothioprine and prednisone,
and Herbalife medications were discontinued. At six months follow-up, his heart function was
normal. We classified this as a possible sentinel event. (297)
A 32-year-old housewife who was noted to be abusing ephedrine, taking up to 450 mg a day,
presented with congestive cardiac failure and received a clinical diagnosis of congestive
cardiomyopathy of unknown etiology. No coronary angiography or myocardial biopsy was
performed, which may have been within the standard of practice at the time of the case (1980).
94

We classified this case as having insufficient information and also note the high level of
ephedrine intake. (260)
A 65-year-old female who had been taking the product Thermolean for two years was
hospitalized with acute congestive heart failure. Evaluation revealed cardiomyopathy with a left
ventricular ejection fraction estimated at 15 percent and atrial fibrillation. It was the treating
physicians opinion that the cardiomyopathy was probably secondary to ephedrine use. There
were no medical records available with this file. Therefore, we classified this case as having
insufficient information. (13793)
Ventricular Tachycardia
A 48-year-old female was taking Metabolife 356 for approximately one month when she
developed a rapid heartbeat that would not subside. She was taken to the emergency department
and found to be in ventricular tachycardia. The file contained no information on diagnosis or
treatment procedures, although the MedWatch form stated that the patient said she was taking
beta-blockers. No medical records were available for this adverse event. Therefore, we classified
this case as having insufficient information. (13945)
Transient Ischemic Attack
A 57-year-old female with prior history of hypothyroidism, gastroesophageal reflux disease,
depression, degenerative joint disease, and fibromyalgia began having nausea and vomiting and
became disoriented. She had been taking Synthroid, Oxycontin, Prozac, Trazodone, Prilosec, and
one other medication whose name was illegible in the file notes. She also took Metabolife 356
for one day and a total of 48 mg of ephedrine prior to the adverse events. Her husband took her
to the emergency department, where an examination was inconclusive. Laboratory values were
essentially normal. A toxicology screen was positive for opiates but negative for amphetamines.
A CT scan was negative. She was seen in consultation by a neurologist who told her that she had
a vasospasm transient ischemic attack, possibly related to ephedrine use, and the discharge
instructions were to stop using the Metabolife 356 supplement. We classified this case as a
possible sentinel event, because the symptoms alone do not confirm that she had a transient
ischemic attack. (13062)
Seizure
Sentinel Events
FDA CaseEphedra
A 19-year-old female who reported using Shape Rite/Shape Fast at half the recommended
dose for three to four weeks had one witnessed episode in which her arms and legs went stiff,
noticeable drool appeared, eyes rolled, [and she] appeared to black out, followed by a postictal
period (period of confusion typically observed following a seizure). She had no prior history of
seizures. Electrolytes were normal; complete blood work was normal; and pregnancy test was
negative. She had a normal CT scan without contrast and a normal electroencephalogram. She
was seen by a consultant neurologist, whose diagnosis was that she had a single-tonic seizure,
and none of the other factors which (sic) are normally associated with seizures, are present.
(10974)

95

Literature CaseEphedrine
A 38-year-old female with no prior seizure history experienced two petit mal seizures (the
authors description) after taking two tablets of over-the-counter ephedrine-containing dietary
supplements in the morning and evening. The following day, she had a generalized tonic-clonic
seizure, during which she required respiratory assistance. Over the next five days, she continued
to have petit mal and generalized tonic-clonic seizures. She was diagnosed with new onset of
tonic-clonic seizures with complex partial seizures. The report (in Morbidity and Mortality
Weekly) stated, Other possible causes of seizures were excluded. After discontinuing the
ephedrine-containing dietary supplement, she had no further seizures. (224)
Possible Sentinel Events
FDA CasesEphedra
A 47-year-old female who had been taking Natures Nutrition-Formula One for three weeks
to lose weight (one pill twice a day) had a tonic-clonic seizure in her sleep at 2 a.m. Her husband
took her to the emergency department, where some evaluation was done, but she was treated and
released on no therapy. Two months later, she had another seizure, which occurred at 5 a.m. At
that time, she was transported by ambulance to the emergency department and was seen by a
neurologist. A random glucose was 90, electrolytes were normal, sedimentation rate was 52,
MRI was normal, and EEG was described as mildly abnormal. She had a remote history of
hysterectomy and ear surgery two years prior to the event. Her only medication was Premarin,
and she used alcohol only socially. It was the neurologists opinion that the patient had new
onset generalized tonic-clonic seizures, and hypoglycemia was suspected but could not be
proved. The patients subsequent course was complicated by rash, fever, mild pancytopenia, and
increased liver function tests, which were thought due to her anticonvulsive therapy. We
classified this case as a possible sentinel event, because other causes were considered and not
effectively excluded. (9534)
A 37-year-old female who was taking Natures Nutrition-Formula One was admitted to the
hospital for symptoms of dizziness, shortness of breath, palpitations, and passing out, and
intermittent episodes of confusion, with one episode of shaking of limbs and saliva coming
out of her mouth. She had no prior history of seizures. She denied using alcohol. She was seen
in consultation by a neurologist. Electrolytes were normal, complete blood count was normal,
arterial blood gas was normal, glucose was 179, magnesium was normal, toxicology screen was
negative, pregnancy test was negative, CT scan with and without contrast was read as a 0.8 by
0.6 centimeter calcification in the frontal region, which may represent dural calcification.
Subsequent MRI was normal and showed no evidence of calcification in the area seen on CT
scan. Mild chronic right sinusitis was noted. EEG was interpreted as mildly abnormal due to
excessive intermittent bi-temporal slowing. No epileptogenic activity was seen. The
impression of the neurologist was complex seizures with generalization. We classified this
case as a possible sentinel event. (10221)
A 62-year-old male who had been taking the product Thermo Slim for three to four months
for weight loss began to have periods of memory loss, confusion, and agitation. He then had a
generalized seizure with lateralizing features characterized by right sided tonic-clonic jerking
and was admitted to the hospital. The patient had a prior history of heavy alcohol consumption
until approximately four years before the event. In the emergency room, the patient was noted to
96

be normotensive and had no fever. He was characterized to be in acute delirium. Although he

had no prior history of diabetes mellitus, his blood sugar on admission was 488, and he was
given the diagnosis of diabetes. Toxicology screen was positive for benzodiazepines. Arterial
blood gases revealed adequate oxygenation. CT scan of the brain without contrast showed
atrophy but was otherwise unremarkable. MRI showed generalized brain atrophy.
Electroencephalogram showed moderate to severe abnormality, with bilateral cerebral
dysfunction; however, no epileptiform activity was identified. He was given a diagnosis of
organic brain syndrome with senile dementia and also the possibility of left temporal lobe
cerebral infarct with secondary seizure. One consultant raised the possibility of acute
encephalopathy of uncertain etiology. Because of the focal nature of the seizures and the
multiple metabolic problems, we classified this case as a possible sentinel event. (10432)
A 23-year-old female was taking Metabolife 356, one pill two to four times a day for four
months. While driving out of a parking lot, she had a generalized tonic-clonic seizure witnessed
by her husband. Her husband controlled the car and prevented a crash. The seizure lasted for two
minutes (during which it was noted that the patient bit her tongue) and was followed by a
postictal period. She had had one seizure two years prior, which had been only partially
evaluated (she had not received a CT scan at that time), and she was not treated. In addition, she
had one sibling who had had a seizure at age 8, and the paternal grandparents were noted to have
had seizures. She reported using alcohol only rarely. On evaluation, CT scan (non-contrast) was
negative, oxygen saturation was 99 percent, toxicology screen was negative, pregnancy test was
negative, and EEG was abnormal, indicative of a potential underlying seizure disorder.
Because of the prior history of seizures, we classified this as a possible sentinel event. (11649)
A 26-year-old male who had been using the product Ripped Fuel for approximately three
years developed a headache that lasted approximately three days and then began experiencing
seizures and was taken to the emergency room. While in the emergency department, he had a
witnessed generalized tonic-clonic seizure. He had taken Ripped Fuel on the day of the event.
Arterial blood gas and glucose were within normal limits. Drug screen was negative. Electrolytes
were normal except for potassium of 3.2 and bicarbonate of 15.8. He had no history of medical
illness, alcoholism, or serious injury. The family had no history of seizures. He had no further
seizures over a three-hour period and was discharged home with referral to neurology. Two days
later, he had another seizure and returned to the emergency department, where he was admitted
to the hospital. Neurological consultation considered this case to be a complex partial seizure
with generalized tonic-clonic seizure of new onset and to have a recent history suggestive of
migraine headaches. Despite being given therapeutic doses of anti-seizure medications, he
continued to have seizures to the point where he was placed in a Phenobarbital coma, requiring
mechanical ventilation. After several weeks of a drug-induced coma, he was weaned off the
Phenobarbital, but remained on anti-seizure medication and was left with a metabolic
encephalopathy. He was transferred to a rehabilitation center. We classified this case as a
possible sentinel event, because the clinical cause was quite complicated and from the records
we had available we could not reach a more definitive conclusion. (13408)
A 30-year-old female was found by her husband to be moaning and making noises. She
became limp and subsequently very confused. Paramedics were called and a blood glucose
measured in the home was normal. She had been taking Metab-O-Lite for two to three months
97

for weight loss (the records are inconsistent on this point). She took one tablet before lunch and
one tablet before dinner. Her last dose was on the day of the episode. She was on no medications,
and had no personal or family history of seizures. She was seen in consultation by a neurologist.
MRI with contrast was normal. An additional CT scan of the head, which was done without
contrast, was interpreted as negative. Serum electrolytes were normal, glucose was recorded in
a dictated note as 19, but this is not commented on anywhere else in the notes and our
presumption is that this is a typographical error. Serum calcium was normal, as was the complete
blood count. The neurologist ordered an EEG, which showed an abnormality due to the presence
of some sharp waves emerging from the left hemisphere, mainly the parietal region. The
neurologists impression was that this was most likely a seizure, and the patient was started
and maintained on Dilantin. No further seizures were noted as of a follow-up three months after
the event. We classified this case as a possible sentinel event, because it was not clear that the
event was a seizure. (14275)
A 31-year-old female who was taking Thin Tabs (one tablet three times a day for
approximately one month) developed a headache over her left eye, which became more severe
after she took aspirin. The headache was followed by visual blurring, nausea, and vomiting, but
no scintillations. She became tremulous, incoherent, and lethargic. She was taken by ambulance
to the emergency room, where she had a generalized tonic-clonic seizure. In the emergency
room, blood pressure was 165/107, pulse was 101, and respiratory rate was 24. Blood glucose
was recorded as slightly elevated, and serum chemistries were normal. Sedimentation rate was
normal. Non-contrast CT scan of the head was normal. Lumbar puncture was unremarkable.
Gram stain was negative. Urine drug screen revealed amphetamines. She had no prior history of
seizure. Physicians notes stated that the patient says she was abusing her diet pills, an
assertion that was later denied in the medical record. Her medical history was remarkable for
depression, for which she was being treated with Depakote (a drug used to treat seizure
disorders, bipolar disorder, and schizophrenia), Trazodone, and Paxil (two antidepressants). It is
also stated that she had an MRI, but those results were not in the material available for review.
She had an electroencephalogram, which was normal, but which did not entirely rule out a
seizure disorder, as no Stage II sleep was seen, and a short record can miss intermittent
phenomenon. Because of her EEG and in light of her taking other medications known to lower
the seizure threshold, we classified this case as a possible sentinel event. (14571)
Insufficient Information
FDA CaseEphedra
A 40-year-old female who had taken Ripped Fuel (two capsules, two times per day) for two
days had a generalized tonic-clonic seizure (witnessed by her husband) while cooking dinner in
her kitchen. She had no history of seizures. During the seizure, she fell, suffering a laceration to
her head and was taken to the emergency department. At that time, glucose was 104, serum
electrolytes were normal, and CT scan with and without contrast was normal. No report of an
electroencephalogram was in the file. We classified this case as having insufficient information.
(9747)

98

Psychiatric Symptoms
Sentinel Events
FDA CasesEphedra
A 21-year-old male took five to seven Natures Nutrition-Formula One capsules in one day
to stay alert while studying for final exams. He became psychotic and did not sleep for five days.
A friend said he ran around campus looking like a homeless crazy person. The patient had no
history of psychiatric or medical problems. (9509)
A 39-year-old female took Diet Now (tested by the manufacturer and said to contain 6 mg
ephedra alkaloids per capsule, 12 mg per dose) for approximately one year at recommended
doses. Her mother reports that the daughter experienced insomnia, hallucinations, psychosis, and
delusions with the onset approximately one year after product initiation. She required
hospitalization in a psychiatric facility for 40 days, with ongoing problems including terror,
panic, and forgetfulness. She has now returned at work. (11678)
A 19-year-old female was taking Hydroxycut 2 pills twice a day to aid muscle definition and
to speed metabolism. She reported dizziness and nausea two hours after use and began having
violent outbursts, nightmares, poor mood, hot flashes, and fatigue. After a few days, she
developed increased anger and rage and fought with boyfriend, mother, father, and sisters. She
also tried to kill her boyfriends sister and herself. After eight days of use, she developed a
migraine and went to the emergency room. She then went home and picked up a knife, with
homicidal intent, but was convinced to return to the hospital voluntarily. She was admitted for 18
hours and was readmitted later that day for a 72-hour involuntary hospitalization. Symptoms
abated four days after Hydroxycut was discontinued. (13809)
A 29-year-old male took Xenadrine (two tablets twice per day) as a diet supplement for over
six months. After six months of use, he was hospitalized three times for hyper-religiosity,
paranoia, delusions, insomnia, and lack of concentration, and displayed some indication of the
onset of bipolar disorder, but had no known history of prior mental health problems. Symptoms
recurred twice more following use. (14529)
FDA CaseEphedrine
A 16-year-old male took MaxAlert and Mini Thin for 11 months, often ingesting up to 40
tablets per day for weight loss and as a stimulant. The patient had episodes of aggressive
behavior, irritability, tachycardia, insomnia, and violent and destructive behavior. When he
visited a physician, it was noted that his symptoms coincided with an increase in dose of
MaxAlert. No significant medical history and no history of other drug use were noted. We
classified this as a sentinel event; however, we note the extraordinary use of product. (1855921)
Literature CaseEphedra
A 45-year-old male who had taken an herbal dietary supplement labeled as ma huang daily
for two months was brought to the emergency room by his wife when, after several weeks of
using greater amounts, he began to display irritability, sleeplessness, and strange religious
preoccupation. He had no medical or psychiatric history. His symptoms disappeared after brief
treatment with Trazodone and discontinuation of ephedra. (48)
99

Literature CasesEphedrine
A 30-year-old female had taken Tedral (which contains ephedrine) for asthma for many
years. She had no other medical problems and no history of psychiatric problems. Her mother
had noticed a marked change in her behavior over the previous two years, which seemed to
coincide with taking more Tedral than medically necessary. The patient became paranoid,
illogical, and hallucinatory. A diagnosis of acute schizophrenic psychosis, either due to or
aggravated by the abuse of ephedrine, was made. She was asked to stop taking Tedral but took it
occasionally until persuaded by her family doctor to switch to cromoglycic acid. At two years
follow-up, her symptoms had disappeared. (238)
A 59-year-old male who had taken ephedrine-containing products for over 25 years to treat
asthma experienced auditory hallucinations, was delusional, and entered a womans home,
believing he was saving her from being tortured. At the time of the event, he had been taking
ephedrine hydrochloride plus Bronchipax (ephedrine resinate 30mg; theophylline 40mg;
salicylamide 250 mg) but had just doubled his daily ephedrine dose to 360 mg ephedrine plus
Bronchipax. The patient had no history of psychiatric problems. The psychotic symptoms
diminished 1013 days after a reduction of the ephedrine dose. (285)
Possible Sentinel Events
FDA CasesEphedra
A 28-year-old female reportedly took one Slim NRG+ (ma huang) three times per day
without incident for over 6 months and lost 30 pounds. After abruptly discontinuing use of the
supplement, she was hospitalized for severe depression and a suicide attempt (gunshot wound to
chest). She took no concomitant medications. Because no information regarding psychiatric or
medical history was available, we classified this case as a possible sentinel event. (9751)
A 19-year-old man who took Ripped Fuel as directed (two capsules, three times per day) for
three weeks was hospitalized with palpitations, increased blood pressure, and psychosis. He had
no previous psychiatric problems or medical conditions. Because no information regarding
psychiatric or medical history was available, we classified this case as a possible sentinel event.
(11157)
A 13-year-old female took Natures Nutrition-Formula One for approximately two weeks at
recommended doses of approximately one tablet twice per day for weight loss. Her first
symptoms were noted approximately one to two weeks after she began to use the product. She
reported auditory hallucinations, disorientation to place, and withdrawal. Her symptoms endured
for approximately two months. No information regarding medical or psychiatric history was
included in the report, so we classified it as a possible sentinel event. (12372)
A 21-year-old male used Ripped Fuel as directed for two weeks. His parents reported
personality changes such as nervousness, anger, and rage, and he went long periods without
sleep. He had no previous psychiatric or medical problems. His parents reported he was sensitive
to caffeine, which is included in the product. His symptoms stopped after the product was
discontinued. (13005)

100

A 52-year-old female took Metab-O-Lite, two tablets three times per day for five months, for
weight loss. She reported hallucinations, psychosis, delusions, and paranoia, which led to
hospitalization in a state psychiatric facility. Her symptoms persisted for two to three days. She
had a history of asthma and high blood pressure and no history of alcohol abuse. The report
included a very confusing indication of a previous episode of hallucinations secondary to surgery
or perhaps to Metab-O-Lite a few months earlier than the identified adverse event. (14436)
A 28-year-old female who took Metab-O-Lite (eight pills per day) for over six months for
weight loss began to experience dizzy spells and headaches almost immediately after starting the
supplement. She later began to experience chest tightness and a racing heart. Approximately one
week after starting, she began to experience auditory hallucinations, delusions, and paranoia.
Auditory hallucinations and delusions endured for over a year after discontinuing the product,
thus we classified this as a possible sentinel event. (14528)
FDA CaseEphedrine
A 31-year-old man used Max Alert for over four years, gradually increasing the dose until he
was consuming 1,250 mg of ephedrine per day. He began to display psychotic behavior,
including paranoia. Over four years, he was hospitalized three times, and at the time of report,
was in a residential rehabilitation center for substance abuse treatment. The report states he had
never used illicit substances and had no significant medical history. We classified this case as a
possible sentinel event, but note the extraordinary dose of ephedrine. (1661966)
Literature CaseEphedra
A 34-year-old male was brought to the emergency room after jumping from a second story
window because he believed he was being chased. While taking ma huang over the previous nine
days, he had experienced paranoid delusions and visual hallucinations. He had no history of
mental illness. Medical history was not contained in the report. The patient was hospitalized for a
number of weeks. After discontinuing ephedra, he remained well. The ephedrine content of the
product was not noted in this case report; investigators contacted the manufacturer; however,
they were unable or unwilling to disclose the amount of ephedrine in each tablet. (79)
Other Adverse Events
The FDA file contained reports of other adverse events associated with ephedra use. We
briefly reviewed these reports in an attempt to more precisely establish the general nature of the
adverse events, but we did not review them in more detail to determine whether they satisfied the
three conditions necessary for a sentinel event. Table 25 presents the list of other adverse
events.
Metabolife File
The MIPER CD-ROMs contained 15,951 files. After removing duplicate, blank, and followup files, we had 18,502 cases for analysis, as indicated in Figure 18. Table 26 presents summary
data regarding the key variables from our abstraction form on these cases. In 57 percent of cases,
the consumers age was not included. The majority of the remaining cases were reported by
persons between the ages of 21 and 50, with a mean age of 38. In 66 percent of all cases, sex was
not recorded. Of the remaining cases, 91 percent were female.
101

A tabulation of the symptoms showed that there were three deaths, 22 cases of myocardial
infarction, three cases of cardiac arrest, 29 cases of stroke, two cases of brain hemorrhage, 46
cases of seizure, three cases of psychosis, and two cases of hallucinations. The files contained
111 cases of hospitalization in addition to those associated with the serious cases just listed.
These hospitalizations were for a variety of reasons, but most were for cardiovascular-related
symptoms.
The MIPER files for death, heart attack, cardiac arrest, stroke, seizure, and certain psychiatric
events were all reexamined by the principal investigator and other physicians and are listed in
Appendix 2 of this report. One case of death occurred as a result of a brain hemorrhage,
according to the notes. Another case involving a death contained a handwritten note that said,
wanted refund (sisters husb died). The third case stated cousin was taking Metabolife last yr,
had stroke, died. No additional information for these cases is present. Two additional deaths,
identified by Metabolife in a document entitled 77 serious AEs as identified by Metabolife (see
below), were not included in the MIPER CD-ROM we received. These additions bring the total
number of Metabolife-related deaths to five. The cases of other serious events range in
documentation from several sentences of clinical information related by the patient, letters from
patients stating that they had a serious adverse event, to simply the words heart attack or
something similar on a sheet of paper in the MIPER file. This level of documentation is
insufficient to make judgments about the possible relation between ephedra use and the event.
The largest proportions of case reports included symptoms of autonomic hyperactivity (14
percent of all cases) and gastrointestinal symptoms (26 percent of all cases). These data are
compatible with the results of our meta-analysis of adverse events in placebo-controlled trials of
ephedra and ephedrine, which demonstrated both autonomic hyperactivity and upper
gastrointestinal symptoms to be causally related to use of ephedra or ephedrine.
As mentioned above, included with the material we received was a document (a sheaf of
paper) entitled 77 serious AEs as identified by Metabolife. This sheaf contained photocopies of
the MIPER files judged by Metabolife to be the most serious in nature. These MIPER files
included reports of three deaths. Two of these deaths had the MIPER number blacked out, were
marked privileged and confidential, were not found on the MIPER CD-ROM by our
abstractors, and were not found using a modified MIPER CD-ROM that allowed text word
searching (prepared for us by FDA).
The documentation on both cases consisted, as did many of the Metabolife files, of a printed
version of an email. The first death was of a 45- to 55-year-old female, who was apparently
initially healthy and continued to take Metabolife 356 for three weeks, despite symptoms of
palpitations and a rapid pulse rate, until she suffered sudden death. An autopsy found no
conclusive cause of death. The email notes that toxicology studies are pending to ascertain if
ephedrine was present in her system at the time of death. No additional clinical information is
available. The second case was that of a 30- to 40-year-old female who was found dead.
According to her father, the autopsy stated that the cause of death was of a cardiac nature of an
unknown origin. Drug analysis found only caffeine. No other clinical information was
available.

102

Some cases identified by Metabolife as serious were not deemed so by us, whereas we
considered a greater number of its cases to be serious than Metabolife did (we did agree on cases
of death). For example, we identified six additional cases of myocardial infarction and nine
additional cases of stroke. Table 27 compares the cases we identified as serious with those
identified by Metabolife, along with a capsule explanation for the coding of discrepant cases.
Review of records with photocopies of medical information. The records varied greatly
from detailed medical records to simply photocopies of medical bills. Table 28 contains a
capsule description of each case and three numbering systems, because, as discussed in the
Methods section, the medical records were numbered in three ways. The first column contains
the number we assigned the records as we removed them from the shipping box. The second
column contains the case number as listed on the Index of Redacted Consumer Medical Records
with Corresponding MIPER Numbers. The third column contains the complaint case number
from the Listing of Key Complaint for the Metabolife Medical Records Submitted. The fourth
column contains the numbers for any related MIPER files that we identified or were identified by
Metabolife on the Index. The column labeled Notes is our capsule description of what we
received.
There were 12 cases with primarily cardiopulmonary symptoms (RAND ID cases 1, 13, 14,
16, 17, 20, 21, 23, 28, 29, 33, 43), two cases with neurologic symptoms (RAND ID cases 18,
38), two cases of seizure (RAND ID cases 32, 36), four cases of allergic reaction (RAND ID
cases 2, 25, 26, 37), and 23 cases of miscellaneous symptoms (RAND ID cases 3, 4, 5, 6, 8, 9,
10, 11, 12, 15, 19, 22, 24, 27, 30, 31, 34, 35, 39, 40, 41, 42). There were no deaths, one case of
myocardial infarction, no strokes, and no severe psychiatric events. The case of myocardial
infarction (RAND ID case 23) was classified as a possible sentinel event, due to the presence of
existing coronary artery disease. The two cases of seizure (RAND ID cases 32 and 36) were
classified as sentinel events. Comparing these cases to the FDA Medwatch data contained in our
Evidence Report demonstrates that neither the case of myocardial infarction nor the two cases of
seizure are reported as sentinel or possible sentinel events in our analysis of the Medwatch file;
thus, we are not double-counting these events.

103

Chapter 4. Limitations
We address the limitations of each set of analyses separately: meta-analysis of the weight
loss and descriptive synthesis of athletic performance randomized controlled trials; analysis of
the adverse events from the randomized controlled trials; and analysis of the case reports of
adverse events.
The systematic reviews of the weight loss and athletic performance randomized controlled
trials have the following potential limitations:

Our search procedures for randomized controlled trials were extensive and included
canvassing experts regarding studies we may have missed. In addition, we observed little
to no evidence of publication bias via visual inspection or formal testing for the weight
loss studies. However, we acknowledge that publication bias may still exist despite our
best efforts to conduct a comprehensive search and the lack of statistical evidence of the
existence of bias. Publication bias may occur for a variety of reasons, including
investigators loss of interest in the study if negative results are found or if results are
obtained that are contrary to the interest of the sponsor or investigator.

An important limitation common to many systematic reviews, whether or not a formal

meta-analysis is conducted, is the quality of the original studies. Many of the weight loss
studies suffered from an attrition rate higher than is normally allowed by FDA when
assessing studies of pharmaceutical products seeking approval. However, recent attempts
to define elements of study design and execution that are related to bias have shown that
in many cases, such efforts are not reproducible and do not distinguish studies based on
their results. Therefore, the current state of the science is to document such
methodological weaknesses and perform sensitivity analyses when possible, but not to
reject studies or use quality criteria to adjust the pooled outcome. We performed a
sensitivity analysis on the subset of studies that had the best quality, according to the only
validated scale available. The results of the sensitivity analysis did not alter the majority
of our findings.

In our meta-analysis of the weight loss studies, we did not observe significant evidence of
heterogeneity. However, the chi-squared test of heterogeneity is underpowered. We did
use a random effects approach to attempt to incorporate any heterogeneity and conducted
sensitivity analyses to assess the robustness of our conclusions.

We were limited by the small number of trials that provided direct comparisons between
treatments of interest in the weight loss meta-analysis. Our meta-regression in this setting
was an attempt to compare treatments across trials, but we acknowledge this approach
does not allow for controlling for confounders within study. Direct comparisons are
needed to draw more definite conclusions. In other words, while our observed results
suggest that the amount of weight loss is approximately the same for ephedrine with
caffeine, herbal ephedra with herbs containing caffeine, and herbal ephedra alone, the
available data do not prove equivalence.
197

The weight loss studies as a group had limited treatment duration; thus, we cannot draw
conclusions about the association between ephedra or ephedrine and weight loss over
longer and more clinically relevant intervals than about four months. Current knowledge
of weight loss is that it generally ceases after six months, irrespective of treatment, and
any weight loss is generally regained. Current recommendations for appropriate clinical
trials in this area include a much longer treatment duration (at least one year) and an
evaluation of what happens after the agent is withdrawn.

The heterogeneity among the athletic performance studies prevented us from conducting
a formal meta-analysis, so we were restricted to a descriptive synthesis.

The results of the clinical trials are directly applicable only to the persons studied in those
trials. In most cases, enrollment was highly selective to avoid certain comorbidities.
Whether efficacy would be equivalent in a more representative population is unknown.

The results of the ephedra studies regarding efficacy cannot be generalized to all ephedracontaining dietary supplements, because these may vary in their constituents from the
concoctions studied and reported on here.

The analysis of the adverse events from the randomized controlled trials have the following
major potential limitations:

In this analysis, we focused only on studies that addressed weight loss or athletic
performance. Although we observed no serious adverse events in these trials, we might
have identified adverse events in trials that tested the efficacy of ephedra for other
conditions, had we included those conditions in our search. However, we did include all
controlled trials of ephedra or ephedrine for weight loss or athletic performance;
therefore, our estimates are relevant to the populations taking those supplements for these
reasons, which certainly constitute the majority of users of ephedrine and ephedra
products in the United States.

As with efficacy, the results of the clinical trials with respect to safety are directly
applicable only to the persons studied in those trials. In most cases, enrollment was
highly selective to avoid certain comorbidities. Whether safety is equivalent in a more
representative population is unknown.

As with efficacy, the results for the ephedra studies with respect to safety cannot be
generalized to all ephedra-containing dietary supplements, because these may vary in
their constituents from those concoctions studied and reported on here.

The analysis of the case reports of adverse events had the following major potential
limitations:

We did not have access to all adverse event files.

198

Many authorities consider MedWatch case reports to underestimate the number of events,
because patients need to suspect an association in order to report an event.

This report did not review in detail other lines of evidence, such as animal studies, basic
neuroscience studies, and adverse event data concerning other sympathomimetic amines
that some authorities consider important when trying to assess causation.

Many of the adverse event reports did not contain all the data that we needed to make
assessments. Therefore, how the cases we classified as insufficient evidence might
have influenced our findings had they contained appropriate documentation is unknown.

An important limitation is that we do not have an estimate of the number of people using
ephedra or ephedrine; that is, we do not have a denominator with which to calculate an
event rate. An additional complication, we believe, is that the use of ephedra and
ephedrine is increasing over time, as is the probability that someone will report an
adverse event due to publicity.

The most important limitation is that the study design (that is, an assessment of case
reports) is insufficient for us to reach conclusions regarding causality.

The major potential limitations of the analysis of the Metabolife files can be classified into
two categories: limitations of the source material and limitations of our methods.
The source material for this review differed in several important ways from source material
used in other EPC projects:

Much of what we reviewed was handwritten. Therefore, when the handwriting was poor
we may not have correctly interpreted what the writer meant to say.

The information was not recorded in an organized fashion, leaving it up to us to interpret

its meaning. A good example of this was MIPER 23695 that we (but not Metabolife)
classified as a death. This file consisted of handwritten notes that stated, migraine HA,
wants refund, sisters husb died. Does this mean the customer is the sisters husband,
who had a migraine headache and then died? Or did the customer have a migraine
headache, perhaps in part because her sisters husband died? Without additional
information it is impossible to tell.

Each file did not attempt to collect the same information, so a recording bias probably
exists.

As already noted, we are not confident we could identify all files associated with a single
case, so some double-counting may have occurred.

199

The methods we used to review the files also had important limitations:

We relied on single-person review to screen cases. In the eight weeks we were given to
review the files, we could not do dual review (which is standard in all our other EPC
work) of over 18,000 cases. Therefore, more coding errors may have occurred than in
situations where we use dual review. Mitigating this limitation is that we did do formal
inter-rater reliability testing and demonstrated excellent reliability among reviewers.
Also, the principal investigator reviewed all cases that were identified as serious.
Furthermore, we identified nearly all the serious events identified by Metabolife, plus
many more that Metabolife did not identify. So, while we acknowledge that there may
still be errors in the data, we do not think they are so numerous or egregious as to
threaten our conclusions.

The Metabolife analysis did not undergo as extensive a review process as did the other
sections of this report. The Metabolife analysis was reviewed by three experts and two
federal agencies, in contrast to the much more extensive review process for the other
sections of the report. Furthermore, because of the timeline necessary to produce the final
report, the time available to the reviewers of the Metabolife analysis was shorter than we
normally afford. How additional peer review may have affected our conclusions from the
Metabolife analysis is unknown.

200

Chapter 5. Conclusions
Efficacy
The efficacy of herbal ephedracontaining dietary supplements has not been extensively
studied in randomized clinical trials. We identified no clinical trials of herbal ephedracontaining
dietary supplements that assessed their effect on athletic performance and only five clinical trials
that assessed their effect on weight loss. Many more studies assessed the effects of ephedrine on
weight loss; however, studies of the effects of ephedrine on athletic performance are still
relatively sparse. The majority of studiesof both ephedra and ephedrineare plagued by
methodological problems known to be associated with bias, particularly high attrition rates. All
of the conclusions on efficacy need to be considered with these methodological limitations in
mind.
Given the above considerations, the evidence we identified and assessed supports the
following conclusions:

Weight Loss

The short-term use of ephedrine, ephedrine plus caffeine, or the assessed dietary
supplements containing ephedra and herbs with caffeine is associated with a statistically
significant increase in short-term weight loss (compared to placebo).

There are no studies assessing the long-term effects of the use of ephedra-containing
dietary supplements or ephedrine on weight loss or maintenance. In order to improve
health outcomes and reduce the risk of morbidities associated with being overweight,
long-term weight maintenance is necessary.

There are no data to indicate that the effects of ephedrine plus caffeine are different from
the effects of ephedra-containing dietary supplements with caffeine-containing herbs.

The effect of either ephedra-containing dietary supplements with caffeine-containing

herbs or ephedrine plus caffeine is a weight loss that is approximately two pounds per
month greater than that of placebo, for up to four to six months in duration.

As a percentage of pretreatment weight, the weight losses in these studies average

between 5 percent and 11 percent in the treatment groups.

The only two studies that compared ephedrine plus caffeine to prescription weight loss
pharmaceutical products reported no differences in effectiveness between products, but
these studies were statistically underpowered to detect differences of moderate size.

The addition of caffeine to ephedrine is associated with a statistically significant increase

in short-term weight loss.
201

One study of ephedra without caffeine-containing herbs reported a statistically significant

increase in short-term weight loss that was comparable to the effects reported by four
studies of ephedra with caffeine-containing herbs.

The data suggest a dose-response relationship with respect to ephedrine and weight loss.

All published studies on herbal ephedra and weight loss have used a medium dose of
ephedra per day; consequently, no dose-response analysis is possible.

Athletic Performance

There are no studies assessing the effect of herbal ephedracontaining dietary

supplements on athletic performance.

The few studies that assessed the effect of ephedrine on athletic performance did so only
in small samples of mostly fit individuals (young male military recruits) and only on very
short-term immediate performance. This model does not reflect the use patterns in the
general population. These data support a modest effect of ephedrine plus caffeine on very
short-term athletic performance.

No studies assessed the effect of the sustained use of ephedrine on performance over
time.

It is probable that ephedrine alone, without the addition of caffeine, has little or no effect
on athletic performance.

In the data we reviewed, the smallest dose of ephedrine that produced a measurable effect
on athletic performance was 0.8 mg per kg of body weight. However, the effect of
smaller doses has not been assessed. Higher doses produced unacceptable gastrointestinal
side effects.

Adverse Consequences
The data we reviewed on adverse consequences came from both clinical trials and case
reports submitted to the FDA. The strongest evidence of causality should come from clinical
trials; however, in most circumstances, such trials do not enroll sufficient numbers of patients to
adequately assess the possibility of rare outcomes. Such was the case with our review of
ephedrine and ephedra-containing dietary supplements. For rare outcomes, we reviewed case
reports. However, we could not determine definite causality from case reports.
With these considerations in mind, the evidence we identified supports the following
conclusions:

There is sufficient evidence from short-term controlled trials to conclude that the use of
ephedrine and/or the use of ephedra or ephedrine plus caffeine is associated with two to
three times the risk of nausea, vomiting, psychiatric symptoms such as anxiety and
202

change in mood, autonomic hyperactivity, and palpitations. It is not possible to separate

out the contribution of caffeine to these events.

There were no reports of serious adverse events in the controlled trials of ephedrine or
ephedra, but these studies are insufficient to assess adverse events that occurred at a rate
of less than 1.0 per 1000.

A large number of adverse event reports regarding herbal ephedracontaining dietary

supplements have been filed with FDA. The majority of FDA case reports are
insufficiently documented to make an informed judgment about the relationship between
the use of ephedra-containing dietary supplements and the adverse event in question.

A very large number of adverse events were reported to one manufacturer of ephedracontaining dietary supplements. Nearly all of the case reports were too poorly
documented to permit us to make any judgments about the potential relationship between
ephedra use and the event.

We identified two deaths, three myocardial infarctions, nine cerebrovascular accidents,

three seizures, and five psychiatric cases as sentinel events with prior ephedra
consumption; and three deaths, two myocardial infarctions, two cerebrovascular
accidents, one seizure, and three psychiatric cases as sentinel events with prior ephedrine
consumption. Classification as a sentinel event does not imply a proven cause and effect
relationship.

We identified 43 additional cases as possible sentinel events with prior ephedra

consumption and seven additional cases as possible sentinel events with prior ephedrine
consumption.

About half of the sentinel events occurred in persons aged 30 years or younger.

Scientific studies (not additional case reports) are necessary in order to assess the
possible association between consumption of ephedra-containing dietary supplements and
these serious adverse events. Given the rarity of such events, a properly designed case
control study would be the appropriate next step. Such a study would need to control for
caffeine consumption.

203

Chapter 6. Future Research

Our analysis of the evidence reveals numerous gaps in the literature regarding the
efficacy and safety of ephedrine and ephedra-containing dietary supplements. The most
important of these gaps are the following:

Long-term assessments of the effectiveness of ephedra or ephedrine at promoting

weight loss. We identified no study having a treatment duration of more than six
months. In order to improve health outcomes and reduce the risk of morbidities
associated with being overweight, sufficient weight loss (5 to 10 percent of body
weight) and long-term weight maintenance are necessary.

A study of the effect of repeated use of ephedra or ephedrine on athletic

performance in a variety of people including women and adolescents who are
known users of these products. If use of ephedra-containing dietary supplements
is going to continue to be promoted for improving athletic performance, then
evidence is needed regarding their efficacy in individuals who represent the
general population.

A proper study to assess the possible association of ephedra or ephedrine

consumption and the occurrence of serious adverse events. Continued analysis of
case reports cannot substitute for a properly designed study to assess causality. A
case-control study would probably be the study design of choice.

A partial list of other possible future research activities includes the following:

Consider a dose-response study that would determine the minimum effective dose
of ephedra and caffeine-containing herbs, or ephedra combined with other
botanicals such as citrus durantium, garcinia cambogia, and other herbal diuretics
and cathartics, for weight loss.

Assess whether ephedra/ephedrine and exercise training interact in their effects on

weight loss and adverse events.

Assess adverse events for ephedrine and other prescription obesity drugs in
Denmark, where doctors began prescribing an ephedrine-containing diet drug
more than 20 years ago.

Conduct studies to determine if the use of ephedrine or ephedrine-containing

alkaloids increases the risk of development of heat-related conditions such as heat
exhaustion, heat stroke, and rhabdomyolysis.

Investigate further the interactions between ephedra/ephedrine and other products

commonly used in the United States for weight loss and/or athletic performance.

205

References
1. McGuffin M. Statement of Michael McGuffin.
Hearings Before the FDA, Docket (Aug. 8,
2000).

13. Centers for Disease Control and Prevention.

Obesity Epidemic Increases Dramatically in
the United States. available at URL:
http://www.cdc.gov/nccdphp/dnpa/obesityepidemic.htm. Accessed 6/11/2002.

2. Abrahamson A. Wheeler used supplement. Los

Angeles Times. Aug. 11, 2001; Sect: Sports.
3. Fernas R, Heckert J. Substances are easily
obtained. Los Angeles Times. Aug. 11,
2001; Sect: Sports.

14. Kuczmarski RJ, Flegal KM, Campbell SM, et

al. Increasing prevalence of overweight
among US adults. JAMA. 1994;272:205211.

4. Springer S. Bobby Bowden says deaths have

him scared. Los Angeles Times. Aug. 11,
2001; Sect: Sports.

15. Katzmarzyk PT. The Canadian obesity

epidemic, 1985-1998. CMAJ.
2002;166(8):1039-40.

5. Carey B. Risks of ephedra usage in spotlight.

Los Angeles Times. Aug. 27, 2001; Sect:
Health.

16. Katzmarzyk PT. Obesity in Canadian children.

CMAJ. 2001;164(11):1563-4; discussion
1564-5.

6. Barnes J., Winter G. Major food companies

using herbal additives. San Francisco
Chronicle. May 27, 2001.

17. World Health Organization. Controlling the

global obesity epidemic. available at URL:
http://www.who.int/nut/obs.htm. Accessed
6/20/2002.

7. Rogers P. A deadly diet aid? People Magazine.

July 1, 2002.

18. Friedrich MJ. Epidemic of obesity expands its

spread to developing countries. JAMA.
2002;287(11):1382-6.

8. Fong H. Ephedra-containing compounds:

Historical and pharmacologic context.
Hearings Before the US Public Health
Service (Aug. 8, 2000).

19. Must A, Spadano J, Coakley EH, Field AE,

Colditz G, Dietz WH. The disease burden
associated with overweight and obesity.
JAMA. 1999;282(16):1523-9.

9. Love LA. Adverse event reports: Database and

clinical studies. Hearings Before the US
Public Health Service (Aug. 8, 2000).

20. Pi SY. Medical Hazards of Obesity. Ann Intern

Med. 1993;119:S655-660.

10. Mokdad AH, Bowman BA, Ford ES, Vinicor F,

Marks JS, Koplan JP. The continuing
epidemics of obesity and diabetes in the
United States. JAMA. 2001;286(10):1195200.

21. Pi-Sunyer XF. Medical hazards of obesity. Ann

Intern Med. 1993;119(7 pt 2):655-660.
22. Allison DB, Fontaine KR, Manson JE, Stevens
J, VanItallie TB. Annual deaths attributable
to obesity in the United States. JAMA.
1999;282(16):1530-8.

11. Mokdad AH, Serdula MK, Dietz WH, Bowman

BA, Marks JS, Koplan JP. The continuing
epidemic of obesity in the United States.
JAMA. 2000;284(13):1650-1.

23. Colditz GA. Economic costs of obesity. Am J

Clin Nutr. 1992;55(2 Suppl):503S-507S.

12. Mokdad AH, Serdula MK, Dietz WH, Bowman

BA, Marks JS, Koplan JP. The spread of the
obesity epidemic in the United States, 19911998. JAMA. 1999;282(16):1519-22.

24. Colditz GA. Economic costs of obesity and

inactivity. Med Sci Sports Exerc.
1999;31(11 Suppl):S663-7.

207

36. Jonnalagadda SS, Rosenbloom CA, Skinner R.

Dietary practices, attitudes, and
physiological status of collegiate freshman
football players. J Strength Cond Res.
2001;15(4):507-13.

25. Allison DB, Zannolli R, Narayan KM. The

direct health care costs of obesity in the
United States. Am J Public Health.
1999;89(8):1194-9.
26. Sturm R. The effects of obesity, smoking, and
drinking on medical problems and costs.
Obesity outranks both smoking and drinking
in its deleterious effects on health and health
costs. Health Aff (Millwood).
2002;21(2):245-53.

37. Baylis A, Cameron-Smith D, Burke LM.

Inadvertent doping through supplement use
by athletes: assessment and management of
the risk in Australia. Int J Sport Nutr Exerc
Metab. 2001;11(3):365-83.
38. Catlin DH, Hatton CK. Use and abuse of
anabolic and other drugs for athletic
enhancement. Adv Intern Med.
1991;36:399-424.

27. National Institutes of Health. NIH guidelines:

Clinical Guidelines on the Identification,
Evaluation, and Treatment of Overweight
and Obesity in Adults. The Evidence Report.
Obes Res. 1998;6(Suppl 2):51S-209S.

39. National College Athletics Association. NCAA

study of substance use habits of college
student-athletes. available at URL:
www.ncaa.org/library/research/htm#substan
ce_use_habits. Accessed 6/25/2002.

28. Williamson DF, Pamuk E, Thun M, Flanders D,

Byers T, Heath C. Prospective study of
intentional weight loss and mortality in
never- smoking overweight US white
women aged 40-64 years. Am J Epidemiol.
1995;141(12):1128-41.

40. Kanayama G, Gruber AJ, Pope HG Jr.,

Borowiecki JJ, Hudson JI. Over-the-counter
drug use in gymnasiums: an
underrecognized substance abuse problem?
Psychother Psychosom. 2001;70(3):137-40.

29. Galuska DA, Will JC, Serdula MK, Ford ES.

Are health care professionals advising obese
patients to lose weight? JAMA.
1999;282(16):1576-8.

41. Gruber AJ, Pope HG Jr. Ephedrine abuse

among 36 female weightlifters. Am J
Addict. 1998;7(4):256-61.

30. Blanck HM, Khan LK, Serdula MK. Use of

nonprescription weight loss products: results
from a multistate survey. JAMA.
2001;286(8):930-5.

42. Mahdihassan S. Ephedra, the oldest medicinal

plant with the history of an uninterrupted
use. In: Ancient Science of Life. 1987. pp.
105-109.

31. Heber D, Greenway FL. Herbal and alternative

approaches to obesity. In: Handbook of
Obesity. in press.

43. Lietava J. Medicinal plants in a middle

paleolithic grave Shanidar IV? J
Ethnopharmacol. 1992;35:263-66.

32. Harnack LJ, Rydell SA, Stang J. Prevalence of

use of herbal products by adults in the
Minneapolis/St Paul, Minn, metropolitan
area. Mayo Clin Proc. 2001;76(7):688-94.

44. Chen KK, Schmidt CF. Ephedrine and related

substances. Medicine. 1930;9:1-117.

33. Bowers LD. Abuse of performance-enhancing

drugs in sport. Ther Drug Monit.
2002;24(1):178-81.

45. Karch SB. Ma huang and the ephedra alkaloids.

In: Cupp MJ. (Ed) Toxicology and Clinical
Pharmacology of Herbal Products. Totowa,
NJ: Humana Press, Inc; 2000. pp. 11-30.

34. Catlin DH, Murray TH. Performance-enhancing

drugs, fair competition, and Olympic sport.
JAMA. 1996;276(3):231-7.

46. Hu SY. Ephedra (ma-huang) in the new

Chinese materia medica. Economic Botany.
1969;23:346-351.

35. Sobal J, Marquart LF. Vitamin/mineral

supplement use among athletes: a review of
the literature. Int J Sport Nutr.
1994;4(4):320-34.

47. Bensky D. Chinese Herbal Medicine: Materia

Medica. Seattle, WA: Eastland Press; 1993.

208

48. Nadkarni KM. Indian Materia Medica, Vol. I.

Bombay, India: Popular Prakashan; 1976.

58. McKenna DJ, Jones K, Hughes K. Botanical

Medicines. New York, NY: The Haworth
Herbal Press; in press.

49. Blumenthal M, Busse WR. Ephedra. In:

Blumenthal M, Busse WR. (Eds)
(Translated by) Klein S, Rister RS.
Complete German Commission E
Monographs: Therapeutic Guide to Herbal
Medicines. Boston, MA: Integrative
Medicine Communications; 1998. pp. 125-6,
414-6, 475-6.

59. Evans WC. Trease and Evans Pharmacognosy.

London: WB Saunders; 1989.
60. Sheu S-J. Identification by chemical analysis of
the botanical sources of commercial samples
of Chinese herbal drugs. Yao Wu Shih P'in
Fen Hsi (J Food Drug Anal). 1997;5(4):285294.

50. Robbers JE, Tyler VE. Tyler's Herbs of Choice.

New York: The Haworth Herbal Press;
1999.

61. Gurley BJ, Gardner SF, White LM, Wang P.

Ephedrine pharmacokinetics after the
ingestion of commercially available herbal
preparations of ephedra sinica (Ma-huang).
Pharmaceut Res. 1997;14:(11
Suppl)S519(Abstract).

51. Bruneton J. Phenethylamines. In: Bruneton J.

Pharmacognosy, Phytochemistry, Medicinal
Plants. Paris, France, Secausuc, NY:
Lavoisier Pub; 1996. pp. 711-15.

62. Chen KK, Wu C-K, Henriksen E. Relationship

between the pharmacological action and the
chemical constitution and configuration of
the optical isomers of ephedrine and related
compounds. J Pharmacol Exper Therapeut.
1929;36:363-400.

52. Leung AY, Foster S. Ephedra. In: Leung AY,

Foster S. Encyclopedia of Common Natural
Ingredients Used in Foods, Drugs, and
Cosmetics. New York, NY: John Wiley;
1996. p. 227.
53. Kamimura K, Iwamoto Y, Yamasaki K.,
Sakagami Y., et al. Variation of growth and
contents in ephedrine type alkaloids in
Ephedra distachya. Natural Medicines.
1994;48:122-125.

63. Chen KK, Schmidt CF. The action of

ephedrine, the active principle of the
Chinese drug Ma Hung. J Pharmacol Exper
Therapeut. 1924;24:339-357.
64. Chen KK, Meek WJ. Further studies on the
effect of ephedrine on the circulation. J Am
Pharmaceut Assoc. 1926;28:31-57.

54. Tanaka T, Obha K, Kawaahara K, Sakai E.

Comparison of the constituents of ephedra
herbs from various countries on ephedrine
type alkaloids. Natural Medicines.
1995;49:418-424.

65. Hardman JG, Limbird LE, Gilman A. (Eds).

Goodman and Gilmans The Pharmacologic
Basis of Disease. New York, NY: McGrawHill; 2001.

55. Kondo N, Mikage M, Idaka K. Medicobotanical studies of Ephedra plants from the
Himalayan region, part III: Causitive factors
of variation of alkaloid content in herbal
stems. Natural Medicines. 1999;53:194-200.

66. Burnham TH, Novak KK, Bell WL, Schweain

SL. Ephedrine. In: Drug Facts and
Comparisons. St Louis, MO: Facts and
Comparisons; 2002. p. 692.

56. Zhang JS, Tian Z, Lou ZC. [Quality evaluation

of twelve species of Chinese Ephedra (ma
huang)]. Yao Xue Xue Bao. 1989
24:(11)865-71. (Abstract).

67. Astrup A, Toubro S. Thermogenic, metabolic,

and cardiovascular responses to ephedrine
and caffeine in man. Int J Obes Relat Metab
Disord. 1993;17(Suppl 1):S41-3.

57. Zhu Y-P. Chinese Materia Medica. Amsterdam,

Netherlands: Hardwood Academic Press;
1998.

209

68. Martin WR, Sloan JW, Sapira JD, Jasinski DR.

Physiologic, subjective, and behavioral
effects of amphetamine, methamphetamine,
ephedrine, phenmetrazine, and
methylphenidate in man. Clin Pharmacol
Ther. 1971;12(2):245-58.

78. Dulloo AG, Seydoux J, Girardier L, Chantre P,

Vandermander J. Green tea and
thermogenesis: interactions between
catechin-polyphenols, caffeine and
sympathetic activity. Int J Obes Relat Metab
Disord. 2000;24(2):252-8.

69. Astrup A, Toubro S, Christensen NJ, Quaade F.

Pharmacology of thermogenic drugs. Am J
Clin Nutr. 1992;55(1 Suppl):246S-248S.

79. Han LK, Takaku T, Li J, Kimura Y, Okuda H.

Anti-obesity action of oolong tea. Int J Obes
Relat Metab Disord. 1999;23(1):98-105.

70. Bukowiecki L, Jahjah L, Follea N. Ephedrine, a

potential slimming drug, directly stimulates
thermogenesis in brown adipocytes via betaadrenoreceptors. Int J Obes. 1982;6(4):34350.

80. Dulloo AG, Seydoux J, Girardier L. Peripheral

mechanisms of thermogenesis induced by
ephedrine and caffeine in brown adipose
tissue. Int J Obes. 1991;15(5):317-26.
81. Greenway F. The safety and efficacy of
pharmaceutical and herbal caffeine and
ephedrine use as a weight loss agent.
Obesity Reviews. 2001;2:199-211.

71. Gurley BJ, Gardner SF, White LM, Wang PL.

Ephedrine pharmacokinetics after the
ingestion of nutritional supplements
containing Ephedra sinica (ma huang). Ther
Drug Monit. 1998;20(4):439-45.

82. Jadad AR, Moore RA, Carroll D, Jenkinson C,

Reynolds DJ, Gavaghan DJ, et al. Assessing
the quality of reports of randomized clinical
trials: is blinding necessary? Control Clin
Trials. 1996;17(1):1-12.

72. Pickup ME, May CS, Ssendagire R, Paterson

JW. The pharmacokinetics of ephedrine
after oral dosage in asthmatics receiving
acute and chronic treatment. Br J Clin
Pharmacol. 1976;3(1):123-34.

83. Moher D, Pham D, Jones A, Cook DJ, Jadad

AR, Moher M, et al. Does quality of reports
of randomized trials affect estimates of
interventions efficacy reported in metaanalyses? Lancet. 1998;352:609-613.

73. Kanfer I, Dowse R, Vuma V. Pharmacokinetics

of oral decongestants. Pharmacotherapy.
1993;13(6 Pt 2):116S-128S; discussion
143S-146S.

84. Moheb MA, Geissler CA, Lancer K. Effect of

ephedrine, caffeine, and aspirin, in
combinations of weight loss in obese
women. Int J Obesity Related Metabol
Disord. 1998;22:(Suppl 3)S264(Abstract).

74. Sever PS, Dring LG, Williams RT. The

metabolism of (-)-ephedrine in man. Eur J
Clin Pharmacol. 1975;9(2-3):193-8.
75. Haller, Jacob P 3rd, Benowitz NL.
Pharmacology of ephedra alkaloids and
caffeine after single-dose dietary supplement
use. Clin Pharmacol Therap. 2002;71:42132.

85. Pasquali R, Baraldi G, Cesari MP, Melchionda

N, Zamboni M, Stefanini C, et al. A
controlled trial using ephedrine in the
treatment of obesity. Int J Obes.
1985;9(2):93-8.

76. Gurley BJ, Gardner SF, White LM, Wang P.

Ephedrine pharmacokinetics after the
ingestion of nutritional supplements
containing Ephedra sinica (ma huang).
Therapeut Drug Monit. 1998;20(4):439-445.

86. Quaade F, Astrup A, Breum L, Toubro S, Hein

P. [The effect of an ephedrine/caffeine
combination as a supplement to a weightreducing diet. A randomized, placebocontrolled, double-blind trial]. Ugeskr
Laeger. 1992;154(18):1258-63.

77. White LM, Gardner SF, Gurley BJ, Marx MA,

Wang PL, Estes M. Pharmacokinetics and
cardiovascular effects of ma-huang (Ephedra
sinica) in normotensive adults. J Clin
Pharmacol. 1997;37(2):116-22.

210

87. Malchow-Moller A, Larsen S, Hey H,

Stokholm KH, Juhl E, Quaade F. Ephedrine
as an anorectic: the story of the 'Elsinore
pill.' Int J Obes. 1981;5(2):183-7.

96. Kalman DS, Colker CM, Shi Q, Swain MA.

Effects of a weight-loss aid in healthy
overweight adults: Double-blind, placebocontrolled clinical trial. Curr Therapeut Res.
2000;61(4):199-205.

88. Jensen KB, Dano P, Draeby N, Hansen SH,

Kanstrup J. [Elsinore tablets and ephedrine
as slimming agents]. Ugeskr Laeger.
1980;142(23):1499-501.

97. Hedges LV, Olkin I. Statistical methods for

meta-analysis. San Diego, CA: Academic
Press, Inc.; 1985.

89. Boozer CN, Daly PA, Homel P, Solomon JL,

Blanchard D, Nasser JA, et al. Herbal
ephedra/caffeine for weight loss: a 6-month
randomized safety and efficacy trial. Int J
Obes Relat Metab Disord. 2002;26(5):593604.

98. Rosenthal R. Meta-analysis: a review.

Psychosom Med. 1991;53(3):247-71.
99. DerSimonian R, Laird N. Meta-analysis in
clinical trials. Control Clin Trials.
1986;7(3):177-88.
100. Mood AM, Graybill FA, Boes DC. Introduction
to the Theory of Statistics. 3rd ed. London:
McGraw-Hill, Inc.; 1974.

90. Kettle R, Toubro S, Astrup A.

Ephedrine/Caffeine enhances abdominal fat
loss in females. Intl J Obesity Related
Metabolic Disorders. 1998;22(Suppl
3):S264.

101. Ray JW, Shadish WR. How interchangeable are

different estimators of effect size? Journal of
Consulting and Clinical Psychology.
1996;64:1316-1325.

91. Van Mil E, Molnar D. Drug treatment in obese

adolescents. Int J Obesity. 2000 24:(Suppl
1)S184(Abstract).

102. Assismos DG, Langenstroer P, Leinbach RF,

Mandel NS, Stern JM, Holmes RP.
Guaifenesin - and ephedrine-induced stones.
J Endourol. 1999;13(9):665-7.

92. Buemann B, Marckmann P, Christensen NJ,

Astrup A. The effect of ephedrine plus
caffeine on plasma lipids and lipoproteins
during a 4.2 MJ/day diet. Int J Obes Relat
Metab Disord. 1994;18(5):329-32.

103. Daly PA, Krieger DR, Dulloo AG, Young JB,

Landsberg L. Ephedrine, caffeine and
aspirin: safety and efficacy for treatment of
human obesity. Int J Obes Relat Metab
Disord. 1993;17(Suppl 1):S73-8.

93. Colker CM, Swain MA, Lynch L. A pilot study

evaluating the effects of an ephedrine and
forskolin-based product on body weight and
body composition in overweight, healthy
women. J Am Coll Nutr.
2001;20(5):a98(Abstract).

104. Berkey CS, Hoaglin DC, Mosteller F, Colditz

GA. A random-effects regression model for
meta-analysis. Stat Med. 1995;14(4):395411.

94. Lumholtz IB, Thorsteinsson B, Wamberg T,

Lehnschau A, Hansen G, Spellerberg S, et
al. [Ephedrine in the treatment of obesity. A
double-blind cross-over trial of the effect of
Elsinore tablets]. Ugeskr Laeger.
1980;142(23):1487-90.

105. Begg CB, Mazumdar M. Operating

characteristics of a rank correlation test for
publication bias. Biometrics.
1994;50(4):1088-101.
106. Egger M, Davey Smith G, Schneider M, Minder
C. Bias in meta-analysis detected by a
simple, graphical test. BMJ.
1997;315(7109):629-34.

95. Roed P, Hansen PW, Bidstrup B, Kaern M,

Helles A, Petersen KP. [Elsinore banting
tablets. A controlled clinical trial in general
practice]. Ugeskr Laeger.
1980;142(23):1491-5.

107. Stata Statistical Software: Release 7.0

[computer program]. College Station, TX:
Stata Corporation; 2001.

211

108. StatXact 4 for Windows [computer program].

Version 4.0.1. Cambridge, MA: Cytel
Software Corporation; 2000.
109. Landis JR, Koch GG. The measurement of
observer agreement for categorical data.
Biometrics. 1977;33(1):159-74.

116. Greenway F, deJonge L, Blanchard D, et al.

Evaluation of a dietary herbal supplement
containing caffeine and ephedrine on
metabolic rate, body composition, serum
lipids and tolerability. Pennington Center,
Louisiana State University. (Unpublished
Work).

110. Robinson TE, Becker JB. Enduring changes in

brain and behavior produced by chronic
amphetamine administration: a review and
evaluation of animal models of
amphetamine psychosis. Brain Res.
1986;396(2):157-98.

117. Fujioka K, Seaton TB, Rowe E, Jelinek CA,

Raskin P, Lebovitz HE, et al. Weight loss
with sibutramine improves glycaemic
control and other metabolic parameters in
obese patients with type 2 diabetes mellitus.
Diabetes Obes Metab. 2000;2(3):175-87.

111. Astrup A, Buemann B, Christensen NJ, Toubro

S, Thorbek G, Victor OJ, et al. The effect of
ephedrine/caffeine mixture on energy
expenditure and body composition in obese
women. Metabolism. 1992;41(7):686-8.

118. Gokcel A, Karakose H, Ertorer EM, Tanaci N,

Tutuncu NB, Guvener N. Effects of
sibutramine in obese female subjects with
type 2 diabetes and poor blood glucose
control. Diabetes Care. 2001;24(11):195760.

112. Molnar D, Torok K, Erhardt E, Jeges S. Safety

and efficacy of treatment with an
ephedrine/caffeine mixture. The first
double-blind placebo-controlled pilot study
in adolescents. Int J Obes Relat Metab
Disord. 2000;24(12):1573-8.

119. Smith IG, Goulder MA. Randomized placebocontrolled trial of long-term treatment with
sibutramine in mild to moderate obesity. J
Fam Pract. 2001;50(6):505-12.
120. Wirth A, Krause J. Long-term weight loss with
sibutramine: a randomized controlled trial.
JAMA. 2001;286(11):1331-9.

113. Breum L, Pedersen JK, Ahlstrom F, FrimodtMoller J. Comparison of an

ephedrine/caffeine combination and
dexfenfluramine in the treatment of obesity.
A double-blind multi-centre trial in general
practice. Int J Obes Relat Metab Disord.
1994;18(2):99-103.

121. Van Gaal LF, Broom JI, Enzi G, Toplak H.

Efficacy and tolerability of orlistat in the
treatment of obesity: a 6-month doseranging study. Orlistat Dose-Ranging Study
Group. Eur J Clin Pharmacol.
1998;54(2):125-32.

114. Donikyan LA. A double-blind randomized,

placebo-controlled, multicenter clinical
study to evaluate the safety and efficacy of a
natural herbal formulation when taken as
recommended for weight control. Technical
Report Version #1. Boca Raton, FL: Rexall
Sundown. 2002 (Unpublished Work).

122. Hill JO, Hauptman J, Anderson JW, Fujioka K,

O'Neil PM, Smith DK, et al. Orlistat, a
lipase inhibitor, for weight maintenance
after conventional dieting: a 1-y study. Am J
Clin Nutr. 1999;69(6):1108-16.

115. Boozer CN, Nasser JA, Heymsfield SB, Wang

V, Chen G, Solomon JL. An herbal
supplement containing Ma Huang-Guarana
for weight loss: a randomized, double-blind
trial. Int J Obes Relat Metab Disord.
2001;25(3):316-24.

123. Karhunen L, Franssila-Kallunki A, Rissanen P,

Valve R, Kolehmainen M, Rissanen A, et al.
Effect of orlistat treatment on body
composition and resting energy expenditure
during a two-year weight-reduction
programme in obese Finns. Int J Obes Relat
Metab Disord. 2000;24(12):1567-72.

212

124. Micic D, Ivkovic-Lazar T, Dragojevic R, Jorga

J, Stokic E, Hajdukovic Z. Orlistat, a
gastrointestinal lipase inhibitor, in therapy of
obesity with concomitant hyperlipidemia.
Med Pregl. 1999;52(9-10):323-33.

132. Pasternak H, Jacobs I, Bell D. Effects of

ingesting caffeine and ephedrine on
muscular endurance. Can J Appl Physiol.
1999;24(5):471.
133. Bell DG, Jacobs I, Ellerington K. Effect of
caffeine and ephedrine ingestion on
anaerobic exercise performance. Med Sci
Sports Exercise. 2001;33:1399-1403.

125. Muls E, Kolanowski J, Scheen A, Van Gaal L.

The effects of orlistat on weight and on
serum lipids in obese patients with
hypercholesterolemia: a randomized,
double-blind, placebo- controlled,
multicentre study. Int J Obes Relat Metab
Disord. 2001;25(11):1713-21.

134. Sidney KH, Lefcoe NM. The effects of

ephedrine on the physiological and
psychological responses to submaximal and
maximal exercise in man. Med Sci Sports.
1977;9(2):95-9.

126. Munro JF, MacCuish AC, Wilson EM, Duncan

LJ. Comparison of continuous and
intermittent anorectic therapy in obesity.
BMJ. 1968;1:352.

135. Horton TJ, Geissler CA. Aspirin potentiates the

effect of ephedrine on the thermogenic
response to a meal in obese but not lean
women. Int J Obes. 1991;15(5):359-66.

127. Oksbjerg N, Meyer T, Hvid-Jacobsen K.

[Aerobic training increases the effect of
ephedrine on energy expenditure]. Ugeskr
Laeger. 1986;148(32):2021-4.

136. Horton TJ, Geissler CA. Post-prandial

thermogenesis with ephedrine, caffeine and
aspirin in lean, pre-disposed obese and
obese women. Int J Obes Relat Metab
Disord. 1996;20(2):91-7.

128. Bell DG, Jacobs I, Zamecnik J. Effects of

caffeine, ephedrine and their combination on
time to exhaustion during high-intensity
exercise. Eur J Appl Physiol Occup Physiol.
1998;77(5):427-33.

137. Shugarman AE. Effect of thermogenic dietary

supplements on resting metabolic rate in
healthy male and female volunteers. Utah:
University of Utah. 1998.

129. Bell DG, Jacobs I, McLellan TM, Zamecnik J.

Reducing the dose of combined caffeine and
ephedrine preserves the ergogenic effect.
Aviat Space Environ Med. 2000;71(4):4159.

138. Toubro S, Astrup A. Randomised comparison

of diets for maintaining obese subjects'
weight after major weight loss: ad lib, low
fat, high carbohydrate diet v fixed energy
intake. BMJ. 1997;314(7073):29-34.

130. Bell DG, Jacobs I. Combined caffeine and

ephedrine ingestion improves run times of
Canadian Forces Warrior Test. Aviat Space
Environ Med. 1999;70(4):325-9.

139. Ofman JJ, MacLean C, Straus WL, et al. A

meta-analysis of dyspepsia and nonsteroidal
anti-inflammatory drugs. Arthritis Care and
Research. In press.

131. Bell DG, Jacobs I, McLellan TM, Miyazaki M,

Sabiston CM. Thermal regulation in the heat
during exercise after caffeine and ephedrine
ingestion. Aviat Space Environ Med.
1999;70(6):583-8.

213

Table 1. Herbs containing caffeine commonly combined with ephedra in products marketed for
weight loss or improved physical performance
Common Name
Cocoa
Coffee
Guarana
Kola nut
Mat leaf
Tea (black, green, oolong)

Botanical Name
Theobroma cacao
Coffea arabica
Paullinia cupana
Cola acuminata
Cola vera
Ilex paraguayensis
Camilla sinensis

11

Table 2. Technical expert panel members

Name
Awang, Dennis V. (PhD)
Benowitz, Neal (MD)
Farnsworth, Norman (PhD)
Fielding, Roger (PhD)
Goldberger, Jeffrey (MD)

Expertise
Natural product chemist
Psychiatry, pharmacology
Pharmacognosy
Exercise
Cardiology

Heber, David (MD, PhD)

Ko, Richard (PharmD, PhD)

Weight loss
Food and drug scientist

Leung, Albert (PhD)

Mills, Simon (FNIMN)

Pharmacognosy
Herbalist

Nestmann, Earle (PhD)

Toxicology

34

Institution
MediPlant Consulting Services
UCSF
University of Illinois at Chicago
Boston University
Northwestern Univ. Medical
School
UCLA School of Medicine
California Department of Health
Services
AYSL Inc.
Center for Complementary
Medicine, Exeter, UK
CANTOX Health Sciences,
Canada

Table 3. Technical expert panel suggestions about data collection

Collection Item
Outcomes of interest when
assessing efficacy

Subpopulations of interest
Risk factors of interest in
assessing possible harmful
effects

Suggestions
Weight = outcome for weight loss
Long-term weight loss = at least six months
Long-term exercise = at least 12 weeks
Change the term exercise enhancement to exercise capacity
VO2 max, metabolism, heart rate = intermediate outcomes for exercise
capacity
Power, strength, endurance = primary outcomes for exercise capacity
Age; gender; race; body composition/BMI; history of (Hx) hypertension;
Hx asthma; Hx diabetes
Existing structural heart disease
Renal function
Use of other drugs, tobacco

35

Table 4. Measures used in assessing causality

Measure
Temporal relationship
De-challenge response
Re-challenge response
Possibility of alternative explanation
Prior reaction to same substance
Dose response
Objective evidence of adverse event
Previous conclusive reports

Example
When the drug was consumed, dosage
Do symptoms disappear when substance is
removed?
Do symptoms appear again if substance is
reintroduced?
Dehydration or consumption of other toxic
substances

Witnesses or medical records

Has this same reaction happened when other
persons consumed substance?

Definition of substance

36

Table 5. Ephedra/ ephedrine search methodology

SEARCH NUMBER
Database searched and time
period covered
Search strategy

Number of items retrieved

SEARCH NUMBER
Database searched and time
period covered
Search strategy
Number of items retrieved
SEARCH NUMBER
Database searched and time
period covered
Search strategy
Number of items retrieved
SEARCH NUMBER
Database searched and time
period covered
Search strategy
Number of items retrieved
SEARCH NUMBER
Database searched and time
period covered
Search strategy

Number of Items Retrieved

SEARCH NUMBER
Database searched and time
period covered
Search strategy
Number of items retrieved

#1A
MEDLINE Via PubMed 1965-2001
Ephedra AND (clinical trial OR clinical trials OR randomized controlled
trials OR meta analysis OR meta-analysis OR review* OR Publication
Type=Meta-Analysis OR Publication Type=Clinical Trial OR
Publication Type=Review OR Publication Type=Randomized
Controlled Trial)
8
#1B
EMBASE 1974-2001
Ephedra AND (clinical trial* OR randomi* OR review* OR metaanalys*
OR meta analys* OR Document Type=Review)
20
#1C
BIOSIS 1969-2001
Ephedra AND (metaanal* OR meta anal* OR trial* OR review* in title
or subject heading field OR Document Type=Review OR Document
Type=Literature Review )
15
#1D
Allied & Complementary Medicine 1984-2001
MANTIS 1880-2000/Apr
Cochrane Library Controlled Clinical Trials Register Database
(CCTR)
ephedra
12
#2A (performed 4/5/01)
MEDLINE via PubMed 1965-2001
ephedrine NOT ephedra AND (review OR meta analysis OR
randomized controlled trials OR clinical trials OR Publication
Type=Review OR Publication Type=Clinical Trial OR Publication
Type=Randomized Controlled Trial OR Publication Type=MetaAnalysis)
704
#2B (performed 4/6/01)
EMBASE 1974-2001
ephedrine NOT ephedra AND (review* OR meta analys* OR
metaanalys* OR random* OR trial*)
1450

Note: *denotes truncated search term.

37

Table 6. Ephedra/ ephedrine search methodology additional databases

SEARCH NUMBER
Database searched and
time period covered

#1A (performed 6/25/01)

International Pharmaceutical Abstracts - 1970-2001/May
Pascal - 1973-2001/June Week 4
SciSearch (Archival File) - 1974-1989
SciSearch (Current File) - 1990-2001/June Week 4

Search strategy

ephedra OR ephedrine AND

trial? OR review? OR rct? OR meta analys? OR metaanal?

Number of items retrieved

167

SEARCH NUMBER
Database searched and
time period covered

#1B (performed 6/25/01)

International Pharmaceutical Abstracts - 1970-2001/May
Pascal - 1973-2001/June Week 4
SciSearch (Archival File) - 1974-1989
SciSearch (Current File) - 1990-2001/June Week 4
ephedra( IN TITLE OR SUBJECT HEADING FIELDS) OR ephedrine (IN
TITLE OR SUBJECT HEADING FIELDS)
AND
adverse OR side effect? OR efficacy OR fail? OR succeed? OR success?
OR effective? OR toxic?
330 (NOTE RESULTS FROM SEARCH 1A WERE NOTTED OUT OF
THESE SEARCH RESULTS)

Search strategy

Number of items retrieved

38

Table 7. Categories of adverse events

Event Type
Death
Stroke (CVA)
Myocardial infarction (heart attack)
Cardiovascular other than MI
Neurological other than stroke
Endocrine
Psychiatric
Pulmonary
Renal/urinary
Musculoskeletal
Gastrointestinal
Hepatic
Rheumatologic
Dermatological
Acid-base/electrolytic disturbances
Pain
Withdrawal symptoms
Gynecological/obstetrical
Hematological
Immunological/allergic reaction
Other rare events
Not described

39

Table 8. Report reviewers

Reviewer
Dr. David Allison
Dr. Arne Astrup

Affiliation
University of Alabama at Birmingham
The Research Department of Human Nutrition
The Royal Veterinary and Agricultural University, Denmark
Dr. Dennis Awang
Mediplant Consulting Services
Dr. Neal Benowitz
University San Francisco, Dept. of Med., SFGH, Clin. Pharm Div.
Dr. Heidi Blanck
Centers for Disease Control and Prevention, Division of Nutrition and Physical
Activity, Chronic Disease Nutrition Branch
Dr. George Bray
Pennington Biomedical Research Center
Hon Dan Burton
U.S. Representative
Mr. John Cardaro
Council for Responsible Nutrition
Ms. Beth Clay
U.S. House of Representatives, Hon Dan Burtons Office
Hon Dick Durbin
U.S. Senator
Dr. Norman Farnsworth Univ. of Illinois Med. Center
Dr. Roger Fielding
Boston University Dept. of Health Services
Dr. Gary Franklin
University of Washington
Dr. Curt Furberg
Wake Forest University
Dr. Frank Greenway
Pennington Biomedical Research Center
Prof. Bill Gurley
University of Arkansas School for Med. Sciences, College of Pharmacy
Dr. Christine Haller
University California San Francisco, Div of Clinical Pharmacology
Dr. Robert Hart
National Institute of Neurological Disorders and Stroke
Dr. David Heber
UCLA Center for Human Nutrition, Obesity and Nutrition
Dr. Steve Heymsfield St. Luke's/Roosevelt Hospital
Mr. Loren Israelsen
Utah Natural Products Alliance
Dr. Steven Karch
Assistant Medical Examiner, San Francisco
Dr. Steve Kimmell
Chair, Ephedra Education Council Expert Panel
Dr. Richard Ko
California Dept. of Health Services, Food and Drug Branch
Dr. Albert Leung
AYSL
Dr. Lori Love
Food and Drug Administration
Mr. Michael McGuffin President, American Herbal Products Association
Dr. Simon Mills
Center for Complementary Health Studies, University of Exeter
Dr. Earle Nestman
CANTOX
Dr. Paul Pentel
Hennepin County Medical Center, Div. of Toxicology, Dept. of Medicine
Mr. Paul Rubin
Patton Boggs
Mr. David Seckman
National Nutritional Foods
Mr. Wes Seigner
Hyman, Phelps, & McNamara
Hon Henry Waxman
U.S. Representative
Dr. Raymond Woosley University of Arizona Health Sciences Center
Ms. Susan Yanovski
Obesity and Eating Disorder Program
National Institute of Diabetes and Digestive and Kidney Diseases
Organizations
National Center for Complementary and Alternative Medicine
National Institute of Diabetes and Digestive and Kidney Diseases
National Heart, Lung and Blood Institute
Office of Dietary Supplements
Center for Science in the Public Interest
Public Citizen Health Research Group

40

Table 9. Weight loss trial inclusion results

Disposition of trials
Total retained in meta-analysis
Total dropped from meta-analysis
Reasons for dropping trials from meta-analysis:
Duration of treatment less than eight weeks
Ephedrine dose did not vary between study arms
Cross-over study without data available prior to the
cross-over point
Insufficient statistics
Inappropriate outcome (weight gain)

104

Number of Trials
20
24
18
1
1
3
1

Table 10. Ephedrine versus placebo

Trial
Jensen88
Lumholtz94
Moheb84
Pasquali85
Pasquali85
Quaade86
Pooled Random Effect Estimate

Total n
17
32
64
19
24
70

Chi-squared test of heterogeneity p-value = 0.185

105

Effect Size
-1.52
-1.03
-0.49
0.00
-0.42
-0.17
-0.501

95% CI
(-2.75, -0.29)
(-1.78, -0.29)
(-0.98, 0.01)
(-0.93, 0.93)
(-1.23, 0.39)
(-0.64, 0.30)
(-0.85, -0.15)

Table 11. Publication bias tests

Adjusted Rank
Correlation Test
p-value
0.45
0.30
N.C.
N.C.
0.73

Trials
Ephedrine vs. placebo
Ephedrine + caffeine vs. placebo
Ephedrine + caffeine vs. ephedrine alone
Ephedrine vs. another weight loss therapy
Ephedra + herbs containing caffeine vs. placebo
N.C. = not calculated due to the small number of trials available.

106

Regression
Asymmetry Test
p-value
0.82
0.12
N.C.
N.C.
0.23

Table 12. Ephedrine + caffeine versus placebo

Trial
Astrup111
Buemann92
Daly103
Jensen88
Kalman96
Kettle90
Malchow-Moll87
Moheb84
Molnar112
Quaade86
Roed95
Van Mil91
Pooled Random Effect Estimate

Total n
12
32
24
18
25
77
69
96
29
70
94
32

Chi-squared test of heterogeneity p-value = 0.073

107

Effect Size
-0.72
-0.55
-0.65
-1.84
-0.46
-0.40
-1.14
-0.76
-1.35
-0.50
-1.38
-1.00
-0.851

95% CI
(-1.88, 0.45)
(-1.26, 0.16)
(-1.47, 0.18)
(-3.10, -0.57)
(-1.25, 0.34)
(-0.85, 0.05)
(-1.65, -0.63)
(-1.20, -0.32)
(-2.16, -0.54)
(-0.98, -0.03)
(-1.83, -0.92)
(-1.74, -0.27)
(-1.08, -0.61)

Table 13. Ephedrine + caffeine versus ephedrine

Trial
Jensen88
Moheb84
Quaade86
Pooled Random Effect Estimate

Total n
27
96
70

Chi-squared test of heterogeneity p-value = 0.966

108

Effect Size
-0.32
-0.27
-0.36
-0.311

95% CI
(-1.07, 0.44)
(-0.70, 0.15)
(-0.83, 0.11)
(-0.60, -0.02)

Table 14. Ephedrine versus another active weight loss therapy

Trial
Breum113
Malchow-Moll87

Total n
81
70

109

Effect Size
-0.29
0.08

95% CI
(-0.73, 0.15)
(-0.36, 0.53)

Table 15. Ephedra versus placebo

Trial
Donikyan114

Total n
154

110

Effect Size
-0.69

95% CI
(-1.02, -0.37)

Table 16. Ephedra + herbs containing caffeine versus placebo

Trial
Boozer115
Boozer89
Colker93
Greenway116
Pooled Random Effect Estimate

Total n
48
83
26
30

Chi-squared test of heterogeneity p-value = 0.689

111

Effect Size
-1.07
-0.63
-0.87
-0.92
-0.811

95% CI
(-1.67, -0.46)
(-1.07, -0.18)
(-1.68, -0.06)
(-1.69, -0.15)
(-1.12, -0.51)

Table 17. Meta-regression results

Comparison Versus Placebo

Ephedrine
Ephedra + herbs containing caffeine
Ephedrine + caffeine

Pooled Monthly
Weight Loss Versus
Placebo (lbs)
-1.3
-2.1
-2.2

N.C. = Not calculated as this is the comparison group.

112

95% CI
(-2.1, -0.43)
(-2.8, -1.3)
(-2.8, -1.7)

p-value for Test

Versus Ephedra
+ Herbs
Containing
Caffeine
0.17
N.C.
0.75

Table 18. Exercise trials by Bell and colleagues

Reference

Compounds

Type of Exercise

Results

Bell, Jacobs &

Zamecnik128

Placebo
1 mg/kg Ephedrine
5 mg/kg Caffeine
1mg/kg Ephedrine + 5
mg/kg Caffeine (E+C)

Cycle ergometer trials

to exhaustion

Bell & Jacobs130

Placebo
75 mg Ephedrine + 375
mg Caffeine (E+C)

Bell, Jacobs,
McLellan,
Miyazakie, and
Sabiston131

Placebo
1 mg/kg Ephedrine + 5
mg/kg Caffeine (E+C)

Canadian Forces
Warrior Test - 3.2 km
run wearing 11 kg
equipment
Treadmill walking at
50% VO2 peak, 40
degrees celsius
climate, 30% relative
humidity

E+C significantly increased time

to exhaustion compared to
placebo. Heart rate during
exercise was significantly
increased for E+C, caffeine
arms.
E+C trial run times were
significantly faster than control
and placebo trials.

Bell, Jacobs,
McLellan &
Zamecnik129

Placebo
5 mg/kg Caffeine + 0.8
mg/kg Ephedrine
4 mg/kg Caffeine + 1
mg/kg Ephedrine
4 mg/kg Caffeine + 0.8
mg/kg Ephedrine
Placebo
0.8 mg/kg Ephedrine
4 mg/kg Caffeine
0.8mg/kg Ephedrine + 4
mg/kg Caffeine (E+C)

Cycle ergometer trials

to exhaustion at 85%
VO2 peak

Placebo
1 mg/kg Ephedrine
5 mg/kg Caffeine
1 mg/kg Ephedrine + 5
mg/kg Caffeine (E+C)

Two different cycle

ergometer tests, one
was to exhaustion at
125% VO2 peak

Pasternak, Jacobs
& Bell132

Bell, Jacobs &

Ellerington133

E+C did not significantly change

tolerance times when compared
to placebo. E+C did not affect
skin or rectal temperature,
sweat rate, or sensation of
thermal comfort.
A lower dose of E+C resulted in
ergogenic effect similar in
magnitude to those reported
previously with a higher dose,
with fewer side effects.

Three supersets of leg Ephedrine, E+C increased

press & bench press, muscular endurance, but only in
to exhaustion
the first set. Systolic blood
pressure was increase with
ephedrine, E+C.

113

Ephedrine improved
performance during Wingate
test of anaerobic power.
Caffeine increased time to
exhaustion in second test.

Table 19. Summary table of meta-analysis of adverse events reported controlled trials

Adverse Events
Psychiatric symptoms

# of
Trials

Placebo
Intervention Groups
# Adverse Sample # Adverse Sample
Events
Size
Events
Size

16

273

59

351

Autonomic hyperactivity

13

39

365

138

587

Palpitations

11

18

386

51

563

257

305

10

46

432

88

568

Headache

123

16

185

Tachycardia

45

90

Hypertension
Upper gastrointestinal
symptoms

N.R. = not reported.

114

Pooled OR
95% CI
3.64
(1.91, 7.31)
3.37
(2.19, 5.31)
2.29
(1.27, 4.32)
2.19
(0.49,13.34)
2.15
(1.39, 3.38)
1.64
(0.62, 4.68)
N.R.

Table 20. Summary table of other of adverse events reported in controlled trials

Other Adverse Events

Bundle branch block
Concentration difficulties
Constipation
Diarrhea
Dry mouth
Fatigue, weakness
Postural hypotension
Syncope
Ventricular events

# of
Trials
1
5
5
3
5
2
1
1
1

Placebo
Intervention Groups
# Adverse Sample # Adverse Sample
Events
Size
Events
Size
0
33
0
49
17
257
18
391
8
139
15
215
3
81
3
114
4
111
22
174
4
49
6
64
1
45
4
90
0
45
1
90
3
84
3
83

115

Table 21. Distribution of adverse events in the FDA file according to the Excel spreadsheet
Data Type
Death

Stroke

Available data

71 (5.3%)*

54 (4.0%)

Data dated after

Sept. 30, 2001

17 (12.4%)

Unavailable data
Total

Event
MI

Other

Total

33 (2.5%)

1,186 (88.2%)

1,344 (100%)

15 (10.9%)

5 (3.7%)

100 (73.0%)

137 (100%)

4 (1.9%)

18 (8.4%)

9 (4.2%)

183 (85.5%)

214 (100%)

92 (5.4%)

87 (5.1%)

47 (2.8%)

1,469 (86.7%)

1,695 (100%)

*Number of events (row percent).

Chi-squared test of independence p-value < 0.001.
Note: summary data were available for AERs beyond Sept 30, 2001. Detailed, redacted records were only available for AERs up
through Sept 30, 2001.

116

Table 22. Evidence table of case reports - Deaths

Event type
Report date
Age, Sex
Constituent
Product
Source
Dose*
(ID)
Timing; Duration
Death
Hydroxycut
11/03/1999
10.0 mg
21 yo Male
<6 hours; < 48 hours
Ephedra
Ripped Fuel
FDA Case
Unknown
(13914)
Not described; not described
Death
Slacker II
09/26/2000
Unknown
22 yo Female Not described; 14-60 days (acute)
Ephedra
FDA Case
(14390)
Death
MiniTabs
30 yo Female 250.0 mg
Ephedrine
Not described; >60 days (chronic)
FDA Case
(3275432)
Death
Max Brand Two-Way
33 yo Male
150.0 mg
Ephedrine
Not described; Not described
FDA Case
(3289590)
Death
Insufficient information
28 yo Male
Unknown
Ephedrine
<24 hours; >60 days (chronic)
Literature Case
(348)
Death
Nature's Nutrition-Formula One
05/19/1994
Unknown
36 yo Female Not described; >60 days (chronic)
Ephedra
FDA Case
(9508)
Death
Nature's Nutrition-Formula One
03/09/1995
Unknown
32 yo Male
Not described; >60 days (chronic)
Ephedra
FDA Case
(10276)

Investigation for Etiology

Autopsy conducted: Yes

RAND Classification
Sentinel event

Autopsy conducted: Yes

Sentinel event

Autopsy conducted: Yes

Sentinel event

Autopsy conducted: Yes

Sentinel event

Autopsy conducted: Yes

Sentinel event

Autopsy conducted: Yes

Possible sentinel
event

Autopsy conducted: Yes

Possible sentinel
event

* dose reported in total daily alkaloids

yo = year old

117

Table 22. Evidence table of case reports Deaths (continued)

Event type
Report date
Age, Sex
Constituent
Product
Source
Dose*
(ID)
Timing; Duration
Death
Ripped Fuel
07/25/1997
43.2 mg
38 yo Male
<6 hours; >60 days (chronic)
Ephedra
FDA Case
(12485)
Death
Thermogenics Plus
12/19/1997
23.1 mg
21 yo Male
Not described; Not described
Ephedra
FDA Case
(12722)
Death
Ripped Fuel
04/11/1998
40.0 mg
15 yo Female <6 hours; Not described
Ephedra
FDA Case
(12843)
Death
Ripped Fuel
08/03/1999
Unknown
26 yo Male
Not described; 2-13 days (acute)
Ephedra
FDA Case
(13906)
Death
Diet Fuel
02/16/2000
26.6 mg
26 yo Female Not described; >60 days (chronic)
Ephedra
FDA Case
(14019)
Death
Hydroxycut
01/09/2001
20.0 mg
35 yo Male
6-24 hours; 2-13 days
Ephedra
FDA Case
(14638)
Death
Ripped Fuel
23 yo Male
50.0 mg
Ephedra
Not described; >60 days (chronic)
Literature Case
(258)

Investigation for Etiology

Autopsy conducted: Yes

RAND Classification
Possible sentinel
event

Autopsy conducted: Yes

Possible sentinel
event

Autopsy conducted: Yes

Possible sentinel
event

Autopsy conducted: Yes

Possible sentinel
event

Autopsy conducted: Yes

Possible sentinel
event

Autopsy conducted: Yes

Possible sentinel
event

Autopsy conducted: Yes

Possible sentinel
event

* dose reported in total daily alkaloids

yo = year old

118

Table 22. Evidence table of case reports Deaths (continued)

Event type
Report date
Age, Sex
Constituent
Product
Source
Dose*
(ID)
Timing; Duration
Death
Street drug (speed)
42 yo Male
306.0 mg
Ephedrine
Not described; >60 days (chronic)
Literature Case
(44)
Death
Unknown
84 yo Female Unknown
Ephedrine
Not described; Not described
Literature Case
(44)
Death
Insufficient information
Literature Case Unknown
Ephedrine
<24 hours; 14-60 days (acute)
44 yo Male
(224)
Death
Ephedrine
31 yo Female Unknown
Ephedrine
<6 hours; <48 hours
FDA Case
(313104)
Death
Insufficient information
30 yo Female Unknown
Literature Case Not described; Not described
(17)
Death
Insufficient information
19 yo Female Unknown
Literature Case <6 hours; Not described
(96)
Death
Insufficient information
21 yo Male
Unknown
Literature Case <6 hours; Not described
(96)
Death
Ripped Fuel
06/11/1999
Unknown
24 yo Male
<6 hours; < 48 hours
Ephedra
FDA Case
(13672)
Death
Max Alert
40 yo Male
Not described
Ephedrine
Not described
FDA Case
(1859087)

Investigation for Etiology

Autopsy conducted: Yes

RAND Classification
Possible sentinel
event

Autopsy conducted: Yes

Possible sentinel
event

Autopsy conducted: Yes

Possible sentinel
event

Autopsy conducted: No

Intraoperative
ephedrine

Autopsy conducted: Yes

Suicide

Autopsy conducted: Yes

Suicide

Autopsy conducted: Yes

Suicide

Autopsy conducted: Yes

Probably not related

Autopsy conducted: No

Probably not related

* dose reported in total daily alkaloids

yo = year old

119

Table 22. Evidence table of case reports Deaths (continued)

Event type
Report date
Age, Sex
Constituent
Source
(ID)
Death
Not described
yo Male
Ephedrine
FDA Case
(1902493)
Death
30 yo Male
Ephedrine
FDA Case
(3491515)
Death
29 yo Male
Ephedrine
FDA Case
(3772362)
Death
03/10/1994
23 yo Male
Ephedra
FDA Case
(9188)
Death
06/09/1994
44 yo Male
Ephedra
FDA Case
(9327)
Death
06/14/1994
43 yo Female
Ephedra
FDA Case
(9395)
Death
06/20/1994
36 yo Female
Ephedra
FDA Case
(9473)

Product
Dose*
Timing; Duration
Unknown
Not described
Not described

Investigation for Etiology

Autopsy conducted: Yes

RAND Classification
Probably not related

Insufficient information
Unknown
Not described; Not described

Autopsy conducted: Yes

Probably not related

Insufficient information
Unknown
Not described; Not described

Autopsy conducted: Yes

Probably not related

Cybergenics Body Builder

Unknown
Not described; 14-60 days (acute)

Autopsy conducted: No

Insufficient information

Asian Herbal High Energy

Unknown
Not described; >60 days (chronic)

Autopsy conducted: No

Insufficient information

Nature's Nutrition-Formula One

Unknown
Not described; 14-60 days (acute)
Nature Nutritional Complex 1
Unknown
Not described; Not described
Nature's Nutrition-Formula One
Unknown
Not described; >60 days (chronic)

Autopsy conducted: No

Insufficient information

Autopsy conducted: No

Insufficient information

* dose reported in total daily alkaloids

yo = year old

120

Table 22. Evidence table of case reports Deaths (continued)

Event type
Report date
Age, Sex
Constituent
Source
(ID)
Death
05/24/1994
43 yo Female
Ephedra
FDA Case
(9506)
Death
09/07/1994
45 yo Male
Ephedra
FDA Case
(9864)
Death
04/07/1995
26 yo Male
Ephedra
FDA Case
(10104)
Death
01/12/1993
43 yo Male
Ephedra
FDA Case
(10251)
Death
06/14/1995
61 yo Female
Ephedra
FDA Case
(10296)
Death
12/19/1994
20 yo Male
Ephedra
FDA Case
(10448)
Death
03/15/1995
17 yo Female
Ephedra
FDA Case
(10849)

Product
Dose*
Timing; Duration
Nature's Nutrition-Formula One
Unknown
Not described; >60 days (chronic)

Investigation for Etiology

Autopsy conducted: No

RAND Classification
Insufficient information

Nature's Nutrition-Formula One

Unknown
Not described; 14-60 days (acute)

Autopsy conducted: Yes

Insufficient information

Natural Trim
Unknown
Not described; 14-60 days (acute)

Autopsy conducted: Yes

Insufficient information

Omnitrition Herbal Tea

39.0 mg
Not described; >60 days (chronic)

Autopsy conducted: Yes

Insufficient information

New Image Plus

Unknown
>24 hours; 14-60 days (acute)

Autopsy conducted: No

Insufficient information

Cybertrim
Unknown
Not described; Not described

Autopsy conducted: Yes

Insufficient information

Unknown Eola Product

Unknown
Not described; Not described

Autopsy conducted: No

Insufficient information

* dose reported in total daily alkaloids

yo = year old

121

Table 22. Evidence table of case reports Deaths (continued)

Event type
Report date
Age, Sex
Constituent
Source
(ID)
Death
03/14/1996
20 yo Male
Ephedra
FDA Case
(10862)
Death
04/08/1996
67 yo Male
Ephedra
FDA Case
(10902)
Death
04/12/1996
29 yo Female
Ephedra
FDA Case
(11018)
Death
02/16/1996
64 yo Female
Ephedra
FDA Case
(11060)
Death
05/20/1996
Not described
yo Male
Ephedra
FDA Case
(11134)
Death
05/13/1996
37 yo Male
Ephedra
FDA Case
(11248)

Product
Dose*
Timing; Duration
The Equillizer- Part B
Unknown
<6 hours; < 48 hours

Investigation for Etiology

Autopsy conducted: No

RAND Classification
Insufficient information

Quickshot
Unknown
Not described; 2-13 days

Autopsy conducted: No

Insufficient information

Omni-Trim (Omni-Trim Intl)

Unknown
Not described; 14-60 days (acute)

Autopsy conducted: No

Insufficient information

Nature's Nutrition-Formula One

Unknown
>24 hours; >60 days (chronic)

Autopsy conducted: No

Insufficient information

Ripped Fuel
60.0 mg
Not described; >60 days (chronic)

Autopsy conducted: Yes

Insufficient information

Nature's Nutrition-Formula One

42.4 mg
Not described; >60 days (chronic)
Equillizer Fast Start
Unknown
>24 hours; >60 days (chronic)
Not described
Unknown
Not described; Not described

Autopsy conducted: No

Insufficient information

* dose reported in total daily alkaloids

yo = year old

122

Table 22. Evidence table of case reports Deaths (continued)

Event type
Report date
Age, Sex
Constituent
Source
(ID)
Death
07/11/1996
59 yo Male
Ephedra
FDA Case
(11307)
Death
06/25/1996
34 yo Female
Ephedra
FDA Case
(11417)
Death
07/23/1996
27 yo Male
Ephedra
FDA Case
(11441)
Death
07/12/1996
24 yo Male
Ephedra
FDA Case
(11444)
Death
10/07/1996
56 yo Female
Ephedra
FDA Case
(11721)
Death
08/25/1997
32 yo Female
Ephedra
FDA Case
(12506)
Death
10/06/1997
0 yo Female
Ephedra
FDA Case
(12594)

Product
Dose*
Timing; Duration
Herbalife Original Green
26.4 mg
Not described; 14-60 days (acute)

Investigation for Etiology

Autopsy conducted: No

RAND Classification
Insufficient information

Herbalife Original Green

Unknown
Not described; Not described

Autopsy conducted: No

Insufficient information

Ripped Fuel
51.4 mg
Not described; >60 days (chronic)

Autopsy conducted: No

Insufficient information

Cybergenic super anti-fatigue

3.3 mg
Not described; 14-60 days (acute)

Autopsy conducted: No

Insufficient information

Easy Trim
Unknown
Not described; Not described

Autopsy conducted: No

Insufficient information

Escalation
Unknown
Not described; Not described

Autopsy conducted: Yes

Insufficient information

Ripped Fuel
20.0 mg
Not described; >60 days (chronic)

Autopsy conducted: No

Insufficient information

* dose reported in total daily alkaloids

yo = year old

123

Table 22. Evidence table of case reports Deaths (continued)

Event type
Report date
Age, Sex
Constituent
Source
(ID)
Death
12/19/1997
22 yo Male
Ephedra
FDA Case
(12720)
Death
04/23/1998
34 yo Male
Ephedra
FDA Case
(12859)
Death
04/24/1998
46 yo Male
Ephedra
FDA Case
(12871)
Death
07/11/1998
43 yo Male
Ephedra
FDA Case
(13021)
Death
09/16/1998
37 yo Female
Ephedra
FDA Case
(13096)
Death
10/08/1998
49 yo Female
Ephedra
FDA Case
(13127)
Death
02/27/1999
18 yo Male
Ephedra
FDA Case
(13380)

Product
Dose*
Timing; Duration
Ripped Fuel
Unknown
>24 hours; Not described

Investigation for Etiology

Autopsy conducted: Yes

RAND Classification
Did not meet temporal
relationship criterion

Herbalife Original Green

42.0 mg
Not described; 14-60 days (acute)

Autopsy conducted: No

Insufficient information

Diet Fuel
20.1 mg
Not described; Not described

Autopsy conducted: No

Insufficient information

Ripped Fuel
63.6 mg
Not described; >60 days (chronic)

Autopsy conducted: No

Insufficient information

Metabolife 356
Unknown
Not described; 2-13 days

Autopsy conducted: Yes

Insufficient information

Thin Tabs
Unknown
Not described; Not described

Autopsy conducted: Yes

Insufficient information

Ultimate Orange
Unknown
<6 hours; Not described

Autopsy conducted: No

Insufficient information

* dose reported in total daily alkaloids

yo = year old

124

Table 22. Evidence table of case reports Deaths (continued)

Event type
Report date
Age, Sex
Constituent
Source
(ID)
Death
05/19/1999
49 yo Male
Ephedra
FDA Case
(13634)
Death
06/04/1999
40 yo Female
Ephedra
FDA Case
(13706)
Death
06/30/1999
37 yo Female
Ephedra
FDA Case
(13762)
Death
08/06/1999
59 yo Female
Ephedra
FDA Case
(13802)
Death
08/03/1999
37 yo Male
Ephedra
FDA Case
(13806)
Death
10/06/1999
42 yo Female
Ephedra
FDA Case
(13901)
Death
10/13/1999
62 yo Female
Ephedra
FDA Case
(13993)

Product
Dose*
Timing; Duration
Metabolife 356
60.0 mg
Not described; < 48 hours

Investigation for Etiology

Autopsy conducted: No

RAND Classification
Insufficient information

Metabolife 356
72.0 mg
>24 hours; 14-60 days (acute)

Autopsy conducted: Yes

Did not meet temporal

relationship criterion

Thermadrene
Unknown
Not described; Not described

Autopsy conducted: No

Insufficient information

Metabolife 356
Unknown
Not described; Not described

Autopsy conducted: No

Insufficient information

Metabolife 356
Unknown
<6 hours; 2-13 days

Autopsy conducted: Yes

Insufficient information

Herbalife Original Green

Unknown
Not described; 14-60 days (acute)

Autopsy conducted: Yes

Insufficient information

Metabolife 356
Unknown
Not described; Not described

Autopsy conducted: No

Insufficient information

* dose reported in total daily alkaloids

yo = year old

125

Table 22. Evidence table of case reports Deaths (continued)

Event type
Report date
Age, Sex
Constituent
Source
(ID)
Death
04/04/2000
29 yo Female
Ephedra
FDA Case
(14113)

Death
08/10/2000
32 yo Female
Ephedra
FDA Case
(14323)
Death
08/31/2000
46 yo Female
Ephedra
FDA Case
(14347)
Death
09/14/2000
40 yo Female
Ephedra
FDA Case
(14370)
Death
03/28/2000
56 yo Male
Ephedra
FDA Case
(14465)
Death
10/16/2000
46 yo Female
Ephedra
FDA Case
(14470)

Product
Dose*
Timing; Duration
Metabolife 356
Unknown
>24 hours; >60 days (chronic)
Omnitrition Herbal Tea
Unknown
Not described; 14-60 days (acute)
Not described
Unknown
Not described; Not described
Metabolift
60.0 mg
Not described; 2-13 days

Investigation for Etiology

Autopsy conducted: Yes

RAND Classification
Did not meet temporal
relationship criterion

Autopsy conducted: Yes

Insufficient information

Metabomax
72.0 mg
Not described; 2-13 days

Autopsy conducted: No

Insufficient information

Metabolife 356
Unknown
Not described; 14-60 days (acute)

Autopsy conducted: No

Insufficient information

Thermogen Plus Liquid

72.0 mg
Not described; 2-13 days

Autopsy conducted: Yes

Insufficient information

Up Your Gas
34.2 mg
Not described; Not described

Autopsy conducted: Not

described

Insufficient information

* dose reported in total daily alkaloids

yo = year old

126

Table 22. Evidence table of case reports Deaths (continued)

Event type
Report date
Age, Sex
Constituent
Source
(ID)
Death
11/14/2000
39 yo Male
Ephedra
FDA Case
(14498)
Death
11/18/2000
45 yo Female
Ephedra
FDA Case
(14509)
Death
12/06/2000
49 yo Female
Ephedra
FDA Case
(14561)
Death
12/24/2000
40 yo Male
Ephedra
FDA Case
(14585)
Death
03/18/2001
28 yo Female
Ephedra
FDA Case
(14747)
Death
03/29/2001
31 yo Female
Ephedra
FDA Case
(14808)
Death
3 yo Male
Ephedrine
FDA Case
(1772115)

Product
Dose*
Timing; Duration
Xenadrine
40.0 mg
Not described; >60 days (chronic)

Investigation for Etiology

Autopsy conducted: Yes

RAND Classification
Insufficient information

Metabolife 356
24.0 mg
>24 hours; 2-13 days

Autopsy conducted: No

Insufficient information

Diet 2X
Unknown
Not described; 14-60 days (acute)

Autopsy conducted: No

Insufficient information

Metabolife 356
72.0 mg
Not described; >60 days (chronic)

Autopsy conducted: Yes

Insufficient information

Mini Thin
75.0 mg
Not described; >60 days (chronic)
Yellow Jacket
Unknown
Not described; >60 days (chronic)
Metabolife 356
Unknown
Not described; >60 days (chronic)

Autopsy conducted: Yes

Insufficient information

Autopsy conducted: Yes

Insufficient information

Ephedrine
Unknown
<6 hours; < 48 hours

Autopsy conducted: No

Insufficient information

* dose reported in total daily alkaloids

yo = year old

127

Table 22. Evidence table of case reports Deaths (continued)

Event type
Report date
Age, Sex
Constituent
Product
Source
Dose*
(ID)
Timing; Duration
Death
Unknown
99 yo Male
Not described
Ephedrine
Not described
FDA Case
(1874879)
Death
MiniTabs
30 yo Female Unknown
Ephedrine
Not described; Not described
FDA Case
(3135225)
Death
Mini 2 Way Action
46 yo Male
Unknown
Ephedrine
Not described; >60 days (chronic)
FDA Case
(3173538)
Death
Metabolift
32 yo Female Unknown
Ephedrine
<6 hours; Not described
FDA Case
(3551127)
Death
Diet 2X
99 yo Female Unknown
Ephedrine
Not described; 14-60 days (acute)
FDA Case
(3623625)
Death
Unknown
44 yo Female Not described
Ephedrine
Not described; >60 days (chronic)
FDA Case
(3768335)
Death
Ultimate Xphoria
20 yo Male
Unknown
Literature Case <6 hours; < 48 Hours
(462)

Investigation for Etiology

Autopsy conducted: Yes

RAND Classification
Insufficient information

Autopsy conducted: Yes

Insufficient information

Autopsy conducted: Yes

Insufficient information

Autopsy conducted: N/A

Insufficient information

Autopsy conducted: No

Insufficient information

Autopsy conducted: No

Insufficient information

Autopsy conducted: Not

described

Insufficient information

* dose reported in total daily alkaloids

yo = year old

128

Table 22. Evidence table of case reports MI

Event type
Report date
Age, Sex
Constituent
Source
(ID)
MI
03/20/1995
45 yo Male
Ephedra
FDA Case
(10024)
MI
23 yo Female
Ephedrine
FDA Case
(3446357)
MI
30 yo Male
Ephedra
Literature Case
(244)
MI
19 yo Male
Ephedra
Literature Case
(516)
MI
35 yo Female
Ephedrine
Literature Case
(224)
MI
04/22/1994
37 yo Male
Ephedra
FDA Case
(9372)
MI
05/23/1994
54 yo Male
Ephedra
FDA Case
(9504)
MI
03/01/1995
35 yo Male
Ephedra
FDA Case
(10009)

Product
Dose*
Timing; Duration
Nature's Nutrition-Formula One
Unknown
<6 hours; 2-13 days

Investigation of Etiology RAND Classification

Angiography: Yes
Sentinel event

Midnight Ecstacy
Unknown
<6 hours; < 48 hours

Angiography: Yes

Sentinel event

Ma huang
Unknown
<24 hours; Not described

Angiography: Yes

Sentinel event

Dymetradine Xtreme
48.0 .
<6 hours; Not described

Angiography: Yes

Sentinel event

Product unknown
Unknown
<24 hours; 14-60 days (acute)

Angiography: Yes

Sentinel event

Eola Amp II Pro Drops

Unknown
<6 hours; 2-13 days

Angiography: Yes

Nature's Nutrition-Formula One

Unknown
6-24 hours; >60 days (chronic)

Angiography: Yes

Possible sentinel event

Note that this product
was removed from the
market- it contained
illegal doses of
ephedrine.
Possible sentinel event

Metabolift
50.0 mg
<6 hours; 14-60 days (acute)

Angiography: Yes

Possible sentinel event

* dose reported in total daily alkaloids

yo = year old

129

Table 22. Evidence table of case reports MI (continued)

Event type
Report date
Age, Sex
Constituent
Source
(ID)
MI
06/15/1998
38 yo Female
Ephedra
FDA Case
(13009)
MI
04/18/2000
37 yo Female
Ephedra
FDA Case
(14114)
MI
11/08/2000
43 yo Female
Ephedra
FDA Case
(14530)
MI
25 yo Male
Literature Case
(64)
MI
04/22/1994
34 yo Female
Ephedra
FDA Case
(9373)
MI
06/17/1994
Not described
yo Male
Ephedra
FDA Case
(9381)
MI
05/24/1994
56 yo Female
Ephedra
FDA Case
(9512)

Product
Dose*
Timing; Duration
Herbalife Original Green
25.6 mg
<6 hours; < 48 hours

Investigation of Etiology RAND Classification

Angiography: Yes
Possible sentinel event

Metabolife 356
Unknown
<6 hours; 14-60 days (acute)

Angiography: Yes

Possible sentinel event

Metab-O-Lite
72.0 mg
<6 hours; >60 days (chronic)

Angiography: Yes

Possible sentinel event

Ephedrine
Unknown
<6 hours; Not described

Angiography: No

Intravenous injection
of ephedrine

Eola Amp II Pro Drops

Unknown
Not described; 2-13 days

Angiography: Yes

Insufficient information

The Edge
Unknown
Not described; Not described

Angiography: No

Insufficient information

Nature's Nutrition-Formula One

Unknown
>24 hours; >60 days (chronic)

Angiography: No

Insufficient information

* dose reported in total daily alkaloids

yo = year old

130

Table 22. Evidence table of case reports MI (continued)

Event type
Report date
Age, Sex
Constituent
Source
(ID)
MI
08/26/1994
49 yo Female
Ephedra
FDA Case
(9572)
MI
03/16/1995
67 yo Female
Ephedra
FDA Case
(10065)
MI
07/03/1997
59 yo Female
Ephedra
FDA Case
(12452)
MI
04/21/1999
39 yo Female
Ephedra
FDA Case
(13532)
MI
08/05/1999
51 yo Male
Ephedra
FDA Case
(13815)
MI
04/06/2000
30 yo Female
Ephedra
FDA Case
(14123)
MI
04/15/2000
53 yo Male
Ephedra
FDA Case
(14222)

Product
Dose*
Timing; Duration
Nature's Nutrition-Formula One
Unknown
Not described; 14-60 days (acute)

Investigation of Etiology RAND Classification

Angiography: Yes
Insufficient information

Nature's Nutrition-Formula One

Unknown
Not described; 2-13 days

Angiography: Yes

Insufficient information

Omnitrition Herbal Tea

60.0 mg
Not described; 14-60 days (acute)

Angiography: Yes

Insufficient information

Metabolife 356
24.0 mg
Not described; >60 days (chronic)

Angiography: Yes

Insufficient information

Metabolife 356
Unknown
Not described; Not described

Angiography: No

Insufficient information

Metabolife 356
Unknown
Not described; Not described

Angiography: No

Insufficient information

Natural Herbal Energizer

Unknown
Not described; Not described

Angiography: Yes

Insufficient information

* dose reported in total daily alkaloids

yo = year old

131

Table 22. Evidence table of case reports MI (continued)

Event type
Report date
Age, Sex
Constituent
Source
(ID)
MI
07/05/2000
39 yo Male
Ephedra
FDA Case
(14259)
MI
11/15/2000
45 yo Male
Ephedra
FDA Case
(14521)
MI
12/02/2000
Not described
yo Not
described
Ephedra
FDA Case
(14555)
MI
01/07/2001
50 yo Female
Ephedra
FDA Case
(14645)

Product
Dose*
Timing; Duration
Diet Fuel
60.0 mg
Not described; 14-60 days (acute)

Investigation of Etiology RAND Classification

Angiography: Yes
Insufficient information

Xenadrine
Unknown
Not described; >60 days (chronic)
Thermocut
Unknown
Not described; >60 days (chronic)
Metabolife 356
Unknown
Not described; 2-13 days

Angiography: Yes

Insufficient information

Angiography: No

Insufficient information

Metabolife 356
Unknown
Not described; 14-60 days (acute)

Angiography: Yes

Insufficient information

* dose reported in total daily alkaloids

yo = year old

132

Table 22. Evidence table of case reports Cerebrovascular Accident/ Stroke

Event type
Report date
Age, Sex
Constituent
Source
(ID)
CVA
01/03/1996
26 yo Female
Ephedra
FDA Case
(10874)
CVA
04/12/1996
42 yo Female
Ephedra
FDA Case
(11062)
CVA
04/17/1996
31 yo Female
Ephedra
FDA Case
(11105)
CVA
09/04/1996
28 yo Male
Ephedra
FDA Case
(11675)
CVA
06/16/1998
39 yo Male
Ephedra
FDA Case
(12980)
CVA
12/31/1998
29 yo Male
Ephedra
FDA Case
(13418)
CVA
09/12/2000
53 yo Female
Ephedra
FDA Case
(14372)

Product
Dose*
Timing; Duration
Thermo Slim
Unknown
<6 hours; 2-13 days

Investigation of Etiology RAND Classification

Implicit review
Sentinel event

Power Trim
Unknown
Not described; >60 days (chronic)

Implicit review

Sentinel event

Trim Easy
72.0 mg
6-24 hours; >60 days (chronic)

Implicit review

Sentinel event

Ripped Fuel
63.6 mg
Not described; >60 days (chronic)

Implicit review

Sentinel event

Ultimate Orange
Unknown
<6 hours; Not described

Implicit review

Sentinel event

Ultimate Orange
62.1 mg
<6 hours; 14-60 days (acute)

Implicit review

Sentinel event

Slim Caps
24.0 mg
6-24 hours; 14-60 days (acute)

Implicit review

Sentinel event

* dose reported in total daily alkaloids

yo = year old

133

Table 22. Evidence table of case reports Cerebrovascular Accident/ Stroke (continued)
Event type
Report date
Age, Sex
Constituent
Source
(ID)
CVA
10/20/2000
46 yo Female
Ephedra
FDA Case
(14473)
CVA
33 yo Male
Ephedra
Literature Case
(552)
CVA
19 yo Female
Ephedrine
Literature Case
(184 )
CVA
20 yo Female
Ephedrine
Literature Case
(514)
CVA
04/17/1992
30 yo Female
Ephedra
FDA Case
(9296)
CVA
04/22/1994
56 yo Female
Ephedra
FDA Case
(9335)
CVA
03/15/1995
24 yo Female
Ephedra
FDA Case
(10094)

Product
Dose*
Timing; Duration
Xenadrine
Unknown
<6 hours; 2-13 days

Investigation of Etiology RAND Classification

Implicit review
Sentinel event

Thermadrene
Unknown
6-24 hours; Not described

Implicit review

Sentinel event

Ephedrine
Unknown
<6 hours; Not described

Implicit review

Sentinel event

Purported amphetamine look-alike

Unknown
<6 hours; < 48 Hours

Implicit review

Sentinel event

E'ola Amp II Pro Drops

75.0 mg
<6 hours; 2-13 days

Implicit review

E'ola Amp II Pro Drops

Unknown
6-24 hours; >60 days (chronic)

Implicit review

Super Fat Burners

Unknown
6-24 hours; <48 hours

Implicit review

Possible sentinel event

Note that this product
was removed from the
market- it contained
illegal doses of
ephedrine.
Possible sentinel event
Note that this product
was removed from the
market- it contained
illegal doses of
ephedrine.
Possible sentinel event

* dose reported in total daily alkaloids

yo = year old

134

Table 22. Evidence table of case reports Cerebrovascular Accident/ Stroke (continued)
Event type
Report date
Age, Sex
Constituent
Source
(ID)
CVA
01/09/1998
64 yo Female
Ephedra
FDA Case
(12713)
CVA
12/23/1997
47 yo Male
Ephedra
FDA Case
(12733)
CVA
04/27/1998
41 yo Female
Ephedra
FDA Case
(12888)
CVA
09/13/2000
25 yo Female
Ephedra
FDA Case
(14378)
CVA
10/12/2000
42 yo Male
Ephedra
FDA Case
(14434)
CVA/
Subarachnoid
hemorrhage
11/16/2000
55 yo Female
Ephedra
FDA Case
(14553)
CVA
33 yo Male
Ephedra
Literature Case
(270)

Product
Dose*
Timing; Duration
Investigation of Etiology RAND Classification
FitAmerica Natural Weight ControlAid Implicit review
Possible sentinel event
100.0 mg
<6 hours; >60 days (chronic)

Purple Blast
Unknown
<6 hours; 14-60 days (acute)

Implicit review

Possible sentinel event

Diet Phen
13.5 mg
6-24 hours; 14-60 days (acute)

Implicit review

Possible sentinel event

Natural Trim
44.0 mg
6-24 hours; 14-60 days (acute)

Implicit review

Possible sentinel event

Slim 'N Up
Unknown
6-24 hours; >60 days (chronic)

Implicit review

Possible sentinel event

Metabolife 356
Unknown
6-24 hours; 14-60 days (acute)

Implicit review

Possible sentinel event

Ma huang
40 mg
6-24 hours; 14-60 days

Implicit review

Possible sentinel event

* dose reported in total daily alkaloids

yo = year old

135

Table 22. Evidence table of case reports Cerebrovascular Accident/ Stroke (continued)
Event type
Report date
Age, Sex
Constituent
Source
(ID)
CVA
37 yo Male
Ephedrine
Literature Case
(44)
CVA
20 yo Male
Ephedrine
Literature Case
(438)
CVA
29 yo Female
Ephedrine
FDA Case
(3720184)
CVA
45 yo Female
Literature Case
(485)
CVA
05/12/1994
36 yo Female
Ephedra
FDA Case
(9521)
CVA
06/22/1994
52 yo Male
Ephedra
FDA Case
(9545)
CVA
10/26/1994
40 yo Female
Ephedra
FDA Case
(9749)
CVA
09/14/1994
49 yo Male
Ephedra
FDA Case
(9865)

Product
Dose*
Timing; Duration
Street drug
153.0 mg
Not described; 14-60 days (acute)

Investigation of Etiology RAND Classification

Implicit review
Possible sentinel event

Speed
Unknown
<6 hours; Not described

Implicit review

Possible sentinel event

Ephedrine
Unknown
<6 hours; < 48 hours

Not relevant

Intraoperative
ephedrine

Ephedrine
Unknown
<6 hours; < 48 hours

Not relevant

Intraoperative
ephedrine

Nature's Nutrition-Formula One

Unknown
<6 hours; >60 days (chronic)

Implicit review

Insufficient information

Nature's Nutrition-Formula One

Unknown
Not described; >60 days (chronic)

Not reviewed

Insufficient information

Equillizer Fast Start

Unknown
>24 hours; 2-13 days

Not reviewed

Did not meet the

temporal relationship
criterion

Nature's Nutrition-Formula One

Unknown
Not described; 14-60 days (acute)

Not reviewed

Insufficient information

* dose reported in total daily alkaloids

yo = year old

136

Table 22. Evidence table of case reports Cerebrovascular Accident/ Stroke (continued)
Event type
Report date
Age, Sex
Constituent
Source
(ID)
CVA
05/12/1995
53 yo Female
Ephedra
FDA Case
(10187)
CVA
10/19/1995
31 yo Female
Ephedra
FDA Case
(10477)
CVA
10/12/1995
19 yo Female
Ephedra
FDA Case
(10508)
CVA
02/07/1996
30 yo Female
Ephedra
FDA Case
(10893)
CVA
04/13/1996
34 yo Female
Ephedra
FDA Case
(10957)
CVA
07/11/1996
55 yo Female
Ephedra
FDA Case
(11306)
CVA
07/18/1996
39 yo Female
Ephedra
FDA Case
(11442)

Product
Dose*
Timing; Duration
Nature's Nutrition-Formula One
Unknown
Not described; >60 days (chronic)

Investigation of Etiology RAND Classification

Not reviewed
Insufficient information

TriChromolean
Unknown
Not described; 14-60 days (acute)

Not reviewed

Insufficient information

Thermoburn
Unknown
Not described; >60 days (chronic)

Not reviewed

Insufficient information

Metabolift
60.0 mg
Not described; 14-60 days (acute)

Not reviewed

Insufficient information

E'ola Amp II Pro Drops

Unknown
Not described; >60 days (chronic)

Not reviewed

Insufficient information

Natural Trim
Unknown
Not described; >60 days (chronic)

Not reviewed

Insufficient information

Herbalife Original Green

Unknown
Not described; >60 days (chronic)

Not reviewed

Insufficient information

* dose reported in total daily alkaloids

yo = year old

137

Table 22. Evidence table of case reports Cerebrovascular Accident/ Stroke (continued)
Event type
Report date
Age, Sex
Constituent
Source
(ID)
CVA
06/18/1996
35 yo Female
Ephedra
FDA Case
(11619)
CVA
08/21/1996
33 yo Female
Ephedra
FDA Case
(11706)
CVA
10/21/1996
69 yo Female
Ephedra
FDA Case
(12340)
CVA
06/05/1997
64 yo Female
Ephedra
FDA Case
(12460)
CVA
08/01/1997
34 yo Female
Ephedra
FDA Case
(12483)
CVA
04/22/1998
43 yo Female
Ephedra
FDA Case
(12861)
CVA
02/11/1999
47 yo Female
Ephedra
FDA Case
(13336)

Product
Dose*
Timing; Duration
E'ola Amp II Pro Drops
21.5 mg
>24 hours; Not described

Investigation of Etiology RAND Classification

Not reviewed
Did not meet the
temporal relationship
criterion

Herbalife Original Green

Unknown
Not described; >60 days (chronic)
AP300
Unknown
Not described; >60 days (chronic)
E'ola Amp II Pro Drops
36.8 mg
Not described; 14-60 days (acute)

Not reviewed

Insufficient information

Not reviewed

Insufficient information

Shape Fast
30.0 mg
Not described; Not described

Not reviewed

Insufficient information

Shape Fast
36.0 mg
Not described; 2-13 days

Not reviewed

Insufficient information

Metabolife 356
Unknown
Not described; >60 days (chronic)

Not reviewed

Insufficient information

Total Control
66.0 mg
>24 hours; >60 days (chronic)

Not reviewed

Did not meet the

temporal relationship
criterion

* dose reported in total daily alkaloids

yo = year old

138

Table 22. Evidence table of case reports Cerebrovascular Accident/ Stroke (continued)
Event type
Report date
Age, Sex
Constituent
Source
(ID)
CVA
02/01/1999
48 yo Female
Ephedra
FDA Case
(13341)
CVA
06/01/1999
16 yo Male
Ephedra
FDA Case
(13661)
CVA
06/23/1999
18 yo Female
Ephedra
FDA Case
(13779)
CVA
08/03/1999
24 yo Female
Ephedra
FDA Case
(13797)
CVA
08/04/1999
30 yo Female
Ephedra
FDA Case
(13829)
CVA
07/01/1999
26 yo Female
Ephedra
FDA Case
(13837)
CVA
10/27/1999
36 yo Female
Ephedra
FDA Case
(13905)

Product
Dose*
Timing; Duration
Metacut
12.3 mg
Not described; 2-13 days

Investigation of Etiology RAND Classification

Not reviewed
Insufficient information

Hydroxycut (Muscle Tech R&D)

160.0 mg
>24 hours; 14-60 days
(acute)

Not reviewed

Did not meet the

temporal relationship
criterion

Metabolife 356
Unknown
Not described; 14-60 days (acute)

Not reviewed

Insufficient information

Metabolife 356
Unknown
Not described; >60 days (chronic)

Not reviewed

Insufficient information

Metabolife 356
48.0 mg
<6 hours; 2-13 days

Implicit review

Insufficient information

Metabolife 356
Unknown
Not described; 14-60 days (acute)

Not reviewed

Insufficient information

Metabolife 356
Unknown
<6 hours; >60 days(chronic)

Implicit review

Insufficient information

* dose reported in total daily alkaloids

yo = year old

139

Table 22. Evidence table of case reports Cerebrovascular Accident/ Stroke (continued)
Event type
Report date
Age, Sex
Constituent
Source
(ID)
CVA
10/08/1998
Not described
yo Female
Ephedra
FDA Case
(14056)
CVA
06/13/2000
46 yo Female
Ephedra
FDA Case
(14231)
CVA
10/03/2000
21 yo Female
Ephedra
FDA Case
(14431)
CVA
01/16/2001
48 yo Female
Ephedra
FDA Case
(14632)
CVA
32 yo Female
Ephedrine
FDA Case
(1823550)
CVA
68 yo Male
Literature Case
(515 )

Product
Dose*
Timing; Duration
Ripped Fuel
Unknown
Not described; Not described

Investigation of Etiology RAND Classification

Not reviewed
Insufficient information

Metabolife 356
Unknown
Not described; 14-60 days (acute)
Xenadrine
Unknown
Not described; 14-60 days (acute)
Slacker II
Unknown
Not described; 2-13 days

Not reviewed

Insufficient information

Not reviewed

Insufficient information

LiquiFit Exercise Drops

75.0 mg
Not described; 14-60 days
(acute)

Not reviewed

Insufficient information

Ephedrine (Maxi Thins)

Unknown
Not described; >60 days (chronic)

Implicit review

Insufficient information

Over-the-counter anti-asthma pill

60.0 mg
<6 hours; >60 days (chronic)

Implicit review

Insufficient information

* dose reported in total daily alkaloids

yo = year old

140

Table 22. Evidence table of case reports Other Cardiovascular

Event type
Report date
Age, Sex
Constituent
Source
(ID)
Cardiac/ Near
sudden death
04/08/1998
22 yo Male
Ephedra
FDA Case
(12851)
Cardiac/
Cardiomyopathy
28 yo Female
Literature Case
(110 )
Cardiac/
Cardiomyopathy
39 yo Male
Literature Case
(297)
Cardiac
59 yo Female
Ephedrine
FDA Case
(3359234)
Cardiac
99 yo Male
Ephedrine
FDA Case
(3537599)
Cardiac
42 yo Male
Literature Case
(174)
Cardiac
07/19/1994
43 yo Male
Ephedra
FDA Case
(9818)
Cardiac
06/02/1995
63 yo Female
Ephedra
FDA Case
(10275)

Product
Dose*
Timing; Duration
Ripped Force
20.4 mg
6-24 hours; >60 days (chronic)

Invesigation of Etiology RAND Classification

Angiography: No
Possible sentinel event

Ephedrine
2000.0 mg
Not described; >60 days
(chronic)

Angiography: Yes

Possible sentinel event

Herbalife Original Green

Unknown
Not described; 14-60 days
(acute)

Angiography: Yes

Possible sentinel event

Ephedrine
Unknown
<6 hours; < 48 hours

Not relevant

Intraoperative
ephedrine

Ephedrine
Unknown
<6 hours; < 48 hours

Not relevant

Intraoperative
ephedrine

Ephedrine
Unknown
<6 hours; < 48 hours

Not relevant

Intraoperative
ephedrine

Power Trim
Unknown
Not described; 14-60 days (acute)

Not reviewed

Insufficient information

Nature's Nutrition-Formula One

Unknown
Not described; 14-60 days (acute)

Not reviewed

Insufficient information

* dose reported in total daily alkaloids

yo = year old

141

Table 22. Evidence table of case reports Other Cardiovascular (continued)

Event type
Report date
Age, Sex
Constituent
Source
(ID)
Cardiac
05/07/1996
47 yo Female
Ephedra
FDA Case
(11133)
Cardiac
05/15/1996
66 yo Female
Ephedra
FDA Case
(11282)
Cardiac
06/24/1996
35 yo Female
Ephedra
FDA Case
(11464)
Cardiac
01/21/1996
48 yo Male
Ephedra
FDA Case
(11782)
Cardiac
01/22/1998
31 yo Female
Ephedra
FDA Case
(12740)
Cardiac/ Near
sudden death
07/29/1998
28 yo Female
Ephedra
FDA Case
(13031)
Cardiac
04/19/1999
57 yo Female
Ephedra
FDA Case
(13516)

Product
Dose*
Timing; Duration
Natural Trim
Unknown
Not described; 14-60 days (acute)

Invesigation of Etiology RAND Classification

Not reviewed
Insufficient information

E'ola Amp II Pro Drops

Unknown
Not described; >60 days (chronic)

Not reviewed

Insufficient information

Shape Fast
80.0 mg
Not described; 2-13 days

Not reviewed

Insufficient information

Pro ripped
Unknown
Not described; >60 days (chronic)

Not reviewed

Insufficient information

Ripped Fuel
Unknown
Not described; 2-13 days

Not reviewed

Insufficient information

Herbalife Original Green

43.2 mg
<6 hours; < 48 Hours

Angiography: Unknown

Insufficient Information

Metabolife 356
Unknown
Not described; >60 days (chronic)

Not reviewed

Insufficient information

* dose reported in total daily alkaloids

yo = year old

142

Table 22. Evidence table of case reports Other Cardiovascular (continued)

Event type
Report date
Age, Sex
Constituent
Source
(ID)
Cardiac/ Near
sudden death
05/19/1999
32 yo Female
Ephedra
FDA Case
(13643)
Cardiac/
Cardiomyopathy
07/23/1999
65 yo Female
Ephedra
FDA Case
(13793)
Cardiac
07/23/1999
39 yo Male
Ephedra
FDA Case
(13796)
Cardiac/
Ventricular
Tachycardia
11/15/1999
48 yo Female
Ephedra
FDA Case
(13945)
Cardiac
12/24/1999
48 yo Female
Ephedra
FDA Case
(13992)
Cardiac
11/08/1999
46 yo Male
Ephedra
FDA Case
(14017)

Product
Dose*
Timing; Duration
Natural Trim
88.0 mg
<6 hours; 14-60 days (acute)

Invesigation of Etiology RAND Classification

Angiography: Unknown
Insufficient Information

Thermolean
Unknown
<6 hours; >60 days (chronic)
Power Trim
84.0 mg
<6 hours; >60 Days (chronic)

Not reviewed

Insufficient Information

Natural Trim
Unknown
Not described; >60 days (chronic)

Not reviewed

Insufficient information

Metabolife 356
Unknown
<6 hours; 14-60 days (acute)

Not reviewed

Insufficient Information

Unknown
45.0 mg
Not described; >60 days (chronic)

Not reviewed

Insufficient information

Metabolife 356
Unknown
Not described; 14-60 days (acute)

Not reviewed

Insufficient information

* dose reported in total daily alkaloids

yo = year old

143

Table 22. Evidence table of case reports Other Cardiovascular (continued)

Event type
Report date
Age, Sex
Constituent
Source
(ID)
Cardiac
03/08/2000
26 yo Male
Ephedra
FDA Case
(14080)
Cardiac
03/23/2000
47 yo Female
Ephedra
FDA Case
(14108)
Cardiac
04/19/2000
41 yo Female
Ephedra
FDA Case
(14143)
Cardiac
04/11/2000
43 yo Female
Ephedra
FDA Case
(14242)
Cardiac
07/19/2000
26 yo Female
Ephedra
FDA Case
(14284)
Cardiac
09/14/2000
39 yo Female
Ephedra
FDA Case
(14383)
Cardiac/
Cardiomyopathy
32 yo Female
Literature Case
(260)

Product
Dose*
Timing; Duration
Ripped Fuel
Unknown
Not described; >60 days (chronic)
Hydroxycut (Muscle Tech R&D)
Unknown
Not described; >60 Days (chronic)
Herbalife Original Green
Unknown
>24 hours; >60 days (chronic)

Invesigation of Etiology RAND Classification

Not reviewed
Insufficient information

Not reviewed

Insufficient information

Metabolife 356
24.0 mg
Not described; >60 days (chronic)

Not reviewed

Insufficient information

Metabolize
Unknown
Not described; >60 days (chronic)
Unknown
Unknown
Not described; Not described
FitAmerica Intl Weight ControlAid
Unknown
>24 hours; 2-13 days

Not reviewed

Insufficient information

Not reviewed

Insufficient information

Biolean
Unknown
Not described; >60 days (chronic)

I Not reviewed

Insufficient information

Ephedrine
450.0 mg
Not described; >60 days (chronic)

Angiography: No

Insufficient information

* dose reported in total daily alkaloids

yo = year old

144

Table 22. Evidence table of case reports Other Cardiovascular (continued)

Event type
Report date
Age, Sex
Constituent
Source
(ID)
Cardiac/
Cardiomyopathy
35 yo Male
Literature Case
(271)

Product
Dose*
Timing; Duration
Insufficient information
Unknown
<24 hours; >60 days (chronic)

Invesigation of Etiology RAND Classification

Angiography: No
Insufficient information

* dose reported in total daily alkaloids

yo = year old

145

Table 22. Evidence table of case reports Other Neurological

Event type
Report date
Age, Sex
Constituent
Source
(ID)
Neurological/
TIA
06/29/1998
57 yo Female
Ephedra
FDA Case
(13062)
Neurological
11/27/1995
54 yo Female
Ephedra
FDA Case
(10573)
Neurological
08/12/1996
39 yo Male
Ephedra
FDA Case
(11900)
Neurological
02/10/2000
59 yo Female
Ephedra
FDA Case
(14018)
Neurological
08/31/2000
31 yo Female
Ephedra
FDA Case
(14352)
Neurological
11/07/2000
35 yo Female
Ephedra
FDA Case
(14495)
Neurological
29 yo Male
Ephedrine
FDA Case
(1535075)

Product
Dose*
Timing; Duration
Metabolife 356
48.0 mg
6-24 hours; < 48 hours

Investigation of Etiology RAND classification

Implicit review
Possible sentinel event

Thermogenic Fat Burner (Joe Weider) Not reviewed

48.0 mg
Not described; 14-60 days (acute)

Insufficient information

Excel Energy
24.0 mg
Not described; >60 days (chronic)

Not reviewed

Insufficient information

Metabolife 356
24.0 mg
>24 hours; >60 days (chronic)

Implicit review

Did not meet temporal

relationship criterion

Ripped Fuel
Unknown
Not described; 14-60 days (acute)

Not reviewed

Insufficient information

Metab-O-Lite
24.0 mg
Not described; >60 days (chronic)

Not reviewed

Insufficient information

Ephedrine
Unknown
Not described; Not described

Implicit review

Insufficient information

* dose reported in total daily alkaloids

yo = year old

146

Table 22. Evidence table of case reports Other Neurological (continued)

Event type
Report date
Age, Sex
Constituent
Source
(ID)
Neurological/
TIA
12 yo Female
Literature Case
(218)

Product
Dose*
Timing; Duration
E'ola
Unknown
<6 hours; < 48 hours

Investigation of Etiology RAND classification

Implicit review
Insufficient information

* dose reported in total daily alkaloids

yo = year old

147

Table 22. Evidence table of case reports Seizure

Event type
Age, Sex
Constituent
Source
(ID)
Seizure
19 YO Female
Ephedra
FDA Case
(10974)

Product
Dose*
Timing; Duration
Shape Fast/Rite
Not described
Not described; 14-60 days (acute)

Investigation for Etiology RAND Classification

CT/MRI of Head: Yes
Sentinel event
Serum Electrolytes: Yes
Glucose: Yes
Calcium: Yes
Magnesium: No
Temperature: Yes
EEG: Yes
Seizure
Ephedrine
CT/MRI of Head: No
Sentinel event
38 YO Female Not described
Serum Electrolytes: No
Ephedrine
6-24 hours; Duration Not described Glucose: No
Literature Case
Calcium: No
(224)
Magnesium: No
Temperature: No
EEG: No
Seizure
Natures Nutrition-Formula One
CT/MRI of Head: Yes
Possible sentinel event
47 YO Female Not described
Serum Electrolytes: Yes
Ephedra
6-24 hours; 14-60 days (acute)
Glucose: Yes
FDA Case
Calcium: No
(9534)
Magnesium: No
Temperature: Yes
EEG: Yes
Seizure
Natures Nutrition-Formula One
CT/MRI of Head: Yes
Possible sentinel event
37 YO Female Not described
Serum Electrolytes: Yes
Ephedra
Not described; >60 days (chronic)
Glucose: Yes
FDA Case
Calcium: Yes
(10221)
Magnesium: Yes
Temperature: Yes
EEG: Yes
Seizure
Thermo Slim
CT/MRI of Head: Yes
Possible sentinel event
62 YO Male
Not described
Serum Electrolytes: Yes
Ephedra
Not described; >60 days (chronic)
Glucose: Yes
FDA Case
Calcium: Yes
(10432)
Magnesium: Yes
Temperature: Yes
EEG: Yes
Seizure
Metabolife 356
CT/MRI of Head: Yes
Possible sentinel event
23 YO Female Not described
Serum Electrolytes: Yes
Ephedra
Not described; >60 days (chronic)
Glucose: Yes
FDA Case
Calcium: Yes
(11649)
Magnesium: No
Temperature: Yes
EEG: Yes

* dose reported in total daily alkaloids

yo = year old

148

Table 22. Evidence table of case reports Seizure (continued)

Event type
Age, Sex
Constituent
Source
(ID)
Seizure
26 YO Male
Ephedra
FDA Case
(13408)

Product
Dose*
Timing; Duration
Ripped Fuel
Not described
<6 hours; >60 days (chronic)

Seizure
Metab-O-Lite
30 YO Female Not described
Ephedra
6-24 hours; >60 days (chronic)
FDA Case
(14275)

Seizure
Thin Tabs
31 YO Female Not described
Ephedra
6-24 hours; 14-60 days (acute)
FDA Case
(14571)

Seizure
Natures Nutrition-Formula One
38 YO Female Not described
Ephedra
Not described; 14-60 days (acute)
FDA Case
(9528 )

Seizure
Natures Nutrition-Formula One
47 YO Female Not described
Ephedra
Not described; 2-13 days
FDA Case
(9547)

Seizure
Ripped Fuel
40 YO Female 25 mg
Ephedra
Not described; 2-13 days
FDA Case
(9747)

Investigation for Etiology RAND Classification

CT/MRI of Head: Yes
Possible sentinel event
Serum Electrolytes: Yes
Glucose: Yes
Calcium: Yes
Magnesium: Yes
Temperature: Yes
EEG: Yes
CT/MRI of Head: Yes
Possible sentinel event
Serum Electrolytes: Yes
Glucose: Yes
Calcium: Yes
Magnesium: No
Temperature: Yes
EEG: Yes
CT/MRI of Head: Yes
Possible sentinel event
Serum Electrolytes: Yes
Glucose: Yes
Calcium: Yes
Magnesium: No
Temperature: Yes
EEG: Yes
CT/MRI of Head: Yes
Insufficient Information
Serum Electrolytes: Yes
Glucose: Yes
Calcium: Yes
Magnesium: No
Temperature: Yes
EEG: Yes
CT/MRI of Head: Yes
Insufficient information
Serum Electrolytes: Yes
Glucose: No
Calcium: Yes
Magnesium: Yes
Temperature: Yes
EEG: Yes
CT/MRI of Head: Yes
Insufficient information
Serum Electrolytes: Yes
Glucose: Yes
Calcium: No
Magnesium: No
Temperature: No
EEG: No

* dose reported in total daily alkaloids

yo = year old

149

Table 22. Evidence table of case reports Seizure (continued)

Event type
Age, Sex
Constituent
Source
(ID)
Seizure
Age Not
described, Male
Ephedra
FDA Case
(9799)

Product
Dose*
Timing; Duration
Eola Amp II Pro Drops
Not described
Not described; 14-60 days (acute)

Seizure
Thermogenics Plus
34 YO Female Not described
Ephedra
Not described; 14-60 days (acute)
FDA Case
(10301)

Seizure
32 YO Male
Ephedra
FDA Case
(10416)

Slim Now
Not described
<6 hours; >60 days (chronic)

Seizure
Herbalife Original Green
55 YO Female Not described
Ephedra
<6 hours; 2-13 days
FDA Case
(10437)

Seizure
Thermochrome 5000
38 YO Female 21 mg
Ephedra
6-24 hours; <48 hours
FDA Case
(10570)

Seizure
Diet Max/Super Diet Max
29 YO Female Not described
Ephedra
Not described; >60 days (chronic)
FDA Case
(10964)

Investigation for Etiology RAND Classification

CT/MRI of Head: No
Insufficient information
Serum Electrolytes: No
Glucose: No
Calcium: No
Magnesium: No
Temperature: No
EEG: Yes
CT/MRI of Head: Yes
Insufficient information
Serum Electrolytes: No
Glucose: No
Calcium: No
Magnesium: No
Temperature: No
EEG: Yes
CT/MRI of Head: Yes
Insufficient information
Serum Electrolytes: Yes
Glucose: Yes
Calcium: Yes
Magnesium: No
Temperature: Yes
EEG: Yes
CT/MRI of Head: Yes
Insufficient information
Serum Electrolytes: Yes
Glucose: Yes
Calcium: Yes
Magnesium: No
Temperature: No
EEG: No
CT/MRI of Head: Yes
Insufficient information
Serum Electrolytes: Yes
Glucose: Yes
Calcium: Yes
Magnesium: Yes
Temperature: Yes
EEG: Yes
CT/MRI of Head: No
Insufficient information
Serum Electrolytes: No
Glucose: No
Calcium: No
Magnesium: No
Temperature: No
EEG: No

* dose reported in total daily alkaloids

yo = year old

150

Table 22. Evidence table of case reports Seizure (continued)

Event type
Age, Sex
Constituent
Source
(ID)
Seizure
41 YO Female
Ephedra
FDA Case
(11001)

Product
Dose*
Timing; Duration
Guarana Plus
Not described
Not described; >60 days (chronic)

Seizure
36 YO Female
Ephedra
FDA Case
(11078)

Quick Start
Not described
Not described; Not described
Natures Nutrition-Formula One
Not described
6-24 hours; >60 days (chronic)

Seizure
19 YO Male
Ephedra
FDA Case
(11181)

Ripped Fuel
Not described
6-24 hours; 2-13 days

Seizure
24 YO Male
Ephedra
FDA Case
(11215)

Ripped Fuel
Not described
Not described; Not described
Ripped Force
Not described
>24 hours; >60 days (chronic)

Seizure
20 YO Male
Ephedra
FDA Case
(11249)

Victory Turbo Pump

Not described
Not described; >60 days (chronic)

Seizure
Eola Amp Pro Drops
38 YO Female Not described
Ephedra
<6 hours; <48 hours
FDA Case
(11304)

Investigation for Etiology RAND Classification

CT/MRI of Head: No
Insufficient information
Serum Electrolytes: No
Glucose: No
Calcium: No
Magnesium: No
Temperature: No
EEG: No
CT/MRI of Head: Yes
Insufficient information
Serum Electrolytes: Yes
Glucose: Yes
Calcium: No
Magnesium: Yes
Temperature: Yes
EEG: Yes
CT/MRI of Head: Yes
Insufficient information
Serum Electrolytes: Yes
Glucose: Yes
Calcium: No
Magnesium: No
Temperature: Yes
EEG: Yes
CT/MRI of Head: No
Insufficient information
Serum Electrolytes: No
Glucose: No
Calcium: No
Magnesium: No
Temperature: No
EEG: No
CT/MRI of Head: Yes
Insufficient information
Serum Electrolytes: No
Glucose: Yes
Calcium: No
Magnesium: No
Temperature: No
EEG: Yes
CT/MRI of Head: No
Insufficient information
Serum Electrolytes: No
Glucose: No
Calcium: No
Magnesium: No
Temperature: No
EEG: No

* dose reported in total daily alkaloids

yo = year old

151

Table 22. Evidence table of case reports Seizure (continued)

Event type
Age, Sex
Constituent
Source
(ID)
Seizure
37 YO Male
Ephedra
FDA Case
(11316)

Product
Dose*
Timing; Duration
Natures Nutrition-Formula One
Not described
6-24 hours; 14-60 days (acute)

Seizure
Fit America Intnl Weight Control Aid
34 YO Female Not described
Ephedra
Not described; 2-13 days
FDA Case
(11594)

Seizure
15 YO Male
Ephedra
FDA Case
(12477)

Up Your Gas
Not described
<6 hours; <48 hours

Seizure
Age Not
described,
Female
Ephedra
FDA Case
(12948)
Seizure
42 YO Female
Ephedra
FDA Case
(13110)

Escalation
Not described
Not described; 2-13 days

E-Z Trim Tablets

24 mg
Not described; 2-13 days

Seizure
Metabolift
53 YO Female Not described
Ephedra
<6 hours; 14-60 days (acute)
FDA Case
(13514)

Investigation for Etiology RAND Classification

CT/MRI of Head: Yes
Insufficient information
Serum Electrolytes: Yes
Glucose: Yes
Calcium: Yes
Magnesium: Yes
Temperature: Yes
EEG: Yes
CT/MRI of Head: No
Insufficient information
Serum Electrolytes: Yes
Glucose: Yes
Calcium: Yes
Magnesium: Yes
Temperature: Yes
EEG: Yes
CT/MRI of Head: No
Insufficient information
Serum Electrolytes: No
Glucose: No
Calcium: No
Magnesium: No
Temperature: No
EEG: No
CT/MRI of Head: No
Insufficient information
Serum Electrolytes: No
Glucose: No
Calcium: No
Magnesium: No
Temperature: No
EEG: No
CT/MRI of Head: Yes
Insufficient information
Serum Electrolytes: Yes
Glucose: Yes
Calcium: Yes
Magnesium: Yes
Temperature: Yes
EEG: Yes
CT/MRI of Head: No
Insufficient information
Serum Electrolytes: No
Glucose: No
Calcium: No
Magnesium: No
Temperature: No
EEG: No

* dose reported in total daily alkaloids

yo = year old

152

Table 22. Evidence table of case reports Seizure (continued)

Event type
Age, Sex
Constituent
Source
(ID)
Seizure
Age Not
described,
Female
Ephedra
FDA Case
(13519)
Seizure
46 YO Female
Ephedra
FDA Case
(13625)

Seizure
25 YO Female
Ephedra
FDA Case
(13715)

Seizure
51 YO Male
Ephedra
FDA Case
(13895)

Product
Dose*
Timing; Duration
Metabolife 356
Not described
Not described; Not described

Metabolife 356
Not described
Not described; 14-60 days (acute)

Diet Fuel
Not described
Not described; Not described
Ripped Fuel
Not described
Not described; Not described
Hydroxycut
Not described
<6 hours; >60 days (chronic)
Metabolife 356
Not described
Not described; 14-60 days (acute)

Seizure
17 YO Male
Ephedra
FDA Case
(13946)

Ripped Fuel
Not described
Not described; Not described
Thermo-Tek
Not described
<6 hours; >60 days (chronic)

Seizure
58 YO Male
Ephedra
FDA Case
(13972)

Metabolife 356
Not described
Not described; >60 days (chronic)

Investigation for Etiology RAND Classification

CT/MRI of Head: No
Insufficient information
Serum Electrolytes: Yes
Glucose: No
Calcium: No
Magnesium: No
Temperature: No
EEG: No
CT/MRI of Head: Yes
Insufficient information
Serum Electrolytes: Yes
Glucose: Yes
Calcium: No
Magnesium: No
Temperature: No
EEG: Yes
CT/MRI of Head: Yes
Insufficient information
Serum Electrolytes: Yes
Glucose: Yes
Calcium: No
Magnesium: No
Temperature: No
EEG: Yes

CT/MRI of Head: Yes

Serum Electrolytes: Yes
Glucose: Yes
Calcium: No
Magnesium: No
Temperature: No
EEG: Yes
CT/MRI of Head: Yes
Serum Electrolytes: Yes
Glucose: Yes
Calcium: No
Magnesium: No
Temperature: Yes
EEG: Yes
CT/MRI of Head: Yes
Serum Electrolytes: Yes
Glucose: Yes
Calcium: No
Magnesium: No
Temperature: No
EEG: Yes

* dose reported in total daily alkaloids

yo = year old

153

Insufficient information

Insufficient information

Insufficient information

Table 22. Evidence table of case reports Seizure (continued)

Event type
Age, Sex
Constituent
Source
(ID)
Seizure
39 YO Female
Ephedra
FDA Case
(14116)

Seizure
23 YO Male
Ephedra
FDA Case
(14258)

Product
Dose*
Timing; Duration
Thermo-Gen
Not described
6-24 hours; >60 days (chronic)

Ripped Fuel
Not described
<6 hours; <48 hours

Seizure
Natural Trim
42 YO Female Not described
Ephedra
6-24 hours; <48 hours
FDA Case
(14297)

Seizure
Age Not
described,
Female
Ephedrine
FDA Case
(3549038)

Ephedrine Plus
Not described
<6 hours; 14-60 days (acute)

Investigation for Etiology RAND Classification

CT/MRI of Head: Yes
Insufficient information
Serum Electrolytes: Yes
Glucose: Yes
Calcium: Yes
Magnesium: Yes
Temperature: Yes
EEG: Yes
CT/MRI of Head: Yes
Insufficient information
Serum Electrolytes: Yes
Glucose: Yes
Calcium: Yes
Magnesium: Yes
Temperature: Yes
EEG: Yes
CT/MRI of Head: Yes
Insufficient information
Serum Electrolytes: No
Glucose: No
Calcium: No
Magnesium: No
Temperature: No
EEG: Yes
CT/MRI of Head: No
Insufficient information
Serum Electrolytes: No
Glucose: No
Calcium: No
Magnesium: No
Temperature: No
EEG: No

* dose reported in total daily alkaloids

yo = year old

154

Table 22. Evidence table of case reports Psychiatric

Event type
Age, Sex
Constituent
Source
(ID)
Psychosis, Sleep
disturbance,
Palpitations, Dizzy
21 YO Male
Ephedra
FDA Case (9509)
Psychosis,
Hallucinations, Sleep
disturbance
39 YO Female
Ephedra
FDA Case (11678)
Mania or severe
agitation, Suicidal
ideation, Violent,
Personality changes,
Headache
19 YO Female
Ephedra
FDA Case (13809)
Psychosis, Mania or
severe agitation,
Severe depression,
Suicidal ideation,
Sleep disturbance,
Homicidal ideation
29 YO Male
Ephedra
FDA Case (14529)
Mania or severe
agitation, Ventricular
tachycardia /
fibrillation, Insomnia,
Violent
16 YO Male
Ephedrine
FDA Case (1855921)
Mania or severe
agitation, Sleep
disturbance
45 YO Male
Ephedra
Literature Case (48)

Product
Dose*
Duration
Addiction Data**
Natures Nutrition-Formula One
Not described
<48 hours
Addiction: No

Investigation for
Etiology***
RAND Classification
Psychiatric History: No
Sentinel Event
Other Substances/Meds:
No

Diet Now
12 mg
Over 1 year
Addiction: No

Psychiatric History: No
Sentinel Event
Other Substances/Meds:
No

Hydroxycut
Not described
2-13 days
Addiction: No

Psychiatric History: No
Sentinel Event
Other Substances/Meds:
No

Xenadrine
Not described
60 days to 1 year
Addiction: No

Psychiatric History: No
Sentinel Event
Other Substances/Meds:
No

Max Alert
Mini Thin
Not described
60 days to 1 year
Addiction: Yes

Psychiatric History: No
Sentinel Event
Other Substances/Meds:
No

Ma huang/Ephedra
Not described
14-60 days (acute)
Addiction: No

Psychiatric History: No
Sentinel Event
Other Substances/Meds:
No

* dose reported in total daily alkaloids

yo = year old
** Addiction: Yes = diagnosis or self-reported addiction to the product
*** Psychiatric history: Yes = recorded psychiatric history
Other substances/meds: Yes = patient taking other substances or medications known to cause psychiatric symptoms

155

Table 22. Evidence table of case reports Psychiatric (continued)

Event type
Age, Sex
Constituent
Source
(ID)
Psychosis, Paranoia
30 YO Female
Ephedrine
Literature Case (238 )
Psychosis,
Hallucinations,
Confusion/Delusional
59 YO Male
Ephedrine
Literature Case (285)
Severe depression,
Suicide attempt
28 YO Female
Ephedra
FDA Case (9751)
Psychosis, Suicidal
ideation, Palpitations,
Increased
hypertension,
19 YO Male
Ephedra
FDA Case (11157)
Psychosis,
Hallucinations,
Memory Loss
13 YO Female
Ephedra
FDA Case (12372)
Mania or severe
agitation, Sleep
disturbance
21 YO Male
Ephedra
FDA Case (13005)
Psychosis,
Hallucinations
52 YO Female
Ephedra
FDA Case (14436)

Product
Dose*
Duration
Addiction Data**
Tedral
144 mg
Over 1 year
Addiction: Yes
Bronchi Pax
360 mg
Over 1 year
Addiction: No

Investigation for
Etiology***
RAND Classification
Psychiatric History: No
Sentinel Event
Other Substances/Meds:
No
Psychiatric History: No
Sentinel Event
Other Substances/Meds:
No

Slim NRG+
Not described
60 days to 1 year
Addiction: No

Psychiatric History: No
Possible Sentinel
Other Substances/Meds: Event
No

Ripped Fuel
Not described
60 days to 1 year
Addiction: No

Psychiatric History: No
Possible Sentinel
Other Substances/Meds: Event
No

Natures Nutrition-Formula One Psychiatric History: No

Possible Sentinel
Not described
Other Substances/Meds: Event
14-60 days (acute)
No
Addiction: No

Ripped Fuel
Not described
14-60 days (acute)
Addiction: No

Psychiatric History: No
Possible Sentinel
Other Substances/Meds: Event
No

Metab-O-Lite
Not described
60 days to 1 year
Addiction: No

Psychiatric History: No
Possible Sentinel
Other Substances/Meds: Event
No

* dose reported in total daily alkaloids

yo = year old
** Addiction: Yes = diagnosis or self-reported addiction to the product
*** Psychiatric history: Yes = recorded psychiatric history
Other substances/meds: Yes = patient taking other substances or medications known to cause psychiatric symptoms

156

Table 22. Evidence table of case reports Psychiatric (continued)

Event type
Age, Sex
Product
Constituent
Dose*
Investigation for
Source
Duration
(ID)
Etiology***
Addiction Data**
Psychosis,
Metab-O-Lite
Psychiatric History: No
Suicide attempt,
Not described
Other Substances/Meds:
Insomnia, Ventricular 60 days to 1 year
No
tachycardia /
Addiction: No
fibrillation, Dizzy
28 YO Female
Ephedra
FDA Case (14528)
Psychosis,
Max Alert
Psychiatric History: No
Addiction/Substance up to 1250 mg / day
Other Substances/Meds:
Abuse, Paranoia
Over 1 year
No
31 YO Male
Addiction: Yes
Ephedrine
FDA Case (1661966)
Psychosis, Mania or Unknown
Psychiatric History: No
severe agitation,
Not described
Other Substances/Meds:
Hallucinations,
2-13 days
No
Paranoia
Addiction: No
34 YO Male
Ephedra
Literature Case (79)
Psychosis, Sleep
Do-Do Tablet or Herbal Balance Psychiatric History: Yes
disturbance, Headache 100 mg
Other Substances/Meds:
40 YO Female
14-60 days (acute)
No
Ephedra
Addiction: No
FDA Case (9060)
Severe depression,
Natures Nutrition-Formula One Psychiatric History: Yes
Suicidal ideation
Not described
Other Substances/Meds:
47 YO Male
14-60 days (acute)
Yes
Ephedra
Addiction: No
FDA Case (9727)
Severe depression,
Natures Nutrition-Formula One Psychiatric History: Yes
Suicidal ideation
Not described
Other Substances/Meds:
38 YO Female
>60 days (chronic)
Yes
Ephedra
Addiction: No
FDA Case (9727)
Severe depression,
Natures Nutrition-Formula One Psychiatric History: Yes
Suicidal ideation
Not described
Other Substances/Meds:
30 YO Female
>60 days (chronic)
Yes
Ephedra
Addiction: No
FDA Case (9727)

RAND Classification
Possible Sentinel
Event

Possible Sentinel
Event

Possible Sentinel
Event

Inconclusive Prior
psychiatric history

Inconclusive Prior
psychiatric history

Inconclusive Prior
psychiatric history

Inconclusive Prior
psychiatric history

* dose reported in total daily alkaloids

yo = year old
** Addiction: Yes = diagnosis or self-reported addiction to the product
*** Psychiatric history: Yes = recorded psychiatric history
Other substances/meds: Yes = patient taking other substances or medications known to cause psychiatric symptoms

157

Table 22. Evidence table of case reports Psychiatric (continued)

Event type
Age, Sex
Product
Constituent
Dose*
Source
Duration
(ID)
Addiction Data**
Psychosis, Insomnia Therachrome
43 YO Female
Not described
Ephedra
60 days to 1 year
FDA Case (9403 )
Addiction: No
Psychosis, Violent
Diet Gel
39 YO Male
Not described
Ephedra
Not described
FDA Case (10042)
Addiction: No
Psychosis, Mania or Ripped Force
severe agitation, Sleep Not described
disturbance
60 days to 1 year
17 YO Male
Addiction: No
Ephedra
FDA Case (10078)
Psychosis, Severe
Mini Thin
depression,
285 mg
Suicide attempt,
60 days to 1 year
Addiction/Substance Addiction: Yes
Abuse
38 YO Female
Ephedra/Ephedrine
FDA Case (11052)
Psychosis, Mania or Up Your Gas
severe agitation,
Not described
Violent,
Not described
Addiction/Substance Addiction: Yes
Abuse
17 YO Male
Ephedra
FDA Case (11096 )
Severe depression,
Natures Nutrition-Formula One
Anxiety, Headache
Not described
31 YO Male
>60 days (chronic)
Ephedra
Addiction: No
FDA Case (11145)
Violent
Up Your Gas
35 YO Male
Not described
Ephedra
14-60 days (acute)
FDA Case (11289)
Addiction: No

Investigation for
Etiology***
RAND Classification
Psychiatric History: Yes Inconclusive Prior
Other Substances/Meds: psychiatric history
No
Psychiatric History: Yes Inconclusive Prior
Other Substances/Meds: psychiatric history
Yes
Psychiatric History: Yes Inconclusive Prior
Other Substances/Meds: psychiatric history
Yes

Psychiatric History: Yes Inconclusive Prior

Other Substances/Meds: psychiatric history
No

Psychiatric History: Yes Inconclusive Prior

Other Substances/Meds: psychiatric history
No

Psychiatric History: Yes Inconclusive Prior

Other Substances/Meds: psychiatric history
Yes

Psychiatric History: Yes Inconclusive Prior

Other Substances/Meds: psychiatric history
Yes

* dose reported in total daily alkaloids

yo = year old
** Addiction: Yes = diagnosis or self-reported addiction to the product
*** Psychiatric history: Yes = recorded psychiatric history
Other substances/meds: Yes = patient taking other substances or medications known to cause psychiatric symptoms

158

Table 22. Evidence table of case reports Psychiatric (continued)

Event type
Age, Sex
Constituent
Source
(ID)
Severe depression,
Addiction/Substance
Abuse, Sleep
disturbance
38 YO Female
Ephedra
FDA Case (11651)
Psychosis
34 YO Female
Ephedra
FDA Case (11717)
Severe depression,
Suicidal ideation
57 YO Female
Ephedra
FDA Case (11828)
Mania or severe
agitation, Suicidal
ideation, Cyclothymia
15 YO Female
Ephedra
FDA Case (13072)
Psychosis, Mania or
severe agitation,
Hallucinations, Sleep
disturbance, Migraine
20 YO Male
Ephedra/Ephedrine
FDA Case (13099)
Severe depression
28 YO Female
Ephedra
FDA Case (14089)
Mania or severe
agitation, Severe
depression,
Addiction/Substance
Abuse
29 YO Female
Ephedra
FDA Case (14276)

Product
Dose*
Duration
Addiction Data**
Natures Nutrition-Formula One
Not described
60 days to 1 year
Addiction: Yes

Investigation for
Etiology***
RAND Classification
Psychiatric History: Yes Inconclusive Prior
Other Substances/Meds: psychiatric history
Yes

M-80 pills
Not described
60 days to 1 year
Addiction: No
Herbalife Original Green
Not described
14-60 days (acute)
Addiction: No

Psychiatric History: Yes Inconclusive Prior

Other Substances/Meds: psychiatric history
No

Caloslim
Not described
>60 days (chronic)
Addiction: No

Psychiatric History: Yes Inconclusive Prior

Other Substances/Meds: psychiatric history
No

Mini Thin
75 mg
Not described
Hydroxycut
Not described
60 days to 1 year
Addiction: No
Xenadrine
Not described
60 days to 1 year
Addiction: No
Up Your Gas
Not described
Over 1 year
Addiction: Yes

Psychiatric History: Yes Inconclusive Prior

Other Substances/Meds: psychiatric history
No

Psychiatric History: Yes Inconclusive Prior

Other Substances/Meds: psychiatric history
No

Psychiatric History: Yes Inconclusive Prior

Other Substances/Meds: psychiatric history
No
Psychiatric History: Yes Inconclusive Prior
Other Substances/Meds: psychiatric history
No

* dose reported in total daily alkaloids

yo = year old
** Addiction: Yes = diagnosis or self-reported addiction to the product
*** Psychiatric history: Yes = recorded psychiatric history
Other substances/meds: Yes = patient taking other substances or medications known to cause psychiatric symptoms

159

Table 22. Evidence table of case reports Psychiatric (continued)

Event type
Age, Sex
Constituent
Source
(ID)
Psychosis, Severe
depression, Insomnia
17 YO Male
Ephedra
FDA Case (14294)
Psychosis, Severe
depression, Motor
vehicle accident,
Paranoia,
Confusion/Delusional
42 YO Female
Ephedra
FDA Case (14394)
Psychosis,
Suicide/Suicide
attempt,
Hallucinations,
Anxiety, Paranoia
36 YO Female
Ephedra
FDA Case (14493)
Mania or severe
agitation,
Hallucinations
32 YO, Sex Not
described
Ephedra
FDA Case (14541)
Psychosis, Severe
depression, Suicidal
ideation,
Hallucinations,
Insomnia
22 YO Female
Ephedra
FDA Case (14543)
Psychosis, Mania or
severe agitation
20 YO Male
Ephedra
Literature Case (136)

Product
Dose*
Duration
Addiction Data**
Hydroxycut
Not described
2-13 days
Addiction: No

Investigation for
Etiology***
RAND Classification
Psychiatric History: Yes Inconclusive Prior
Other Substances/Meds: psychiatric history
Yes

Fen-Chi
Not described
14-60 days (acute)
Addiction: No

Psychiatric History: Yes Inconclusive Prior

Other Substances/Meds: psychiatric history
Yes

Herbalife Original Green

Not described
>60 days (chronic)
Addiction: No

Psychiatric History: Yes Inconclusive Prior

Other Substances/Meds: psychiatric history
Yes

Metabolife 356
Not described
60 days to 1 year
Addiction: No

Psychiatric History: Yes Inconclusive Prior

Other Substances/Meds: psychiatric history
Yes

Metabolift
Not described
14-60 days (acute)
Addiction: No

Psychiatric History: Yes Inconclusive Prior

Other Substances/Meds: psychiatric history
Yes

Metabolife 356
Not described
>60 days (chronic)
Addiction: No

Psychiatric History: Yes Inconclusive Prior

Other Substances/Meds: psychiatric history
Yes

* dose reported in total daily alkaloids

yo = year old
** Addiction: Yes = diagnosis or self-reported addiction to the product
*** Psychiatric history: Yes = recorded psychiatric history
Other substances/meds: Yes = patient taking other substances or medications known to cause psychiatric symptoms

160

Table 22. Evidence table of case reports Psychiatric (continued)

Event type
Age, Sex
Constituent
Source
(ID)
Severe depression,
Suicidal ideation,
Violent
27 YO Male
Ephedra
Literature Case (136)
Mania or severe
agitation, Severe
depression
40 YO Female
Ephedra
Literature Case (519)
Psychosis, Suicidal
ideation, Anxiety,
Dizzy, Increased
hypertension
33 YO Female
Ephedra
FDA Case (9516)
Mania or severe
agitation, Suicidal
ideation, Sleep
disturbance
45 YO Male
Ephedra
FDA Case (10233)
Psychosis, Catatonia
36 YO Male
Ephedra
FDA Case (12488)

Product
Dose*
Duration
Addiction Data**
Metabolife 356
Not described
Over 1 year
Addiction: No

Investigation for
Etiology***
RAND Classification
Psychiatric History: Yes Inconclusive Prior
Other Substances/Meds: psychiatric history
No

Product Not described

Not described
Over 1 year
Addiction: No

Psychiatric History: Yes Inconclusive Prior

Other Substances/Meds: psychiatric history
Yes

Natures Nutrition-Formula One Psychiatric History: No

Inconclusive Other
Not described
Other Substances/Meds: meds / substances
60 days to 1 year
Yes
Addiction: No

Natures Nutrition-Formula One Psychiatric History: No

Inconclusive Other
Not described
Other Substances/Meds: meds / substances
14-60 days (acute)
Yes
Addiction: No

Nature's Super Cap

99mg
Thermadrene
Just Be Natural
Gorilla Nitro Plus
Mega Creatine Fuel
Bolt
Not described
Over 1 year
Addiction: No

Psychiatric History: No
Inconclusive Other
Other Substances/Meds: meds / substances
Yes

* dose reported in total daily alkaloids

yo = year old
** Addiction: Yes = diagnosis or self-reported addiction to the product
*** Psychiatric history: Yes = recorded psychiatric history
Other substances/meds: Yes = patient taking other substances or medications known to cause psychiatric symptoms

161

Table 22. Evidence table of case reports Psychiatric (continued)

Event type
Age, Sex
Product
Constituent
Dose*
Source
Duration
(ID)
Addiction Data**
Mania or severe
Metacuts
agitation,
Not described
Hallucinations,
>60 days (chronic)
Headache, Sleep
Addiction: No
disturbance, Irregular
heart rate
19 YO Male
Ephedra/Ephedrine
FDA Case (13370)
Psychosis, Mania or Xenadrine
severe agitation,
40 mg
Hallucinations, Motor 2-13 days
vehicle accident, Sleep Addiction: No
disturbance
27 YO Female
Ephedra
FDA Case (13526)
Severe depression,
Ripped Force
Suicidal ideation,
Not described
Sleep disturbance
60 days to 1 year
15 YO Male
Addiction: No
Ephedra
FDA Case (14082)
Psychosis,
Thermogenics Plus
Suicide attempt
Not described
Age Not described,
Over 1 year
Female
Addiction: No
Ephedra
FDA Case (14213 )
Mania or severe
Diet Fuel
agitation, Severe
Not described
depression, Aneurysm, 60 days to 1 year
ruptured cerebra,
Addiction: No
Encephalopathy
Age Not described,
Male
Ephedra
FDA Case (14287)
Psychosis,
Metabolife 356
Hyperkalemia
Not described
37 YO Female
14-60 days (acute)
Ephedra
Addiction: No
FDA Case (14300)

Investigation for
Etiology***
RAND Classification
Psychiatric History: No
Inconclusive Other
Other Substances/Meds: meds / substances
Yes

Psychiatric History: No
Inconclusive Other
Other Substances/Meds: meds / substances
Yes

Psychiatric History: No
Inconclusive Other
Other Substances/Meds: meds / substances
Yes

Psychiatric History: No
Inconclusive Other
Other Substances/Meds: meds / substances
Yes

Psychiatric History: No
Inconclusive Other
Other Substances/Meds: meds / substances
Yes

Psychiatric History: No
Inconclusive Other
Other Substances/Meds: meds / substances
Yes

* dose reported in total daily alkaloids

yo = year old
** Addiction: Yes = diagnosis or self-reported addiction to the product
*** Psychiatric history: Yes = recorded psychiatric history
Other substances/meds: Yes = patient taking other substances or medications known to cause psychiatric symptoms

162

Table 22. Evidence table of case reports Psychiatric (continued)

Event type
Age, Sex
Constituent
Source
(ID)
Psychosis, Sleep
disturbance,
Confusion/Delusional
36 YO Female
Ephedra
FDA Case (14546)
Psychosis
39 YO Female
Ephedra
FDA Case (14575)
Psychosis, Mania or
severe agitation,
Confusion/Delusional,
Headache
54 YO Female
Ephedra
FDA Case (14582)
Psychosis,
Hallucinations
Age Not described,
Female
Ephedra
FDA Case (10019)
Psychosis, Irregular
heart rate, Insomnia,
Dizzy, Gastrointestinal
problems
24 YO Female
Ephedra
FDA Case (10614)
Hallucinations,
Nausea, Dizzy,
Headache
20 YO Male
Ephedra
FDA Case (11131)
Anxiety, Palpitations
25 YO Male
Ephedra
FDA Case (11354)

Product
Dose*
Duration
Addiction Data**
Metabolife 356
Not described
>60 days (chronic)
Addiction: No

Investigation for
Etiology***
RAND Classification
Psychiatric History: No
Inconclusive Other
Other Substances/Meds: meds / substances
Yes

Up Your Gas
Not described
Over 1 year
Addiction: No
Metabolife 356
Not described
14-60 days (acute)
Addiction: No

Psychiatric History: No
Inconclusive Other
Other Substances/Meds: meds / substances
Yes

Ripped Fuel
Not described
60 days to 1 year
Addiction: No

Psychiatric History: No
Insufficient Information
Other Substances/Meds:
No

Diet Max
Not described
14-60 days (acute)
Addiction: No

Psychiatric History: No
Insufficient Information
Other Substances/Meds:
No

Herbal Ecstasy
Not described
<48 hours
Addiction: No

Psychiatric History: No
Insufficient Information
Other Substances/Meds:
Yes

Ripped Fuel
Not described
14-60 days (acute)
Addiction: No

Psychiatric History: No
Insufficient Information
Other Substances/Meds:
No

Psychiatric History: No
Inconclusive Other
Other Substances/Meds: meds / substances
Yes

* dose reported in total daily alkaloids

yo = year old
** Addiction: Yes = diagnosis or self-reported addiction to the product
*** Psychiatric history: Yes = recorded psychiatric history
Other substances/meds: Yes = patient taking other substances or medications known to cause psychiatric symptoms

163

Table 22. Evidence table of case reports Psychiatric (continued)

Event type
Age, Sex
Constituent
Source
(ID)
Hallucinations
Age Not described,
Sex Not described
Ephedra
FDA Case (12368)
Psychosis, Severe
depression, Seizure
32 YO Female
Ephedra
FDA Case (14105)
Psychosis, Seizure,
Transient ischemic
attack
37 YO Female
Ephedra
FDA Case (14615)
Suicide attempt
15 YO Female
Ephedra
FDA Case (10378)
Suicide attempt,
Headache
16 YO Male
Ephedra
FDA Case (13331)
Severe depression,
Suicide attempt,
Addiction/Substance
Abuse
25 YO Male
Ephedrine
FDA Case (11103)
Mania or severe
agitation
50 YO Female
Ephedra
FDA Case (11780 )

Product
Dose*
Duration
Addiction Data**
Power Trim
Not described
Not described
Addiction: No

Investigation for
Etiology***
RAND Classification
Psychiatric History: No
Insufficient Information
Other Substances/Meds:
No

Metab-O-Lite
Not described
60 days to 1 year
Addiction: No

Psychiatric History: No
Insufficient Information
Other Substances/Meds:
No

Metab-O-Lite
Not described
60 days to 1 year
Addiction: No

Psychiatric History: No
Insufficient Information
Other Substances/Meds:
No

Thermogenic Fat Burner

Not described
<48 hours
Addiction: No
Metabolife 356
Not described
<48 hours
Addiction: No

Psychiatric History: No
Product taken solely
Other Substances/Meds: as suicide attempt
No

Ephedrine
2500 mg
Over 1 year
Addiction: Yes

Psychiatric History: Yes Inconclusive Prior

Other Substances/Meds: psychiatric history
Yes

Psychiatric History: No
Product taken solely
Other Substances/Meds: as suicide attempt
No

Ma Huang/Ephedra + Caffeine Psychiatric History: Yes Inconclusive Prior

5.6 mg
Other Substances/Meds: psychiatric history
14-60 days (acute)
Yes
Addiction: No

* dose reported in total daily alkaloids

yo = year old
** Addiction: Yes = diagnosis or self-reported addiction to the product
*** Psychiatric history: Yes = recorded psychiatric history
Other substances/meds: Yes = patient taking other substances or medications known to cause psychiatric symptoms

164

Table 22. Evidence table of case reports Psychiatric (continued)

Event type
Age, Sex
Constituent
Source
(ID)
Mania or severe
agitation, Violent,
Addiction/Substance
Abuse
30 YO Male
Ephedrine
FDA Case (185564)
Mania or severe
agitation,
Encephalopathy,
Rhabdomyolysis,
Hyperthermia
28 YO Female
Ephedrine
Literature Case (69)
Psychosis, Severe
depression,
Hallucinations
54 YO Female
Ephedrine
Literature Case (120)
Mania or severe
agitation,
Confusion/Delusional,
Insomnia, Palpitations
21 YO Male
Ephedrine
Literature Case (157)
Psychosis,
Hallucinations,
Confusion/Delusional,
Violent
26 YO Male
Ephedrine
Literature Case (238)
Psychosis, Mania or
severe agitation,
Hallucinations,
Confusion/Delusional,
Violent
26 YO Male
Ephedrine
Literature Case (238)

Product
Dose*
Duration
Addiction Data**
Mini Thin
Not described
Over 1 year
Addiction: Yes

Investigation for
Etiology***
RAND Classification
Psychiatric History: Yes Inconclusive Prior
Other Substances/Meds: psychiatric history
Yes

Do-Do Tablet or Herbal Balance Psychiatric History: Yes Inconclusive Prior

18.31 mg
Other Substances/Meds: psychiatric history
<48 hours
Yes
Addiction: No

Ephedrine
Not described
Over 1 year
Addiction: Yes

Psychiatric History: Yes Inconclusive Prior

Other Substances/Meds: psychiatric history
Yes

Black Beauty
Not described
<48 hours
Addiction: No

Psychiatric History: Yes Inconclusive Prior

Other Substances/Meds: psychiatric history
Yes

Ephedrine
Not described
2-13 days
Addiction: No

Psychiatric History: Yes Inconclusive Prior

Other Substances/Meds: psychiatric history
No

Ephedrine
300 mg
14-60 days (acute)
Addiction: No

Psychiatric History: Yes Inconclusive Prior

Other Substances/Meds: psychiatric history
No

* dose reported in total daily alkaloids

yo = year old
** Addiction: Yes = diagnosis or self-reported addiction to the product
*** Psychiatric history: Yes = recorded psychiatric history
Other substances/meds: Yes = patient taking other substances or medications known to cause psychiatric symptoms

165

Table 22. Evidence table of case reports Psychiatric (continued)

Event type
Age, Sex
Product
Constituent
Dose*
Source
Duration
(ID)
Addiction Data**
Suicide attempt,
Product Not described
Ventricular tachycardia 22500 gm
/ fibrillation
<48 hours
20 YO Female
Addiction: No
Ephedrine
Literature Case (250)
Psychosis,
Vicks inhaler
Hallucinations, Sleep Not described
disturbance,
Over 1 year
Confusion/Delusional Addiction: No
61 YO Male
Ephedrine
Literature Case (488)
Psychosis,
Ephedrine
Addiction/Substance Not described
Abuse
Not described
35 YO Male
Addiction: Yes
Ephedrine
FDA Case (130741)
Psychosis, Mania or Marax
severe agitation
125 mg
19 YO Male
14-60 days (acute)
Ephedrine
Addiction: No
Literature Case (490)
Psychosis, Paranoia, Ephedrine
Violent, Impotence
Not described
65 YO Male
Over 1 year
Ephedrine
Addiction: Yes
Literature Case (120)
Psychosis, Mania or Thermolift
severe agitation,
Not described
Severe depression,
14-60 days (acute)
Hallucinations
Addiction: No
27 YO Female
Ephedrine
FDA Case (14542)
Psychosis,
Mini Thin
Suicide attempt,
Not described
Paranoia,
14-60 days (acute)
Confusion/Delusional Addiction: Yes
46 YO Male
Ephedrine
FDA Case (94799)

Investigation for
Etiology***
RAND Classification
Psychiatric History: Yes Inconclusive Prior
Other Substances/Meds: psychiatric history
No

Psychiatric History: Yes Inconclusive Prior

Other Substances/Meds: psychiatric history
No

Psychiatric History: No
Inconclusive Other
Other Substances/Meds: substances involved
Yes

Psychiatric History: No
Inconclusive Other
Other Substances/Meds: meds / substances
Yes

Psychiatric History: No
Inconclusive Other
Other Substances/Meds: substances involved
Yes

Psychiatric History: No
Not related
Other Substances/Meds: exacerbation of
No
previously
undiagnosed bipolar
disorder

Psychiatric History: No
Product taken solely
Other Substances/Meds: as suicide attempt
No

* dose reported in total daily alkaloids

yo = year old
** Addiction: Yes = diagnosis or self-reported addiction to the product
*** Psychiatric history: Yes = recorded psychiatric history
Other substances/meds: Yes = patient taking other substances or medications known to cause psychiatric symptoms

166

Table 22. Evidence table of case reports Psychiatric (continued)

Event type
Age, Sex
Constituent
Source
(ID)
Suicide attempt
19 YO Male
Ephedrine
FDA Case (1454817)
Suicide attempt,
Arrhythmia (NOS)
14 YO Female
Ephedrine
Literature Case (281)
Psychosis, Severe
depression,
Palpitations
40 YO Male
Ephedrine
FDA Case (1761109)
Psychosis, Severe
depression,
Hallucinations,
Addiction/Substance
Abuse
38 YO Female
Ephedrine
FDA Case (1834206)
Severe depression,
Suicide/Suicide
attempt
43 YO Female
Ephedra/Ephedrine
FDA Case (9568)
Mania or severe
agitation,
Confusion/Delusional,
Jumped out of car,
Cardiac Enlargement
47 YO Female
Ephedra
FDA Case (12486)
Psychosis,
Hallucinations,
Rhabdomyolysis
54 YO Male
Ephedra/Ephedrine
FDA Case (10894)

Product
Dose*
Duration
Addiction Data**
Max Alert
Not described
Not described
Addiction: No
RJ8
Not described
<48 hours
Addiction: No

Investigation for
Etiology***
RAND Classification
Psychiatric History: No
Product taken solely
Other Substances/Meds: as suicide attempt
No
Psychiatric History: No
Product taken solely
Other Substances/Meds: as suicide attempt
No

Max Alert
Not described
60 days to 1 year
Addiction: No

Psychiatric History: No
Insufficient Information
Other Substances/Meds:
No

Mini Thin
Not described
Not described
Excel Energy
Not described
Not described
Addiction: Yes

Psychiatric History: No
Insufficient Information
Other Substances/Meds:
No

Eola Amp Pro Drops

38 mg
14-60 days (acute)
Addiction: No

Psychiatric History: Yes Product removed from

Other Substances/Meds: market- Contained
Yes
illegal doses of
ephedrine

LiquiThin
Not described
Not described
Eola Amp Pro Drops
97.2 mg
2-13 days
Addiction: No

Psychiatric History: Yes Product removed from

Other Substances/Meds: market- Contained
Yes
illegal doses of
ephedrine

Eola Amp Pro Drops

Not described
14-60 days (acute)
Addiction: No

Psychiatric History: No
Product removed from
Other Substances/Meds: market- Contained
Yes
illegal doses of
ephedrine

* dose reported in total daily alkaloids

yo = year old
** Addiction: Yes = diagnosis or self-reported addiction to the product
*** Psychiatric history: Yes = recorded psychiatric history
Other substances/meds: Yes = patient taking other substances or medications known to cause psychiatric symptoms

167

Table 22. Evidence table of case reports Psychiatric (continued)

Event type
Age, Sex
Constituent
Source
(ID)
Psychosis,
Hallucinations,
Paranoia,
Confusion/Delusional
54 YO Female
Ephedra/Ephedrine
Literature Case (275)

Product
Dose*
Duration
Addiction Data**
Eola Amp Pro Drops
28000 mg
Over 1 year
Addiction: No

Investigation for
Etiology***
Psychiatric History: No
Other Substances/Meds:
Yes

RAND Classification
Product removed from
market- Contained
illegal doses of
ephedrine

* dose reported in total daily alkaloids

yo = year old
** Addiction: Yes = diagnosis or self-reported addiction to the product
*** Psychiatric history: Yes = recorded psychiatric history
Other substances/meds: Yes = patient taking other substances or medications known to cause psychiatric symptoms

168

Table 23. Summary of adverse events with ephedra consumption*

Type of
Demographics Event
Total Events
Sentinel Events
Possible Sentinel Events
Events by Sex
Female
Sentinel Events
Possible Sentinel Events
Male
Sentinel Events
Possible Sentinel Events
Events by Age
1330
Sentinel Events
Possible Sentinel Events
3150
Sentinel Events
Possible Sentinel Events
5170
Sentinel Events
Possible Sentinel Events

Death

MI

Other
Cardiac

CVA /
Stroke

Other
Neurological

Seizure

Psychiatric
Symptoms

2
9

3
7

0
2

9
10

0
1

3
7

5
7

1
3

0
4

0
0

5
7

0
1

3
5

2
4

1
6

3
3

0
2

4
3

0
0

0
2

3
3

2
5

2
0

0
1

3
2

0
0

2
3

3
5

0
4

1
6

0
1

5
5

0
0

1
3

2
1

0
0

0
1

0
0

1
3

0
1

0
1

0
1

* includes three events from Metabolife analysis.

169

Table 24. Summary of adverse events with ephedrine consumption

Type of
Demographics Event
Total Events
Sentinel Events
Possible Sentinel Events
Events by Sex
Female
Sentinel Events
Possible Sentinel Events
Male
Sentinel Events
Possible Sentinel Events
Events by Age
1330
Sentinel Events
Possible Sentinel Events
3150
Sentinel Events
Possible Sentinel Events
5170
Sentinel Events
Possible Sentinel Events

Death

MI

Other
Cardiac

CVA /
Stroke

Other
Neurological

Seizure

Psychiatric
Symptoms

3
3

2
0

0
1

2
2

0
0

1
0

3
1

1
1

2
0

0
1

2
0

0
0

1
0

1
0

2
1

0
0

0
1

0
2

0
0

0
0

2
1

2
0

1
0

0
1

2
1

0
0

0
0

2
0

1
1

1
0

0
1

0
1

0
0

1
0

0
1

0
1

0
0

0
0

0
0

0
0

0
0

1
0

170

Table 25. Summary of adverse events not reviewed in detail

Number of Events
Reported
39
45
51
7
7
3
1
1

Adverse Event
Fainting/ loss of consciousness
Heart rate >120 or <50
Hypertension, systolic >180 or diastolic >120
Paralysis
Liver failure,ALT/AST >200
Rhabdomyolysis, CPK >400
Coma
Miscarriage

171

Table 26. Summary data of key variables from Metabolife file analysis
Age
10
1120
2130
3140
4150
5160
>60
No Data
Average Age

N
5
340
2163
2369
1598
912
343
10627

%
<1
2
12
13
9
5
2
57

38

Gender
N
%
Male
707
4
Female
6792
36*
No Data
11032
30
*91 percent of files with reported gender are female.
Adverse Event
No adverse event reported
Death
Cardiovascular: Heart rate, >120 or <50
Cardiovascular: Heart rate, 50-120, or not otherwise unspecified
Cardiovascular: Hypertension, Systolic>180 or Diastolic>105
Cardiovascular: Hypertension, Systolic<180 or Diastolic<105, not otherwise specified
Cardiovascular: Myocardial Infarction/ Heart Attack
Cardiovascular: Cardiac Dysrythmia, Other/ Palpitations
Cardiovascular: Cardiac arrest
Cardiovascular: Ventricular Tachycardia/ Fibrillation
Cardiovascular: Chest Pain, not specified as MI
Pulmonary: Respiratory arrest
Neurological: Transient Ischemic Attack
Neurological: CVA/ Stroke, not known to be hemorrhage
Neurological: Brain Hemorrhage
Neurological: Fainting / Loss of consciousness
Neurological: Coma
Neurological: Seizure
Psychiatric: Depression
Psychiatric: Hallucinations
Psychiatric: Mania or severe agitation
Psychiatric: Psychosis
Psychiatric: Suicide attempt
Autonomic Hyperactivity (Includes: Tremor, twitching, jitteriness, insomnia, increased
sweating, agitation, nervousness, and irritability)
Changes in glucose <40 or >400
Liver failure ALT/AST >200
Liver abnormality, not otherwise specified
Rhabdomyolysis CPK >400
Rhabdomyolysis, not otherwise specified
Miscarriage

172

N
2019
3
23
584
45
405
22
630
3
0
582
2
5
29
2
41
0
46
57
2
1
3
0
2536

%
11
<1
<1
3
<1
2
<1
3
<1
0
3
<1
<1
<1
<1
<1
0
<1
<1
<1
<1
<1
0
14

56
5
46
1
0
6

<1
<1
<1
<1
0
<1

Table 26. Summary data of key variables from Metabolife file analysis (continued)
Adverse Event
Allergic Reaction
Anesthesia complication
Fatigue/Fever/ Chills
Abnormal lab values, not otherwise specified
Ear, Eye, Nose, or Throat
Respiratory System
Cardiovascular System
Gastrointestinal System
Hepatobiliary System
Musculoskeletal System
Genitourinary System
Gynecologic (includes breast and menstrual symptoms)
Sexual Dysfunction
Neurological System (includes headache)
Mental Health
Skin (Includes Pruritis)
Hematologic System
Oncologic System
Other symptoms not specified above

173

N
614
2
724
216
795
374
255
4680
25
1136
395
1009
115
2475
462
1385
126
4
396

%
3
<1
4
1
4
2
1
26
<1
6
2
5
1
13
3
7
1
<1
2

Table 27. Comparison of serious cases identified by RAND and by Metabolife

RAND #
DEATH
23695

Metabolife #

Explanation

No #
No #

Not on our MIPER CD-ROM

Not on our MIPER CD-ROM
Only notation is migraine HA, wanted refund (sisters husb died).
Unclear if this death is the consumer or a relative

35062
35062
MYOCARDIAL INFARCTION/ HEART ATTACK
16006
16006
17002
17002
20416
20416
20918
20918
21010
the man who was taking them [Metabolife 356] has now suffered a
heart attack
22492
28 yrs old had a heart attack
22584
customer had a heart attack thinks it was Met
22779
heart attack, gall bladder surgery, cholecystectomy
23877
13 heart att, 3 strokes
24166
24166
24236
24236
24383
cold sweat ht attack
24448
24448
24859
24859
27941
27941
28168
28168
28488
28488
28835
28835
35532
Not on our MIPER CD-ROM
CARDIAC ARREST
15409
15409
27600
27600
35063
35063
STROKE
16593
16593
17196
vision disturbance
18199
18199
19474
short of breath, tachycardia
20763
20763
22308
pain in chest, took NTG, stood up, ?stroke, ?side wont move, CAT
scan negative
22325
had ministroke$50 refundwill see neurologist
22479
legal customer that had 2 strokes lawyer
22496
BP and Premarin caffeine stroke
23002
23002
23663
stroke that cousin suffered
23877
13 heart att, 3 strokes
24825
24825
24945
24945
25011
stroke written on note, but remainder of notes are about skin and
gastrointestinal symptoms.

174

Table 27. Comparison of serious cases identified by RAND and by Metabolife (continued)
RAND #
25147
25482
25495
25521
27791

Metabolife #
25482
25495
25521

28156
28157
28201
28281
28281
28321
28321
29424
29424
29469
29469
30391
30391
30407
30407
BRAIN HEMORRHAGE
27754
35062
SEIZURE
15281
15281
15345
15345
16461
16461
16653
16653
16703
16703
16897
16897
16970
16970
17369
17369
17752
18335
18962
19149
20812
20864
20979

18335
18962
19149
20812
20864
20979
22150
22238

22364
22539
22800
23029

22800
23029
23440

23468
24172
24344
24482
24711
24839

Explanation
client had a stroke

24209
24344
24482
24711
24839

wife 1997 had stroke

mild stroke symptoms
facial numbness
muscle weakness

brain bleeding?
Recorded by Metabolife under death

[redacted] and her sister both take Met [redacted] reports [redacted]
had a seizure recently

Definitely a seizure, but contained in the section of the main file that
has refund requests and we inferred these cases were also recorded
elsewhere in the MIPER file
black out while driving had hot flashes also had 2 screwdrivers
seizures like activity
friend of a friend had seizure
s/es dad
sister grand mal seizure
seizure
Same case as 16897

175

Table 27. Comparison of serious cases identified by RAND and by Metabolife (continued)
RAND #
24947
25371
27487
27523
28183
28329
28442
29882

Metabolife #

Explanation
seizures

25371
27487
27523
28183
28329
28442
35568

seizure checked off on list of symptoms on standardized form

Not on our MIPER CD-ROM

176

Table 28. Summary of Metabolife medical records

RAND
Case #
1

Index
Case #
1

Complaint
Case #
1

MIPER#(s)

Notes

20867
20868

20871
20872

16287
20873-75
21033
24047
24051

20876-78

This is a 42-year-old male who took two Metabolife pills

for the first time and presented with chest pain, chesttightness and shortness of breath. He ended up in the
emergency department where he was found to have a
blood pressure of 140/82 with a pulse of 111. The
electrocardiogram showed him to be in atrial fibrillation.
A discharge summary is not included among the records
received. However, the patients note to Metabolife said
that he was discharged after one day and that his
doctors were convinced that his heart was back to
normal. Of note is that his laboratory values established
that he did not have thyroid disease and did not have
any evidence of a myocardial infarction.
This is a 28-year-old female who had shortness of
breath, dyspnea and wheezing. She was seen in the
emergency department and was said to be having an
anaphylactoid reaction. She was treated with
epinephrine, steroids and Benadryl with a complete
response.
This is a 38-year-old female who was admitted to the
hospital with acute pancreatitis. The hospital record
notes that she is quite obese. The record also notes
that she had a prior total abdominal hysterectomy with
bilateral salpingo-oopherectomy and at that time was
found to have ovarian cancer with involvement of the
bowel. This resulted in partial colectomy with a diverting
colostomy, and subsequently she had a renastomosis.
She also had a prior cholycystectomy. It is noted that
she did not drink alcohol. Her admission records note an
elevated white blood cell count with a value of 14,000
but no elevation in amylase or lipase. A subsequent
note states that these laboratory tests did become
elevated and then returned to normal. She was
discharged after recovery. Actual laboratory values are
not included with the records. There is no mention of a
measurement of serum triglycerides.
This case consists of a handwritten note from the patient
and a medical care bill for $34. The age and gender of
the patient are unknown. The complaint is of headache,
dizziness and tingling.

Index Case # taken from Index of Redacted Consumer Medical Records with Corresponding MIPER Numbers.
Complaint Case # taken from Listing of Key Complaint for the Metabolife Medical Records Submitted .
MIPER #(s) taken from Index of Redacted Consumer Medical Records with Corresponding MIPER Numbers.
No MIPER located: this is the text from the Index of Redacted Consumer Medical Records as it pertains to the Index Case #.
n/a: not available, no match found.

177

Table 28. Summary of Metabolife medical records (continued)

RAND
Case #
5

Index
Case #
5

Complaint
Case #
5

MIPER#(s)

Notes

17895
20879
23365

20880
20883-85

15998
20886

16166
20887
24083-84

20888-89

10

10

10

20890

This file consists of a single physician note of a female

of unstated age who came in with the complaints of
pain over the joints, gums would bleed, veins seemed
to be thrombosing, some itching, easy bruisability and
pain in the back over her kidneys. The physical
examination was normal, clotting studies were normal,
sedimentation rate was seven, chemistry panel was
normal. The patient left in good condition.
This is a 28-year-old female who presented with 2
weeks of stomach pain, mostly after eating food, along
with explosive diarrhea. Her laboratory work-up was
essentially normal with a normal white blood cell count,
liver enzymes and amylase. She was diagnosed as
having acute gastritis. She had both upper and lower
endoscopy that did not reveal a clear diagnosis. Stool
for ova and parasites was negative. Stool was positive
for occult blood. Stool culture was negative, abdominal
series was negative.
This is a 53-year-old female who presented with emesis
and diarrhea after eating a hamburger at a fast food
restaurant. Her examination was essentially
unremarkable. The diagnostic impression was acute
gastroenteritis. She was treated with antibiotics and
Kaopectate.
This is a 61-year-old female with a history of asthma
who presented with headaches. She was found to have
a potassium of 3.3 and a sodium of 118. It was noted
that she drinks eight glasses of water a day. The
diagnosis given was headache, possibly due to low
sodium, and a viral upper respiratory tract infection.
This is a 53-year-old female who was seen for increased
intraoccular pressure. She was under the care of an
ophthalmologist for what she called the
iridocorneoendothelial syndrome. It was treated by her
ophthalmologist.
This is a female of unstated age who presented with a
headache. The records note that she had migraines
eight years ago. She had photophobia and emesis. Her
blood pressure was 153/73. She was treated with
Imitrex with mild relief and she also received Demerol
and Phenergan.

Index Case # taken from Index of Redacted Consumer Medical Records with Corresponding MIPER Numbers.
Complaint Case # taken from Listing of Key Complaint for the Metabolife Medical Records Submitted .
MIPER #(s) taken from Index of Redacted Consumer Medical Records with Corresponding MIPER Numbers.
No MIPER located: this is the text from the Index of Redacted Consumer Medical Records as it pertains to the Index Case #.
n/a: not available, no match found.

178

Table 28. Summary of Metabolife medical records (continued)

RAND
Case #
11

Index
Case #
11

Complaint
Case #
11

MIPER#(s)

Notes

No MIPER
located

This is a male of unstated age who presented with

abdominal indigestion without vomiting. The records
note the patient had a prior vagotomy and pyloroplasty
with gastroenterostomy. He received an ultrasound, CT
scan and endoscopy. There was no indication of his
treatment or response. In addition, there is no mention
of taking Metabolife anywhere in the medical records.
This is a 53-year-old female. The complaint that is listed
is hyponatremia. However there is no medical record
documentation of this. All that is included is a single
copy of lab tests showing normal thyroid function and
normal complete blood count. Whether these data apply
to this patient is unclear, as the patient age on the lab
slip is listed as 33. A doctors note in the MIPER file
states she required hospitalization.
This is a 61-year-old female who presented with
supraventricular tachycardia which required
cardioversion and subsequent treatment with atenolol.
This is documented in a note, possibly from her doctor,
however there are no medical records included with this
case.
This is a 60-year-old female who presented with
palpitations and was found to be in atrial flutter.
According to the discharge summary, she was
electrically cardioverted and then given Digoxin and
Cardizem. Subsequent clinic notes showed her to
continue to be in sinus rhythm. There is no indication or
records that other diagnostic studies were done.
This is a 21-year-old female. The complaint is a rash.
The only documentation provided is the bill of an
emergency department visit. There are no medical
records.
This is a 49-year-old female, noted to weigh 160 lbs.,
who presented with chest pain and had an overnight
hospitalization to evaluate myocardial infarction. CPK
was elevated but the MB fractions were negative. The
patient was discharged with the diagnosis of chest wall
pain.

12

12

13

16995
20892
25503

13

13

14

15996
20893-95
23828
21035-37

14

n/a

12

n/a

15

14

15

16

15

n/a

16642
20897-99
21034
23859
17569
20900-01

n/a

16

n/a

15351
23010

No medical record received

Index Case # taken from Index of Redacted Consumer Medical Records with Corresponding MIPER Numbers.
Complaint Case # taken from Listing of Key Complaint for the Metabolife Medical Records Submitted .
MIPER #(s) taken from Index of Redacted Consumer Medical Records with Corresponding MIPER Numbers.
No MIPER located: this is the text from the Index of Redacted Consumer Medical Records as it pertains to the Index Case #.
n/a: not available, no match found.

179

Table 28. Summary of Metabolife medical records (continued)

RAND
Case #
17

Index
Case #
17

Complaint
Case #
16

MIPER#(s)

Notes

17605
20904

This is a 36-year-old female who claims to be only 20

lbs. overweight who stated that she had elevations in
blood pressure, now requiring treatment with Maxzide.
However there are no medical records accompanying
this complaint, only a copy of bills.
(Listing of Key Complaints states chest pain, shortness
of breath)
(Listing of Key Complaints states elevated blood
pressure/ racing pulse)
This is a 73-year-old female who was evaluated for near
syncope that occurred while eating in a restaurant. In
the emergency department, blood pressure was noted
to be 132/37 with a pulse of 64 and glucose was 90. The
discharge diagnosis was syncope related to
hypoglycemia vs. Metabolife vs. vasovagal episode.
Exercise treadmill test performed later was normal but
there was a submaximal heart rate achieved. Carotid
ultrasound was normal. Many additional notes cover
healthcare judged to be irrelevant to the use of ephedra,
including a podiatry consult, breast biopsies, pap smear
and an endometrial biopsy.
This is a 24-year-old female who presented with blood in
the urine for one day. The records consist of a urine
culture which was negative, a urinalysis which showed
2+ blood and a hemoglobin and hematocrit of 18 and
50, respectively.
This is a 47-year-old male who presented in atrial
fibrillilation, was shown not to have had a myocardial
infarction, and who had an echo and exercise treadmill
test that were both normal. There was no evidence of
thyroid disease. The patient converted to sinus rhythm
with medication and was then treated with Digoxin. A
followup doctors note stated that the patient was in
sinus rhythm and implied that he was off Digoxin.
This is a 31-year-old female who is noted to weigh 261
lbs. She presented with heart palpitations, shortness of
breath and heart flutter. She had a history of
hypertension with pregnancy. A consultants note
reported T-wave inversions in V1 and V3 with an
elevated CPK but the MB fraction was normal and the
Troponin test was negative. It is unclear exactly what
happened, but this apparently resolved.

n/a

n/a

17

n/a

n/a

n/a

18

n/a

18

18

19 (?)
(Listing of
Key
Complaints
states
Fainting)

16199

19

19

20

No MIPER
located

20

20

21

20905-06
25529

21

21

22

17028
20907-08
24154

Index Case # taken from Index of Redacted Consumer Medical Records with Corresponding MIPER Numbers.
Complaint Case # taken from Listing of Key Complaint for the Metabolife Medical Records Submitted .
MIPER #(s) taken from Index of Redacted Consumer Medical Records with Corresponding MIPER Numbers.
No MIPER located: this is the text from the Index of Redacted Consumer Medical Records as it pertains to the Index Case #.
n/a: not available, no match found.

180

Table 28. Summary of Metabolife medical records (continued)

RAND
Case #
22

Index
Case #
22

Complaint
Case #
23

MIPER#(s)

Notes

17277
20914-15

This is a 36-year-old female who presented with

nausea, dizziness and vomiting, headache and
abdominal pain. Blood pressure was noted to be 134/87
and the pulse was 85. Abdominal ultrasound was
normal, pregnancy test was normal. Urinalysis showed
moderate ketones. The discharge diagnosis was
abdominal pain of uncertain etiology.
No medical record received

n/a

23

n/a

n/a

n/a

24

23

24

25

20918-21
21032

24

25

26

20950
20953-54
20958-59
20961

25

26

27

26

27

28

20962-66
21006-07
18445

n/a

28

n/a

27

29

29

22408
20916-17
n/a

16521
17536
20967-68

(Listing of Key Complaints states difficulty breathing/

anxiety)
This is a 38-year-old female who made three visits to
the emergency room over four days for epigastric and
chest pain, initially being diagnosed as having
esophageal reflux, then gastritis and then finally being
recognized as having coronary artery disease with an
80% left anterior descending stenosis. This was treated
with a coronary stent. Her cardiologist notes that she
had a very positive family history of coronary artery
disease and that her mother had an early heart attack.
There was no indication in the record that a cholesterol
test was done.
This file contains no medical records, only medical bills
documenting prescriptions for hydrochlorothiazide and
phenazopyridine, along with a urinalysis. The MIPER file
indicates the patient said she was diagnosed with
hemorrhagic cystitis, and later hypertension.
There are no medical records in this file, only bills. On
one of the bills is written drug reaction.
This is a 39-year-old female. The complaint is an allergic
reaction. There are no medical records in this file, only
bills.
No medical record received
This is a 40-year-old female who developed transient
elevations of liver enzymes with an ALT of 125. Albumin
and bilirubin were normal. Multiple tests for possible
etiologies of this were performed, all of which were
negative. Metabolife was discontinued and the liver
function abnormalities drifted down to normal over time;
the last note said that she had recovered totally.

Index Case # taken from Index of Redacted Consumer Medical Records with Corresponding MIPER Numbers.
Complaint Case # taken from Listing of Key Complaint for the Metabolife Medical Records Submitted .
MIPER #(s) taken from Index of Redacted Consumer Medical Records with Corresponding MIPER Numbers.
No MIPER located: this is the text from the Index of Redacted Consumer Medical Records as it pertains to the Index Case #.
n/a: not available, no match found.

181

Table 28. Summary of Metabolife medical records (continued)

RAND
Case #
28

Index
Case #
30

Complaint
Case #
30

MIPER#(s)

Notes

20969
20971-75
20977-78

There are no medical records in this file. The only thing

that is listed is a complaint from the patient about a
heart rate being 188 and the blood pressure being high,
that the patient was treated in the emergency room and
that there were blood tests to assess heart damage.
The MIPER includes a long letter from the patient that
relates much the same thing.
No medical record received

n/a

32

n/a

n/a

33

n/a

n/a

n/a

32

No MIPER
located
No MIPER
located
n/a

n/a

n/a

33

n/a

29

35

35

16376
21030

30

34

34

21027
21029

31

31

31

21000-01

No medical record received

Listing of Key Complaints states nothing identified- just
a bill
Listing of Key Complaints states nothing identified- list of
medications
This is a 54-year-old male with chest pain and a
headache who also complained of high blood pressure
and lightheadedness. There are minimal records
associated with this report of August 11, 1999, other
than that the patient was diagnosed with accelerated
hypertension. Of note, however, is that there are
numerous clinic visit notes dating back to 1997,
documenting that the patient had a history of
hypertension, diabetes and hyperlipidemia, with a blood
pressure on one occasion 154/92. It is noted that this
was taken with a large cuff. In addition, there are clinic
visits with chest pain as far back as 1997.
This is a male of unstated age, possibly 40 years old,
who wrote a note saying that he had stomach problems,
kidney stones, colon problems and anxiety problems.
There are no medical records associated with this file,
only bills.
There are no medical records with this file, only some
discharge instructions that say that the diagnosis was
acute nausea. The MIPER file states the patient is a
37-year-old female and that hypoglycemia was likely.

Index Case # taken from Index of Redacted Consumer Medical Records with Corresponding MIPER Numbers.
Complaint Case # taken from Listing of Key Complaint for the Metabolife Medical Records Submitted .
MIPER #(s) taken from Index of Redacted Consumer Medical Records with Corresponding MIPER Numbers.
No MIPER located: this is the text from the Index of Redacted Consumer Medical Records as it pertains to the Index Case #.
n/a: not available, no match found.

182

Table 28. Summary of Metabolife medical records (continued)

RAND
Case #
32

Index
Case #
36

Complaint
Case #
36

MIPER#(s)

Notes

20979
24840
25498
25501

This is a 50-year-old female who had a witnessed grand

mal seizure while driving. Later that day, after
undergoing a CT scan in the emergency room, she had
a 2nd witnessed seizure and, according to an attorneys
letter, she then had a 3rd seizure at some point. An
evaluation included a CT of the brain, which was
normal, and an electroencephalogram, which was also
normal. She had no history of alcoholism. Serum
sodium was normal and glucose was normal. Pulse
oximetry was 99%. Toxicology screen was positive for
amphetamines. There was no prior history of seizure
disorder or neurologic disease.
This is a 21-year-old female who is noted to weigh 200
pounds and on whom the MIPER file will say shortness
of breath and tachycardia. There are minimal records
associated with this, only a discharge diagnosis of
hyperventilation, with a notation saying that a friend died
two days ago. There is a listing of medications and, by
implication, these are being taken by the patient. These
are Darvocet (which may have been discontinued),
Flexeril, Reglan, Cytotec, Dicyclomine, Viokase,
Sudafed, Lopid, Citracel, Pariodel, Benadryl, DDVAP,
Zantac, Trilisat, Carafate. Of note, the MIPER may also
say the complaint includes aphasia, paralysis, and
shortness of breath.
There are no medical records in this file, only a bill.
This is a 39-year-old female who had the complaint of
abdominal pain. The medical records submitted with this
consist of a clinic note which says that the patient has
classic gastrointestinal illness with mild nausea and no
diarrhea, progressing to diarrhea with no vomiting.
This is a 29-year-old female who had a witnessed tonic
clonic seizure. There is no history of alcoholism. The
grandmother had a history of seizures but was also
noted to be an alcoholic. Blood pressure was normal at
120/80. Brain MRI was normal. EEG was normal.
Toxicology screen by report had large amount
ephedrine and pseudoephedrines.
No medical record received
(Listing of Key Complaints states lower back pain/GI)
This is a patient of unknown age and unknown gender
who presented for an allergic reaction. There are no
medical records and only bills and medications in this
file.

33

37

37

19473
23970

34
35

n/a
40

39
n/a

n/a
19350

36

38

38

20864-66

n/a
n/a
37

39
n/a
n/a

n/a
40
41

24495
n/a
n/a

Index Case # taken from Index of Redacted Consumer Medical Records with Corresponding MIPER Numbers.
Complaint Case # taken from Listing of Key Complaint for the Metabolife Medical Records Submitted .
MIPER #(s) taken from Index of Redacted Consumer Medical Records with Corresponding MIPER Numbers.
No MIPER located: this is the text from the Index of Redacted Consumer Medical Records as it pertains to the Index Case #.
n/a: not available, no match found.

183

Table 28. Summary of Metabolife medical records (continued)

RAND
Case #
n/a

Index
Case #
41

Complaint
Case #
n/a

MIPER#(s)

Notes
No medical record received

42

No MIPER
located
n/a

38

n/a

39

42

43

19604

40

n/a

n/a

41

n/a

44 (?)
(Listing of
Key
Complaints
states
intracranial
hemorrhage,
which is
mentioned
in patient
history)
45

n/a

42

n/a

46

n/a

43

n/a

n/a

n/a

There are only bills which have a diagnostic code 780.2

which is syncope and collapse, along with indications
that an echocardiogram and duplex sonography were
done. There are no other medical records.
This is a 29-year-old female, who is noted to weigh 230
lbs., who presented for menstrual irregularity, numbness
and tingling. Evaluation was unremarkable and no
diagnosis was given.
This is a 36-year-old female who complained of
menstrual irregularity. The records document that she
recently had a right posterior parietal intracranial
hemorrhage with extension into the ventricular system
requiring neurosurgery with a drain. This was
subsequently shown by angiography to be due to an
arteriovenous malformation, which was then
subsequently resected. After this neurosurgery she had
not had resumption of her menstrual period. In the notes
available there was no work up of this symptom.
This is a 26-year-old female, who is noted to weigh 155
lbs., who had chest pains after using Metabolife for two
months. She also had asthma and a brother who died of
myocardial infarction at age 33. Her discharge
diagnoses were asthma and chest pain.
This is a 27-year-old female who presented with sudden
abdominal pain which was found to be due to a rupture
of a splenic artery aneurysm which required emergency
laporatomy and resection. The records note she had a
history of congenital multiple ureters which had been
surgically repaired at age 7 and she was left with some
renal insufficiency as a result. There is no mention of the
use of Metabolife in the medical records that are
provided.
This is a 49-year-old male who had symptoms of chest
pressure and pain along with shortness of breath. He
had been a cigarette smoker but the record notes he
quit. He had an exercise treadmill test that showed a
normal electrocardiogram response but he had
scintigraphic evidence of ischemia. He underwent
coronary angiography that revealed normal coronary
arteries. Three months later he was continuing to have
unexplained chest pressure and pain.

Index Case # taken from Index of Redacted Consumer Medical Records with Corresponding MIPER Numbers.
Complaint Case # taken from Listing of Key Complaint for the Metabolife Medical Records Submitted .
MIPER #(s) taken from Index of Redacted Consumer Medical Records with Corresponding MIPER Numbers.
No MIPER located: this is the text from the Index of Redacted Consumer Medical Records as it pertains to the Index Case #.
n/a: not available, no match found.

184

Table 28. Summary of Metabolife medical records (continued)

RAND
Case #
n/a

Index
Case #
43

Complaint
Case #
n/a

n/a

44

n/a

n/a

45

n/a

n/a

46

n/a

MIPER#(s)

Notes

No MIPER
located
No MIPER
located
No MIPER
located
No MIPER
located

No medical record received

No medical record received
No medical record received
No medical record received.

Index Case # taken from Index of Redacted Consumer Medical Records with Corresponding MIPER Numbers.
Complaint Case # taken from Listing of Key Complaint for the Metabolife Medical Records Submitted .
MIPER #(s) taken from Index of Redacted Consumer Medical Records with Corresponding MIPER Numbers.
No MIPER located: this is the text from the Index of Redacted Consumer Medical Records as it pertains to the Index Case #.
n/a: not available, no match found.

185

Figure 1. Screening form for literature

RAND EPC EPHEDRA PROJECT

SCREENER FORM

1. Article ID: ______________________

Notes:

2. First Author: _____________________

(LAST NAME OF FIRST AUTHOR)
3. Reviewer: _______________________
4. Research topic:
CHECK ALL THAT APPLY
Ephedra .......................................
Ephedrine ....................................
Pseudoephedrine ........................ (STOP)
Unclear ........................................
Other ( _________________ ) .. (STOP)
5. Subject of article:
CHECK ALL THAT APPLY
Weight Loss .................................
Athletic Performance ...................
Adverse Events ...........................
Other ( _________________ ) .. (STOP)
6. Study population:
CHECK ALL THAT APPLY
Human .........................................
Animal.......................................... (STOP)
Unclear ........................................
Other (specify: ___________ ) .. (STOP)
7. Study design:
CHECK ALL THAT APPLY
Descriptive (historical, editorial etc.)
Review/meta-analysis.................
Randomized Clinical Trial............
Controlled Clinical Trial................
Case Series .................................
Case Report: medical literature ...
Case Report: popular literature ...
Other (specify: ___________ ) ..
8. Does the intervention contain caffeine or
caffeine-containing herbs?
CIRCLE ONE
Yes................................................ 1
No ................................................. 2
Unclear ......................................... 7
Not applicable ............................... 8
9. Language of article:
CIRCLE ONE
English .......................................... 1
Chinese......................................... 2
Japanese ...................................... 3
Other (specify: ___________ ) ... 4

41

Figure 2. Quality review form for literature

RAND EPC EPHEDRA PROJECT
Article ID:___________

QUALITY REVIEW FORM

Reviewer:_____________________________

First Author: ______________________________________________

(Last Name Only)
Study Number: ___of____Description:__________________________
(Enter 1 of 1 if only one) (If more than one study)
1. Design:
CIRCLE ONE
RCT.................................................................. 1
CCT.................................................................. 2
Other ................................................................ 3 (STOP)
(IF NOT RCT OR CCT, CHANGE STUDY DESIGN ON COVER SHEET AND STOP)
2. Were any adverse events mentioned?
CHECK ALL THAT APPLY
CHECK OR CODE
CHECK IF SERIOUS
Cardiovascular ................................ (01) ........................
Death............................................... (02) ........................
Endocrine ........................................ (03) ........................
Neurologic ....................................... (04) ........................
Psychiatric....................................... (05) ........................
Pulmonary ....................................... (06) ........................
Renal............................................... (07) ........................
Other: .............................................(___ ___, ____ ___, ____ ____)
No adverse events .......................... (96) ...........................
None mentioned.............................. (97) ...........................
Mentioned but not described........... (98) ...........................
3. For articles on weight loss, is there a follow up of at least 8 weeks?
CIRCLE ONE
Yes ................................................................... 1
No..................................................................... 2 (STOP)
Not applicable .................................................. 9
4. Is the study described as randomized?
CIRCLE ONE
Yes ................................................................... 1
No..................................................................... 2
5. If the study was randomized, was method of randomization appropriate?
CIRCLE ONE
Yes ................................................................... 1
No..................................................................... 2
Method not described ...................................... 8
Not applicable .................................................. 9

42

Figure 2. Quality review form for literature (continued)

RAND EPC EPHEDRA PROJECT

QUALITY REVIEW FORM

6. Is the study described as:

CIRCLE ONE
Double blind....................................................... 1
Single blind, patient ........................................... 2
Single blind, outcome assessment ................... 3
Open .................................................................. 4
Blinding not described ....................................... 8
Not applicable .................................................... 9
7. If reported, was the method of double blinding appropriate?
CIRCLE ONE
Yes..................................................................... 1
No ...................................................................... 2
Double blinding method not described .............. 8
Not applicable .................................................... 9
8. If study was randomized, did the method of randomization provide for
concealment of allocation?
CIRCLE ONE
Yes..................................................................... 1
No ...................................................................... 2
Concealment not described............................... 8
Not applicable .................................................... 9
9. Are withdrawals (W) and dropouts (D) described?
CIRCLE ONE
Yes, reason described for all W and D ............. 1
Yes, reason described for some W and D ........ 2
Not described .................................................... 8
Not applicable .................................................... 9
10. Is this a cross-over study design?
CIRCLE ONE
Yes..................................................................... 1
No ...................................................................... 2
Not described .................................................... 8
11. Are outcome data reported separately for or primarily on over 75% of any
of the following populations?
CHECK ALL THAT APPLY
Race:
African-Americans...................................... (01)
Hispanic ..................................................... (02)
Asian .......................................................... (03)
Gender:
Male ........................................................... (04)
Female ....................................................... (05)
Age:
Adolescents (12-17) ................................... (06)
Children (0-11) ........................................... (07)
Misc.:
Athletes ...................................................... (08)
Military........................................................ (09)
Other:
(Enter code: ____ ____, ____ ____, ____ ___, ____ ____)

43

Figure 2. Quality review form for literature (continued)

RAND EPC EPHEDRA PROJECT
12. What types of comorbidities are described in the groups?
CHECK ALL THAT APPLY
Overweight/ Obesity (BMI > 27) ....................... (01)
Coronary Artery Disease .................................. (02)
Hypertension..................................................... (03)
Neurological...................................................... (04)
Psychiatric ........................................................ (05)
Asthma.............................................................. (06)
Gastrointestinal................................................. (07)
Diabetes............................................................ (08)
Renal ................................................................ (09)
Other:
(Enter code: ____ ____, ____ ____, ____ ____, ____ ____)
Not described ................................................... (98)

44

QUALITY REVIEW FORM

Figure 2. Quality review form for literature (continued)

RAND EPC EPHEDRA PROJECT

QUALITY REVIEW FORM

Arm ___ of ___ Description ____________________

If the study has a control/usual care arm, enter that data in arm 1.
Otherwise, enter data for the groups in order of first mention.
13. What type of arm is this?
CIRCLE ONE
Placebo.............................................................1
Usual care.........................................................2
Primary intervention..........................................3
Other active treatment ......................................4
14. Is there a significant co-intervention?
CHECK ALL THAT APPLY OR ENTER CODE
Diet .................................................................. (01)
Exercise ........................................................... (02)
Education......................................................... (03)
Other: (enter code ___ ___, ___ ___, ___ ___)
No co-interventions.......................................... (97)
15. What was the sample size in this arm?
___ ___ ___ , ___ ___ ___
___ ___ ___ , ___ ___ ___
Entering
Completing
(ENTER 999,999 IF NOT REPORTED.)
16. What is the common, proprietary, and/or scientific (genus, genus/species)
name of the product?
ENTER CODE OR CIRCLE ONE OF THE BELOW
Code: ____ ____
None ................................................................ 97
Not described .................................................. 98
Not applicable .................................................. 99
17. Of which main constituents is the product made?
ENTER CODE OR CIRCLE ONE OF THE BELOW
Code: ____ ____, ____ ____, ____ ____
None .................................................................... 97
Not described ...................................................... 98
Not applicable ...................................................... 99
18. Was chemical analysis performed on ephedrine alkaloids?
CIRCLE ONE
Yes....................................................................... 1
No ........................................................................ 2
Not described ...................................................... 8
Not applicable ...................................................... 9

45

Figure 2. Quality review form for literature (continued)

RAND EPC EPHEDRA PROJECT

QUALITY REVIEW FORM

19. Intervention:
TOTAL DAILY AMOUNT
DOSE
PER DOSE
INTERVENTION

UNITS

ROUTE OF
ADMINISTRATION

DURATION

UNITS

EPHEDRINE
ALKALOIDS

1 ____

_____

_____

_____

_____

_____

____

_____

2 ____

_____

_____

_____

_____

_____

____

_____

3 ____

_____

_____

_____

_____

____

____

_____

4 ____

_____

_____

_____

_____

_____

____

_____

Enter a
number

1. Hour
2. Day
3. Week

1. Included in
total
ephedrine
alkaloids
2. In addition to
ephedrine
alkaloids
3. Unclear
8. ND
9. NA

Enter code

Enter a
number
998. ND
999. NA

1. g
2. mg
998. ND 3. gm
1
999. NA 4. mg kg
8. ND
9. NA

Enter a
number

1. PO
2. IV

998. ND
999. NA

8. ND
9. NA

8. ND
9. NA

20. Type of outcomes measured:

ENTER THE CODE FOR EACH OUTCOME MEASURED
____ ____
____ ____
____ ____
21. When, relative to the start of the intervention, were outcomes reported?
ENTER THE NUMBER AND LETTERS IN THE APPROPRIATE BOX
NUMBER
UNIT
Use the following
1st follow-up
abbreviations for units:
2nd follow-up
rd

3 follow-up
4th follow-up
5th follow-up
th

6 follow-up
Additional
follow-ups:
END

46

MI

minute

HR

hour

DY

day

WK

week

MO

month

YR

year

ND

not described

NA

not applicable

Figure 3a. Adverse events analysis form for death, MI, stroke cases
RAND EPC EPHEDRA PROJECT
ADVERSE EVENTS ANALYSIS FORM

Article ID:

Reviewer:________________

FDA Case Number: ___________________________________

Form Number: _____of______(Fill out one form for each subject)

1. Does adverse event form report on ephedra or ephedrine?

CIRCLE ONE
Yes..................................................................... 1
No/ Unsure..........................................................2 (STOP)
(IF NOT EPHEDRA/EPHEDRINE THEN STOP)
2. Are there adequate data available to analyze this report?
CIRCLE ONE
Yes..................................................................... 1
No .......................................................................2 (STOP)
(IF NOT ADEQUATE DATA THEN STOPMUST BE A SERIOUS ADVERSE EVENT AND PRODUCT SPECIFICALLY IDENTIFIED)
3. What additional sources of data are available?
CHECK ALL THAT APPLY AND/OR ENTER CODE
FDA affidavit ..................................................... (01)
Medical records ................................................ (02)
Legal documents .............................................. (03)
Labels ............................................................... (04)
Other (_______________________________) (96)
None of the above ........................................... (97)
4. What was the adverse event? CHECK ALL THAT APPLY AND/OR ENTER CODE
(Start codes at 40)
Death ................................................................ (01)
MI...................................................................... (02)
CVA .................................................................. (03)
Other serious adverse event (enter code: _________)
Other (_______________________________) (96)
None of the above ........................................... (97) (STOP)
5. IF MI, what procedures were done?
CHECK ALL THAT APPLY
Coronary angiography ...................................... (01)
Revascularization ............................................. (02)
6. IF MI, what was(were) the outcome of the procedure(s)?
No significant CAD ........................................... (01)
< 3V CAD.......................................................... (02)
3V or LMD......................................................... (03)
Low LVEF ( 40%) ........................................... (04)

47

Figure 3a. Adverse events analysis form for death, MI, stroke cases (continued)
RAND EPC EPHEDRA PROJECT
ADVERSE EVENTS ANALYSIS FORM
7. IF STROKE, what is the outcome?
CIRCLE ONE
Complete resolution........................................... 1
Minimally affected (still able to work)................. 2
Moderately affected (more than one limb)......... 3
Severely affected............................................... 4
Not described .................................................... 8
8. Who completed the adverse events form? CIRCLE ONE
Physician / Health care provider........................ 1
Subject............................................................... 2
Subject surrogate ............................................. 3
Government agency .......................................... 4
9. What was the age of the subject on the date report was made?
Enter number: ______ ______
10. What is the gender of the subject?
CIRCLE ONE
Male ................................................................... 1
Female............................................................... 2
Not described .................................................... 8
11. Why was the subject taking the product?
CHECK ALL THAT APPLY AND/OR ENTER CODE
(Start codes at 4)
Weight loss ....................................................... (01)
Improved athletic performance ......................... (02)
Psychological effect.......................................... (03)
Other: (enter code _______ , _______ , _______)
Not described ................................................... (98)
12. What was the source of the product?
CIRCLE ONE
Retail market...................................................... 1
Multi-level marketing/ out of home .................... 2
Direct from manufacturer................................... 3
Health care provider .......................................... 4
Other (__________________________) .......... 6
Not described .................................................... 8
13. Was the product specifically identified?
CIRCLE ONE
Yes..................................................................... 1
No ...................................................................... 2
(IF NO THEN SKIP TO QUESTION 18)
14. What is the common, proprietary, and/or scientific (genus, genus/species)
name of the product?
ENTER CODE OR CIRCLE ONE OF THE BELOW
Code:________
None ................................................................. 97
Not described ................................................... 98
Not applicable ................................................... 99

48

Figure 3a. Adverse events analysis form for death, MI, stroke cases (continued)
RAND EPC EPHEDRA PROJECT
ADVERSE EVENTS ANALYSIS FORM

15. Of which main constituents is the product made?

ENTER CODE FOR EACH OR CIRCLE ONE OF THE BELOW
Code: ________ ________, ________ ________ , ________ , ________
None ................................................................. 97
Not described ................................................... 98
Not applicable ................................................... 99
16. Was chemical analysis on ephedra alkaloids data presented?
CIRCLE ONE
Yes..................................................................... 1
No ...................................................................... 2
Not described .................................................... 8
Not applicable .................................................... 9
17. Please fill in the following information on dosage data.
This information is from analysis: ( ENTER THE NUMBER AND CODES IN THE APPROPRIATE BOXES.)
Unit
Codes for units:
Dosage data
Number
(code)
g
1
Total daily dose of
mg
2
ephedrine alkaloids
Single dose of
gm
3
ephedrine alkaloids
-1
mgkg 4
Total daily dose of caffeine
Ratio caffeine/ephedrine
alkaloids

ND

NA

18. Please fill in the following information on dosage data.

This information is from label: ( ENTER THE NUMBER AND CODES IN THE APPROPRIATE BOXES.)
Unit
Codes for units:
Dosage data
Number
(code)
g
1
Total daily dose of
mg
2
ephedrine alkaloids
gm
3
Single dose of
-1
ephedrine alkaloids
mgkg 4
Total daily dose of caffeine
Ratio caffeine/ephedrine
alkaloids

ND

NA

19. What was the duration of ephedrine use? CIRCLE ONE

<48 hours .......................................................... 1
2-13 days ........................................................... 2
14-60 days (acute)............................................. 3
>60 days (chronic) ............................................ 4
Not described .................................................... 8

49

Figure 3a. Adverse events analysis form for death, MI, stroke cases (continued)
RAND EPC EPHEDRA PROJECT
ADVERSE EVENTS ANALYSIS FORM
20. What was the timing of the last ephedrine dose? CIRCLE ONE
<6 hours............................................................. 1
6-24 hours.......................................................... 2
>24 hours........................................................... 3
Not described .................................................... 8
21. Was the product used again after first adverse event? CIRCLE ONE
Yes..................................................................... 1
No ...................................................................... 2
Not described .................................................... 8
Not applicable .................................................... 9
22. If product was used again after first adverse event, did the adverse event reoccur?
CIRCLE ONE
Yes..................................................................... 1
No ...................................................................... 2
Not described .................................................... 8
Not applicable .................................................... 9
23. Was the subject actively involved in exercise at or immediately before the
occurrence of the adverse event?
CIRCLE ONE
Yes .................................................................... 1
No ...................................................................... 2
Not described .................................................... 8
Not applicable .................................................... 9
24. Did form report on use of any other substances? (CHECK ALL THAT APPLY AND ENTER CODE)
Caffeine (in addition to product) ..............................................
Illicit drugs:...............................................................................
Code: _________ , _________ , _________ , _________ ,
_________ , _________ , _________ , _________
Other Herbs: ............................................................................
Code: _________ , _________ , _________ , _________ ,
_________ , _________ , _________ , _________
Prescribed or OTC medication: ...............................................
Code: _________ , _________ , _________ , _________ ,
_________ , _________ , _________ , _________
Other substance: .....................................................................
Code: _________ , _________ , _________ , _________ ,
_________ , _________ , _________ , _________
Not described ..........................................................................
None ........................................................................................

50

Figure 3a. Adverse events analysis form for death, MI, stroke cases (continued)
RAND EPC EPHEDRA PROJECT
ADVERSE EVENTS ANALYSIS FORM

25. Which of the following conditions were evaluated?

CHECK ALL THAT APPLY AND/OR ENTER CODE
(Start codes at 15)
Pre-existing condition:
PRESENTEXCLUDED
Asthma................................................................. .........
CAD ..................................................................... .........
DM ....................................................................... .........
HTN ..................................................................... .........
Obesity................................................................. .........
Renal disease ...................................................... .........
Substance abuse................................................. .........
Syncope............................................................... .........
Thyroid condition ................................................. .........
TIA History ........................................................... .........
Other vascular disease (________________) .... .........
Rheumatological diseases................................... .........
Other (Enter code: _________) ......................... .........
Other (Enter code: _________) ......................... .........
Other (Enter code: _________) ......................... .........
Other (Enter code: _________) ......................... .........
Other (Enter code: _________) ......................... .........
Other (Enter code: _________) ......................... .........
26. Was a drug screen performed?
(CIRCLE ONE)
Yes.........................................................................1
No ..........................................................................2 (STOP)
27. Results of URINE screen:
(start codes at 03)
(CHECK ALL THAT APPLY)
No substance found............................................. (01)
Substance(s) found and identified: (Enter code(s)):
( ______ , ______ , ______ , ______ , ______ , _______ )
Not described ...................................................... (98)
28. Results of BLOOD screen:
(start codes at 03)
(CHECK ALL THAT APPLY)
No substance found............................................. (01)
Substance(s) found and identified: (Enter code(s) below)
( ______ , ______ , ______ , ______ , ______ , _______ )
Not described ...................................................... (98)
END

51

Figure 3b. Adverse events analysis form for seizure cases

RAND EPC EPHEDRA PROJECT
ID/ FDA Case Number:

ADVERSE EVENTS ANALYSIS FORM

Reviewer:________________

First Author: ___________________________________________________________

(Last Name Only)
Form Number: _____of______(Fill out one form for each subject)
1. Does this adverse event report use of ephedra or ephedrine?
CIRCLE ONE

Ephedra only .............................................................1

No/ Unsure ................................................................2 (STOP)
Ephedrine only ..........................................................3
Ephedra and Ephedrine ............................................4
(IF NOT EPHEDRA/ OR EPHEDRINE THEN STOP)
2. Is a generalized (tonic-clonic) seizure reported as an adverse event
(synonym = grandmal seizure)?
CIRCLE ONE
Yes ........................................................... ................1
No, another type of seizure is reported.... ................2 (STOP)
No, seizure unspecified is reported...........................3 (STOP)
No, seizure is not reported as an adverse event ......9 (STOP)
(IF NO SEIZURE REPORTED THEN STOP)
3. For which evaluations are results reported as part of the evaluation
of the seizure?
CHECK ALL THAT APPLY
Serum electrolytes (must include Na) .......................
Calcium .....................................................................
Magnesium................................................................
Glucose .....................................................................
CT/ MRI of head........................................................
EEG...........................................................................
Temperature..............................................................
4. Were the following pre-existing conditions specifically mentioned as present
or excluded?
NOT DESCRIBED PRESENT EXCLUDED
Pre-existing condition:
Alcoholism............................. ............... ............
Substance Abuse .................. ............... ............
Seizure Disorder ................... ............... ............
5. What was the age of the subject on the date the report was made?
Enter number: ______ ______ (No Data = 99)
6. What is the gender of the subject?
Male............................................................................ 1
Female ....................................................................... 2
Not described ............................................................. 8

52

Figure 3b. Adverse events analysis form for seizure cases (continued)
RAND EPC EPHEDRA PROJECT
ADVERSE EVENTS ANALYSIS FORM
7. Why was the subject taking the product? ......
(Start codes at 04)
(CHECK ALL THAT APPLY AND/OR ENTER CODE)
Weight loss..................................................................... (01)
Improved athletic performance ...................................... (02)
Psychological effect ....................................................... (03)
Other:. (enter code _______ , _______ , _______)
Not described ................................................................. (98)

53

Figure 3b. Adverse events analysis form for seizure cases (continued)
RAND EPC EPHEDRA PROJECT
ADVERSE EVENTS ANALYSIS FORM
Product: ________ of ________
Description:____________________________________________________
8. What is the common, proprietary, and/or scientific (genus, genus/species)
name of the product?
(ENTER CODE OR CIRCLE ONE OF THE BELOW)
Code:________
None.............................................................................. 97
Not applicable ............................................................... 99
.
9. Of which main constituents is the product made?
(ENTER CODE FOR EACH OR CIRCLE ONE OF THE BELOW)
Code: ______ , ______ , ______ , ______ , ______
______ , ______ , ______ , ______ , ______
None.............................................................................. 97
Not applicable ............................................................... 99
10. Was chemical analysis on ephedra alkaloids data presented?
(CIRCLE ONE)
Yes .................................................................................. 1
No.................................................................................... 2
Ordered but not reported ................................................ 3
Not described .................................................................. 8
Not applicable ................................................................. 9
11. Please fill in the following information on dosage data.
This information is from analysis: ( ENTER THE NUMBER AND CODES IN THE APPROPRIATE BOXES.)
Unit
Codes for units:
Dosage data
Number
(code)
g
1
Total daily dose of
mg
2
ephedrine alkaloids
Single dose of
gm
3
ephedrine alkaloids
-1
mgkg 4
Total daily dose of caffeine
ND
8
Ratio caffeine/ephedrine
:
NA
9
alkaloids
12. Please fill in the following information on dosage data.
This information is from label: ( ENTER THE NUMBER AND CODES IN THE APPROPRIATE BOXES.)
Unit
Codes for units:
Dosage data
Number
(code)
g
1
Total daily dose of
mg
2
ephedrine alkaloids
gm
3
Single dose of
-1
ephedrine alkaloids
mgkg 4
Total daily dose of caffeine
Ratio caffeine/ephedrine
alkaloids

54

ND

NA

Figure 3b. Adverse events analysis form for seizure cases (continued)
RAND EPC EPHEDRA PROJECT
ADVERSE EVENTS ANALYSIS FORM
13. What was the duration of ephedrine use?
(CIRCLE ONE)
<48 hours ....................................................................... 1
2-13 days ........................................................................ 2
14-60 days (acute) .......................................................... 3
>60 days (chronic) ......................................................... 4
Not described .................................................................. 8
14. What was the timing of the last ephedrine dose?
(CIRCLE ONE)
<6 hours .......................................................................... 1
6-24 hours ....................................................................... 2
>24 hours ........................................................................ 3
Not described .................................................................. 8
15. Was/were the product(s) discontinued after problematic symptoms emerged?
(CIRCLE ONE)
Yes .................................................................................. 1
No.................................................................................... 2
Not described .................................................................. 8
Not applicable ................................................................. 9
16. If product(s) was/were used again after discontinuation, did the problematic
symptoms reoccur?
(CIRCLE ONE)
Yes .................................................................................. 1
No.................................................................................... 2
Not described .................................................................. 8
Not applicable ................................................................ 9
17. Did form report on use of any other substances?
(ENTER CODE OR CIRCLE)
Code: ______ , ______ , ______ , ______ , ______
______ , ______ , ______ , ______ , ______
None.............................................................................. 97
Not described ................................................................ 98
Not applicable ............................................................... 99
18. Which of the following conditions were evaluated?

(Start codes at 15)

(CHECK ALL THAT APPLY AND/OR ENTER CODE)
Pre-existing condition:
PRESENT EXCLUDED
Asthma .................................................................... ...........
CAD......................................................................... ...........
DM........................................................................... ...........
HTN ......................................................................... ...........
Obesity .................................................................... ...........
Prior psychiatric history ........................................... ...........
Renal disease ......................................................... ...........
Syncope .................................................................. ...........
Thyroid condition..................................................... ...........
TIA History .............................................................. ...........
Other vascular disease (__________________).... ...........
Rheumatological diseases ...................................... ...........
Not described .......................................................... (98)

55

Figure 3b. Adverse events analysis form for seizure cases (continued)
RAND EPC EPHEDRA PROJECT
ADVERSE EVENTS ANALYSIS FORM
19. Was a drug screen performed?

(CIRCLE ONE)
Yes .................................................................................. 1
No.................................................................................... 2 (STOP)

20. Results of URINE screen:

(start codes at 03)

(CHECK ALL THAT APPLY)
No substance found ....................................................... (01)
Substance(s) found and identified: (Enter code(s)):
( ______ , ______ , ______ , ______ , ______ , _______ )
Not described ................................................................. (98)
21. Results of BLOOD screen:
(start codes at 03)
(CHECK ALL THAT APPLY)
No substance found ....................................................... (01)
Substance(s) found and identified: (Enter code(s) below)
( ______ , ______ , ______ , ______ , ______ , _______ )
Not described ................................................................. (98)

END

56

Figure 3c. Adverse events analysis form for psychiatric cases

RAND EPC EPHEDRA PROJECT
ADVERSE EVENTS ANALYSIS FORM
ID/ FDA Case Number:

Reviewer:________________

First Author: ___________________________________________________________

(Last Name Only)
Form Number: _____of______(Fill out one form for each subject)
1. Does this adverse event report use of ephedra or ephedrine?
(CIRCLE ONE)
Ephedra only ................................................................... 1
No/ Unsure ...................................................................... 2 (STOP)
Ephedrine only ................................................................ 3
Ephedra and Ephedrine .................................................. 4
(IF NOT EPHEDRA/ OR EPHEDRINE THEN STOP)
2. Is there an adverse event?
(CIRCLE ONE)
Yes .................................................................................. 1
No.................................................................................... 2 (STOP)
(IF NO ADVERSE EVENT THEN STOP)
3. Was the product specifically identified?
(CIRCLE ONE)
Yes .................................................................................. 1
No.................................................................................... 2 (STOP)
(MUST BE A SERIOUS ADVERSE EVENT AND
PRODUCT SPECIFICALLY IDENTIFIED OR STOP)
4. What was the adverse event?
(CHECK ALL THAT APPLY AND/OR ENTER TEXT)
Psychosis ....................................................................... (06)
Mania or severe agitation............................................... (07)
Severe depression ......................................................... (08)
Suicidal ideation ............................................................. (09)
Suicide attempt/ Suicide................................................. (146)
Hallucinations................................................................. (138)
Other serious psychiatric events: (enter below)
_______________________________________...... ( )

_______________________________________...... (

_______________________________________...... (

_______________________________________...... (

Other non-serious event: ____________________) ..... ( 96) (STOP)

None of the above ......................................................... (97) (STOP)

57

Figure 3c. Adverse events analysis form for psychiatric cases (continued)
RAND EPC EPHEDRA PROJECT
ADVERSE EVENTS ANALYSIS FORM
5. Is there a presence or history of the following conditions?
(CHECK ALL THAT APPLY AND/OR ENTER TEXT)
PRESENCE
HISTORY (CODES)
Psychosis ................................................... ............... .(01)
Mania or severe agitation........................... ............... .(02)
Hallucinations............................................. ............... .(03)
Severe depression ..................................... ............... .(04)
Suicide attempt .......................................... ............... .(05)
Suicide ideation.......................................... ............... .(06)
Schizophrenia ............................................ ............... .(07)
Acute confusion.......................................... ............... .(08)
Delusions ................................................... ............... .(09)
Aggression/threatened violence................. ............... .(10)
Substance abuse ....................................... ............... .(11)
Other conditions:
___________________________ ............. ................ .( )
___________________________ ............. ................ .(

___________________________ ............. ................ .(

___________________________ ............. ................ .(

___________________________ ............. ................ .(

___________________________ ............. ................ .(

___________________________ ............. ................ .(

___________________________ ............. ................ .(

None described .................................... ......................... (98)

6. What was the outcome of the event?
(CHECK ALL THAT APPLY)
Death .............................................................................
Harm to self/others.........................................................
Hospitalization................................................................
ER Visit ..........................................................................
On-going adverse event/disability..................................
Resolved ........................................................................
Other:__________________________________.........
Not described .................................................................

58

Figure 3c. Adverse events analysis form for psychiatric cases (continued)
RAND EPC EPHEDRA PROJECT
ADVERSE EVENTS ANALYSIS FORM
7. What was intervention was prescribed after adverse event occurred?
(CHECK ALL THAT APPLY)
No procedure .................................................................
Discontinue Ephedra......................................................
Change existing medication...........................................
New medication..............................................................
Initiate/change frequency/intensity of outpatient visits...
Hospitalization................................................................
Involuntary hospitalization..............................................
Legal action....................................................................
Not described .................................................................
Not applicable ................................................................
8. What was the age of the subject on the date report was made?
Enter number: ______ ______

(No Data=99)

9. What is the gender of the subject?

(CIRCLE ONE)
Male................................................................................. 1
Female ............................................................................ 2
Not described .................................................................. 8
10. Why was the subject taking the product?
(CHECK ALL THAT APPLY)
Weight loss.....................................................................
Improved athletic performance ......................................
Psychological effect .......................................................
Addiction ........................................................................
Other:____________________________________.....
Not described .................................................................
11. Did report describe the use of any other substances or medications
taken prior to/or during the event?
_______________________________________ (

_______________________________________ (

_______________________________________ (

_______________________________________ (

_______________________________________ (

None described .............................................................. 98

59

Figure 3c. Adverse events analysis form for psychiatric cases (continued)
RAND EPC EPHEDRA PROJECT
ADVERSE EVENTS ANALYSIS FORM
12. What is the common, proprietary, and/or scientific (genus,
genus/species) name of the product?
(ENTER TEXT OR CIRCLE ONE BELOW)
Name:_____________________________________( )
None.............................................................................. 97
Not described ................................................................ 98
Not applicable ............................................................... 99
13. Of which main constituents is the product made?
(Enter text or circle one below)
__________________________________________(

__________________________________________(

__________________________________________(

__________________________________________(

__________________________________________(

__________________________________________(

__________________________________________(

__________________________________________(

None.............................................................................. 97
Not described ................................................................ 98
Not applicable ............................................................... 99
14. Was chemical analysis on ephedra alkaloids data presented?
(CIRCLE ONE)
Yes .................................................................................. 1
No.................................................................................... 2
Ordered but not presented.............................................. 3
Not described .................................................................. 8
Not applicable ................................................................. 9

60

Figure 3c. Adverse events analysis form for psychiatric cases (continued)
RAND EPC EPHEDRA PROJECT
ADVERSE EVENTS ANALYSIS FORM
15. Please fill in the following information on dosage data.
This information is from analysis: ( ENTER THE NUMBER AND UNITS IN THE APPROPRIATE BOXES.)
Unit
Codes for units:
Dosage data
Number
Unit
Code
g
1
Total daily dose of
mg
2
ephedrine alkaloids
Single dose of
gm
3
ephedrine alkaloids
-1
mgkg 4
Total daily dose of
ND
8
caffeine
Ratio
NA
9
caffeine/ephedrine
:
alkaloids
16. This information is from label: ( ENTER THE NUMBER AND UNITS IN THE APPROPRIATE BOXES.)
Unit
Dosage data
Number
Unit
Codes for units:
Code
g
1
Total daily dose of
ephedrine alkaloids
mg
2
Single dose of
gm
3
ephedrine alkaloids
-1
mgkg 4
Total daily dose of
caffeine
ND
8
Ratio
NA
9
caffeine/ephedrine
:
alkaloids
17. What was the duration of ephedra/ephedrine use?
(CIRCLE ONE)
<48 hour .......................................................................... 1
2-13 days ........................................................................ 2
14-60 days (acute) .......................................................... 3
>60 days (chronic) ......................................................... 4
60 days to 1 year............................................................. 5
Over 1 year ..................................................................... 6
Not described .................................................................. 8
18. What was the timing of the last ephedra/ephedrine dose?(CIRCLE ONE)
<6 hours .......................................................................... 1
6-24 hours ....................................................................... 2
>24 hours ........................................................................ 3
Not described .................................................................. 8
19. Was/were the product(s) discontinued after problematic
symptoms emerged?
(CIRCLE ONE)
Yes .................................................................................. 1
No.................................................................................... 2
Not described .................................................................. 8
Not applicable ................................................................. 9

61

Figure 3c. Adverse events analysis form for psychiatric cases (continued)
RAND EPC EPHEDRA PROJECT
ADVERSE EVENTS ANALYSIS FORM
20. If product(s) was/were used again after discontinuation, did the
problematic symptoms reoccur?
(CIRCLE ONE)
Yes .................................................................................. 1
No.................................................................................... 2
Not described .................................................................. 8
Not applicable ................................................................. 9
21. Was autopsy performed?
(CIRCLE ONE)
Yes .................................................................................. 1
No.................................................................................... 2
Not Applicable ................................................................. 9
22. Was drug screen performed?
(CIRCLE ONE)
Yes .................................................................................. 1
No.................................................................................... 2 (STOP)
23. Results of URINE screen:
(CHECK ALL THAT APPLY AND/OR ENTER TEXT)
No substance found ....................................................... (01)
Substance(s) found and identified:
__________________________________________...... (

__________________________________________...... (

__________________________________________...... (

__________________________________________...... (

__________________________________________...... (

__________________________________________...... (

Not described ................................................................. (98)

21. Results of BLOOD screen:(check all that apply and/or enter text)

No substance found ....................................................... (01)

Substance(s) found and identified:
__________________________________________...... (

__________________________________________...... (

__________________________________________...... (

__________________________________________...... (

__________________________________________...... (

__________________________________________...... (

Not described ................................................................. (98)

END

62

Figure 4. Brief data collection form for case reports

RAND EPC EPHEDRA PROJECT
Article ID:
FDA Case Number:
Form Number:

BRIEF FORM FOR CASE REPORTS

Reviewer:
of

(Fill out one form for each subject)

1. Does adverse event form report on ephedra or ephedrine?

CIRCLE ONE
Yes .............................................................................. 1
No/ Unsure .................................................................. 2 (STOP)
(IF NOT EPHEDRA/EPHEDRINE THEN STOP)
2. What was the adverse event?
CHECK ALL THAT APPLY
Death (01)
Cardiovascular:
Heart rate, >120 or <50........................................... (02)
Hypertension, Systolic >180 or Diastolic >105 ....... (03)
MI ............................................................................ (04)
Ventricular tachycardia/ fibrillation .......................... (05)
Cardiac arrest.......................................................... (06)
Pulmonary:
Respiratory arrest.................................................... (07)
Neurological:
TIA........................................................................... (08)
CVA ......................................................................... (09)
Brain Hemorrhage, not CVA ................................... (10)
Fainting / Loss of consciousness ............................ (11)
Coma....................................................................... (12)
Seizure .................................................................... (13)
Paralysis.................................................................. (14)
Psychiatric:
Severe depression .................................................. (15)
Hallucinations.......................................................... (16)
Mania or severe agitation........................................ (17)
Psychosis ................................................................ (18)
Suicide .................................................................... (19)
Other adverse events:
Changes in glucose <40 or >400 ............................ (20)
Liver failure ALT/AST >200..................................... (21)
Rhabdomyolysis CPK >400 .................................... (22)
Miscarriage.............................................................. (23)
Serious renal event ................................................. (25)
Autonomic Hyperactivity.......................................... (26)
None of the above .................................................. (24)

63

Figure 5. Examples of MIPER Files

5a. Email record of a telephone conversation

64

Figure 5. Examples of MIPER Files (continued)

5b. Typed or handwritten letter from the consumer to the company

65

Figure 5. Examples of MIPER Files (continued)

5c. Handwritten note of telephone conversation with consumer written on a rudimentary form. Note more
than one case is recorded on a single MIPER file.

66

Figure 5. Examples of MIPER Files (continued)

5d. Handwritten note of telephone conversation with consumer written on a piece of paper

67

Figure 5. Examples of MIPER Files (continued)

5e. A form developed for systematically collecting information about possible adverse events

68

Figure 6. Example of duplicate case

69

Figure 6. Example of duplicate case (continued)

70

Figure 7. Metabolife record screener form

Case Number:

Reviewer:________________

Form Number: _____of______(Fill out one form for each subject)

1. Subjects age: _________ (Not Described =999)
2. What is the subjects gender?
(CIRCLE ONE)
Male................................................................................. 1
Female ............................................................................ 2
Not described/ Not reported............................................ 3
3. What was the adverse event?
(CHECK ALL THAT APPLY)
No adverse event reported ............................................ (01)
(IF NO ADVERSE EVENT THEN STOP.)
Death.............................................................................. (02)
Cardiovascular:
Heart rate, >120 or <50.................................................. (03)
Heart rate, 50-120, or not otherwise unspecified........... (04)
Hypertension, Systolic >180 or Diastolic >105 .............. (05)
Hypertension, Systolic <180 or Diastolic <105, or
not otherwise specified .............................................. (06)
Myocardial Infarction/ Heart Attack ................................ (07)
Cardiac Dysrythmia, Other/ Palpitations ........................ (08)
Cardiac arrest................................................................. (09)
Ventricular Tachycardia/ Fibrillation............................... (10)
Chest Pain, not specified as MI ..................................... (11)
Pulmonary:
Respiratory arrest........................................................... (12)
Neurological:
Transient Ischemic Attack .............................................. (13)
CVA/ Stroke, not known to be hemorrhage ................... (14)
Brain Hemorrhage.......................................................... (15)
Fainting / Loss of consciousness ................................... (16)
Coma.............................................................................. (17)
Seizure ........................................................................... (18)
Psychiatric:
Depression ..................................................................... (19)
Hallucinations................................................................. (20)
Mania or severe agitation............................................... (21)
Psychosis ....................................................................... (22)
Suicide attempt .............................................................. (23)
Autonomic Hyperactivity (includes: tremor, twitching,
jitteriness, insomnia, increased sweating, agitation,
nervousness, and irritability) ....................................... (24)

71

Figure 7. Metabolife record screener form (continued)

3. What was the adverse event?
(CHECK ALL THAT APPLY)
(continued)
Other adverse events:
Changes in glucose <40 or >400 ................................... (25)
Liver failure ALT/AST >200............................................ (26)
Liver abnormality, not otherwise specified ..................... (27)
Rhabdomyolysis CPK >400 ........................................... (28)
Rhabdomyolysis, not otherwise specified ...................... (29)
Miscarriage..................................................................... (30)
Allergic Reaction ............................................................ (31)
Anesthesia complication ................................................ (32)
Fatigue/Fever/ Chills ...................................................... (33)
Abnormal lab values, not otherwise specified................ (34)
Other adverse events not already specified:
Ear, Eye, Nose, or Throat .............................................. (35)
Respiratory System........................................................ (36)
Cardiovascular System .................................................. (37)
Gastrointestinal System ................................................. (38)
Hepatobiliary System ..................................................... (39)
Musculoskeletal System................................................. (40)
Genitourinary System..................................................... (41)
Gynecologic (includes breast and menstrual
symptoms) .................................................................. (42)
Sexual Dysfunction ........................................................ (43)
Neurological System (includes headache)..................... (44)
Mental Health ................................................................. (45)
Skin (includes Pruritis) ................................................... (46)
Hematologic System ...................................................... (47)
Oncologic System .......................................................... (48)
Other symptoms not specified above ............................ (49)
4. Did the adverse event result in a hospital stay (at least one night;
do not include emergency room visits)?
(CIRCLE ONE)
Yes .................................................................................. 1
No/ No Data .................................................................... 2
5. Is there additional information (medical records or similar) available
for more detailed review regarding past health history, current,
problems, toxicology results, etc?
(CIRCLE ONE)
Yes .................................................................................. 1
No.................................................................................... 2
END

72

Figure 8. Literature flow

Library Search
(n=455)

Identified by
Outreach to Expert
(n=64)

Reference Lists
(n=34)

553 Articles Requested

20 Not found
533 Articles Screened

66 Case
reports/
Case series
articles

57 RCT/
CCT
On weight loss or
Athletic performance

57 Articles,
corresponding
to 52 trials,
assessed

To Adverse Events
Seriousness Review
(see Figure 17c)

410 Rejected
158
124
48
22
20
18

Not Ephedra/ephedrine
No adverse events
Not Human
Study design: Descriptive
Study design: Review/Meta-analysis
Study Design: RCT/CCT (not weight
loss or athletic performance)
12 Study design: Other
7 Duplicate of article already included
1 No translator available

8 Athletic
performance
trials

44 Weight loss
trials

52 Trials evaluated for

adverse events analysis

46 Case Reports
contributing to
adverse events
analysis

52 Trials in adverse
events analysis**

24 Trials excluded from

meta-analysis
18 Follow-up < 8 wk
6 Other*

20 Trials in weight
loss meta-analysis

* Various reasons, see table 9.

** Two studies had no placebo group, and therefore, contribute to the power calculations, but not
to the odds-ratio meta-analysis.

186

Figure 9. Ephedrine versus placebo forest plot

Jensen(88)
Lumholtz(94)
Moheb(84)
Pasquali(85)
Pasquali(85)
Quaade(86)

Combined
-2.7

Effect Size

-.50

Favors Ephedrine

Favors Placebo

Figure 10. Ephedrine versus placebo funnel plot

Effect Size

-1

-2
0.00

0.20

0.40
Standard Error of Effect Size

187

0.60

Figure 11. Ephedrine + caffeine versus placebo forest plot

Astrup(111)
Buemann(92)
Daly(103)
Jensen(88)
Kalman(96)
Kettle(90)
Malchow-Moll(87)
Moheb(84)
Molnar(112)
Quaade(86)
Roed(95)
Van Mil(91)
Combined
-.85

-3.1

Effect Size

Figure 12. Ephedrine + caffeine versus placebo funnel plot

Effect Size

-1

-2
0.00

0.20

0.40
Standard Error of Effect Size

188

0.60

Figure 13. Ephedrine + caffeine versus ephedrine alone forest plot

Jensen(88)

Moheb(84)

Quaade(86)

Combined
-1

-.31
Effect Size

189

.5

Figure 14. Ephedra + herbs containing caffeine versus placebo forest plot

Boozer(115)

Boozer(89)

Colker(93)

Greenway(116)

Combined
-1.7

-.81
Effect Size

Figure 15. Ephedra + herbs containing caffeine versus placebo funnel plot

Effect Size

-.5

-1

-1.5
0.00

0.20
Standard Error of Effect Size

190

0.40

Figure 16. Effect sizes by comparison group

Monthly Weight Loss (lbs) vs Placebo (lbs)

-1

-2

-3

-4

-5
Ephedrine

Ephedra + Herbs
containing Caffeine

191

Ephedrine
+ Caffeine

Figure 17a. Flow of evidence for adverse events analysis, part 1

FDA Batch 1

1783 Adverse Events report IDs

from the FDA

Review of IDs on
Excel Master Sheet

88 Rejected for multiple

products per report

1695 unique Adverse Events report IDs

214 AE full reports
unavailable from the FDA:
Death(4)
Stroke(18)
MI(9)
Psychiatric(143)
All others(40)

158 AE reports sent

on to detailed review:
Death, Stroke, MI

137 IDs not reviewed:

post-9/30/01 report received date:
Death(17)
Stroke(15)
MI(5)
Other adverse event(100)

1186 AE reports
available on other
adverse events

251 AE reports on
other events

935 AE reports on to
screening:
Other cardiovascular
Other neurological
Psychiatric
Seizure

Continued on Figure 17b

192

Figure 17b. Flow of evidence for adverse events analysis, part 2

From Figure 17a, part 1

FDA Batch 1

160 subjects
reported on in
158 death, MI,
stroke AE
reports

FDA Batch 2

965 subjects reported

on in 935 other
cardiovascular, other
neurological, seizure,
psychiatric AE reports

130 subjects
reported in 125
AE reports

43 subjects not
reporting on
ephedra
24 subjects not
reporting on ephedra
or ephedrine

922 subjects
reporting on
ephedra

764 subjects reporting

other adverse events
(see Table 28)

6 subjects reporting on
ephedra
100 subjects reporting
on ephedrine

73 subjects reporting
other adverse events
(see Table 28)

33 subjects reporting
adverse events sent on
to detailed review:
death(15); MI(3);
stroke(2); cardiac
arrest(1); TIA(1); brain
hemorrhage(2);
seizure(5); psychiatric
symptoms(9)

158 subjects reporting adverse

events* sent on to detailed
review: vtach/ vfib(25); cardiac
arrest(10); pulmonary arrest(6);
TIA(14); brain hemorrhage(4)
seizure(69); psychiatric
symptoms(51)
*adverse events are not
mutually exclusive

160 subjects
detailed review
completed

158 subjects
detailed review
completed

33 subjects
detailed review
completed

Continued on Figure 17c

193

Figure 17c. Flow of evidence for adverse events analysis, part 3

Adverse Events Seriousness Review
on published articles
(from Figure 8)
66 Case Reports/
Case Series articles

6 reports
rejected as
duplicates

4 subjects not
reporting ephedra
or ephedrine

49 subjects
reporting other
adverse events

60 Case
Reports/ Case
Series articles
reporting on
99 subjects

95 subjects reporting on
ephedra or ephedrine

46 subjects reporting adverse

events sent to detailed review;
death(9); MI(4); stroke(8);
cardiac(6); neurological(1);
seizure(1); psychiatric(17)

From Figure 17b, part 2

FDA Batch 1
160 subjects
detailed review
completed

FDA Batch 1
158 subjects
detailed review
completed

FDA Batch 2
33 subjects
detailed review
completed

Literature Case Reports

46 subjects detailed
review completed

397 detailed
review
completed

68 subjects rejected for

inadequate data to
complete the review
329 subjects
contributing to adverse
events analysis
(Table 23)

194

Figure 18. Flow of MIPER ID Numbers

15951 MIPER files on CD-ROMs

329 files excluded

from analysis
(blank files,
duplicate cases,
etc)

13907 files of single cases

13907 cases

1715 files of multiple cases

4595 cases

15622 MIPER files contributing

18502 cases to the analysis

195

Evidence Tables

Evidence Table 1 RCTs and CCTs reporting on Athletic Performance Enhancement with Ephedra

First Author
Year
Bell DG &
Jacobs I
1999 #24

Design
Study Quality
Population (>75%)
Comorbidities
CCT
Jadad Score:
1
Population:
Male
athletes
Comorbidities:
N/A

Bell DG,
CCT
Jacobs I, et al. Jadad Score:
Population:
1999 #25
Comorbidities:

Intervention
Total Daily Dose
Route of Administration
Arm # Duration
1 Placebo
Placebo for 2 days
2 Ephedrine
75 mg orally for 2 days
Caffeine
375 mg orally for 2 days
1

1
Male
N/A

2
3

217
Bell DG,
CCT
Jacobs I, et al. Jadad Score:
Population:
2000 #26
Comorbidities:

1
3
Male
N/A

N/A = not available or not applicable

Sample Size
n Entered:
n Analyzed:
n Entered:
n Analyzed:

9
9
9
9

Control
No dosage data reported
Placebo
Placebo for 1 day
Ephedrine
1 mgkg-1 orally for 1 day
Caffeine
5 mgkg-1 orally for 1 day

n Entered:
n Analyzed:
n Entered:
n Analyzed:
n Entered:
n Analyzed:

10
10
10
10
10
10

Placebo
Placebo for 1 day
Ephedrine
0.8 mgkg-1 orally for 1 day
Caffeine
5 mgkg-1 orally for 1 day
Caffeine
1 mgkg-1 orally for 1 day
Caffeine
4 mgkg-1 orally for 1 day
Ephedrine
0.8 mgkg-1 orally for 1 day
Caffeine
4 mgkg-1 orally for 1 day

n Entered:
n Analyzed:
n Entered:
n Analyzed:

12
12
12
12

n Entered:
n Analyzed:

12
12

n Entered:
n Analyzed:

N/A
N/A

Summary of Results
VO2 maximum during the treadmill runs, VO2 at
standard running velocities, and the relationship
between the heart rate and the VO2 were similar in
both the Caffeine and Ephedrine (C+E, Arm 2) and
the Placebo (Arm 1) groups. Run times of the
performance test for subjects in the C+E group (Arm
2) was significantly faster (p < 0.05) than for subjects
in the Placebo group (Arm 1).
Individuals in the Caffeine and Ephedrine (C+E)
group (Arm 3) experienced a significant VO2
increase of 7.5% compared to individuals in the
Placebo group (Arm 2), but similar to individuals in
the Control group (Arm 1). Tolerance times were
similar for the C+E (Arm 3, 121.3 +/- 33.9 minutes)
and Placebo (Arm 2, 120.0 +/- 28.4) groups, but
significantly longer than the Control group (Arm 1,
106.6 +/- 24.0).
VO2 maximum was similar among all groups.
Endurance ride times to exhaustion for all Caffeine
and Ephedrine groups with different dosages (Arm 2,
27.5 +/- 12.4 minutes; Arm 3, 27.6 +/-10.9; and Arm
4, 28.2 +/- 9.3) were similar, and significantly greater
than Placebo (Arm 1, 17.0 +/- 3.0) with an
approximated 64% improvement.

Evidence Table 1 RCTs and CCTs reporting on Athletic Performance Enhancement with Ephedra (continued)
Design
Study Quality
First Author Population (>75%)
Year
Comorbidities
Bell DG,
CCT
Jacobs I, et al. Jadad Score:
4
Population:
Male
1998 #27
Comorbidities:
N/A

1
Military
N/A

Oksbjerg N,
CCT
Meyer T, et al. Jadad Score:
Population:
1986 #214
Comorbidities:

1
Male
N/A

218

Bell DG,
CCT
Jacobs I, et al. Jadad Score:
Population:
2001 #512
Comorbidities:

Pasternak 1999 CCT

Jadad Score:
#511
Population:
Comorbidities:

Intervention
Total Daily Dose
Route of Administration
Arm # Duration
1 Placebo
Placebo for 1 day
2 Ephedrine
1 mgkg-1 orally for 1 day
Caffeine
1 mgkg-1 orally for 1 day
3 Caffeine
5 mgkg-1 orally for 1 day
4 Ephedrine
1 mgkg-1 orally for 1 day
1 Placebo
Placebo for 1 day
2 Caffeine
5 mgkg-1 orally for 1 day
3 Ephedrine
1 mgkg-1 orally for 1 day
4 Ephedrine
1 mgkg-1 orally for 1 day
Caffeine
1 mgkg-1 orally for 1 day
1 Ephedrine
40 mg orally for 1 day
2 Placebo
No dosage data reported

1
1
Male
athletes
N/A

2
3
4

N/A = not available or not applicable

Placebo
Placebo for 1 day
Caffeine
4 mgkg-1 orally for 1 day
Ephedrine
0.8 mgkg-1 orally for 1 day
Caffeine
4 mgkg-1 orally for 1 day
Ephedrine
0.8 mgkg-1 orally for 1 day

Sample Size
n Entered:
n Analyzed:
n Entered:
n Analyzed:
n Entered:
n Analyzed:
n Entered:
n Analyzed:
n Entered:
n Analyzed:
n Entered:
n Analyzed:
n Entered:
n Analyzed:
n Entered:
n Analyzed:

Summary of Results
VO2 maximum increased progressively during
exercise in all trials (Arms 1, 2, 3, and 4, p < 0.05),
but no significant difference was found among them.
Time to exhaustion was significantly longer for the
Caffeine and Ephedrine trial ((Arm 2) when compared
to Placebo (Arm1) and Caffeine (Arm 3) trials (p <
12 0.05).
8
12
8
24 Accumulated VO2 was similar between all groups.
24 The Ephedrine (Arm 3) and Caffeine plus Ephedrine
24 (Arm 4) treatments increased power output
24 significantly (p < 0.05) early in the Wingate test
24 compared to the Placebo (Arm 1) and Caffeine (Arm
24 2) treatments. Caffeine-containing treatments (Arms 2
24 and 4) significantly improved times to exhaustion by
24 8% compared to non-caffeine treatments (Arms 1 and
3).
12
8
12
8

n Entered:
n Analyzed:
n Entered:
n Analyzed:

6
6
6
6

n Entered:
n Analyzed:
n Entered:
n Analyzed:
n Entered:
n Analyzed:
n Entered:
n Analyzed:

13
13
13
13
13
13
13
13

A thermogenic effect of 4.3 +/- 1.3 watt was

established for the Ephedrine group (Arm 1), the
effect in the Placebo group (Arm 2) was only 1.6 +/1.6. The thermogenic effect in the Ephedrine group
(Arm 1) increased by 100% ( p < 0.05) following
aerobic training. Overall, aerobic training increased
VO2 maximum by 7 % ( p < 0.05).
For muscular endurance outcomes, mean number of
leg and bench press repetitions only in the first set
increased significantly (p < 0.05) for individuals in the
Caffeine and Ephedrine ( Arm 4) and the Ephedrine
(Arm 3) groups compared to the Caffeine (Arm 2) and
Placebo (Arm 1) groups. The mean number for all 3
sets of leg and bench repetitions was similar among
all groups.

Evidence Table 1 RCTs and CCTs reporting on Athletic Performance Enhancement with Ephedra (continued)
First Author
Year
Sidney KH &
Lefcoe NM
1977 #247

Design
Study Quality
Population (>75%)
Comorbidities
CCT
Jadad Score:
2
Population:
Male
Comorbidities:
N/A

219
N/A = not available or not applicable

Intervention
Total Daily Dose
Route of Administration
Arm # Duration
1 Placebo
Placebo for 1 day
2 Ephedrine
24 mg orally for 1 day

Sample Size
n Entered:
n Analyzed:
n Entered:
n Analyzed:

Summary of Results
21 No significant difference was seen between the
21 Placebo (Arm 1) and Ephedrine (Arm 2) groups for
21 any variable including VO2 maximum, and endurance.
21

Evidence Table 2 RCTs and CCTs reporting on Weight Loss

Design
Study Quality
First Author Population (>75%)
Year
Comorbidities
Astrup A,
CCT
Buemann B, et Jadad Score:
2
Population:
Female
al. 1992 #9
Comorbidities:
Obesity

220

Belfie L, Petrie CCT

Jadad Score:
H, et al.
2001 #317
Population:
Comorbidities:

1
N/A
Obesity

Boozer CN,
RCT
Daly PA, et al. Jadad Score:
Population:
2000 #34
Comorbidities:

5
Female
Obesity

Boozer CN,
Nasser JA, et
al. 2001 #333

RCT
Jadad Score:
Population:
Comorbidities:

5
Female
Obesity

Breum L,
RCT
Pedersen JK, Jadad Score:
et al. 1994 #41 Population:
Comorbidities:

4
Female
Obesity

Buemann B,
RCT
Marckmann P, Jadad Score:
et al. 1994 #45 Population:
Comorbidities:

3
Female
Obesity

Intervention
Total Daily Dose
Route of Administration
Arm # Duration
1 Placebo
Placebo for 8 weeks
2 Ephedrine
60 mg orally for 8 weeks
Caffeine
600 mg orally for 8 weeks
1 Placebo
Placebo for 12 weeks
2 Ephedrine from Ma Huang
60 mg orally for 12 weeks
Caffeine from Guarana
600 mg orally for 12 weeks
1 Placebo
Placebo for 24 weeks
2 Ephedrine from Ma Huang
86.4 mg orally for 24 weeks
Caffeine from Kola nut
196 mg orally for 24 weeks
1 Placebo
Placebo for 8 weeks
2 Ephedrine from Ma Huang
77.4 mg orally for 8 weeks
Caffeine from Guarana
300 mg orally for 8 weeks
1 Dexfenfluramine
30 mg orally for 15 weeks
2 Ephedrine
60 mg orally for 15 weeks
Caffeine
600 mg orally for 15 weeks
1 Placebo
Placebo for 8 weeks
2 Ephedrine
60 mg orally for 8 weeks
Caffeine
600 mg orally for 8 weeks

N/A = not available or not applicable

* Meta-analysis data reports standard deviation in parentheses.

Sample Size
n Entered:
n Analyzed:
n Entered:
n Analyzed:
n Entered:
n Analyzed:
n Entered:
n Analyzed:

Meta-analysis Data*
Or Summary of Results
8 Average weight loss at 2 months in kg:
6
Arm 1 = 8.4 (2.9)
Arm 2 = 10.1 (1.0)
8
6
N/A
10
N/A
11

Excluded from meta-analysis due to Insufficient

statistics. At follow up, decreases were seen only in
the Ma Huang Supplement group (Arm 2) for mass
(106.0 +/-11.5 to 96.9 +/- 12.1 kg), fat mass (31.3 +/5.3 to 25.8 +/- 5.8 kg, p < 0.05), and percent body fat
(29.4 +/- 3.1 to 26.4 +/- 3.0 %, p < 0.05).

n Entered:
n Analyzed:
n Entered:
n Analyzed:

84 Average weight loss at 6 months in kg:

38
Arm 1 = 2.6 (3.2)
Arm 2 = 5.3 (5.0)
83
45

n Entered:
n Analyzed:
n Entered:
n Analyzed:

32 Average weight loss at 2 months in kg:

24
Arm 1 = 0.8 (2.4)
Arm 2 = 4.0 (3.4)
35
24

n Entered:
n Analyzed:
n Entered:
n Analyzed:

53 Average weight loss at 3.75 months in kg:

43
Arm 1 = 6.9 (4.3)
Arm 2 = 8.3 (5.2)
50
38

n Entered:
n Analyzed:
n Entered:
n Analyzed:

N/A Average weight loss at 2 months in kg:

16
Arm 1 = 7.1 (2.4)
Arm 2 = 8.4 (2.4)
N/A
16

Evidence Table 2 RCTs and CCTs reporting on Weight Loss (continued)

Design
Study Quality
First Author Population (>75%)
Year
Comorbidities
Colker, Swain, RCT
et al. 2001
Jadad Score: 2
Population:
Female
#548
Comorbidities: Obesity

Colker, Torina, RCT

Jadad Score: 1
et al. 1999
Population:
N/A
#549
Comorbidities: Obesity

221

Daly PA,
RCT
Krieger DR, et Jadad Score:
Population:
al. 1993 #68
Comorbidities:

Intervention
Total Daily Dose
Route of Administration
Arm # Duration
1
Placebo
Placebo for 8 weeks
2
Ephedrine from Ma Huang
Taken orally for 8 weeks
Coleus forksohlli
Taken orally for 8 weeks
1
Placebo
Placebo for 8 Weeks
2
Ephedrine from Ma Huang
60 mg orally for 8 weeks
Caffeine from unspecified herb
600 mg orally for 8 weeks
Aspirin
45 mg orally for 8 weeks
1

2
Female
Obesity

Placebo
Placebo for 8 weeks
Ephedrine
75 mg orally for 4 weeks
Second round of previous
intervention
150 mg orally for 4 weeks
Caffeine
150 mg orally for 8 weeks
Aspirin
330 mg orally for 8 weeks

N/A = not available or not applicable

* Meta-analysis data reports standard deviation in parentheses.

Sample Size
n Entered:
n Analyzed:
n Entered:
n Analyzed:
n Entered:
n Analyzed:
n Entered:
n Analyzed:

n Entered:
n Analyzed:
n Entered:
n Analyzed:

Meta-analysis Data*
Or Summary of Results
12 Average weight loss at 2 months in kg:
Arm 1 = 0.49 (2.35)
12
Arm 2 = 2.56 (2.35)
14
14
8
8
8
8

Excluded from meta-analysis because of insufficient

statistics: study reports weight loss for one group
only. The Ephedra, Caffeine, Aspirin, and Exercise
(E+C+A+E) group (Arm 3) had a significant reduction
in body weight (-3.8 kg, p<0.01) compared to the
Ephedra, Caffeine, and Aspirin (E+C+A, Arm 2) and
Placebo groups (Arm 1). The E+C+A (Arm 2) group
experienced a significant reduction in caloric intake (680.2 kcal, p<0.05) compared to the other groups.

15 Average weight loss at 2 months in kg:

13
Arm 1 = 0.7 (2.2)
Arm 2 = 2.2 (2.3)
14
11

Evidence Table 2 RCTs and CCTs reporting on Weight Loss (continued)

First Author
Year
Donikyan LA
2002 #509

222

Greenway F,
deJonge L, et
al.
Unpublished
#475

Design
Study Quality
Population (>75%)
Comorbidities
RCT
Jadad Score:
4
Population:
Male and
female
Comorbidities:
Obesity

RCT
Jadad Score:
Population:
Comorbidities:

2
N/A
Obesity

Jensen, Dano, RCT

Jadad Score:
et al. 1980
Population:
#536
Comorbidities:

1
N/A
Obesity

Intervention
Total Daily Dose
Route of Administration
Arm # Duration
1 Placebo
Placebo for 12 weeks
2 Ephedrine from Ma Huang
72 mg orally for 8 weeks
Chromium picolinate
450 mcq orally for 8 weeks
Placebo
Placebo for 4 weeks
3 Ephedrine from Ma Huang
72 mg orally for 12 weeks
Chromium picolinate
450 mcq orally for 12 weeks
1 Placebo
Placebo for 12 weeks
2 Ephedrine from Ma Huang
72 mg orally for 12 weeks
Caffeine from unspecified herb
210 mg orally for 12 weeks
Phenylalanine
300 mg orally for 12 days
1 Ephedrine
100 mg orally for 16 weeks
Caffeine
275 mg orally for 16 weeks
2 Ephedrine
100 mg orally for 16 weeks
3 Placebo
No dosage data reported

N/A = not available or not applicable

* Meta-analysis data reports standard deviation in parentheses.

Meta-analysis Data*
Or Summary of Results
Average weight loss at 3 months in kg:
Arm 1 = 3.0 (6.0)
Arm 2 = excluded
Arm 3 = 7.4 (6.8)

Sample Size
n Entered:
n Analyzed:
n Entered:
n Analyzed:

94
78
93
75

n Entered:
n Analyzed:

92
76

n Entered:
n Analyzed:
n Entered:
n Analyzed:

20 Average weight loss at 3 months in kg:

18
Arm 1 = 0.8 (2.6)
Arm 2 = 3.9 (4.0)
20
12

n Entered:
n Analyzed:

23 Average weight loss at 4 months in kg:

14
Arm 1 = 9.4 (4.7)
Arm 2 = 7.9 (4.7)
Arm 3 = 0.5 (4.7)
24
13
17
4

n Entered:
n Analyzed:
n Entered:
n Analyzed:

Evidence Table 2 RCTs and CCTs reporting on Weight Loss (continued)

Design
Study Quality
First Author Population (>75%)
Year
Comorbidities
Kalman DS,
RCT
Colker CM, et Jadad Score:
3
Male
al. 2000 #140 Population:
Comorbidities:
Obesity

Kalman,
Colker, et al.
2000 #550

RCT
Jadad Score: 3
Population:
N/A
Comorbidities: Obesity

223
Kettle R,
CCT
Toubro S, et al. Jadad Score:
Population:
1998 #510
Comorbidities:

0
N/A
Obesity

Lumholtz IB,
Thorsteinsson
B, et al. 1980
#173

2
N/A
Obesity

RCT
Jadad Score:
Population:
Comorbidities:

Intervention
Total Daily Dose
Route of Administration
Arm # Duration
1 Placebo
Placebo for 8 weeks
2 Ephedrine
40 mg orally for 8 weeks
Synephrine
10 mg orally for 8 weeks
Caffeine
400 mg orally for 8 weeks
Aspirin
30 mg orally for 8 weeks
1
Placebo
Placebo for 8 weeks
2
Ma Huang/Ephedra
20 mg orally for 8 weeks
28
5 mg orally for 8 weeks
Caffeine from unspecified herb
200 mg orally for 8 weeks
Aspirin
15 mg orally for 8 weeks
1
2

1
2

Placebo
Placebo for 6 months
Ephedrine
20 mg orally for 6 months
Caffeine
200 mg orally for 6 months
Ephedrine
120 mg orally for 18 weeks
Placebo
No dosage data reported

N/A = not available or not applicable

* Meta-analysis data reports standard deviation in parentheses.

Sample Size
n Entered:
n Analyzed:
n Entered:
n Analyzed:

n Entered:
n Analyzed:
n Entered:
n Analyzed:

Meta-analysis Data*
Or Summary of Results
14 Average weight loss at 2 months in kg:
Arm 1 = 2.1 (2.4)
13
Arm 2 = 3.1 (2.4)
16
12

15
15
15
15

Excluded from meta-analysis because of insufficient

statistics: study only reports weight loss in percent.
Subjects in the Ephedrine, Synephrine, Caffeine, and
Aspirin (E+S+C+A) group (Arm 2) experienced a
significant reduction in body weight (-9%, p0.05) as
well as in percent of body fat (-16%, p<0.001)
compared to the Placebo group (Arm 1, -3.8% and 1% respectively). An intragroup difference in fat free
mass was seen in both groups: -0.92 kg (p<0.01) in
the E+S+C+A group (Arm 2) and -3.47 kg (p<0.05) in
the Placebo group (Arm 1).

n Entered:
n Analyzed:
n Entered:
n Analyzed:

45 Average weight loss at 6 months in kg:

37
Arm 1 = 12.8 (6.7)
Arm 2 = 15.6 (7.1)
45
40

n Entered:
n Analyzed:
n Entered:
n Analyzed:

63 Average weight loss at 4.5 months in kg:

18
Arm 1 = 9.5 (5.3)
Arm 2 = 4.0 (5.3)
63
14

Evidence Table 2 RCTs and CCTs reporting on Weight Loss (continued)

Design
Study Quality
First Author Population (>75%)
Year
Comorbidities
Malchow-Moller CCT
A, Larsen S, et Jadad Score:
3
N/A
al. 1981 #177 Population:
Comorbidities:
Obesity

Moheb MA,
RCT
Geissler CA, et Jadad Score:
al. 1998 #193 Population:
Comorbidities:

2
Female
Obesity

224

Intervention
Total Daily Dose
Route of Administration
Arm # Duration
1 Placebo
Placebo for 12 weeks
2 Ephedrine
60 mg orally for 12 weeks
Caffeine
150 mg orally for 12 weeks
3 Diethylpropion
37.5 mg orally for 12 weeks
1 Placebo
Placebo for 12 weeks
2 Ephedrine
150 mg orally for 12 weeks
3 Ephedrine
150 mg orally for 12 weeks
Aspirin
330 mg orally for 12 weeks
4 Ephedrine
150 mg orally for 12 weeks
Caffeine
150 mg orally for 12 weeks
5 Ephedrine
150 mg orally for 12 weeks
Caffeine
150 mg orally for 12 weeks
Aspirin
330 mg orally for 12 weeks

N/A = not available or not applicable

* Meta-analysis data reports standard deviation in parentheses.

Sample Size
n Entered:
n Analyzed:
n Entered:
n Analyzed:

33
31
49
38

Meta-analysis Data*
Or Summary of Results
Average weight loss at 3 months in kg:
Arm 1 = 4.1 (3.5)
Arm 2 = 8.1 (3.5)
Arm 3 = 8.4 (3.5)

n Entered:
n Analyzed:
n Entered:
n Analyzed:
n Entered:
n Analyzed:
n Entered:
n Analyzed:

50
39
N/A Average weight loss at 3 months in kg:
32
Arm 1 = 6.2 (3.5)
Arm 2 = 7.9 (3.5)
N/A
Arm 3 = 9.6 (3.5)
32
Arm 4 = 8.8 (3.5)
N/A
Arm 5 = 8.9 (3.5)
32

n Entered:
n Analyzed:

N/A
32

n Entered:
n Analyzed:

N/A
32

Evidence Table 2 RCTs and CCTs reporting on Weight Loss (continued)

225

Design
Intervention
Study Quality
Total Daily Dose
First Author Population (>75%)
Route of Administration
Year
Comorbidities
Arm # Duration
Molnar D,
RCT
1 Placebo
Torok K, et al. Jadad Score: 4
Placebo for 20 weeks
2000 #195
Population:
Adolescents
2 Ephedrine
(12-17)
10 mg orally for 1 weeks
Second round of previous
Comorbidities: Obesity
intervention
30-60 mg orally for 19 weeks
Caffeine
100 mg orally for 1 weeks
Second round of previous
intervention
300-600 mg orally for 19 weeks
Norregaard J, RCT
1 Placebo
3
Placebo for 9 months
Jorgensen S, et Jadad Score:
N/A
al. 1996 #210 Population:
2 Ephedrine
Comorbidities:
Obesity,
60 mg orally for 3 months
hypertension, pulmonary,
Second round of previous
AVD.
intervention
40 mg orally for 3 months
Third round of previous
intervention
20 mg orally for 3 months
Caffeine
600 mg orally for 3 months
Second round of previous
intervention
400 mg orally for 3 months
Third round of previous
intervention
200 mg orally for 3 months
Pasquali R,
RCT
1 Placebo
3
Placebo for 3 months
Baraldi G, et al. Jadad Score:
Population:
N/A
1985 #220
2 Ephedrine
Comorbidities:
Obesity
75 mg orally for 3 months
3 Ephedrine
150 mg orally for 3 months
N/A = not available or not applicable
* Meta-analysis data reports standard deviation in parentheses.

Sample Size
n Entered:
n Analyzed:
n Entered:
n Analyzed:

n Entered:
n Analyzed:
n Entered:
n Analyzed:

n Entered:
n Analyzed:
n Entered:
n Analyzed:
n Entered:
n Analyzed:

Meta-analysis Data*
Or Summary of Results
Average
weight loss at 5 months in kg:
16
Arm 1 = 0.5 (4.3)
13
Arm 2 = 7.9 (6.0)
16
16

80
73
167
152

Excluded from meta-analysis because there was no

weight loss outcome, this study addressed weight
gain. Subjects in the Ephedrine plus Caffeine group
(Arm 2) gained significantly less weight during the first
12 weeks (Week 3 = p<0.001;
Week 6 = p<0.01; Week 12 = p<0.05) than subjects
in the Placebo group (Arm 1). Weight gain was similar
for both groups after 1 year.

21 Average weight loss at 3 months in kg:

12
Arm 1 = 8.7 (3.5)
Arm 2 = 8.7 (2.4)
19
Arm 3 = 10.2 (3.5)
7
22
12

Evidence Table 2 RCTs and CCTs reporting on Weight Loss (continued)

First Author
Year
Pasquali R,
Cesari MP, et
al. 1987 #223

Design
Study Quality
Population (>75%)
Comorbidities
RCT
Jadad Score:
2
Population:
Female
Comorbidities:
Obesity

Quaade F,
RCT
Astrup A, et al. Jadad Score:
1992 #230
Population:
Comorbidities:

Intervention
Total Daily Dose
Route of Administration
Arm # Duration
1 Placebo
Placebo for 2 months
2 Ephedrine
150 mg orally for 2 months
1

3
Male and
female
Obesity

2
3
4

226

Roed, Hansen, RCT

et al. 1980
Jadad Score:
Population:
#535
Comorbidities:

1
3
Male and
female
Obesity
2

3
Toubro S &
RCT
Astrup A 1997 Jadad Score:
Population:
#261
Comorbidities:

1
2
Female
Obesity
2

Ephedrine
60 mg orally for 24 weeks
Caffeine
600 mg orally for 24 weeks
Ephedrine
60 mg orally for 24 weeks
Caffeine
600 mg orally for 24 weeks
Placebo
No dosage data reported
Ephedrine
60 mg orally for 12 weeks
Caffeine
150 mg orally for 12 weeks
Phenobarbital
60 mg orally for 12 weeks
Ephedrine
60 mg orally for 12 weeks
Caffeine
150 mg orally for 12 weeks
Placebo
No dosage data reported
Ephedrine
60 mg orally for 8 weeks
Caffeine
600 mg orally for 8 weeks
Ephedrine
60 mg orally for 17 weeks
Caffeine
600 mg orally for 17 weeks

N/A = not available or not applicable

* Meta-analysis data reports standard deviation in parentheses.

Meta-analysis Data*
Or Summary of Results
Excluded from meta-analysis because crossover
study design. Patients' weight loss was significantly
(p<0.05) more during the Ephedrine treatment (Arm 2,
2.41 +/- 0.6 kg.) than during the Placebo treatment
(Arm 1, 0.64 +/- 0.05 kg.).
Average weight loss at 3 months in kg:
Arm 1 = 11.7 (5.3)
Arm 2 = 10.3 (4.0)
Arm 3 = 9.0 (3.6)
Arm 4 = 10.2 (5.7)
Average weight loss at 6 months in kg:
Arm 1 = 16.6 (6.8)
Arm 2 = 14.3 (5.9)
Arm 3 = 11.5 (6.0)
Arm 4 = 13.2 (6.6)
Average weight loss at 3 months in kg:
Arm 1 = excluded
Arm 2 = 10.0 (3.5)
Arm 3 = 5.2 (3.5)

Sample Size
n Entered:
n Analyzed:
n Entered:
n Analyzed:

10
10
10
10

n Entered:
n Analyzed:

45
35

n Entered:
n Analyzed:
n Entered:
n Analyzed:
n Entered:
n Analyzed:
n Entered:
n Analyzed:

45
35
45
36
45
35
70
49

n Entered:
n Analyzed:

69
52

n Entered:
n Analyzed:
n Entered:
n Analyzed:

69
42
21 Excluded from meta-analysis due to study design:
19 ephedrine dose did not vary between arms.
The mean weight loss achieved during the reduction
phase was 12.6 kg (95% CI: 10.9-14.3) for the Low
22 Energy Diet (LED) group (Arm1) and 12.6 kg (CI: 9.919 15.3) for the Conventional Diet (CD) group (Arm 2).
The rate of weight loss was twice as high in the CD
group (Arm 2, 1.6 kg/week, CI: 1.4 -1.8) than in the
LED group (Arm 1, 0.8 kg/week, CI: 0.7-1.0).

n Entered:
n Analyzed:

Evidence Table 2 RCTs and CCTs reporting on Weight Loss (continued)

Design
Intervention
Study Quality
Total Daily Dose
First Author Population (>75%)
Route of Administration
Year
Comorbidities
Arm # Duration
Van Mil E &
RCT
1 Placebo
Molnar D 2000 Jadad Score: 1
Placebo for 20 weeks
Population:
Adolescents
#272
2 Ephedrine
(12-17)
60 mg orally for 20 weeks
Comorbidities: Obesity
Caffeine
600 mg orally for 20 weeks

227
N/A = not available or not applicable
* Meta-analysis data reports standard deviation in parentheses.

Sample Size
n Entered:
n Analyzed:
n Entered:
n Analyzed:

Meta-analysis Data*
Or Summary of Results
16 Average weight loss at 5 months in kg:
Arm 1 = 1.5 (8.1)
16
Arm 2 = 8.7 (5.7)
16
16

Acronyms
AEA
AHRQ
ARMS
BMI
CCT
CI
CPK isozymes
CPR
CT scan
CVA
CVD
DF
DSHEA
EPC
FDA
GAO
HHS
IOC
kg
MB fractions
MI
mg
MRI
NCAA
NHANES
NIH
ODS
OTC
PDF
QRF
RCT
TEP
VCO2
VO2

Adverse events analysis

Agency for Healthcare Research and Quality
Adverse Reaction Monitoring System
Body Mass Index
Controlled clinical trial
Confidence interval
Creatine phosphokinase isoenzyme
Cardio-plumonary resuscitation
Computerized tomography scan
Cerebral vascular accident
Cardiovascular diseases
Dexfenfluramine
Dietary Supplement Health and Education Act
Evidence-based Practice Center
US Food and Drug Administration
General Accounting Office
US Department of Health and Human Services
International Olympic Committee
kilograms
Myocardial band fractions (of CPK isoenzymes)
Myocardial infarction
milligrams
Magnetic resonance imagery
National Collegiate Athletic Association
National Health and Nutrition Examination Survey
National Institutes of Health
Office of Dietary Supplements
Over-the-counter
Portable document format
Quality review form
Randomized controlled trial
Technical Expert Panel
Volume of carbon dioxide production
Volume of oxygen consumption

228

Appendix 1. Bibliography
Articles Accepted for Efficacy Analysis
7. Bell DG, Jacobs I, Ellerington K. Effect of
caffeine and ephedrine ingestion on
anaerobic exercise performance. Med Sci
Sports Exercise. 2001;33:1399-1403. [Rec#:
512]

1. Astrup A, Breum L, Toubro S, Hein P, Quaade

F. The effect and safety of an
ephedrine/caffeine compound compared to
ephedrine, caffeine and placebo in obese
subjects on an energy restricted diet. A
double blind trial. Int J Obes Relat Metab
Disord. 1992;16:(4)269-77. [Rec#: 7]

8. Bell DG, Jacobs I, McLellan TM, Miyazaki

M, Sabiston CM. Thermal regulation in the
heat during exercise after caffeine and
ephedrine ingestion. Aviat Space Environ
Med. 1999;70:(6)583-8. [Rec#: 25]

2. Astrup A, Breum L, Toubro S, Hein P, Quaade

F. Ephedrine and weight loss. Int J Obes
Relat Metab Disord. 1992;16:(9)715. [Rec#:
8]

9. Bell DG, Jacobs I, McLellan TM, Zamecnik J.

Reducing the dose of combined caffeine and
ephedrine preserves the ergogenic effect.
Aviat Space Environ Med. 2000;71:(4)4159. [Rec#: 26]

3. Astrup A, Buemann B, Christensen NJ,

Toubro S, Thorbek G, Victor OJ, et al. The
effect of ephedrine/caffeine mixture on
energy expenditure and body composition in
obese women. Metabolism. 1992;41:(7)6868. [Rec#: 9]

10. Bell DG, Jacobs I, Zamecnik J. Effects of

caffeine, ephedrine and their combination on
time to exhaustion during high-intensity
exercise. Eur J Appl Physiol Occup Physiol.
1998;77:(5)427-33. [Rec#: 27]

4. Astrup A, Toubro S, Cannon S, Hein P,

Madsen J. Thermogenic synergism between
ephedrine and caffeine in healthy
volunteers: a double-blind, placebocontrolled study. Metabolism.
1991;40:(3)323-9. [Rec#: 13]

11. Boozer CN, Daly PA, Homel P, et al. Herbal

ephedra/caffeine for weight loss: a 6-month
randomized safety and efficacy trial.
International Journal of Obesity. (In Press).
[Rec#: 474]

5. Belfie L, Petrie H, Chown S, Duncan AM,

Gawron SJ, Conquer J, et al. Safety and
effectiveness of an herbal dietary
supplement containing ephedra (ma huang)
and caffeine (guarana extract) when used in
combination with a supervised diet and
exercise intervention. Obesity Research
Program Abstracts. 2001;9:(Suppl 3)26.
[Rec#: 317]

12. Boozer CN, Nasser JA, Heymsfield SB, Wang

V, Chen G, Solomon JL. An herbal
supplement containing Ma Huang-Guarana
for weight loss: a randomized, double-blind
trial. Int J Obes Relat Metab Disord.
2001;25:(3)316-24. [Rec#: 333]

6. Bell DG, Jacobs I. Combined caffeine and

ephedrine ingestion improves run times of
Canadian Forces Warrior Test. Aviat Space
Environ Med. 1999;70:(4)325-9. [Rec#: 24]

13. Breum L, Pedersen JK, Ahlstrom F, FrimodtMoller J. Comparison of an

ephedrine/caffeine combination and
dexfenfluramine in the treatment of obesity.
A double-blind multi-centre trial in general
practice. Int J Obes Relat Metab Disord.
1994;18:(2)99-103. [Rec#: 41]

A1-1

Articles Accepted for Efficacy Analysis (continued)

14. Buemann B, Marckmann P, Christensen NJ,
Astrup A. The effect of ephedrine plus
caffeine on plasma lipids and lipoproteins
during a 4.2 MJ/day diet. Int J Obes Relat
Metab Disord. 1994;18:(5)329-32. [Rec#:
45]

22. Horton TJ, Geissler CA. Aspirin potentiates

the effect of ephedrine on the thermogenic
response to a meal in obese but not lean
women. Int J Obes. 1991;15:(5)359-66.
[Rec#: 123]
23. Horton TJ, Geissler CA. Post-prandial
thermogenesis with ephedrine, caffeine and
aspirin in lean, pre-disposed obese and
obese women. Int J Obes Relat Metab
Disord. 1996;20:(2)91-7. [Rec#: 124]

15. Daly PA, Krieger DR, Dulloo AG, Young JB,

Landsberg L. Ephedrine, caffeine and
aspirin: safety and efficacy for treatment of
human obesity. Int J Obes Relat Metab
Disord. 1993;17:(Suppl 1)S73-8. [Rec#: 68]

24. Ingerslev J, Svendsen TL, Mork A. Is an

ephedrine caffeine treatment contraindicated
in hypertension? Int J Obesity Related
Metabol Disord. 1997;21:(8)666-673.
[Rec#: 132]

16. de Jonge L, Frisard M, Blanchard D,

Greenway F. Safety and efficacy of an
herbal dietary supplement containing
caffeine and ephedra for obesity treatment.
Obesity Research Program Abstracts.
2001;9:(Suppl 3)20. [Rec#: 316]

25. Jaedig S, Henningsen NC. Increased metabolic

rate in obese women after ingestion of
potassium, magnesium- and phosphateenriched orange juice or injection of
ephedrine. Int J Obes. 1991;15:(6)429-36.
[Rec#: 391]

17. Donikyan LA. A double-blind randomized,

placebo-controlled, multicenter clinical
study to evaluate the safety and efficacy of a
natural herbal formulation when taken as
recommended for weight control. Boca
Raton, FL: Rexall Sundown. 2002. [Rec#:
509]

26. Jensen KB, Dano P, Draeby N, Hansen SH,

Kanstrup J. [Elsinore tablets and ephedrine
as slimming agents]. Ugeskr Laeger.
1980;142:(23)1499-501. [Rec#: 536]

18. Dulloo AG, Miller DS. The thermogenic

properties of ephedrine/methylxanthine
mixtures: human studies. Int J Obes.
1986;10:(6)467-81. [Rec#: 83]

27. Jonderko K, Kucio C. Effect of anti-obesity

drugs promoting energy expenditure,
yohimbine and ephedrine, on gastric
emptying in obese patients. Aliment
Pharmacol Ther. 1991;5:(4)413-8. [Rec#:
137]

19. Geissler CA. Effects of weight loss, ephedrine

and aspirin on energy expenditure in obese
women. Int J Obes Relat Metab Disord.
1993;17(Suppl 1):S45-8. [Rec#: 97]
20. Greenway F, deJonge L, Blanchard D, et al.
Evaluation of a dietary herbal supplement
containing caffeine and ephedrine on
metabolic rate, body composition, serum
lipids and tolerability. (Unpublished).
[Rec#: 475]

28. Kaats GR, Aldeman JA. Effects of multiple

herbal formulation on body composition,
blood chemistry, vital signs and selfreported energy levels and appetite control.
Int J Obesity Related Metabol Disord.
1994;18:(Suppl 2)145(Abstract). [Rec#:
139]

21. Greenway FL, Raum WJ, DeLany JP. The

effect of an herbal dietary supplement
containing ephedrine and caffeine on
oxygen consumption in humans. J Altern
Complement Med. 2000;6:(6)553-5. [Rec#:
106]

29. Kalman D, Incledon T, Gaunaurd I, et al. An

acute clinical trial evaluating the
cardiovascular effects of an herbal
ephedra/caffeine weight loss product in
healthy overweight adults. International
Journal of Obesity. (In Press). [Rec#: 476]

A1-2

Articles Accepted for Efficacy Analysis (continued)

30. Kalman DS, Colker CM, Shi Q, Swain MA.
Effects of a weight-loss aid in healthy
overweight adults: Double-blind, placebocontrolled clinical trial. Curr Therapeut Res.
2000;61:(4)199-205. [Rec#: 140]

38. Molnar D. Effects of ephedrine and

aminophylline on resting energy expenditure
in obese adolescents. Int J Obes Relat Metab
Disord. 1993;17:(Suppl 1)S49-52. [Rec#:
194]

31. Kettle R, Toubro S, Astrup A. Ephedrine,

caffeine enhances abdominal fat loss in
females. Int J Obesity Related Metabol
Disord. 1998;22(Suppl 3):S264(Abstract).
[Rec#: 510]

39. Molnar D, Torok K, Erhardt E, Jeges S. Safety

and efficacy of treatment with an
ephedrine/caffeine mixture. The first
double-blind placebo-controlled pilot study
in adolescents. Int J Obes Relat Metab
Disord. 2000;24:(12)1573-8. [Rec#: 195]

32. Krieger DR, Daly PA, Dulloo AG, Ransil BJ,

Young JB, Landsberg L. Ephedrine, caffeine
and aspirin promote weight loss in obese
subjects. Trans Assoc Am Physicians.
1990;103:307-12. [Rec#: 151]

40. Morgan JB, York DA, Wasilewska A,

Portman J. A study of the thermic responses
to a meal and to a sympathomimetic drug
(ephedrine) in relation to energy balance in
man. Br J Nutr. 1982;47:(1)21-32. [Rec#:
197]

33. Liu YL, Toubro S, Astrup A, Stock MJ.

Contribution of beta 3-adrenoceptor
activation to ephedrine-induced
thermogenesis in humans. Int J Obes Relat
Metab Disord. 1995;19:(9)678-85. [Rec#:
169]

41. Nasser JA, Wang V, Chen GC, Solomon JL,

Heymsfield SB, Boozer CN. Efficacy trial
for weight loss of an herbal supplement of
Ma huang and guarana. FASEB J. 1999;13:
(5,Prt.2)A874(Abstract). [Rec#: 204]

34. Lumholtz IB, Thorsteinsson B, Wamberg T,

Lehnschau A, Hansen G, Spellerberg S, et
al. [Ephedrine in the treatment of obesity. A
double-blind cross-over trial of the effect of
Elsinore tablets]. Ugeskr Laeger.
1980;142:(23)1487-90. [Rec#: 173]

42. Norregaard J, Jorgensen S, Mikkelsen KL,

Tonnesen P, Iversen E, Sorensen T, et al.
The effect of ephedrine plus caffeine on
smoking cessation and postcessation weight
gain. Clin Pharmacol Ther. 1996;60:(6)67986. [Rec#: 210]

35. Malchow-Moller A, Larsen S, Hey H,

Stokholm KH, Juhl E, Quaade F. Ephedrine
as an anorectic: the story of the 'Elsinore
pill.' Int J Obes. 1981;5:(2)183-7. [Rec#:
177]

43. Oksbjerg N, Meyer T, Hvid-Jacobsen K.

[Aerobic training increases the effect of
ephedrine on energy expenditure]. Ugeskr
Laeger. 1986;148:(32)2021-4. [Rec#: 214]

36. Martinet A, Hostettmann K, Schutz Y.

Thermogenic effects of commercially
available plant preparations aimed at
treating human obesity. Phytomedicine.
1999;6:(4)231-8. [Rec#: 335]

44. Pasquali R, Baraldi G, Cesari MP, Melchionda

N, Zamboni M, Stefanini C, et al. A
controlled trial using ephedrine in the
treatment of obesity. Int J Obes.
1985;9:(2)93-8. [Rec#: 220]

37. Moheb MA, Geissler CA, Lancer K. Effect of

ephedrine, caffeine, and aspirin, in
combinations of weight loss in obese
women. Int J Obesity Related Metabol
Disord. 1998;22:(Suppl 3)S264(Abstract).
[Rec#: 193]

45. Pasquali R, Casimirri F, Melchionda N, Grossi

G, Bortoluzzi L, Morselli Labate AM, et al.
Effects of chronic administration of
ephedrine during very-low-calorie diets on
energy expenditure, protein metabolism and
hormone levels in obese subjects. Clin Sci
(Colch). 1992;82:(1)85-92. [Rec#: 222]

A1-3

Articles Accepted for Efficacy Analysis (continued)

54. Toubro S, Astrup A V, Breum L, Quaade F.
Safety and efficacy of long-term treatment
with ephedrine, caffeine and an
ephedrine/caffeine mixture. Int J Obes Relat
Metab Disord. 1993;17: (Suppl 1)S69-72.
[Rec#: 263]

46. Pasquali R, Cesari MP, Melchionda N,

Stefanini C, Raitano A, Labo G. Does
ephedrine promote weight loss in lowenergy-adapted obese women? Int J Obes.
1987;11:(2)163-8. [Rec#: 223]
47. Pasternak H, Jacobs I, Bell D. Effects of
ingesting caffeine and ephedrine on
muscular endurance. Can J Appl Physiol.
1999;24:(5)471. [Rec#: 511]

55. Van Mil E, Molnar D. Drug treatment in obese

adolescents. Int J Obesity. 2000. 24:(Suppl
1)S184(Abstract). [Rec#: 272]

48. Quaade F, Astrup A, Breum L, Toubro S, Hein

P. [The effect of an ephedrine/caffeine
combination as a supplement to a weightreducing diet. A randomized, placebocontrolled, double-blind trial]. Ugeskr
Laeger. 1992;154:(18)1258-63. [Rec#: 230]

56. Waluga M, Janusz M, Karpel E, Hartleb M,

Nowak A. Cardiovascular effects of
ephedrine, caffeine and yohimbine measured
by thoracic electrical bioimpedance in obese
women. Clin Physiol. 1998;18:(1)69-76.
[Rec#: 276]

49. Roed P, Hansen PW, Bidstrup B, Kaern M,

Helles A, Petersen KP. [Elsinore banting
tablets. A controlled clinical trial in general
practice]. Ugeskr Laeger.
1980;142:(23)1491-5. [Rec#: 535]
50. Shannon JR, Gottesdiener K, Jordan J, Chen
K, Flattery S, Larson PJ, et al. Acute effect
of ephedrine on 24-h energy balance. Clin
Sci (Colch). 1999;96:(5)483-91. [Rec#: 243]
51. Shugarman AE. Effect of thermogenic dietary
supplements on resting metabolic rate in
healthy male and female volunteers. Utah:
University of Utah. 1998. [Rec#: 305]
52. Sidney KH, Lefcoe NM. The effects of
ephedrine on the physiological and
psychological responses to submaximal and
maximal exercise in man. Med Sci Sports.
1977;9:(2)95-9. [Rec#: 247]
53. Toubro S, Astrup A. Randomised comparison
of diets for maintaining obese subjects'
weight after major weight loss: ad lib, low
fat, high carbohydrate diet v fixed energy
intake. BMJ. 1997;314:(7073)29-34. [Rec#:
261]

A1-4

Articles Rejected from Efficacy Analysis

11. Astrup A. Commentary on safety of products
containing ephedrine alkaloids/caffeine.
(Unpublished). [Rec#: 477]

1. Abdallah AH, Tye A, Lapidus JB, Patil P.

Effects of ephedrine isomers on myocardial
catecholamines. Life Sci. 1967;6:(1)39-43.
[Rec#: 420]

12. Astrup A. Thermogenesis in human brown

adipose tissue and skeletal muscle induced
by sympathomimetic stimulation. Acta
Endocrinol Suppl (Copenh). 1986;278:1-32.
[Rec#: 389]

2. Abernethy DR, Todd EL, Schwartz JB.

Caffeine disposition in obesity. Br J Clin
Pharmacol. 1985;20:(1)61-6. [Rec#: 378]
3. Acheson KJ, Zahorska-Markiewicz B, Pittet P,
Anantharaman K, Jequier E. Caffeine and
coffee: their influence on metabolic rate and
substrate utilization in normal weight and
obese individuals. Am J Clin Nutr.
1980;33:(5)989-97. [Rec#: 371]

13. Astrup A. Thermogenic drugs as a strategy for

treatment of obesity. Endocrine.
2000;13:(2)207-12. [Rec#: 5]
14. Astrup A, Breum L, Toubro S.
Pharmacological and clinical studies of
ephedrine and other thermogenic agonists.
Obes Res. 1995;3:(Suppl 4)537S-540S.
[Rec#: 6]

4. Ahrendt DM. Ergogenic aids: counseling the

athlete. American Family Physician.
2001;63:(5)913-922. [Rec#: 303]
5. al-Khalil S, Alkofahi A, el-Eisawi D, al-Shibib
A. Transtorine, a new quinoline alkaloid
from Ephedra transitoria. J Nat Prod.
1998;61:(2)262-3. [Rec#: 442]

15. Astrup A, Lundsgaard C. What do

pharmacological approaches to obesity
management offer? Linking
pharmacological mechanisms of obesity
management agents to clinical practice. Exp
Clin Endocrinol Diabetes. 1998;106:(Suppl
2)29-34. [Rec#: 10]

6. Altschuler S, Conte A, Sebok M, Marlin RL,

Winick C. Three controlled trials of weight
loss with phenylpropanolamine. Int J Obes.
1982;6:(6)549-56. [Rec#: 399]

16. Astrup A, Lundsgaard C, Madsen J,

Christensen N J. Enhanced thermogenic
responsiveness during chronic ephedrine
treatment in man. Am J Clin Nutr.
1985;42:(1)83-94. [Rec#: 11]

7. Anastasio GD, Harston P. Fetal tachycardia

associated with maternal use of
pseudoephedrine, an OTC oral decongestant.
J Am Board Fam Pract. 1992;5:(5)527-528.
[Rec#: 3]

17. Astrup A, Madsen J, Holst JJ, Christensen NJ.

The effect of chronic ephedrine treatment on
substrate utilization, the sympathoadrenal
activity, and energy expenditure during
glucose-induced thermogenesis in man.
Metabolism. 1986;35:(3)260-5. [Rec#: 387]

8. Anon. Herbs to handle with care. New York:

Newsweek. 1996.
9. Anselme F, Collomp K, Mercier B, Ahmaidi
S, Prefaut C. Caffeine increases maximal
anaerobic power and blood lactate
concentration. Eur J Appl Physiol Occup
Physiol. 1992;65:(2)188-91. [Rec#: 541]

18. Astrup A, Toubro S. Thermogenic, metabolic,

and cardiovascular responses to ephedrine
and caffeine in man. Int J Obes Relat Metab
Disord. 1993;17:(Suppl 1)S41-3. [Rec#: 12]

10. Assismos DG, Langenstroer P, Leinbach RF,

Mandel NS, Stern JM, Holmes RP.
Guaifenesin - and ephedrine-induced stones.
J Endourol. 1999;13:(9)665-7. [Rec#: 4]

A1-5

Articles Excluded from Efficacy Analysis (continued)

28. Backer R, Tautman D, Lowry S, Harvey C M,
Poklis A. Fatal ephedrine intoxication. J
Forensic Sci. 1997;42:(1)157-9. [Rec#: 17]

19. Astrup A, Toubro S, Cannon S, Hein P, Breum

L, Madsen J. Caffeine: a double-blind,
placebo-controlled study of its thermogenic,
metabolic, and cardiovascular effects in
healthy volunteers. Am J Clin Nutr.
1990;51:(5)759-67. [Rec#: 499]

29. Badiei B, Faciane J, Sly RM. Effect of

theophylline, ephedrine and their
combination upon exercise-induced airway
obstruction. Ann Allergy. 1975;35:(1)32-6.
[Rec#: 18]

20. Astrup A, Toubro S, Christensen NJ, Quaade

F. Pharmacology of thermogenic drugs. Am
J Clin Nutr. 1992;55:(1 Suppl)246S-248S.
[Rec#: 397]

30. Bailey CJ, Thornburn CC, Flatt PR. Effects of

ephedrine and atenolol on the development
of obesity and diabetes in ob/ob mice. Gen
Pharmacol. 1986;17:(2)243-6. [Rec#: 388]

21. Astrup AV. Treatment of obesity with

thermogenic agents. Nutrition. 1989;5:(1)70.
[Rec#: 403]

31. Bale JF Jr., Fountain MT, Shaddy R.

Phenylpropanolamine-associated CNS
complications in children and adolescents.
Am J Dis Child. 1984;138:(7)683-5. [Rec#:
19]

22. Aswal BS, Bhakuni DS, Goel AK, Kar K,

Mehrotra BN. Screening of Indian plants for
biological activity - Part XI. Indian J Exp
Biol. 1984;22:(9)487-504. [Rec#: 14]
23. Atkinson RL. The herbal ephedra and caffeine
debate continues. International Journal of
Obesity. 2002;26:589. [Rec#: 531]

32. Barnes J, Mills SY, Abbot NC, Willoughby M,

Ernst E. Different standards for reporting
ADRs to herbal remedies and conventional
OTC medicines: face-to-face interviews
with 515 users of herbal remedies. Br J Clin
Pharmacol. 1998;45:(5)496-500. [Rec#:
439]

24. Atkinson RL, Blank RC, Loper JF,

Schumacher D, Lutes RA. Combined drug
treatment of obesity. Obes Res.
1995;3:(Suppl 4)497S-500S. [Rec#: 15]

33. Baselt RC. Disposition of toxic drugs and

chemicals in man. Davis, CA: Biomedical
Publications. 1982. pp. 108-111. [Rec#:
433]

25. Audicana M, Urrutia I, Echechipia S, Munoz

D, Fernandez de Corres L. Sensitization to
ephedrine in oral anticatarrhal drugs.
Contact Dermatitis. 1991;24:(3)223. [Rec#:
366]

34. Baselt RC. Ephedra. In: Baselt RC. (Ed)

Disposition of Toxic Drugs and Chemicals
in Man (2nd). Foster City, CA: Chemical
Toxicology Institute. 1982a. [Rec#: 21]

26. Aviado DM, Wnuck AL, De Beer EJ.

Cardiovascular effects of sympathomimetic
bronchodilators: epinephrine, ephedrine,
pseudoephedrine, isoproterenol,
mehtoxyphenamine and isoprophenamine. J
Pharmacology and Experimental
Therapeutics. 1958;122:(407-417). [Rec#:
405]

35. Baselt RC. Caffeine. In: Baselt RC. (Ed)

Drugs and Chemicals in Man (2nd). Foster
City, CA: Chemical Toxicology Institute.
1982b. pp. 108-111. [Rec#: 20]
36. Battig K. Acute and chronic cardiovascular
and behavioural effects of caffeine, aspirin
and ephedrine. Int J Obes Relat Metab
Disord. 1993;17:(Suppl 1)S61-4. [Rec#: 22]

27. Axelrod J. Studies on sympathomimetic

amines: I. The biotransformation and
physiological disposition of 1-ephedrine and
1-norephedrine. J Pharmacol Exper
Therapeut. 1953;109:(1)62-73. [Rec#: 16]

A1-6

Articles Excluded from Efficacy Analysis (continued)

46. Blenkinsopp A, Bradley C. Patients, society,
and the increase in self medication. BMJ.
1996;312:(7031)629-32. [Rec#: 533]

37. Beck RA, Mercado DL, Seguin SM, Andrade

WP, Cushner HM. Cardiovascular effects of
pseudoephedrine in medically controlled
hypertensive patients. Arch Intern Med.
1992;152:(6)1242-5. [Rec#: 23]

47. Blumenthal M. FDA Holds Expert Advisory

Committee Hearing on Ma Huang: Experts
Recommend Appropriate Labeling and
Warnings - Not Banning the Herb.
HERBALGRAM. 1996;36:21-3, 73. [Rec#:
32]

38. Bell DG, Jacobs I, Ellerington K. Effect of

caffeine and ephedrine ingestion on
anaerobic exercise performance. Med Sci
Sports Exerc. 2001;33:(8)1399-403. [Rec#:
481]

48. Blumenthal M, Busse WR. Ephedra. In:

Blumenthal M, Busse WR. (Eds)
(Translated by) Klein S, Rister RS.
Complete German Commission E
Monographs: Therapeutic Guide to Herbal
Medicines. Boston, MA: Integrative
Medicine Communications; 1998. p. 125-6,
414-6, 475-6. [Rec#: 33]

39. Benichou C, Danan G, Flahault A. Causality

assessment of adverse reactions to drugs--II.
An original model for validation of drug
causality assessment methods: case reports
with positive rechallenge. J Clin Epidemiol
1993 46:(11)1331-6. [Rec#: 325]
40. Betz JM, Gay ML, Mossoba MM, Adams S,
Portz BS. Chiral gas chromatographic
determination of ephedrine-type alkaloids in
dietary supplements containing Ma Huang. J
AOAC Int. 1997;80:(2)303-15. [Rec#: 28]

49. Bondi M, Grugni G, Velardo A, Biella O,

Venneri MG, Morabito F, et al.
Adrenomedullary response to caffeine in
prepubertal and pubertal obese subjects. Int
J Obes Relat Metab Disord. 1999;23:(9)9926. [Rec#: 376]

41. Bierman CW, Pierson WE, Shapiro GG.

Exercise-induced asthma. Pharmacological
assessment of single drugs and drug
combinations. JAMA. 1975;234:(3)295-8.
[Rec#: 29]

50. Boozer CN, Daly PA, Blanchard D, Nasser

JA, Solomon JL, Homel P. Herbal
ephedra/caffeine for weight loss: A 6-month
safety and efficacy trial. 2000. (Abstract).
[Rec#: 34]

42. Blanc PD, Kuschner WG, Katz PP, Smith S,

Yelin EH. Use of herbal products, coffee or
black tea, and over-the-counter medications
as self-treatments among adults with asthma.
J Allergy Clin Immunol. 1997;100:(6 Pt
1)789-91. [Rec#: 30]

51. Borum M. Fulminant exacerbation of

autoimmune hepatitis after the use of ma
huang. American Journal of
Gastroenterology. 2001;94:2430-4. [Rec#:
521]

43. Blanc S, Leuenberger P, Berger JP, et al.

Judgments of trained observers on adverse
drug reactions. Clinical Pharmacology and
Therapeutics. 1979;25:(5)493-498. [Rec#:
461]

52. Boston LN. Dysuria following ephedrine

therapy. Med Times. 1928;56:94-95. [Rec#:
35]
53. Boston LN. Dysuria following ephedrine
therapy. The Medical Times.
1928;56:(Apr)94-95. [Rec#: 302]

44. Blaszyk HJ, Edwards WD. Sudden cardiac

death in a healthy 28-year-old male truck
driver abusing ephedrine. J Forensic Sci.
1997;42:(1)169-. [Rec#: 348]

54. Boullata JI, Nace AM. Safety issues with

herbal medicine. Pharmacotherapy.
2000;20:(3)257-69. [Rec#: 36]

45. Blau JJ. Ephedrine nephrolithiasis associated

with chronic ephedrine abuse. J Urol.
1998;160:(3 Pt 1)825. [Rec#: 31]

A1-7

Articles Excluded from Efficacy Analysis (continued)

55. Bowyer JF, Newport GD, Slikker W Jr.,
Gough B, Ferguson SA, Tor-Agbidye J. An
evaluation of l-ephedrine neurotoxicity with
respect to hyperthermia and
caudate/putamen microdialysate levels of
ephedrine, dopamine, serotonin, and
glutamate. Toxicol Sci. 2000;55:(1)133-42.
[Rec#: 37]

63. Britton MG, Empey DW, John GC,

McDonnell KA, Hughes DT. Histamine
challenge and anterior nasal rhinometry:
their use in the assessment of
pseudoephedrine and triprolidine as nasal
decongestants in subjects with hayfever. Br
J Clin Pharmacol. 1978;6:(1)51-8. [Rec#:
419]

56. Bracco D, Ferrarra JM, Arnaud MJ, Jequier E,

Schutz Y. Effects of caffeine on energy
metabolism, heart rate, and methylxanthine
metabolism in lean and obese women. Am J
Physiol. 1995;269:(4 Pt 1)E671-8. [Rec#:
375]

64. Brown NJ, Ryder D, Branch RA. A

pharmacodynamic interaction between
caffeine and phenylpropanolamine. Clin
Pharmacol Ther. 1991;50:(4)363-71. [Rec#:
42]
65. Bruneton J. Phenethylamines. In: Bruneton J.
Pharmacognosy, Phytochemistry, Medicinal
Plants. Paris, France, Secausuc, NY:
Lavoisier Pub; 1996. pp. 711-15. [Rec#: 43]

57. Bradley JG. Nonprescription drugs and

hypertension. Which ones affect blood
pressure? Postgrad Med. 1991;89:(6)195-7,
201-2. [Rec#: 38]

66. Bruno A, Nolte KB, Chapin J. Stroke

associated with ephedrine use. Neurology.
1993;43:(7)1313-6. [Rec#: 44]

58. Bradley JG, Kallail KJ, Dorsch JN, Fox J. The

effects of pseudoephedrine on blood
pressure in patients with controlled,
uncomplicated hypertension: a randomized,
double-blind, placebo-controlled trial. J Am
Board Fam Pract. 1991;4:(4)201-6. [Rec#:
39]

67. Bucci LR. Selected herbals and human

exercise performance. Am J Clin Nutr.
2000;72:(2 Suppl)624S-36S. [Rec#: 479]
68. Bukowiecki L, Jahjah L, Follea N. Ephedrine,
a potential slimming drug, directly
stimulates thermogenesis in brown
adipocytes via beta-adrenoreceptors. Int J
Obes. 1982;6:(4)343-50. [Rec#: 367]

59. Brater DC, Kaojarern S, Benet LZ, Lin ET,

Lockwood T, Morris RC, et al. Renal
excretion of pseudoephedrine. Clin
Pharmacol Ther. 1980;28:(5)690-4. [Rec#:
443]

69. Burkhart KK. Intravenous propranolol

reverses hypertension after
sympathomimetic overdose: two case
reports. J Toxicol Clin Toxicol.
1992;30:(1)109-14. [Rec#: 46]

60. Bray GA. Use and abuse of appetitesuppressant drugs in the treatment of
obesity. Ann Intern Med. 1993;119:(7 Pt
2)707-13. [Rec#: 306]
61. Bray GA. Drug treatment of obesity. Baillieres
Best Pract Res Clin Endocrinol Metab.
1999;13:(1)131-48. [Rec#: 40]

70. Bye C, Dewsbury D, Peck AW. Effects on the

human central nervous system of the two
isomers of ephedrine and triprolidine and
their interaction. Br J Clin Pharmacol.
1974;1:(1)71-78. [Rec#: 47]

62. Bright TP, Sandage BW Jr., Fletcher HP.

Selected cardiac and metabolic responses to
pseudoephedrine with exercise. J Clin
Pharmacol. 1981;21:(11-12 Pt 1)488-92.
[Rec#: 416]

A1-8

Articles Excluded from Efficacy Analysis (continued)

81. Chappel CI, Rona G, Balazs T, Gaudry R.
Comparison of cardiotoxic actions of certain
sympathomimetic amines. Cdn J Biochem
Physiol. 1959;37:35-42. [Rec#: 52]

71. Bye C, Hill HM, Hughes DT, Peck AW. A

comparison of plasma levels of L(+)
pseudoephedrine following different
formulations, and their relation to
cardiovascular and subjective effects in man.
Eur J Clin Pharmacol. 1975;8:(1)47-53.
[Rec#: 415]

82. Chen KK. The effect of ephedrine on

experimental shock and hemorrhage.
Proceedings of the Society for Experimental
Biology and Medicine. 1925;22:(3)203-206.
[Rec#: 406]

72. Capwell RR. Ephedrine-induced mania from

an herbal diet supplement. Am J Psychiatry.
1995;152:(4)647. [Rec#: 48]

83. Chen KK. The effect of repeated

administration of ephedrine. Proc Soc Exper
Biol Med. 1925;22:(8)568-570. [Rec#: 54]

73. Caraco Y, Zylber-Katz E, Berry EM, Levy M.

Caffeine pharmacokinetics in obesity and
following significant weight reduction. Int J
Obes Relat Metab Disord. 1995;19:(4)234-9.
[Rec#: 380]

84. Chen KK. The acute toxicity of ephedrine.

Proc Soc Exper Biol Med. 1925a;22:(7)404405. [Rec#: 53]

74. Case B, Oszko MA. Use of an algorithm to

evaluate published reports of adverse drug
reactions. Am J Hosp Pharm.
1991;48:(1)121-2. [Rec#: 331]

85. Chen KK. A comparative study of ephedrine,

pseudoephedrine and beta-phenylethylamine. Archives of Internal Medicine.
1927;39:404-411. [Rec#: 407]

75. Catlin DH, Hatton CK. Use and abuse of

anabolic and other drugs for athletic
enhancement. Adv Intern Med.
1991;36:399-424. [Rec#: 49]

86. Chen KK. A comparative study of synthetic

and natural ephedrines. J Pharmacol Exper
Therapeutics. 1928;33:237-258. [Rec#: 408]

76. CDC. Adverse events associated with

ephedrine containing products. MMWR.
1995;45:689-93. [Rec#: 434]

87. Chen KK, Kao CH. Ephedrine and

pseudoephedrine, their isolation,
constitution, isomerism, properties,
derivatives and synthesis. J Am Pharmaceut
Assoc. 1926;15:(8)625-639. [Rec#: 409]

77. CDC. Adverse events associated with

ephedrine-containing products. JAMA.
1996;276:(21)1711-2. [Rec#: 436]

88. Chen KK, Meek WJ. A comparative study of

ephedrine, tyramine and epinephrine with
special reference to the circulation. J
Pharmacol Exper Therapeut. 1926;28:59-76.
[Rec#: 411]

78. Cetaruk E W, Aaron C K. Hazards of

nonprescription medications. Emerg Med
Clin North Am. 1994;12:(2)483-510. [Rec#:
50]

89. Chen KK, Meek WJ. Further studies on the

effect of ephedrine on the circulation. J Am
Pharmaceut Assoc. 1926;28:31-57. [Rec#:
410]

79. Chait L. Factors influencing the reinforcing

and subjective effects of ephedrine in
humans. Psychopharmacology.
1994;113:381-7. [Rec#: 520 ]

90. Chen KK, Poth EJ. The racial difference of the

mydriatic action of ephedrines, cocaine, and
euphthalmine. Proc Soc Exper Biol Med.
1927;25:(2)150-151. [Rec#: 413]

80. Chance MRA. Aggregation as a factor

influencing the toxicity of sympathomimetic
amines in mice. J Pharmacol Exper
Therapeut. 1946;87:214-19. [Rec#: 51]

A1-9

Articles Excluded from Efficacy Analysis (continued)

99. Clarkson PM, Thompson HS. Drugs and sport.
Research findings and limitations. Sports
Med. 1997;24:(6)366-84. [Rec#: 60]

91. Chen KK, Schmidt CF. The action of

ephedrine, the active principle of the
Chinese drug Ma Hung. J Pharmacol Exper
Therapeut. 1924;24:339-357. [Rec#: 414]

100. Clemons JM, Crosby SL. Cardiopulmonary

and subjective effects of a 60 mg dose of
pseudoephedrine on graded treadmill
exercise. J Sports Med Phys Fitness.
1993;33:(4)405-12. [Rec#: 61]

92. Chen KK, Schmidt CF. Ephedrine and related

substances. Medicine. 1930;9:1-117. [Rec#:
55]
93. Chen KK, Wu C-K, Henriksen E. Ephedrine
homologs and isomers: Relationship
between their pharmacological action and
their chemical constitution and
stereoisomerism. Proc Soc Exper Biol Med.
1929;26:(4)348. [Rec#: 56]

101. Clovis WL. Mania and cough syrup. J Clin

Psychiatry. 1993;54:(5)200. [Rec#: 62]
102. Coates ML, Rembold CM, Farr BM. Does
pseudoephedrine increase blood pressure in
patients with controlled hypertension? J Fam
Pract. 1995;40:(1)22-6. [Rec#: 63]

94. Chen KK, Wu C-K, Henriksen E. Relationship

between the pharmacological action and the
chemical constitution and configuration of
the optical isomers of ephedrine and related
compounds. J Pharmacol Exper Therapeut.
1929;36:363-400. [Rec#: 57]

103. Cockings JG, Brown M. Ephedrine abuse

causing acute myocardial infarction. Med J
Aust. 1997;167:(4)199-200. [Rec#: 64]
104. Colditz GA. Economic costs of obesity. Am J
Clin Nutr. 1992;55:(2 Suppl)503S-507S.
[Rec#: 543]

95. Chestnut DH, Weiner CP, Wang JP, Herrig JE,

Martin JG. The effect of ephedrine upon
uterine artery blood flow velocity in the
pregnant guinea pig subjected to terbutaline
infusion and acute hemorrhage.
Anesthesiology. 1987;66:(4)508-12. [Rec#:
370]

105. Costello JF, May CS, Paterson JW, Pickup

ME. Pharmacokinetics of ephedrine in
asthmatics receiving acute and chronic
treatment. Br J Clin Pharmacol.
1975;2:(2)180-181. [Rec#: 65]

96. Chicharro M, Zapardiel A, Bermejo E, Perez

JA, Hernandez L. Direct determination of
ephedrine and norephedrine in human urine
by capillary zone electrophoresis. J
Chromatogr. 1993;622:(1)103-8. [Rec#:
444]

106. Cowley G. Herbal warning. New York:

Newsweek. 1996.
107. Cupp MJ. Herbal remedies: adverse effects
and drug interactions. Am Fam Physician.
1999;59:(5)1239-45. [Rec#: 66]

97. Chua SS, Benrimoj SI. Non-prescription

sympathomimetic agents and hypertension.
Med Toxicol Adverse Drug Exp.
1988;3:(5)387-417. [Rec#: 58]

108. Dalton R. Mixed bipolar disorder precipitated

by pseudoephedrine hydrochloride. South
Med J. 1990;83:(1)64-5. [Rec#: 67]

98. Chua SS, Benrimoj SI, Gordon RD, Williams

G. A controlled clinical trial on the
cardiovascular effects of single doses of
pseudoephedrine in hypertensive patients.
Br J Clin Pharmacol. 1989;28:(3)369-72.
[Rec#: 59]

109. Danan G, Benichou C. Causality assessment of

adverse reactions to drugs - I. A novel
method based on the conclusions of
international consensus meetings:
application to drug-induced liver injuries. J
Clin Epidemiol. 1993;46:(11)1323-30.
[Rec#: 328]

A1-10

Articles Excluded from Efficacy Analysis (continued)

110. Dangoumau J, Evreux JC, Jouglard J. [Method
for determination of undesirable effects of
drugs]. Therapie. 1978;33:(3)373-81. [Rec#:
323]

119. Dollery C. Phenylpropanolamine

(hydrochloride). In: Dollery C. (Ed)
Therapeutic Drugs. New York, NY:
Churchill Livingstone; 1991. pp. P91-P93.
[Rec#: 77]

111. Dawson JK, Earnshaw SM, Graham CS.

Dangerous monoamine oxidase inhibitor
interactions are still occurring in the 1990s. J
Accid Emerg Med. 1995;12:(1)49-51.
[Rec#: 69]

120. Dollery C. Pseudophedrine hydrochloride. In:

Dollery C. (Ed) Therapeutic Drugs. New
York, NY: Churchill Livingstone; 1991. pp.
P297-P299. [Rec#: 78]
121. Dougherty R. Pseudo-speed: look-alikes or
peashooters. NY State Journal of Medicine.
1982;1:74-75. [Rec#: 524]

112. Derreza H, Fine MD, Sadaniantz A. Acute

myocardial infarction after use of
pseudoephedrine for sinus congestion. J Am
Board Fam Pract. 1997;10:(6)436-8. [Rec#:
70]

122. Doyle H, Kargin M. Herbal stimulant

containing ephedrine has also caused
psychosis. BMJ. 1996;313:(7059)756.
[Rec#: 79]

113. Dewsnap P, Libby G. A case of affective

psychosis after routine use of proprietary
cold remedy containing
phenylpropanolamine. Hum Exp Toxicol.
1992;11:(4)295-6. [Rec#: 71]

123. Drew CD, Knight GT, Hughes DT, Bush M.

Comparison of the effects of D-(-)ephedrine and L-(+)-pseudoephedrine on the
cardiovascular and respiratory systems in
man. Br J Clin Pharmacol. 1978;6:(3)221-5.
[Rec#: 80]

114. Diaz JL. Ethnopharmacology of sacred

psychoactive plants used by the Indians of
Mexico. Annu Rev Pharmacol Toxicol.
1977;17:647-75. [Rec#: 72]

124. Dulloo AG, Miller DS. The thermogenic

properties of ephedrine/methylxanthine
mixtures: animal studies. Am J Clin Nutr.
1986;43:(3)388-94. [Rec#: 501]

115. Dickerson J, Perrier D, Mayersohn M, Bressler

R. Dose tolerance and pharmacokinetic
studies of L (+) pseudoephedrine capsules in
man. Eur J Clin Pharmacol. 1978;14:(4)2539. [Rec#: 73]

125. Dulloo AG. Ephedrine, xanthines and

prostaglandin-inhibitors: actions and
interactions in the stimulation of
thermogenesis. Int J Obes Relat Metab
Disord. 1993;17:(Suppl 1)S35-40. [Rec#:
81]

116. Dingemanse J, Guentert T, Gieschke R, Stabl

M. Modification of the cardiovascular
effects of ephedrine by the reversible
monoamine oxidase A-inhibitor
moclobemide. J Cardiovasc Pharmacol.
1996;28:(6)856-61. [Rec#: 74]

126. Dulloo AG. Herbal simulation of ephedrine

and caffeine in treatment of obesity.
International Journal of Obesity.
2002;26:590-2. [Rec#: 532]

117. Dollery C. Caffeine. In: Dollery C. (Ed)

Therapeutic Drugs. New York, NY:
Churchill Livingstone; 1991. pp. C3-C6.
[Rec#: 75]

127. Dulloo AG, Duret C, Rohrer D, Girardier L,

Mensi N, Fathi M, et al. Efficacy of a green
tea extract rich in catechin polyphenols and
caffeine in increasing 24-h energy
expenditure and fat oxidation in humans.
Am J Clin Nutr. 1999;70:(6)1040-5. [Rec#:
396]

118. Dollery C. Ephedrine (Hydrochloride). In:

Dollery C. (Ed) Therapeutic Drugs. New
York, NY: Churchill Livingstone; 1991. p.
E26-E29. [Rec#: 76]

A1-11

Articles Excluded from Efficacy Analysis (continued)

137. Dvorak R, Starling RD, Calles-Escandon J, et
al. Drug therapy for obesity in the elderly.
Drugs Aging. 1997;11:(5)338-51. [Rec#: 84]

128. Dulloo AG, Geissler CA, Horton T, et al.

Normal caffeine consumption: influence on
thermogenesis and daily energy expenditure
in lean and postobese human volunteers. Am
J Clin Nutr. 1989;49:(1)44-50. [Rec#: 374]

138. Dyck DJ. Dietary fat intake, supplements, and

weight loss. Can J Appl Physiol.
2000;25:(6)495-523. [Rec#: 85]

129. Dulloo AG, Miller DS. Thermogenic drugs for

the treatment of obesity: sympathetic
stimulants in animal models. Br J Nutr.
1984;52:(2)179-96. [Rec#: 385]

139. Edwards IR, Aronson JK. Adverse drug

reactions: definitions, diagnosis, and
management. Lancet. 2000;356:(9237)12559. [Rec#: 480]

130. Dulloo AG, Miller DS. Aspirin as a promoter

of ephedrine-induced thermogenesis:
potential use in the treatment of obesity. Am
J Clin Nutr. 1987;45:(3)564-9. [Rec#: 386]

140. Edwards M, Russo L, Harwood-Nuss A.

Cerebral infarction with a single oral dose of
phenylpropanolamine. Am J Emerg Med.
1987;5:(2)163-4. [Rec#: 86]

131. Dulloo AG, Miller DS. Ephedrine, caffeine

and aspirin: "over-the-counter" drugs that
interact to stimulate thermogenesis in the
obese. Nutrition. 1989;5:(1)7-9. [Rec#: 82]

141. Egger G, Cameron-Smith D, Stanton R. The

effectiveness of popular, non-prescription
weight loss supplements. Med J Aust.
1999;171:(11-12)604-8. [Rec#: 402]

132. Dulloo AG, Seydoux J, Girardier L. Dietary

and pharmacological effectiveness of
thermogenic stimulation in obesity
treatment. In: Oomura et al. (eds.) Progress
in Obesity Research, 1990. London: John
Libbey and Company, Lt.; 1991. pp. 135144. [Rec#: 504]

142. Eldor J. Ephedrine in the initial treatment of

haemorrhagic shock. Med Hypotheses.
1991;35:(3)250-2. [Rec#: 87]
143. Ellenhorn MJ, Barceloux DG. Over-thecounter products. In: Ellenhorn MJ,
Barceloux DG. Medical Toxicology:
Diagnosis and Treatment of Human
Poisoning. New York, NY: Elsevier; 1988.
pp. 501-559. [Rec#: 88]

133. Dulloo AG, Seydoux J, Girardier L. Peripheral

mechanisms of thermogenesis induced by
ephedrine and caffeine in brown adipose
tissue. Int J Obes. 1991;15:(5)317-26.
[Rec#: 364]

144. Emanueli A, Sacchetti G. An algorithm for the

classification of untoward events in large
scale clinical trials. Agents Actions Suppl.
1980;7:318-22. [Rec#: 321]

134. Dulloo AG, Seydoux J, Girardier L, et al.

Green tea and thermogenesis: interactions
between catechin-polyphenols, caffeine and
sympathetic activity. Int J Obes Relat Metab
Disord. 2000;24:(2)252-8. [Rec#: 395]

145. Emmanuel N, Jones C, Lydiard R. Use of

herbal products and symptoms of bipolar
disorder. American Journal of Psychiatry.
1998;155:1627. [Rec#: 519]

135. Dulloo AG, Stock MJ. Ephedrine in the

treatment of obesity. Int J Obes Relat Metab
Disord. 1993;17:(Suppl 1)S1-2. [Rec#: 404]

146. Emmanuel NP, Jones C, Lydiard RB. Use of

herbal products and symptoms of bipolar
disorder. Am J Psychiatry.
1998;155:(11)1627. [Rec#: 486]

136. Dunagan N, Greenleaf JE, Cisar CJ.

Thermoregulatory effects of caffeine
ingestion during submaximal exercise in
men. Aviat Space Environ Med.
1998;69:(12)1178-81. [Rec#: 542]

A1-12

Articles Excluded from Efficacy Analysis (continued)

147. Empey DW, Frosolono MF, Hughes DT,
Perkins JG. Comparison of pseudoephedrine
and triprolidine, alone and in combination in
preventing nasal congestion in subjects with
allergic rhinitis using nasal histamine
challenge. Br J Clin Pharmacol.
1984;18:(1)86-9. [Rec#: 418]

156. Finer N. Present and future pharmacological

approaches [obesity]. British Medical
Bulletin. 1997;53:(2)409-432. [Rec#: 304]
157. Fisher CR, Veronneau SJ. Herbal preparations:
a primer for the aeromedical physician.
Aviat Space Environ Med. 2000;71:(1)4560. [Rec#: 92]

148. Empey DW, Young GA, Letley E, John GC,

Smith P, McDonnell KA, et al. Doseresponse study of the nasal decongestant and
cardiovascular effects of pseudoephedrine.
Br J Clin Pharmacol. 1980;9:(4)351-8.
[Rec#: 417]

158. Floch A, Lagente V, Freslon JL, Blanc M,

Advenier C. [Bronchomotor effects of (-)
ephedrine, (+/-) ephedrine and (+)
pseudoephedrine in guinea pigs]. Ann
Pharm Fr. 1985;43:(1)31-8. [Rec#: 369]

149. Escobar JI, Karno M. Chronic hallucinosis

from nasal drops. JAMA.
1982;247:(13)1859-60. [Rec#: 488]

159. Flordal PA, Svensson J. Hemostatic effects of

ephedrine. Thromb Res. 1992;68:(3)295302. [Rec#: 93]

150. Fairchild M D, Alles G A. The central

locomotor stimulatory activity and acute
toxicity of the ephedrine and norephedrine
isomers in mice. J Pharmacol Exp Ther.
1967;158:(1)135-9. [Rec#: 89]

160. Flurer CL, Lin LA, Satzger RD, Wolnik KA.

Determination of ephedrine compounds in
nutritional supplements by cyclodextrinmodified capillary electrophoresis. J
Chromatogr B Biomed Appl 1995
669:(1)133-9. [Rec#: 447]

151. FDA. Center for Food Safety and Applied

Nutrition. Assessment of Public Health
Risks Associated with the Use of Ephedrine
Alkaloid-Containing Dietary Supplements.
2000. [Rec#: 94]

161. Frick PA, Cohen LG, Rovers JP. Algorithms

used in adverse drug event reports: a
comparative study. Ann Pharmacother.
1997;31:(2)164-7. [Rec#: 330]

152. FDA. Dietary supplements containing ephedra

alkaloids. FDA docket no. 00N-1200. In:
Rockville, MD: Food and Drug
Administration. 2000. [Rec#: 440]

162. Fugh-Berman A. Clinical trials of Herbs.

Primary Care - Clinics Office Pract.
1997;24:(4)889-903. [Rec#: 95]
163. Fugh-Berman A, Allina A. Ephedra for
Weight Loss. Alternative Therapies in
Women's Health. 2000;2:81-84. [Rec#: 312]

153. FDA. Spontaneous adverse events reports.

available at URL:
www.fda.gov/ohrms/dockets/ac/slides/3558s
le/tsld003.htm.

164. Garcia Ortiz JC, Terron M, Bellido J.

Nonpigmenting fixed exanthema from
ephedrine and pseudoephedrine. Allergy.
1997;52:(2)229-230. [Rec#: 522]

154. Feller D R, Malspeis L. Biotransformation of

D(-)-ephedrine and L(+)-ephedrine in the
rabbit, in vivo and in vitro. Drug Metab
Dispos. 1977;5:(1)37-46. [Rec#: 90]

165. Garriott JC, Simmons LM, Poklis A, Mackell

MA. Five cases of fatal overdose from
caffeine-containing "look-alike" drugs. J
Anal Toxicol. 1985;9:(3)141-3. [Rec#: 96]

155. Fillmore CM, Bartoli L, Bach R, Park Y.

Nutrition and dietary supplements. Phys
Med Rehabil Clin N Am. 1999;10:(3)673703. [Rec#: 91]

166. Gaultieri J. Cardiomyopathy in a heavy

ephedrine abuser. J Toxicol Clin Toxicol.
1996;34:581-582. [Rec#: 446]

A1-13

Articles Excluded from Efficacy Analysis (continued)

167. Gilbert-Barness E, Drut RM. Association of
sympathomimetic drugs with malformations.
Vet Human Toxicol. 2000;42:(3)168-171.
[Rec#: 98]

176. Gordiecheva NV, Shashkov VS, Kaplan EY,

Lakota NG, Bliznets NI, Mikheeva LN, et
al. Effectiveness of drugs on physical
exercises. Kosm Ciol Aviakosm Med.
1975;9:(5)6-. [Rec#: 346]

168. Gilbert EF, Matsuoka R, Bruyere H, Optiz JM.

Teratogenic effect of Tedral (theophylline,
ephedrine and phenobarbital) on cardiac
development in chick embryos. Circulation.
1981;64:(Suppl 4)233(Abstract). [Rec#: 99]

177. Graham TE, Spriet LL. Performance and

metabolic responses to a high caffeine dose
during prolonged exercise. J Appl Physiol.
1991;71:(6)2292-8. [Rec#: 539]
178. Greening GK. The effect of an antihistaminepseudoephedrine drug on the blood pressure
of moderate hypertensives. Curr Ther Res
Clin Exp. 1969;11:(5)252-5. [Rec#: 427]

169. Gill ND, Shield A, Blazevich AJ, Zhou S,

Weatherby RP. Muscular and
cardiorespiratory effects of pseudoephedrine
in human athletes. Br J Clin Pharmacol.
2000;50:(3)205-13. [Rec#: 100]

179. Greenway F. The safety and efficacy of

pharmaceutical and herbal caffeine and
ephedrine use as a weight loss agent.
Obesity Reviews. 2001;2:199-211. [Rec#:
314]

170. Gillies H, Derman WE, Noakes TD, Smith P,

Evans A, Gabriels G. Pseudoephedrine is
without ergogenic effects during prolonged
exercise. J Appl Physiol. 1996;81:(6)2611-7.
[Rec#: 101]
171. Glennon RA, Young R. (+)Amphetaminestimulus generalization to an herbal
ephedrine product. Pharmacol Biochem
Behav. 2000;65:(4)655-8. [Rec#: 363]

180. Greenway F, Herber D, Raum W, Herber D,

Morales S. Double-blind, randomized,
placebo-controlled clinical trials with nonprescription medications for the treatment of
obesity. Obes Res. 1999;7:(4)370-8. [Rec#:
105]

172. Glick R, Hoying J, Cerullo L, Perlman S.

Phenylpropanolamine: an over-the-counter
drug causing central nervous system
vasculitis and intracerebral hemorrhage.
Case report and review. Neurosurgery.
1987;20:(6)969-74. [Rec#: 102]

181. Greenway FL. Clinical studies with

phenylpropanolamine: a meta-analysis. Am
J Clin Nutr. 1992;55:(1 Suppl)203S-205S.
[Rec#: 307]

173. Glidden RS, DiBona FJ. Urinary retention

associated with ephedrine. J Pediatr.
1977;90:(6)1013-4. [Rec#: 103]

182. Greenway FL, Ryan DH, Bray GA, Rood JC,

Tucker EW, Smith SR. Pharmaceutical cost
savings of treating obesity with weight loss
medications. Obes Res. 1999;7:(6)523-31.
[Rec#: 398]

174. Goldberg LI, deV Cotton M, Darby TD,

Howell EV. Comparative heart contractile
force effects of equipressor doses of several
sympathomimetic amines. J Pharmacol Exp
Ther. 1953;108:177. [Rec#: 526]

183. Griffith JD. Structure-activity relationships of

several amphetamine drugs in man. In:
Ellinwood EH Jr., Kilbey MM. (Eds)
Cocaine and other stimulants. New York,
NY: Plenum Press; 1977. pp. 705-15. [Rec#:
107]

175. Goldfrank LR, Hoffman RS. The

cardiovascular effects of cocaine. Ann
Emerg Med. 1991;20:(2)165-75. [Rec#:
104]

184. Gruber AJ, Pope HG Jr. Ephedrine abuse

among 36 female weightlifters. Am J
Addict. 1998;7:(4)256-61. [Rec#: 108]

A1-14

Articles Excluded from Efficacy Analysis (continued)

185. Grunberg E, Randall LO, Schnitzer RJ. The
chemotherapeutic and pharmacological
properties of the l-ephedrine salt of
penicillin G (Tersavin). J Pharmacol Exper
Therapeut. 1949;95:336-346. [Rec#: 109]

193. Hall RC, Beresford TP, Stickney SK, Nasdahl

CS, Coleman JH. Psychiatric reactions
produced by respiratory drugs.
Psychosomatics. 1985;26:(7)605-8, 615-6.
[Rec#: 489]

186. Gualtieri J, Harris C. Dilated cardiomyopathy

in a heavy ephedrine abuser. J Toxicol (Clin
Toxicol). 1996;34:(5)581-582(Abstract).
[Rec#: 110]

194. Haller CA, Benowitz NL. Adverse

cardiovascular and central nervous system
events associated with dietary supplements
containing ephedra alkaloids. N Engl J Med.
2000;343:(25)1833-8. [Rec#: 116]

187. Gurley BJ, Gardner SF, White LM, Wang P.

Ephedrine pharmacokinetics after the
ingestion of commercially available herbal
preparations of ephedra sinica (Ma-huang).
Pharmaceut Res. 1997;14:(11
Suppl)S519(Abstract). [Rec#: 111]

195. Han LK, Takaku T, Li J, Kimura Y, Okuda H.

Anti-obesity action of oolong tea. Int J Obes
Relat Metab Disord. 1999;23:(1)98-105.
[Rec#: 393]

188. Gurley BJ, Gardner SF, White LM, Wang P.

Ephedrine pharmacokinetics after the
ingestion of nutritional supplements
containing Ephedra sinica (ma huang).
Therapeut Drug Monit. 1998;20:(4)439-445.
[Rec#: 112]

196. Hanzlick R. National Association of Medical

Examiners Pediatric Toxicology (PedTox)
Registry Report 3. Case submission
summary and data for acetaminophen,
benzene, carboxyhemoglobin,
dextromethorphan, ethanol, phenobarbital,
and pseudoephedrine. Am J Forensic Med
Pathol. 1995;16:(4)270-7. [Rec#: 449]

189. Gurley BJ, Wang P, Gardner SF. Ephedrine

alkaloid content of five commercially
available herbal products containing ephedra
sinica (Ma huang). Pharmaceut Res.
1997;14:(11 Suppl 1)S582-S582(Abstract).
[Rec#: 113]

197. Haseman JK. Patterns of tumor incidence in

two-year cancer bioassay feeding studies in
Fischer 344 rats. Fundam Appl Toxicol.
1983;3:(1)1-9. [Rec#: 117]
198. Hellriegel ET, Bjornsson TD, Hauck WW.
Interpatient variability in bioavailability is
related to the extent of absorption:
implications for bioavailability and
bioequivalence studies. Clin Pharmacol
Ther. 1996;60:(6)601-7. [Rec#: 118]

190. Gurley BJ, Wang P, Gardner SF. Ephedrinetype alkaloid content of nutritional
supplements containing Ephedra sinica (Ma
huang) as determined by high performance
liquid chromatography. J Pharmaceut Sci.
1998;87:(12)1547-1553. [Rec#: 114]

199. Herridge CF, Brook MF. Ephedrine psychosis.

Br Med J. 1968;2:(598)160. [Rec#: 120]

191. Gurley DJ, Gardner SF, Hubbard MA. Content

versus label claims in ephedra-containing
dietary supplements. Am J Healthy Syst
Pharm. 2000;57. [Rec#: 115]

200. Hirabayashi Y, Saitoh K, Fukuda H, Mitsuhata

H, Shimizu R. Coronary artery spasm after
ephedrine in a patient with high spinal
anesthesia. Anesthesiology. 1996;84:(1)2214. [Rec#: 121]

192. Haffner SM, Knapp JA, Stern MP, Hazuda

HP, Rosenthal M, Franco LJ. Coffee
consumption, diet, and lipids. Am J
Epidemiol. 1985;122:(1)1-12. [Rec#: 381]

201. Hirsch MS, Walter RM, Hasterlik RJ.

Subarachnoid hemorrhage following
ephedrine and MAO inhibitor. JAMA.
1965;194:(11)1259. [Rec#: 485]

A1-15

Articles Excluded from Efficacy Analysis (continued)

211. Hull KM, Zanzillari R, Maher TJ. Comparison
between phenylpropanolamine and
structurally related compounds on gastric
transit in the rat. Pharmacol Biochem Behav.
1993;46:(2)411-4. [Rec#: 471]

202. Hoffman BB, Lefkowitz RJ. Catecholamines,

sympathomimetic drugs, and adrenergic
receptors antagonists. In: Hardman JG,
Limbird LE, Molinoff PB, Ruddon RW,
Goodman Gilman A. (Eds) Goodman and
Gillman's The Pharmacological Basis of
Therapeutics. New York, NY, Toronto, ON:
McGraw Hill; 1996. pp. 199-248. [Rec#:
122]

212. Hurlbut JA, Carr JR, Singleton ER, Faul KC,

Madson MR, Storey JM, et al. Solid-phase
extraction cleanup and liquid
chromatography with ultraviolet detection of
ephedrine alkaloids in herbal products. J
AOAC Int. 1998;81:(6)1121-7. [Rec#: 448]

203. Horn J, Anderson K. Who in America is trying

to lose weight? Ann Intern Med.
1993;119:672-676. [Rec#: 467]

213. Hutchins GM. Dietary supplements containing

ephedra alkaloids. N Engl J Med.
2001;344:(14)1095-6; discussion 1096-7.
[Rec#: 353]

204. Howard HJ, Lee TP. The effect of instillations

of ephedrine solution upon the eye. Proc Soc
Exper Biol Med. 1927;24:(7)700-702.
[Rec#: 421]

214. Hutchinson TA, Leventhal JM, Kramer MS,

Karch FE, Lipman AG, Feinstein AR. An
algorithm for the operational assessment of
adverse drug reactions. II. Demonstration of
reproducibility and validity. JAMA.
1979;242:(7)633-8. [Rec#: 326]

205. Hu SY. Ephedra (ma-huang) in the new

Chinese materia medica. Economic Botany.
1969;23:346-351. [Rec#: 505]
206. Hughes DT, Empey DW, Land M. Effects of
pseudoephedrine in man. J Clin Hosp
Pharm. 1983;8:(4)315-21. [Rec#: 128]

215. Huxtable RJ. The myth of beneficent nature:

the risks of herbal preparations. Ann Intern
Med. 1992;117:(2)165-6. [Rec#: 130]

207. Hughes SC, Ward MG, Levinson G, Shnider

SM, Wright RG, Gruenke LD, et al.
Placental transfer of ephedrine does not
affect neonatal outcome. Anesthesiology.
1985;63:(2)217-9. [Rec#: 129]

216. Hwang EJ, Kan EM, Yang KS, Park SB, Kim
TH. Effects of ephedrae herba extract on
cellular and humoral immune responses.
Korean J Pharmacog. 1990;21:(3)256-.
[Rec#: 350]

208. Hull KM, Maher TJ. L-tyrosine potentiates the

anorexia induced by mixed-acting
sympathomimetic drugs in hyperphagic rats.
J Pharmacol Exp Ther. 1990;255:(2)403-9.
[Rec#: 470]

217. Insel PA. Beta-adrenergic receptors in

pathophysiologic states and in clinical
medicine. In: Perkins JP. (Ed) The BetaAdrenergic Receptors. Clifton, NJ: Humana
Press; 1991. pp. 295-343. [Rec#: 133]

209. Hull KM, Maher TJ. Effects of L-tyrosine on

mixed-acting sympathomimetic-induced
pressor actions. Pharmacol Biochem Behav.
1992;43:(4)1047-52. [Rec#: 469]

218. Insel PA, Motusky HJ. Physiologic and

pharmacologic regulation of adrenergic
receptors. In: Insel PA. (Ed) Adrenergic
Receptors in Man. New York, NY: Dekker;
1987. pp. 201-236. [Rec#: 134]

210. Hull KM, Sanderson WD, Maher TJ. Effect of

phenylpropanolamine and related
compounds on beta- adrenoceptor-induced
activation of adenylyl cyclase. Life Sci.
1993;52:(15)1287-94. [Rec#: 468]

A1-16

Articles Excluded from Efficacy Analysis (continued)

219. Ishigooka J, Yoshida Y, Murasaki M. Abuse
of "BRON": a Japanese OTC cough
suppressant solution containing
methylephedrine, codeine, caffeine and
chlorpheniramine. Prog
Neuropsychopharmacol Biol Psychiatry.
1991;15:(4)513-21. [Rec#: 135]

228. Kalix P. The pharmacology of psychoactive

alkaloids from ephedra and catha. J
Ethnopharmacol. 1991;32:(1-3)201-8.
[Rec#: 452]
229. Kalmar JM, Cafarelli E. Effects of caffeine on
neuromuscular function. J Appl Physiol.
1999;87:(2)801-8. [Rec#: 540]

220. Jacobs I, Bell DG. Acute physical performance

effects after ingestion of ephedrine or
ephedrine combined with caffeine. 2002.
[Rec#: 530]

230. Kanai T, Nishikawa T, Satoh A, Kajita A.

Cardiovascular teratogenicity of ephedrine
in rats. Senten Ijo (Cong Anom).
1986;26:(3)246. [Rec#: 141]

221. Jacobs KM, Hirsch KA. Psychiatric

complications of Ma-huang.
Psychosomatics. 2000;41:(1)58-62. [Rec#:
136]

231. Kandil A, Shukrallah I, Hussein L, Azmy A.

Biological and chemical evaluation of
ephedrine. J Drug Res (Egypt).
1983;14:(1/2)119-122. [Rec#: 349]

222. James LP, Farrar HC, Komoroski EM, et al.

Sympathomimetic drug use in adolescents
presenting to a pediatric emergency
department with chest pain. Journal of
Toxicology - Clinical Toxicology.
1998;36:321-8. [Rec#: 517]

232. Kane FJ Jr., Florenzano R. Psychosis

accompanying use of bronchodilator
compound. JAMA. 1971;215:(13)2116.
[Rec#: 490]
233. Kanfer I, Dowse R, Vuma V.
Pharmacokinetics of oral decongestants.
Pharmacotherapy. 1993;13:(6 Pt 2)116S128S; discussion 143S-146S. [Rec#: 142]

223. Jawad SS, Eccles R. Effect of

pseudoephedrine on nasal airflow in patients
with nasal congestion associated with
common cold. Rhinology. 1998;36:(2)73-6.
[Rec#: 451]

234. Karch FE, Lasagna L. Toward the operational

identification of adverse drug reactions. Clin
Pharmacol Ther. 1977;21:(3)247-54. [Rec#:
322]

224. Jequier E. Energy utilization in human obesity.

Ann N Y Acad Sci. 1987;499:73-83. [Rec#:
400]

235. Karch SB, Green GS, Young S. Myocardial

hypertrophy and coronary artery disease in
male cocaine users. J Forensic Sci.
1995;40:(4)591-5. [Rec#: 143]

225. Jones JK. Approaches to evaluation causation

of suspected drug reactions. In: Strom BL,
Velo G. (Eds) Drug Epidemiology and PostMarketing Surveillance. New York, NY:
Plenum Press; 1992. pp. 103-113. [Rec#:
138]

236. Karch SB. Synthetic stimulants

<Fenfluramine, mescaline analogs>. In:
Karch SB. The Pathology of Drug Abuse.
Boca Raton, FL: CRC Press; 1993. pp. 199218. [Rec#: 145]

226. Josefson D. Herbal stimulant causes US

deaths. BMJ. 1996;312:(7044)1441. [Rec#:
309]

237. Karch SB. Synthetic stimulants

<Phenylpropanolamine>. In: Karch SB. The
Pathology of Drug Abuse. Boca Raton, FL:
CRC Press; 1993. pp. 165, 177-198. [Rec#:
146]

227. Jung RT, Shetty PS, James WP, Barrand MA,

Callingham BA. Caffeine: its effect on
catecholamines and metabolism in lean and
obese humans. Clin Sci (Colch).
1981;60:(5)527-35. [Rec#: 372]

A1-17

Articles Excluded from Efficacy Analysis (continued)

248. Knight DC, Eden JA. Phytoestrogens: a short
review. Maturitas. 1995;22:167-175. [Rec#:
311]

238. Karch SB. Ma huang and the ephedra

alkaloids. In: Cupp MJ. (Ed) Toxicology and
Clinical Pharmacology of Herbal Products.
Totowa, NJ: Humana Press, Inc; 2000. pp.
11-30. [Rec#: 144]

249. Kolata G. Appetite suppressants aid obese.

The New York Times. July 5, 1992; Sect:
1:9(N), P12(L) (col. 4).

239. Karch SB. Comments on "ma huang toxicity"

letter by Dr. Theoharides. J Clin
Psychopharmacol. 1999;19:(2)196-9. [Rec#:
454]

250. Kramer MS, Leventhal JM, Hutchinson TA,

Feinstein AR. An algorithm for the
operational assessment of adverse drug
reactions. I. Background, description, and
instructions for use. JAMA.
1979;242:(7)623-32. [Rec#: 327]

240. Kasahara Y, HIkino H, Tsurufuji S, Watanabe

M, Ohuchi K. Anti-inflammatory actions of
ephedrines in acute inflammations. Planta
Med. 1985;51:(4)325-331. [Rec#: 351]

251. Kuczmarski RJ, Flegal KM, Campbell SM, et

al. Increasing prevalence of overweight
among US adults. JAMA. 1994;272:205211. [Rec#: 465]

241. Keld DB, Bendsen AK. [Doping - a medical

problem in Denmark?]. Ugeskr Laeger.
1998;160:(33)4749-53. [Rec#: 147]

252. Kuitunen T, Karkkainen S, Ylitalo P.

Comparison of the acute physical and
mental effects of ephedrine, fenfluramine,
phentermine and prolintane. Methods Find
Exp Clin Pharmacol. 1984;6:(5)265-70.
[Rec#: 152]

242. Kernan WN, Viscoli CM, Brass LM,

Broderick JP, Brott T, Feldmann E, et al.
Phenylpropanolamine and the risk of
hemorrhagic stroke. N Engl J Med.
2000;343:(25)1826-32. [Rec#: 148]
243. Kim TH, Yang KS, Hwang EZ, Park SB.
Effect of ephedrae herba on the immune
response in mice. Korean J Pharmacog.
1991;22:(3)183-91. [Rec#: 340]

253. Kustrak D, Ramic A. Ephedra species and the

alkaloids of this plants. Farm Glas.
1994;50:(12)321-339. [Rec#: 153]
254. Laitinen LA, Empey DW, Bye C, Britton MG,
McDonnell K, Hughes DT. A comparison of
the bronchodilator action of
pseudoephedrine and ephedrine in patients
with reversible airway obstruction. Eur J
Clin Pharmacol. 1982;23:(2)107-9. [Rec#:
154]

244. Kimmel SE, Hutchins GM. Commentary on

the article: "Adverse cardiovascular events
temporally associated with Mahuang, an
herbal source of ephedrine," published in
Mayo Clinic Proceedings, January 2002.
(Unpublished). [Rec#: 478]
245. Kitagawa H, Saito H. Differences in the effect
of ephedrine isomers on blood pressure in
rats. Jpn J Pharmacol. 1975;25:(4)473-5.
[Rec#: 356]

255. Lake CR, Gallant S, Masson E, Miller P.

Adverse drug effects attributed to
phenylpropanolamine: a review of 142 case
reports. Am J Med. 1990;89:(2)195-208.
[Rec#: 155]

246. Kizer KW. Intracranial hemorrhage associated

with overdose of decongestant containing
phenylpropanolamine. Am J Emerg Med.
1984;2:(2)180-1. [Rec#: 149]

256. Lake CR, Rosenberg DB, Gallant S, Zaloga G,

Chernow B. Phenylpropanolamine increases
plasma caffeine levels. Clin Pharmacol
Ther. 1990;47:(6)675-85. [Rec#: 156]

247. Klepser TB, Klepser ME. Unsafe and

potentially safe herbal therapies. Am J
Health Syst Pharm. 1999;56:(2)125-38; quiz
139-41. [Rec#: 150]

A1-18

Articles Excluded from Efficacy Analysis (continued)

257. Lake CR, Tenglin R, Chernow B, Holloway
HC. Psychomotor stimulant-induced mania
in a genetically predisposed patient: a
review of the literature and report of a case.
J Clin Psychopharmacol. 1983;3:(2)97-100.
[Rec#: 157]

266. Levine B, Jones R, Klette K, Smith M L,

Kilbane E. An intoxication involving BRON
and verapamil. J Anal Toxicol.
1993;17:(6)381-3. [Rec#: 164]
267. Levitsky DA, Strupp BJ. Direct and indirect
thermogenic effects of anorectic drugs. Adv
Nutr Res. 1985;7:187-201. [Rec#: 165]

258. Lake CQR. CNS stimulants and the look-alike

drugs. Psychiatr Clin North Am.
1984;7:689-701. [Rec#: 525]

268. Li Q, Li BH. [Comparison of central

stimulating effects between ephedrine and
apomorphine]. Zhongguo Yao Li Xue Bao.
1991;12:(5)468-71. [Rec#: 365]

259. Lanctot KL, Naranjo CA. Comparison of the

Bayesian approach and a simple algorithm
for assessment of adverse drug events. Clin
Pharmacol Ther. 1995;58:(6)692-8. [Rec#:
324]

269. Liljestrand SH. The action of

pseudoephedrine. I. On the isolated uterus
and bladder. Chung-Kuo Sheng Li Hsueh
Tsa Chih. 1929;3:249-. [Rec#: 343]

260. Law MY, Pedersen GH, Hennen WJ,

McCausland CW, Sidwell RW. Subacute
toxicity study of ma-huang in mice. Fundam
Appl Toxicol. 1996;30:(1-Prt.
2)111(Abstract). [Rec#: 158]

270. Lin ET, Brater DC, Benet LZ. Gas-liquid

chromatographic determination of
pseudoephedrine and norpseudoephedrine in
human plasma and urine. J Chromatogr.
1977;140:(3)273-9. [Rec#: 431]

261. Lee EB. Teratogenicity of the extracts of crude

drugs. Saengyak Hakhoe Chi (Korean J
Pharmacognosy). 1982;13:116-121. [Rec#:
159]

271. Lindahl A, Sandstrom R, Ungell AL,

Abrahamsson B, Knutson TW, Knutson L,
et al. Jejunal permeability and hepatic
extraction of fluvastatin in humans. Clin
Pharmacol Ther. 1996;60:(5)493-503.
[Rec#: 166]

262. Lee MK; Cheng BW; Che CT, et al.

Cytotoxicity assessment of Ma-huang
(Ephedra) under different conditions of
preparation. Toxicol Sci 2000 56:424-30.
[Rec#: 513]

272. Lindqvist A, Jalonen J, Laitinen LA, Seppala

T, Stromberg C. The effects of midazolam
and ephedrine on post-exercise autonomic
chronotropic control of the heart in normal
subjects. Clin Auton Res. 1996;6:(6)343-9.
[Rec#: 167]

263. Lee MK, Cheng BW, Che CT, Hsieh DP.

Cytotoxicity assessment of Ma-huang
(Ephedra) under different conditions of
preparation. Toxicol Sci. 2000;56:(2)424-30.
[Rec#: 160]

273. Lindsay BD. Are serious adverse

cardiovascular events an unintended
consequence of the Dietary Supplement
Health and Education Act of 1994? Mayo
Clin Proc. 2002;77:(1)7-9. [Rec#: 472]

264. Lefebvre RA, Surmont F, Bouckaert J,

Moerman E. Urinary excretion of ephedrine
after nasal application in healthy volunteers.
J Pharm Pharmacol. 1992;44:(8)672-5.
[Rec#: 450]

274. Ling M, Piddlesden SJ, Morgan BP. A

component of the medicinal herb ephedra
blocks activation in the classical and
alternative pathways of complement. Clin
Exp Immunol. 1995;102:(3)582-8. [Rec#:
334]

265. Leung AY, Foster S. Ephedra. In: Leung AY,

Foster S. Encyclopedia of Common Natural
Ingredients Used in Foods, Drugs, and
Cosmetics. New York, NY: John Wiley;
1996. p. 227. [Rec#: 163]

A1-19

Articles Excluded from Efficacy Analysis (continued)

284. Maertens P, Lum G, Williams J P, White J.
Intracranial hemorrhage and cerebral
angiopathic changes in a suicidal
phenylpropanolamine poisoning. South Med
J. 1987;80:(12)1584-6. [Rec#: 175]

275. Liu XX. [Pharmacologic action and clinical

use of herbal Ephedrae]. Zhong Xi Yi Jie He
Za Zhi. 1989;9:(4)255-6. [Rec#: 168]
276. Liu YM, Sheu SJ, Chiou SH, Chang HC, Chen
YP. A comparative study on commercial
samples of Ephedra Herba. Planta Medica.
1993;59:376-378. [Rec#: 506]

285. Mahdihasan S. Ephedra the oldest medicinal

plant with the history of an uninterrupted
use. Ancient Sci Life. 1987;7:(2)105-9.
[Rec#: 176]

277. Loizou LA, Hamilton JG, Tsementzis SA.

Intracranial haemorrhage in association with
pseudoephedrine overdose. J Neurol
Neurosurg Psychiatry. 1982;45:(5)471-2.
[Rec#: 170]

286. Maitai CK. D-norpseudoephedrine as an antiasthmatic: comparison with l-ephedrine, dpseudoephedrine. East Afr Med J.
1975;52:(6)330-2. [Rec#: 425]

278. Loosmore S, Armstrong D. Do-Do abuse. Br J

Psychiatry. 1990;157:278-81. [Rec#: 171]

287. Martin WR, Sloan JW, Sapira JD, Jasinski

DR. Physiologic, subjective, and behavioral
effects of amphetamine, methamphetamine,
ephedrine, phenmetrazine, and
methylphenidate in man. Clin Pharmacol
Ther. 1971;12:(2)245-58. [Rec#: 178]

279. Lowell BB, Napolitano A, Usher P, Dulloo

AG, Rosen BS, Spiegelman BM, et al.
Reduced adipsin expression in murine
obesity: effect of age and treatment with the
sympathomimetic-thermogenic drug mixture
ephedrine and caffeine. Endocrinology.
1990;126:(3)1514-20. [Rec#: 392]

288. Martinez FV, Badas AJ, Goitia JFG, Aguirre I.

Generalized dermatitis due to oral
ephedrine. Contact Dermatit.
1993;29:(4)215-216. [Rec#: 179]

280. LSRO. Alternative and Traditional Models for

Safety Evaluation of Food Ingredients.
Prepared for: Center for Food Safety and
Applied Nutrition Food and Drug
Administration Department of Health and
Human Services, Washington DC 20204,
under FDA Contract No. 223-92-2195, Task
Order #9. Bethesda, MD: Life Sciences
Research Office, American Society for
Nutritional Sciences. 1998. [Rec#: 172]

289. Marvola M. Effect of acetylated derivatives of

some sympathomimetic amines on the acute
toxicity, locomotor activity and barbiturate
anaesthesia time in mice. Acta Pharmacol
Toxicol (Copenh). 1976;38:(5)474-89.
[Rec#: 180]
290. Marvola M, Kivirinta R. Pharmacokinetics and
locomotor activity increasing effect of
ephedrine in mice. Acta Pharmacol Toxicol
(Copenh). 1978;43:(5)381-6. [Rec#: 181]

281. Lustik SJ, Chhibber AK, van Vliet M,

Pomerantz RM. Ephedrine-induced coronary
artery vasospasm in a patient with prior
cocaine use. Anesth Analg. 1997;84:(4)9313. [Rec#: 174]

291. Massoudi M, Evans E, Miller DS.

Thermogenic drugs for the treatment of
obesity: screening using obese rats and
mice. Ann Nutr Metab. 1983;27:(1)26-37.
[Rec#: 384]

282. Lyon CC, Turney JH. Pseudoephedrine

toxicity in renal failure. Br J Clin Pract.
1996;50:(7)396-7. [Rec#: 453]

292. Massoudi M, Miller DS. Ephedrine, a

thermogenic and potential slimming drug.
Proc Nutr Soc. 1977;36:(3)135A. [Rec#:
182]

283. Machii S. Pharmacological studies on the

optical isomers of ephedrine and
pseudoephedrine. Nippon Yakurigaku
Zasshi. 1957;53:638-. [Rec#: 344]

A1-20

Articles Excluded from Efficacy Analysis (continued)

303. Miller DS. A controlled trial using ephedrine
in the treatment of obesity. Int J Obes.
1986;10:(2)159-60. [Rec#: 191]

293. Matsuoka R, Gilbert EF, Bruyers H Jr., Optiz

JM. An aborted human fetus with truncus
arteriosus communis--possible teratogenic
effect of Tedral. Heart Vessels.
1985;1:(3)176-8. [Rec#: 183]

304. Miller TG. A consideration of the clinical

value of ephedrine. American Journal of
Medical Science. 1925;170:157-181. [Rec#:
507]

294. Matthews G, Smolinske S, White S.

Ephedrine-related stroke in a teenager. J
Toxicol (Clin Toxicol).
1997;35:(5)555(Abstract). [Rec#: 184]

305. Minimatsu S, Kobayashi Y, Kobayashi N,

Juhii Y, Aburada M, Yamushita M. Acute
ephedra herba and ephedrine poisoning in
mice. Chudoku Kenkyu. 1991;4:143-149.
[Rec#: 192]

295. May CS, Paterson JW, Spiro SG, Johnson AJ.

Effect of intravenous injection of salbutamol
in asthma. Br J Clin Pharmacol.
1975;2:(6)495-501. [Rec#: 185]

306. Molenaar PC, Biewenga JE, Van Kempen GT,

De Priester JA. Effect of ephedrine on
muscle weakness in a model of myasthenia
gravis in rats. Neuropharmacology.
1993;32:(4)373-6. [Rec#: 355]

296. McCleave DJ, Phillips PJ, Vedig AE.

Compartmental shift of potassium - a result
of sympathomimetic overdose. Aust N Z J
Med. 1978;8:(2)180-3. [Rec#: 482]

307. Moolenaar M, Desmond PA, Mascord DJ,

Starmer GA, Tattam B, Volkerts ER. The
effects of ephedrine on the development of
fatigue in a prolonged driving related task.
Hum Psychopharmacol Clin Exp.
1999;14:(6)415-427. [Rec#: 345]

297. McKenna D. Ma huang. INPR. 2001. [Rec#:

508]
298. McLaurin JQ, Shipman WF, Rosedale R. Oral
decongestants. A double blind comparison
study of the effectiveness of four
sympathomimetic drugs: Objective and
subjective. Laryngoscope. 1961;71:54-67.
[Rec#: 186]

308. Mores N, Campia U, Navarra P, Cardillo C,

Preziosi P. No cardiovascular effects of
single-dose pseudoephedrine in patients with
essential hypertension treated with betablockers. Eur J Clin Pharmacol.
1999;55:(4)251-4. [Rec#: 196]

299. McNamara SH. FDA Has Adequate Power

and Authority to Protect the Public From
Unsafe Dietary Supplements.
HERBALGRAM. 1996;38:25-7. [Rec#:
187]

309. Morgan JP. Phenylpropanolamine: A Critical

Analysis of Reported Adverse Reactions and
Overdosage. Fort Lee, NJ: Jack K. Burgess,
Inc.; 1986. [Rec#: 198]

300. Mendez MF. Mania self-induced with cough

syrup. J Clin Psychiatry. 1992;53:(5)173-4.
[Rec#: 188]

310. Morgan JP, Funderburk FR.

Phenylpropanalomine and blood pressure: a
review of prospective studies. Am J Clin
Nutr. 1992;55:(1 Suppl)206S-210S. [Rec#:
463]

301. Midha KK, Cooper JK, McGilveray IJ. Simple

and specific electron-capture GLC assay for
plasma and urine ephedrine concentrations
following single doses. J Pharm Sci.
1979;68:(5)557-60. [Rec#: 189]

311. Morimoto I, Watanabe F, Osawa T, Okitsu T,

Kada T. Mutagenicity screening of crude
drugs with Bacillus subtilis rec-assay and
Salmonella/microsome reversion assay.
Mutat Res. 1982;97:(2)81-102. [Rec#: 199]

302. Miller DK, McMahon LR, Green TA, Nation

JR, Wellman PJ. Repeated administration of
ephedrine induces behavioral sensitization in
rats. Psychopharmacology (Berl).
1998;140:(1)52-6. [Rec#: 190]

A1-21

Articles Excluded from Efficacy Analysis (continued)

321. National Academy of Sciences (NAS).
National Academy of Sciences Model:
Dietary Reference Intakes: A Risk
Assessment Model for Establishing Upper
Intake Levels for Nutrients. Washington
DC: Institute of Medicine, National
Academy Press (NAP). 1998. [Rec#: 205]

312. Morris HG, Rusnak SA, Selner JC, Barzens K,

Barnes J. Comparative effects of ephedrine
on adrenergic responsiveness in normal and
asthmatic subjects. J Allergy Clin Immunol.
1978;61:(5)294-302. [Rec#: 200]
313. Mortimer EA Jr. Drug toxicity from breast
milk? Pediatrics. 1977;60:(5)780-1. [Rec#:
201]

322. Nelson HS, Black JW, Branch LB, Pfuetze B,

Spaulding H, Summers R, et al.
Subsensitivity to epinephrine following the
administration of epinephrine and ephedrine
to normal individuals. J Allergy Clin
Immunol. 1975;55:(5)299-309. [Rec#: 206]

314. Mueller SM, Solow EB. Seizures associated

with a new combination "pick-me-up" pill.
Ann Neurol. 1982;11:(3)322. [Rec#: 487]
315. Nadeau SE. Intracerebral hemorrhage and
vasculitis related to ephedrine abuse. Ann
Neurol. 1984;15:(1)114-5. [Rec#: 484]

323. Nesheim MC. What is the research base for

the use of dietary supplements? Public
Health Nutr. 1999;2:(1)35-8. [Rec#: 538]

316. Nadeau SE. Intracerebral hemorrhage and

vasculitis related to ephedrine abuse. Ann
Neurol. 1984;15:(1)114-5. [Rec#: 455]

324. Neve KA, Molinoff PB. Effects of chronic

administration of agonists and antagonists
on the density of beta-adrenergic receptors.
Am J Cardiol. 1986;57:(12)17F-22F. [Rec#:
207]

317. Nadir A, Agrawal S, King PD, Marshall JB.

Acute hepatitis associated with the use of a
Chinese herbal product, ma-huang. Am J
Gastroenterol. 1996;91:(7)1436-8. [Rec#:
202]

325. Nightingale SL. From the Food and Drug

Administration. JAMA. 1996;275:(20)1534.
[Rec#: 495]

318. Nakahara Y, Kikura R. Hair analysis for drugs

of abuse. XIX. Determination of ephedrine
and its homologs in rat hair and human hair.
J Chromatogr B Biomed Sci Appl.
1997;700:(1-2)83-91. [Rec#: 456]

326. Nightingale SL. From the Food and Drug

Administration. JAMA. 1996;276:(21)1710.
[Rec#: 502]
327. NIOSH. Registry of Toxic Effects of Chemical
Substances, Centers for Disease Control and
Prevention. Ephedrine. 1997. [Rec#: 208]

319. Namba T, Kubo M, Kanai Y, Namba K,

Nishimura H, Qazilbash NA.
Pharmacognostical studies of Ephedra plants
Part I - The comparative histological studies
on Ephedra rhizomes from Pakistan and
Afghanistan and Chinese crude drugs "MaHung-Gen". Planta Med. 1976;29:(3)21625. [Rec#: 203]

328. Nishikawa T, Bruyere HJ Jr., Gilbert EF,

Takagi Y. Potentiating effects of caffeine on
the cardiovascular teratogenicity of
ephedrine in chick embryos. Toxicol Lett.
1985;29:(1)65-8. [Rec#: 360]
329. Nishikawa THJ, Bruyere HJ Jr., Takagi Y,
Gilbert EF, Uno H. Cardiovascular
teratogenicity of ephedrine in chick
embryos. Toxicol Lett. 1985;29:(1)59-63.
[Rec#: 209]

320. Naranjo CA, Busto U, Sellers EM, Sandor P,

Ruiz I, Roberts EA, et al. A method for
estimating the probability of adverse drug
reactions. Clin Pharmacol Ther.
1981;30:(2)239-45. [Rec#: 319]

A1-22

Articles Excluded from Efficacy Analysis (continued)

339. Pace S. Ma huang food supplement toxicity in
two adolescents. J Toxicol (Clin Toxicol).
1996;34:(5)598(Abstract). [Rec#: 218 ]

330. Noble R. A controlled clinical trial of the

cardiovascular and psychological effects of
phenylpropanolamine and caffeine. Drug
Intell Clin Pharm. 1988;22:(4)296-9. [Rec#:
382]

340. Pantelis C, Hindler CG, Taylor JC. Use and

abuse of khat (Catha edulis): a review of the
distribution, pharmacology, side effects and
a description of psychosis attributed to khat
chewing. Psychol Med. 1989;19:(3)657-68.
[Rec#: 219]

331. NTP. Toxicology and Carcinogenesis Studies

of Ephedrine Sulfate (CAS No. 134-72-5) in
F344/N Rats and B6C3F1 Mice (Feed
Studies) National Toxicology Program
(NTP), Research Triangle Park, NC, (NTP
Technical Reports Series) No. 307. 1986.
[Rec#: 211]

341. Pasquali R, Casimirri F. Clinical aspects of

ephedrine in the treatment of obesity. Int J
Obes Relat Metab Disord. 1993;17:(Suppl
1)S65-8. [Rec#: 221]

332. Nuotto E. Psychomotor, physiological and

cognitive effects of scopolamine and
ephedrine in healthy man. Eur J Clin
Pharmacol. 1983;24:(5)603-9. [Rec#: 212]

342. Pasquali R, Cesari MP, Besteghi L,

Melchionda N, Balestra V. Thermogenic
agents in the treatment of human obesity:
preliminary results. Int J Obes.
1987;11:(Suppl 3)23-6. [Rec#: 390]

333. Nyland H. [Ephedrine, abuse and psychosis].

Tidsskr Nor Laegeforen. 1973;93:(27)20279. [Rec#: 213]

343. Patil PN, Tye A, Lapidus JB. A

pharmacological study of the ephedrine
isomers. J Pharmacol Exper Therapeut.
1965;148:(2)158-168. [Rec#: 422]

334. Oberman Z, Herzberg M, Jaskolka H, Harell

A, Hoerer E, Laurian L. Changes in plasma
cortisol, glucose and free fatty acids after
caffeine ingestion in obese women. Isr J
Med Sci. 1975;11:(1)33-6. [Rec#: 373]

344. Pentel P. Toxicity of over-the-counter

stimulants. JAMA.1984;252:(14)1898-903.
[Rec#: 368]

335. Ogden JA. Congenital anomalies of the

coronary arteries. Am J Cardiol.
1970;25:(4)474-9. [Rec#: 545]

345. Pere JC, Begaud B, Haramburu F, Albin H.

Computerized comparison of six adverse
drug reaction assessment procedures. Clin
Pharmacol Ther. 1986;40:(4)451-61. [Rec#:
329]

336. Onuigbo M, Alikhan M. Over-the-counter

sympathomimetics: a risk factor for cardiac
arrhythmias in pregnancy. South Med J.
1998;91:(12)1153-5. [Rec#: 215]

346. Perrota DM, Coody G, Culmo C. Adverse

events associated with ephedrine-containing
products: Texas, December 1993-September
1995. JAMA. 1996;276:(21)1711-1712.
[Rec#: 224]

337. Ooms TG, Khan SA, Means C. Suspected

caffeine and ephedrine toxicosis resulting
from ingestion of an herbal supplement
containing guarana and ma huang in dogs:
47 cases (1997-1999). J Am Vet Med Assoc.
2001;218:(2)225-9. [Rec#: 216]

347. Peterson DI, Hardinge MG. The effect of

various environmental factors on cocaine
and ephedrine toxicity. J Pharm Pharmacol.
1967;19:(12)810-4. [Rec#: 225]

338. Oosterbaan R, Burns MJ. Myocardial

infarction associated with
phenylpropanolamine. J Emerg Med.
2000;18:(1)55-9. [Rec#: 217]

348. Pi-Sunyer XF. Medical hazards of obesity.

Ann Intern Med. 1993;119:(7 pt 2)655-660.
[Rec#: 464]

A1-23

Articles Excluded from Efficacy Analysis (continued)

359. Rang HP, Dale MM. Pharmacology. New
York, NY: Churchill Livingstone Inc.; 1995.
[Rec#: 233]

349. Pickup ME, May CS, Ssendagire R, Paterson

JW. The pharmacokinetics of ephedrine
after oral dosage in asthmatics receiving
acute and chronic treatment. Br J Clin
Pharmacol. 1976;3:(1)123-34. [Rec#: 226]

360. Read BE, Feng CT. The alleged ephedrine

action of two California species of ephedra.
Proc Soc Exp Biol Med. 1927;24:819-21.
[Rec#: 342]

350. Pickupp ME, Paterson JW. The determination

of ephedrine plasma levels by a gas
chromatographic method. J Pharm
Pharmacol. 1974;26:561-62. [Rec#: 432]

361. Renwick AG. The use of an additional safety

or uncertainty factor for nature of toxicity in
the estimation of acceptable daily intake and
tolerable daily intake values. Regul Toxicol
Pharmacol. 1995;22:(3)250-61. [Rec#: 234]

351. Pomerantz B, O'Rourke RA. The StokesAdams syndrome. Am J Med.

1969;46:(6)941-60. [Rec#: 227]

362. Renwick AG, Lazarus NR. Human variability

and noncancer risk assessment- An analysis
of the default uncertainty factor. Regul
Toxicol Pharmacol. 1998;27:(1 Pt 2)3-20.
[Rec#: 235]

352. Porta M, Jick H, Habakangas JA. Follow-up

study of pseudoephedrine users. Ann
Allergy. 1986;57:(5)340-2. [Rec#: 426]
353. Powell T, et al. Dangers of Ma-huang.
Alternative Medicine Alert. 1998;1:119.
[Rec#: 313]

363. Robertson D, Frolich JC, Carr RK, Watson JT,

Hollifield JW, Shand DG, et al. Effects of
caffeine on plasma renin activity,
catecholamines and blood pressure. N Engl J
Med. 1978;298:(4)181-6. [Rec#: 236]

354. Powell T, Hsu FF, Turk J, Hruska K. Mahuang strikes again: ephedrine
nephrolithiasis. Am J Kidney Dis.
1998;32:(1)153-9. [Rec#: 228]

364. Rogers P, Eftimiades M, Aunapu G, Keller M,

Jones R. Lethal but legal: an over-thecounter high takes the life of a college
student. People. 1996;45:(20)105-109.
[Rec#: 462]

355. Puar HS. Acute memory loss and nominal

aphasia caused by phenylpropanolamine.
South Med J. 1984;77:(12)1604-5. [Rec#:
229]

365. Rosene JM, Rosene JA, Koczajowski DL,

Paolone VJ. Decongestant effects on
hemodynamics at rest, exercise, and
recovery from exercise during--6 degrees of
head down tilt. Aviat Space Environ Med.
1999;70:(1)15-21. [Rec#: 237]

356. Raadstroem M, Bengtssonk J, Ederberg S,

Bengtson A, Loswick AC, Bengtson JP.
Effects of ephedrine on oxygen consumption
and cardiac output. Acta Anaesthesiol
Scand. 1995;39:(8)1084-87. [Rec#: 347]
357. Radstrom M, Bengtsson J, Ederberg S,
Bengtsson A, Loswick AC, Bengtson JP.
Effects of ephedrine on oxygen consumption
and cardiac output. Acta Anaesthesiol
Scand. 1995;39:(8)1084-7. [Rec#: 231]

366. Roth RP, Cantekin EI, Bluestone CD, Welch

RM, Cho YW. Nasal decongestant activity
of pseudoephedrine. Ann Otol Rhinol
Laryngol. 1977;86:(2 pt. 1)235-42. [Rec#:
428]

358. Ramsey JJ, Colman RJ, Swick AG, Kemnitz

JW. Energy expenditure, body composition,
and glucose metabolism in lean and obese
rhesus monkeys treated with ephedrine and
caffeine. Am J Clin Nutr. 1998;68:(1)42-51.
[Rec#: 232]

367. Roxanas MG, Spalding J. Ephedrine abuse

psychosis. Med J Aust. 1977;2:(19)639-40.
[Rec#: 238]

A1-24

Articles Excluded from Efficacy Analysis (continued)

368. Rutstein HR. Ingestion of pseudoephedrine.
Arch Otolaryngol. 1963;77:145-47. [Rec#:
423]

378. Sheu S-J. Identification by chemical analysis

of the botanical sources of commercial
samples of Chinese herbal drugs. Yao Wu
Shih P'in Fen Hsi (J Food Drug Anal).
1997;5:(4)285-294. [Rec#: 246]

369. Ryan DH. Medicating the obese patient.

Endocrinol Metab Clin North Am.
1996;25:(4)989-1004. [Rec#: 239]

379. Shu H, Jin B, Yang Z. The mechanism of

inhibitory action of ephedrine on the twitch
response of mouse vas deferens to field
stimulation. Sheng Li Hsueh Pao.
1986;38:(6)583-88. [Rec#: 352]

370. Sagara K, Oshima T, Misaki T. A

simultaneous determination of norephedrine,
pseudoephedrine, ephedrine and
methylephedrine in Ephedrae Herba and
oriental pharmaceutical preparations by ionpair high-performance liquid
chromatography. Chem Pharm Bull
(Tokyo). 1983;31:(7)2359-65. [Rec#: 240]

380. Sieb JP, Engel AG. Ephedrine: effects on

neuromuscular transmission. Brain Res.
1993;623:(1)167-71. [Rec#: 358]
381. Sinsheimer JE, Dring LG, Williams RT.
Species differences in the metabolism of
norephedrine in man, rabbit and rat.
Biochem J. 1973;136:(3)763-71. [Rec#:
248]

371. Samenuk D, Link MS, Homoud MK,

Contreras R, Theohardes TC, Wang PJ, et
al. Adverse cardiovascular events
temporally associated with ma huang, an
herbal source of ephedrine. Mayo Clin Proc.
2002;77:(1)12-6. [Rec#: 473]

382. Smith DA, Perry PJ. The efficacy of ergogenic

agents in athletic competition. Part II: Other
performance-enhancing agents. Ann
Pharmacother. 1992;26:(5)653-9. [Rec#:
249]

372. Saunders A, Elliott P. Pseudoephedrine in the

hypertensive patient. Australian J Hospit
Pharm. 1987;17:(2)150-151. [Rec#: 424]

383. Snook C, Otten M, Hassan M. Massive

ephedrine overdose: Case report and
toxicokinetic analysis. Vet Hum Toxicol.
1992;34:(4)335(Abstract). [Rec#: 250]

373. Sawynok J. Pharmacological rationale for the

clinical use of caffeine. Drugs.
1995;49:(1)37-50. [Rec#: 383]
374. Schryver T. The buzz about ephedra and
caffeine. Dietitian's Edge. 2001;March
Apr:30-31, 33-34. [Rec#: 315]

384. Sobal J, Marquart LF. Vitamin/mineral

supplement use among athletes: a review of
the literature. Int J Sport Nutr.
1994;4:(4)320-34. [Rec#: 537]

375. Sever PS, Dring LG, Williams RT. The

metabolism of (-)-ephedrine in man. Eur J
Clin Pharmacol. 1975;9:(2-3)193-8. [Rec#:
242]

385. Sprague JE, Harrod AD, Teconchuk AL.

Pharmacology and abuse potential of
ephedrine. Pharm Times. 1998;64:72-80.
[Rec#: 251]

376. Sharma AN, Cinci PM, Hoffman RS. Ma

huang-induced myocardial infarction with
normal coronary arteries. J Toxicol (Clin
Toxicol). 2000;38:(5)521(Abstract). [Rec#:
244]

386. Stoessl AJ, Young GB, Feasby TE.

Intracerebral haemorrhage and angiographic
beading following ingestion of
catecholaminergics. Stroke. 1985;16:734-6.
[Rec#: 514]

377. Shenfield GM. Combination bronchodilator

therapy. Drugs. 1982;24:(5)414-39. [Rec#:
245]

A1-25

Articles Excluded from Efficacy Analysis (continued)

387. Stromberg C, Vanakoski J, Olkkola KT,
Lindqvist A, Seppala T, Laitinen LA.
Exercise alters the pharmacokinetics of
midazolam. Clin Pharmacol Ther.
1992;51:(5)527-32. [Rec#: 252]

396. Taylor DM, O'Toole KS, Auble TE, Ryan CM,

Sherman DR. The psychometric and cardiac
effects of pseudoephedrine in the hyperbaric
environment. Pharmacotherapy.
2000;20:(9)1045-50. [Rec#: 257]

388. Swain RA, Harsha DM, Baenziger J, Saywell

RM. Do pseudoephedrine or
phenylpropanolamine improve maximum
oxygen uptake and time to exhaustion? Clin
J Sport Med. 1997;7:(3)168-73. [Rec#: 253]

397. Tekol Y, Tercan E, Esmaoglu A. Ephedrine

enhances analgesic effect of morphine. Acta
Anaesthesiol Scand. 1994;38:(4)396-7.
[Rec#: 457]
398. Theoharides TC. Sudden death of a healthy
college student related to ephedrine toxicity
from a ma huang-containing drink. J Clin
Psychopharmacol. 1997;17:(5)437-9. [Rec#:
258]

389. Tagliabue A, Terracina D, Cena H, Turconi G,

Lanzola E, Montomoli C. Coffee induced
thermogenesis and skin temperature. Int J
Obes Relat Metab Disord. 1994;18:(8)53741. [Rec#: 377]

399. Tinsley JA, Watkins DD. Over-the-counter

stimulants: abuse and addiction. Mayo Clin
Proc. 1998;73:(10)977-82. [Rec#: 493]

390. Takaku T, Ming J, Okuda H, Maeda N.

Isolation of insulin-like substances from
ephedra sinica stapf. Wakan Iyakugaku
Zasshi. 1997;14:(4)358-9. [Rec#: 332]

400. Tinsley JA. The hazards of psychotropic herbs.

Minn Med. 1999;82:(5)29-31. [Rec#: 259]

391. Takayasu T, Ohshima T, Ohtsuji M,

Takayama N, Kondo T, Lin Z, et al. False
identification of urinary ephedrine as
methamphetamine by gas
chromatography/mass spectrometry with a
DB-17 column. Nippon Hoigaku Zasshi.
1997;51:(3)235-40. [Rec#: 459]

401. To LB, Sangster JF, Rampling D, Cammens I.

Ephedrine-induced cardiomyopathy. Med J
Aust. 1980;2:(1)35-6. [Rec#: 260]
402. Toubro S, Astrup A, Breum L, Quaade F. The
acute and chronic effects of
ephedrine/caffeine mixtures on energy
expenditure and glucose metabolism in
humans. Int J Obes Relat Metab Disord.
1993;17:(Suppl 3)S73-7; discussion S82.
[Rec#: 262]

392. Tamada M, Endo K, Hikino H. Maokonine,

hypertensive principle of Ephedra roots.
Planta Med. 1978;34:(3)291-3. [Rec#: 336]
393. Tamada M, Endo K, Hikino H, Kabuto C.
Structure of Ephedradine A, a hypotensive
principle of ephedra roots. Planta Med.
1978;34:291. [Rec#: 341]

403. Traub SJ; Hoyek W; Hoffman RS. Dietary

supplements containing ephedra alkaloids.
New England Journal of Medicine.
2001;344:1096. [Rec#: 516]

394. Tanaka M, Dohi S. [The effects of ephedrine

and phenylephrine on arterial partial
pressure of oxygen during one-lung
ventilation]. Masui. 1994;43:(8)1124-9.
[Rec#: 254]

404. Turner JE, Cravey RH. A fatal ingestion of

caffeine. Clin Toxicol. 1977;10:(3)341-4.
[Rec#: 264]
405. Tyler VE. The Honest Herbal: A Sensible
Guide to the Use of Herbs and Related
Remedies. New York, NY: Pharmaceutical
Products Press (PPP); 1993. pp. 119-121.
[Rec#: 265]

395. Tashkin DP, Meth R, Simmons DH, Lee YE.

Double-blind comparison of acute bronchial
and cardiovascular effects of oral terbutaline
and ephedrine. Chest. 1975;68:(2)155-61.
[Rec#: 256]

A1-26

Articles Excluded from Efficacy Analysis (continued)

415. Vanakoski J, Stromberg C, Seppala T. Effects
of a sauna on the pharmacokinetics and
pharmacodynamics of midazolam and
ephedrine in healthy young women. Eur J
Clin Pharmacol. 1993;45:(4)377-81. [Rec#:
274]

406. Ueng TH, Kang JJ, Wang HW, Cheng YW,

Chiang LY. Suppression of microsomal
cytochrome P450-dependent
monooxygenases and mitochondrial
oxidative phosphorylation by fullerenol, a
polyhydroxylated fullerene C60. Toxicol
Lett. 1997;93:(1)29-37. [Rec#: 266]

416. Vansal SS, Feller DR. Direct effects of

ephedrine isomers on human beta-adrenergic
receptor subtypes. Biochem Pharmacol.
1999;58:(5)807-10. [Rec#: 361]

407. Ueng T-H, Kang J-J, Wang H-W, Lin P-C. An

overview of the toxicology of commonly
used traditional Chinese medicine. Yao Wu
Shih P'in Fen Hsi (J Food Drug Anal).
1997;5:(4)241-264. [Rec#: 267]

417. Villas Martinez F, Badas AJ, Garmendia

Goitia JF, Aguirre I. Generalized dermatitis
due to oral ephedrine. Contact Dermatitis.
1993;29:(4)215-6. [Rec#: 503]

408. Unger DL. Blood pressure and pulse rate

changes in hypertensive asthmatic patients:
effects of an ephedrine compound. Ann
Allergy. 1968;26:(12)637-8. [Rec#: 429]

418. Wahl M, Grant C, Jesperson R. Acute

psychosis from a dietary weight control
supplement. J Toxicol (Clin Toxicol).
2000;38:(5)522-523.(Abstract). [Rec#: 275]

409. United States General Accounting Office.

Dietary Supplements: Uncertainties in
Analyses Underlying FDA's Proposed Rule
on Ephedrine Alkaloids. Washington, DC:
United States General Accounting Office.
1999. [Rec#: 268]

419. Walker RB, Fitz LD, Williams LM, Linton H

Jr., Smith CC. The effect of ephedrine
isomers and their oxazolidines on locomotor
activity in rats. Gen Pharmacol.
1993;24:(3)669-73. [Rec#: 359]

410. Upton RA. Pharmacokinetic interactions

between theophylline and other medication
(Part I). Clin Pharmacokinet. 1991;20:(1)6680. [Rec#: 269]

420. Warot D, Berlin I, Aymard G, Acquaviva E,

Legrand M, Labarthe B, et al. Comparative
pharmacodynamics and pharmacokinetics of
oral and intranasal ephedrine. Fun & Clin
Pharma. 2000. Abstract 237-297. [Rec#:
277]

411. USDA. Dietary supplements containing

ephedrine alkaloids; proposed rule. 21 CFR
111. Federal Register. 1997;62:(107)30677724. [Rec#: 445]

421. Warren MR, Werner HW. Acute toxicity of

vasopressor amines: II. Comparative data. J
Pharmacol Exper Therapeut. 1945;86:284286. [Rec#: 278]

412. Vahedi K, Domigo V, Amarenco P, Bousser

M G. Ischaemic stroke in a sportsman who
consumed MaHuang extract and creatine
monohydrate for body building. J Neurol
Neurosurg Psychiatry. 2000;68:(1)112-3.
[Rec#: 270]

422. Warren MR, Werner HW. The central

stimulant action of some vasopressor
amines. J Pharmacol Exper Therapeut.
1945;85:119-121. [Rec#: 279]

413. Van Mieghem W, Stevens E, Cosemans J.

Ephedrine-induced cardiopathy. Br Med J.
1978;1:(6116)816. [Rec#: 271]

423. Warren MR, Werner HW. Effect of

temperature on the toxicity of some
vasopressor amines. Fed Proc (FASEB).
1945;4:139(Abstract). [Rec#: 280]

414. Vanakoski J, Seppala T. Heat exposure and

drugs. A review of the effects of
hyperthermia on pharmacokinetics. Clin
Pharmacokinet. 1998;34:(4)311-22. [Rec#:
273]

A1-27

Articles Excluded from Efficacy Analysis (continued)

432. WHO. Principles for the Safety Assessment of
Food Additives and Contaminants in Food.
World Health Organization (WHO),
International Programme on Chemical
Safety (IPCS), Geneva, Switz.
Environmental Health Criteria No. 70. 1987.
[Rec#: 286]

424. Webb AA, Shipton EA. Re-evaluation of i.m.

ephedrine as prophylaxis against
hypotension associated with spinal
anaesthesia for Caesarean section. Can J
Anaesth. 1998;45:(4)367-9. [Rec#: 458]
425. Weesner KM, Denison M, Roberts RJ. Cardiac
arrhythmias in an adolescent following
ingestion of an over-the- counter stimulant.
Clin Pediatr (Phila) 1982 21:(11)700-1.
[Rec#: 281]

433. Wiener ITAPM. Coronary artery spasm and

myocardial infarction in a patient with
normal coronary arteries: temporal
relationship to pseudoephedrine ingestion.
Catheterization & Cardiovascular Diagnosis.
1990;20:51-3. [Rec#: 518]

426. Weinberger M, Bronsky E, Bensch GW, Bock

GN, Yecies JJ. Interaction of ephedrine and
theophylline. Clin Pharmacol Ther.
1975;17:(5)585-92. [Rec#: 282]

434. Wilkinson GR, Beckett AH. Absorption

metabolism and excretion of the ephedrines
in man. I. The influence of urinary pH and
urine volume output. J Pharmacol Exp Ther.
1968;162:(1)139-47. [Rec#: 287]

427. Weiner I, Tilkian AG, Palazzolo M. Coronary

artery spasm and myocardial infarction in a
patient with normal coronary arteries:
temporal relationship to pseudoephedrine
ingestion. Cathet Cardiovasc Diagn.
1990;20:(1)51-3. [Rec#: 437]

435. Wilkinson GR, Beckett AH. Absorption,

metabolism, and excretion of the ephedrines
in man. II. Pharmacokinetics. J Pharm Sci.
1968;57:(11)1933-8. [Rec#: 288]

428. Welling PG, Lee KP, Patel JA, Walker JE,

Wagner JG. Urinary excretion of ephedrine
in man without pH control following oral
administration of three commercial
ephedrine sulfate preparations. J Pharm Sci.
1971;60:(11)1629-34. [Rec#: 283]

436. Williams R, Caldwell J, Dring L. Comparative

metabolism of some amphetamines in
various species. In: Usdin E, Snyder S. (Eds)
Frontiers in Catecholamine Research. New
York, NY: Pergamon Press; 1973. pp. 927932. [Rec#: 289]

429. Wellman PJ, Miller DK, Livermore CL, Green

TA, McMahon LR, Nation JR. Effects of ()-ephedrine on locomotion, feeding, and
nucleus accumbens dopamine in rats.
Psychopharmacology (Berl).
1998;135:(2)133-40. [Rec#: 354]

437. Williamson DF, Pamuk E, Thun M, Flanders

D, Byers T, Heath C. Prospective study of
intentional weight loss and mortality in
never-smoking overweight US white women
aged 40-64 years. Am J Epidemiol.
1995;141:(12)1128-41. [Rec#: 544]

430. White LM, Gardner SF, Gurley BJ, Marx MA,

Wang PL, Estes M. Pharmacokinetics and
cardiovascular effects of ma-huang (Ephedra
sinica) in normotensive adults. J Clin
Pharmacol. 1997;37:(2)116-22. [Rec#: 284]

438. Wilson AF, Novey HS, Cloninger P, Davis J,

White D. Cardiopulmonary effects of longterm bronchodilator administration. J
Allergy Clin Immunol. 1976;58:(1 PT.
2)204-12. [Rec#: 290]

431. Whitehouse AM, Duncan JM. Ephedrine

psychosis rediscovered. Br J Psychiatry.
1987;150:258-61. [Rec#: 285]

439. Wolf HH, Schuster MA, Lanciault G.

'Aggregation' phenomenon and some
ephedrine isomers. J Pharmaceut Sci.
1964;53:987-988. [Rec#: 291]

A1-28

Articles Excluded from Efficacy Analysis (continued)

450. Yoshida T, Sakane N, Umekawa T, Kondo M.
Relationship between basal metabolic rate,
thermogenic response to caffeine, and body
weight loss following combined low calorie
and exercise treatment in obese women. Int J
Obes Relat Metab Disord. 1994;18:(5)34550. [Rec#: 379]

440. Wood KA. Nasal decongestant and psychiatric

disturbance. Br J Psychiatry.
1994;164:(4)566-7. [Rec#: 492]
441. Wooten MR, Khangure MS, Murphy MJ.
Intracerebral hemorrhage and vasculitis
related to ephedrine abuse. Ann Neurol.
1983;13:(3)337-40. [Rec#: 438]

451. Yoshida T, Sakane N, Wakabayashi Y,

Umekawa T, Kondo M. Thermogenic, antiobesity effects of bofu-tsusho-san in MSGobese mice. Int J Obes Relat Metab Disord.
1995;19:(10)717-22. [Rec#: 394]

442. Xue-Jun Y, De-xiang L, Hechuan W, Yu Z. A

study on the mutagenicity of 102 raw
pharmaceuticals used in Chinese traditional
medicine. Mutat Res. 1991;260:(1)73-82.
[Rec#: 292]

452. Young R, Gabryszuk M, Glennon RA. ()Ephedrine and caffeine mutually potentiate
one another's amphetamine- like stimulus
effects. Pharmacol Biochem Behav.
1998;61:(2)169-73. [Rec#: 357]

443. Yacobi A, Stoll RG, Chao GC, Carter JE,

Baaske DM, Kamath BL, et al. Evaluation
of sustained-action chlorpheniraminepseudoephedrine dosage form in humans. J
Pharm Sci. 1980;69:(9)1077-81. [Rec#: 430]

453. Young R, Glennon RA. Discriminative

stimulus properties of (-)ephedrine.
Pharmacol Biochem Behav. 1998;60:(3)7715. [Rec#: 494]

444. Yale Hemorrhagic Stroke Project Final Study

Report. The safety of phenylpropanolamine
hydrochloride (PPA) used in OTC weight
control and nasal decongestant drug
products. 2000. [Rec#: 293]

454. Yuan D, Komatsu K, Cui Z, Kano Y.

Pharmacological properties of traditional
medicines. XXV. Effects of ephedrine,
amygdalin, glycyrrhizin, gypsum and their
combinations on body temperature and body
fluid. Biol Pharm Bull. 1999;22:(2)165-71.
[Rec#: 362]

445. Yates KM, O'Connor A, Horsley CA. "Herbal

Ecstasy": a case series of adverse reactions.
N Z Med J. 2000;113:(1114)315-7. [Rec#:
294]
446. Yen TT, McKee MM, Bemis KG. Ephedrine
reduces weight of viable yellow obese mice
(Avy/a). Life Sci. 1981;28:(2)119-28.
[Rec#: 295]

455. Zaacks SM, Klein L, Tan CD, Rodriguez ER,

Leikin JB. Hypersensitivity myocarditis
associated with ephedra use. J Toxicol Clin
Toxicol. 1999;37:(4)485-9. [Rec#: 297]

447. Yin P. Ephedrine-induced intracerebral

hemorrhage and central nervous system
vasculitis. Stroke. 1990;21:1641. [Rec#:
515]

456. Zahn KA, Li RL, Purssell RA. Cardiovascular

toxicity after ingestion of "herbal ecstacy." J
Emerg Med. 1999;17:(2)289-91. [Rec#:
298]

448. Yin PA. Ephedrine-induced intracerebral

hemorrhage and central nervous system
vasculitis. Stroke. 1990;21:(11)1641. [Rec#:
483]

457. Zeisel SH. Is there a metabolic basis for

dietary supplementation? Am J Clin Nutr.
2000;72:(2 Suppl)507S-11S. [Rec#: 299]

449. Yoovathaworn KC, Sriwatanakul K,

Thithapandha A. Influence of caffeine on
aspirin pharmacokinetics. Eur J Drug Metab
Pharmacokinet. 1986;11:(1)71-6. [Rec#:
296]

A1-29

Articles Excluded from Efficacy Analysis (continued)

458. Zgourides GD, Warren R, Englert ME.
Ephedrine-induced thermogenesis as an
adjunct to cognitive restructuring and covert
conditioning: a proposal for treatment of
obese individuals. Percept Mot Skills.
1989;69:(2)563-72. [Rec#: 401]
459. Zhang J, Tian Z, Lou ZC. Simultaneous
determination of six alkaloids in Ephedrae
herba by high performance liquid
chromatography. Planta Med.
1988;54:(1)69-71. [Rec#: 300]
460. Zhang JS, Tian Z, Lou ZC. Quality evaluation
of twelve species of Chinese Ephedra (mahuang). Chung Kuo Yao Li Hsueh Pao (Acta
Pharmacologica Sinica). 1993;22:419-425.
[Rec#: 301]

A1-30

Case Report and Case Series Articles Identified

ID#: 17

Backer R, Tautman D, Lowry S, Harvey

CM, Poklis A. Fatal ephedrine intoxication.
J Forensic Sci. 1997;42(1):157-9.

ID#: 44

ID#: 48

ID#: 64

ID#: 69

ID#: 79

ID#: 96

ID#: 110

ID#: 120

ID#: 136

ID#: 157

ID#: 174

ID#: 184

Matthews G, Smolinske S, White S.

Ephedrine-related stroke in a teenager. J
Toxicol (Clin Toxicol).
1997;35:(5)555(Abstract).

Bruno A, Nolte KB, Chapin J. Stroke

associated with ephedrine use. Neurology.
1993;43(7):1313-6.

ID#: 218

Pace S. Ma huang food supplement toxicity

in two adolescents. J Toxicol (Clin Toxicol).
1996;34:(5)598(Abstract).

Capwell RR. Ephedrine-induced mania from

an herbal diet supplement. Am J Psychiatry.
1995;152(4):647.

ID#: 224

Perrota DM, Coody G, Culmo C. Adverse

events associated with ephedrine-containing
products: Texas, December 1993-September
1995. JAMA. 1996;276(21):1711-1712.

ID#: 238

Roxanas MG, Spalding J. Ephedrine abuse

psychosis. Med J Aust. 1977;2(19):639-40.

ID#: 244

Sharma AN, Cinci PM, Hoffman RS. Ma

huang-induced myocardial infarction with
normal coronary arteries. J Toxicol (Clin
Toxicol). 2000;38:(5)521(Abstract).

ID#: 250

Snook C, Otten M, Hassan M. Massive

ephedrine overdose: Case report and
toxicokinetic analysis. Vet Hum Toxicol.
1992;34:(4)335(Abstract).

ID#: 258

Theoharides TC. Sudden death of a healthy

college student related to ephedrine toxicity
from a ma huang-containing drink. J Clin
Psychopharmacol. 1997;17(5):437-9.

ID#: 260

To LB, Sangster JF, Rampling D, Cammens

I. Ephedrine-induced cardiomyopathy. Med
J Aust. 1980;2(1):35-6.

Herridge CF, Brook MF. Ephedrine

psychosis. Br Med J. 1968;2(598):160.

ID#: 271

Jacobs KM, Hirsch KA. Psychiatric

complications of Ma-huang.
Psychosomatics. 2000;41(1):58-62.

Van Mieghem W, Stevens E, Cosemans J.

Ephedrine-induced cardiopathy. Br Med J.
1978;1(6116):816.

ID#: 281

Lake CR, Tenglin R, Chernow B, Holloway

HC. Psychomotor stimulant-induced mania
in a genetically predisposed patient: a
review of the literature and report of a case.
J Clin Psychopharmacol. 1983;3(2):97-100.

Weesner KM, Denison M, Roberts RJ.

Cardiac arrhythmias in an adolescent
following ingestion of an over-the-counter
stimulant. Clin Pediatr (Phila).
1982;21(11):700-1.

ID#: 285

Whitehouse AM, Duncan JM. Ephedrine

psychosis rediscovered. Br J Psychiatry.
1987;150:258-61.

ID#: 297

Zaacks SM, Klein L, Tan CD, Rodriguez

ER, Leikin JB. Hypersensitivity myocarditis
associated with ephedra use. J Toxicol Clin
Toxicol. 1999;37(4):485-9.

Cockings JG, Brown M. Ephedrine abuse

causing acute myocardial infarction. Med J
Aust. 1997;167(4):199-200.
Dawson JK, Earnshaw SM, Graham CS.
Dangerous monoamine oxidase inhibitor
interactions are still occurring in the 1990s. J
Accid Emerg Med. 1995;12(1):49-51.
Doyle H, Kargin M. Herbal stimulant
containing ephedrine has also caused
psychosis. BMJ. 1996;313(7059):756.
Garriott JC, Simmons LM, Poklis A,
Mackell MA. Five cases of fatal overdose
from caffeine-containing "look-alike" drugs.
J Anal Toxicol. 1985;9(3):141-3.
Gualtieri J, Harris C. Dilated
cardiomyopathy in a heavy ephedrine
abuser. J Toxicol (Clin Toxicol).
1996;34:(5)581-582.(Abstract).

Lustik SJ, Chhibber AK, van Vliet M,

Pomerantz RM. Ephedrine-induced coronary
artery vasospasm in a patient with prior
cocaine use. Anesth Analg. 1997;84(4):9313.

A1-31

Case Report and Case Series Articles Identified (continued)

ID#: 438

Wooten MR, Khangure MS, Murphy MJ.

Intracerebral hemorrhage and vasculitis
related to ephedrine abuse. Ann Neurol.
1983;13(3):337-40.

ID#: 462

Rogers P, Eftimiades M, Aunapu G, Keller

M, Jones R. Lethal but legal: an over-thecounter high takes the life of a college
student. People. 1996;45(20):105-109.

ID#: 485

Hirsch MS, Walter RM, Hasterlik RJ.

Subarachnoid hemorrhage following
ephedrine and MAO inhibitor. JAMA.
1965;194(11):1259.

ID#: 488

Escobar JI, Karno M. Chronic hallucinosis

from nasal drops. JAMA.
1982;247(13):1859-60.

ID#: 490

Kane FJ Jr., Florenzano R. Psychosis

accompanying use of bronchodilator
compound. JAMA. 1971;215(13):2116.

ID#: 493

Tinsley JA, Watkins DD. Over-the-counter

stimulants: abuse and addiction. Mayo Clin
Proc. 1998;73(10):977-82.

ID#: 514

Stoessl A, Young G, Feasby T. Intracerebral

haemorrhage and angiographic beading
following ingestion of catecholaminergics.
Stroke. 1985;16:734-6.

ID#: 515

Yin P. Ephedrine-induced intracerebral

hemorrhage and central nervous system
vasculitis. Stroke. 1990;21:1641.

ID#: 516

Traub S, Hoyek W, Hoffman R. Dietary

supplements containing ephedra alkaloids.
New England Journal of Medicine.
2001;344:1096.

ID#: 519

Emmanuel N, Jones C, Lydiard R. Use of

herbal products and symptoms of bipolar
disorder. American Journal of Psychiatry.
1998;155:1627.

ID#: 552

Kaberi-Otarod J, Conetta R, Kundo KK,

Farkash A. Ischemic stroke in a user of
Thermadrene: A case study in alternative
medicine. Clin Pharmacol Ther.
2002;72:343-6.

A1-32

Appendix 2. Metabolife Serious Adverse Events

A2-1

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-2

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-3

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-4

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-5

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-6

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-7

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-8

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-9

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-10

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-11

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-12

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-13

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-14

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-15

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-16

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-17

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-18

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-19

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-20

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-21

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-22

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-23

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-24

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-25

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-26

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-27

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-28

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-29

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-30

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-31

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-32

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-33

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-34

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-35

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-36

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-37

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-38

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-39

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-40

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-41

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-42

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-43

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-44

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-45

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-46

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-47

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-48

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-49

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-50

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-51

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-52

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-53

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-54

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-55

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-56

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-57

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-58

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-59

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-60

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-61

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-62

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-63

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-64

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-65

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-66

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-67

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-68

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-69

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-70

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-71

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-72

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-73

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-74

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-75

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-76

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-77

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-78

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-79

Appendix 2. Metabolife Serious Adverse Events (continued)

A2-80

Appendix 3. Reviewer Comments

Reviewer Comment

Rand Response

Title is not very informative. Should include something about Change Made
the conditions under study. Example: Efficacy and Safety of
Ephedra for Weight Management and Athletic Performance
Enhancement.
Since the stated overall objective is to assess he efficacy of We think this is more appropriate for the
herbal ephedra and synthetic ephedrine on weight loss and text, and the title reflects the uses for
athletic performance and since there is stated too few which we attempted to find evidence.
studies and data available to conduct an analysis of herbal
ephedra on athletic performance, should not the title of the
study be altered or the reported at least noted to reflect this
limitation?
I think you did an excellent job. Having reviewed this subject No Response
in more superficial fashion in the past, I can appreciate,
more than most, what fine job you have done.
The overall purpose of the evaluation, including the
questions, methods, findings and conclusions are clearly
and succinctly written and easy to understand.

No Response

The search for relevant data appears to have been thorough No Response
and encompassed a broad range of literature resources.
The study selection appears to be appropriate for an
evidence-based review of this type.

No Response

Data collection and data synthesis appear to be reasonable. No Response

This is an excellent comprehensive review, and it will make
an important contribution to the literature. Strong points are
a clear description of review criteria, rigorous assessment
methods, and straightforward data presentation. The
questions formulated are relevant and appropriate, search
strategies seem reasonable, and study selection is well
justified. The meta-analyses are useful.

No Response

The Evidence Report utilizes modern methods of metaanalysis of clinical trials. However, it ignores a great deal of
scientific evidence that can augment the interpretation of
data from the clinical trials and has a major structural flaw
and several weaknesses that are discussed below.

No Response. A specific response to the

great deal of scientific evidence ignored
is presented where such evidence is
specifically referred to.

In my area of expertise (clinical studies of obesity), the

findings were consistent with my understanding of the
literature.

No Response

The overall evaluation is clear, and the purpose of the report No Response
is well stated.
Overall I found the report well-researched and written.

No Response

The questions were adequately formulated and easily

understood.

No Response

A3-1

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

Reasons for inclusion or exclusion of studies were clear.

No Response

In evaluating the obesity weight loss clinical studies, the

report acknowledges some of the problems [small numbers
of subjects, short durations of treatment, etc.] and states
that long-term assessments of effectiveness are lacking. It
would be useful to put these statements in the context of
current knowledge in this area: that weight loss generally
ceases after about 6 months irrespective of the treatment
and any weight lost is generally regained. Current
recommendations for appropriate clinical trials in this area
include a much longer duration of treatment [1 2 years]
and an evaluation of what happens after the agent is
withdrawn . Both of these are very important in evaluating
the efficacy and the risk to benefit ratio of a particular
substance. Although ephedrine plus caffeine combinations
[pharmaceutical and dietary supplement sources] are being
compared to certain prescription drugs, to date no
ephedrine plus caffeine product has undergone the
equivalent types of efficacy and safety studies that are
required prior to marketing of a prescription drug in the US.

This information was added to the

limitations.

The purpose of the study and the means for arriving at its
conclusions were clear and relatively easy to follow. The
Meta analysis approach was appropriate and the criteria
well defined. I believe some discussion should be given to
the purported mechanisms of action (i.e. anoretic versus
thermogenic) behind the " statistically significant "weight
loss attributed to synthetic ephedrine/caffeine/ or ephedracontaining dietary supplements. The impression given by
the meta analysis results is that, while statistically
significant, these types of products also provide clinically
relevant weight reductions. Given the results of the case
report analyses, I don't believe the benefit of minimal weight
loss (e.g. 1 to 3 pounds per month) outweighs the potential
risk of serious adverse health effects exemplified by the
case report analysis. Despite the study's inability to assign
causality to most, if not all, of the serious adverse events,
the authors, in their conclusion, seem to downplay the
"potential" risks associated with these products.

This communicates a value judgment

about the balance of evidence that is
beyond the scope of the EPC. The
concern about the report downplaying the
potential risks is, as later peer review
comments will indicate, shared by some
other reviewers, but directly contradicted
by others.

The appraisal of ephedra studies for weight loss could

include a stronger statement about the unusually high
attrition rates as compared to many drug studies. Although
this is mentioned in the Limitations section, it also might be
included in the results section where the data is interpreted.
Can you expand on whether attrition rates differed between
treatments a placebo groups? In my view, this is a major
weakness of the recent efficacy studies involving ephedra.

Attrition rates did not differ between

treatment and placebo groups. This has
now been added to the results.

It is stated under Findings [p4] and elsewhere that in

Change Made
aggregate the clinical trials only enrolled a sufficient number
of patients to detect a serious adverse event rate of one per
one thousand or three per thousand in the case of

A3-2

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

botanical sources of ephedrine It would be useful to put

these numbers in the context of the frequencies of adverse
events [common, infrequent, rare, etc.] Using commonly
accepted definitions , all of the current clinical trials in
aggregate, irrespective of source, lack the power to detect
any rare adverse event [defined as greater than 1 per 1000
rate or frequency].
Throughout the report, reference is made to "synthetic
ephedrine". I suggest deleting "synthetic", since ephedrine
is ephedrine. Some is extracted from plants and some is
synthesized.
The term "synthetic ephedrine" is ambiguous due to the
meaning of the terms "natural" and "synthetic" with respect
to natural products chemistry. What could be meant are
synthetically derived ephedrine alkaloids because these are
natural products by virtue of their existence as naturally
occurring compounds regardless of how they are produced.
Ephedrine is by definition always a natural product unless
one is referring to the racemate that is produced during
some synthetic production processes because the specific
optical isomer that is identical to naturally occurring
ephedrine is itself is itself often synthesized through chiral
specific processes. The fact of the matter is that what the
draft means when referring to "synthetic ephedrine" could
be either naturally or synthetically derived. It may be
preferable then, in the interest of clarity throughout the
document, to use some consistent terminology, such as:
"ephedra" as the name of the crude raw material (with
parenthetical identification of the pinyin name: ma huang
one time, but not as a substitute common name) which
consist of the dried stem of the plant; "ephedra extract"
when referring to raw materials or ingredients that are
processed extracts of ephedra; "ephedrine " when referring
specifically to those one alkaloids as found in the plant or
wherever the term "synthetic ephedrine" now occurs in the
draft.

As these contradictory comments indicate,

there is no agreement among experts
about standardized terminology. In this
report, for simplicitys sake, we use the
term ephedra to mean the herb or herb
abstract, and ephedrine to mean the
chemical, regardless of source.

I would place synthetic in front of all mention of ephedrine,

or ephedrine alkaloids; for policy experts and others it is
important to make the distinction between herbal and
synthetic. I would use herbal ephedra when possible. I
would also state more directly and more often why the
synthetic ephedrine use is not reviewed as part of the
AERs.
Also, contrary to the phytochemical section of the report,
ephedra is know to contain (-)-norephedrine but not (+)norephedrine. Phenylpropanolamine consists of (-/+)norephedrine, while ephedra does not contain (+)norephedrine. The parenthetical identification of
norephedrine as phenylpropanolamine should therefore be
removed.

A3-3

Change made

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

A minor point in phytochemistry (page 13) is that only (-)norephedrine occurs naturally in ephedra, whereas the
synthetic drug, phelypropanolamine is the racemic mixture
of (+/-) -norephedrine. So, it is more precise to state that
ephedra contains norephedrine, as opposed to containing
PPA.
After reading the RAND report, my first impression is the
following: What are we evaluating ephedrine or herb
ephedra? The latter is not a single-chemical entity and
cannot be assumed to be ephedrine. Even assuming the
herb ephedra in the literature is defined to contain specific
dosage levels of ephedrine, what efforts were made to
ascertain that this ephedrine is indeed ephedrine and not a
mixture of ephedrine-type alkaloids, or, worse, different
types of alkaloids that are also present in ephedra? Any
study or report on a natural product (not just a singlechemical compound) must clearly define what the material
under study or being reported is. I dont see such a
definition in this report. Despite the limited availability of
useful data, this reports conclusions regarding the efficacy
of ephedrine (the single-chemical drug), in the presence and
the absence of caffeine, in short-term weight loss and
athletic performance, appears to be sound.
However, this cannot be said of the herb ephedra that
contains ephedrine but is not equivalent to ephedrine.
Hence, the conclusion regarding ephedras efficacy
Ephedrine, ephedrine + caffeine, and ephedra-containing
dietary supplements + herbs containing caffeine all promote
modest amounts of weight loss over the short term... lacks
supporting data, unless all the limited number of clinical
studies employing ephedra-containing dietary
supplements had clearly defined ephedra, including
amounts of ephedrine and related alkaloids (not just
ephedrine and inert herb carrier).

We agree that the lack of specificity is a

problem. We have modified the
conclusions to be more specific to only
these herbal combinations studied. In the
RCTs of herbal ephedra included in the
efficacy analysis, the dose of ephedrine
alkaloid was stated.

The ODS and AHRQ contracted with RAND (Dr. Paul

Shekelle as Task Director) to conduct a thorough synthesis
of the clinical efficacy and adverse effects of ephedra. It
was clear to me that the objective of this contract was met.
The review was complete and the researchers used the
systematic review/meta-analysis tool to review the
published controlled clinical studies on ephedra-containing
dietary supplements.

No Response

There was a mention of 157 articles that were case reports

of adverse events published in medical journal, however,
they are not included in the case report and there are no
mention of the finding in the Limitation section on page 110.
Would those case reports provide more information than
what are available from FDA? Should a statement be made
on why those published case reports not included in the
analysis (e g potential duplication with FDA time and

These case reports are now included in

this revision.

A3-4

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

resources...etc.)?
The report has been carried out and is free from bias. It is
objective. I cannot comment on some areas of the report
that are not my expertise.

No response

The draft report is incomplete since it does not include a

review of studies of two types: Toxicology in laboratory
animals, and published case reports.

Published case reports are now included.

Toxicology and animal studies were not
included, as this (and most all EPC
reports) focus on clinical studies in
humans.

This draft emphasizes the subjective judgments of the

authors over the objective findings of the clinical studies and
therefore appears from the outset to have a slant against
the safety of ephedra products.

We disagree that the report is slanted

against the safety of ephedra products,
and note the peer review comments we
received with exactly the opposite opinion
(i.e. that we were too conservative in our
conclusions regarding possible adverse
events from ephedra).

Given the observations and comments above, one is left

with the impression that this draft report has a tone or tenor
that leans toward an apparently preconceived conclusion
that ephedra supplements are not safe. The tone is
established in the abstract by reference to the FDA's AERs
"related to herbal ephedra" and "available reports of herbal
ephedra-related death, myocardial infarction (heart attack)
and cerebral vascular accident (stroke)." the abstract goes
on to describe "our causality algorithm and later to use
terms "probably causally related" and "possibly casually
related". Nowhere in the abstract is it suggested that these
purported AERs were looked at objectively and found (to
quote page 112) that "definite causality cannot be
determined from case reports".
The statements in the abstract strike the reader as definite
scientific conclusions rather than subjective observations
that is not consistent with other objective data. Nowhere in
the abstract are the major limitations described, nor is there
any mention that "scientific studies (not additional case
reports) are necessary" (from page 113). On page 5 it is
stated that "Continued analysis of case studies cannot
substitute for a properly designed study to assess
causality", yet this is precisely what this draft report has
done.
Additional statements and references point to a lack of
objectivity and a bent toward sensationalism. For example,
page 5 a comparison is drawn with phenylpropanolamine
and it's "reported association and cerebral hemorrhage"
without nothing that the report at issues is highly
controversial, or that the report found no association
between ephedrine as an over-the-counter drug. The
mention of cerebral hemorrhage at the conclusion of the
abstract is presumably also a result of some unfounded

We have endeavored to keep the

language of the report as factual as
possible. We note that other reviewers
criticized the report for exactly the opposite
reason being too soft and down
playing the risks of ephedra use. We do
not think we can revise the report to
reconcile these two divergent opinions.
With regard to format, this report adheres
to EPC format requirements. With regards
to phenylpropanolamine, we note other
reviewers critiqued us for not making more
of possible similarities. In this case, we
deleted the phenylpropanolamine
sentence.

A3-5

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

conclusion that phenylpropanolamine have been

conclusively tied to cerebral hemorrhage when this is not
the case. In counterpoint to a description of the extent of
present use and to the long history of use in China,
references are made to media attention, lawsuits, a citizens'
petition, and a ban by the National Football League, and a
Canadian Warning. These references are not helpful in a
scientific review that should be evidence based, but instead
give an impression of the slant toward a view that ephedra
products are not safe.
Although the draft report contains much factual information
about both the benefits and risks of ephedra, ephedrine,
and combinations of one or the other with caffeine sources,
certain critical components of a full analysis are missing.
Specifically, there is much proper emphasis on examining
the data for evidence of causality, but little or no attention to
the dose-response relationship within any possible causal
case. This is a critical limitation that prevents the safety
component of the report from being fully useful.

A dose response analysis has been added

to the RCT analysis. We indicate that we
do not feel such an analysis is justified on
the case report data.

The strength of this report is that it is not only

comprehensive, but also objectively performed. Another
strength of this report is defining the areas that need further
research. The limitations are those imposed by the data.

No response

It is clear what was done.

No response

The major strength of study was the statistical approach

utilized for assessment of efficacy and the incidence of
minor adverse effects.

No response

The major limitation was the coupling of conservative

causality assessment criteria with limited medical records
and toxicology data while interpreting the case reports.
While the case reports do not offer mechanism for
assessing the incidence of serious adverse events, they
shouldn't be dismissed completely owing to an overly
conservative set of exclusion criteria. Case control studies
are definitely warranted, but it would be especially tragic if
their outcome, when determined three of four years from
now, confirm what is strongly suspected at the moment.

We acknowledge our criteria are

conservative. We note the great deal of
discussion among peer reviewers
regarding whether a case report analysis
was biased toward or against the safety of
these products.

There are some nomenclature issues in the draft that should We have endeavored to keep the
be corrected or clarified. The term "herbal ephedra"
nomenclature clear.
contains a redundancy, as by definition, all ephedra is
herbal or herbally derived. Also, and this goes beyond
nomenclature to ingredient definition, the term ephedra is
often used in the draft when in fact what is being discussed
is an extract of ephedra with a specified percentage of
ephedrine alkaloids. This sort of misrepresentation of
material identity leads to confusion between ephedra as a
crude botanical, an extract of ephedra (the form most often
used in dietary supplements) with a specified percentage of
constituent ephedrine alkaloids (usually 8%) and the

A3-6

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

ephedrine alkaloids themselves.

Perhaps it could be made clearer that, overall, a very small
number of people have been studied in controlled trials of
any duration. This is an issue as regards to safety, rather
than efficacy where the studies, though small, are quite
consistent.

We emphasize the limited power of the

RCTs to assess safety.

This review reflects my perspective as a neurologist and

stroke researcher. It is a very valuable collection of data
assembled to address clear, relevant clinical questions.

No response

It was clear how the report was developed.

No response

The major strength of this report is its collection of data

systematically on one report for review and assessment.

No response

The major limitation is the way in which the conclusion are

stated and failure to distinguish for the lay reader the
difference in strength of evidence of adverse reports vs.
intervention studies.

We have tried to make this distinction

clearer in this revision.

This well-done report takes a conservative approach without No response, other than causality has
extrapolating the interpretation beyond the available data . It been removed from this revision.
clearly describes the methods used, limitations of the
methodology, and results. The text under Future Research
describing identification of gaps in knowledge is particularly
useful. The presentation of the analysis of adverse events
reports (AERs) might be made clearer by using different
terminology or a narrative explanation of the causality
designations.
We agree this outcome is important.
Quality of Life. As I view the field of obesity, there are two
However, we did not find it reported in the
reasons people want to lose weight. One is for the healthclinical trials we identified.
related benefits. For most physicians, of whom I am one,
this is often the major focus of our support for efforts to lose
weight. However, over the years, I have come to realize
that the major reason people want to lose weight is because
obesity is a "stigmatized" condition. The fact that 75% or so
of the people volunteering for treatment are women, and
that obesity carries such a negative social view stimulates
people, particularly women, to use over-the-counter
medications. Yet there is no mention that I can find of
quality of life in this report.
Body Composition. One of the interesting responses to
treatment with ephedrine and caffeine in the reports of
Astrup and his colleagues is the increase in lean body
mass, or loss of less lean body mass. The implications of
this for use of these medications and in the future research
is not even mentioned that I can find.

This distinction is not one that was

included as an outcome of interest by our
TEP. We agree it is a potential area for
future study.

Performance. There is quite a literature on caffeine and

We were not requested to assess the
performance that certainly plays into the ephedra/caffeine
literature on caffeine and performance.
use by athletes. Yet none of this literature is dealt with here.

A3-7

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

Drop Outs. The issue of drop-outs is considered with the

<20% vs. >20%. From a therapeutic effect, the
"completers" in a trial are much more informative to me than
using the data on those who drop-out in a last observation
carried forward analysis. We are certain that drop-outs are
likely to regain weight - We aren't curing obesity and weight
gain during the adult life is the "expected". Moreover, if we
do not use the LOCF approach, the impact depends
strongly on when people drop out. If they drop-out at month
5 of a 6 month trial it has essentially no effect. If they drop
out in the first month it has a major effect.

We agree that knowing when dropouts

occurred might make it possible to better
understand the results of weight loss trials.
However, when dropouts leave a study is
not routinely reported and hence we did
not have access to these data.

On page 3 there is mention that the studies have

"particularly high attrition rates." What is considered a high
attrition rate? How do these studies compare to other
studies on obesity? There is no explanation as to whether
there is a particular challenge in all obesity studies or
research in general, or whether this attrition rate appears to
be unique to the ephedra studies.

The attrition rate issue is explained in more

detail on page 27, where 20% is identified
as a threshold. A high attrition rate is not
unique to studies of ephedra, but
regardless of study question a high attrition
rate increases the concern regarding bias.

Long-Term Trials. In the Future Research area you call for

"longer" term trials. For all reported drugs the maximal
weight loss is achieved by 6 months. Continuing treatment
usually maintains an effect, but because weight losses of
10% (20 lbs for someone weighing 200 lbs) does not often
get them to a satisfactory weight, people drop-out because
of perceived "failure" of the medication. I thus have limited
enthusiasm for long term studies with agents that don't
produce weight losses of more than 10%. On p. 4 you
indicate that there are "no long term" studies. As noted
above, I think the 6 month studies that reach a plateau tell
us about all we can expect from these trials. Do you
disagree?

We clarified this to indicate both longer

duration of treatment and maintenance of
weight loss.

The report should be reorganized to focus on the

conclusions about the need for further research. The section
on safety should address expected effects at intended
doses and comment on adverse effects of higher doses.
The transient nature of the events observed in the clinical
studies should be discussed. The FDA AER database
unfortunately is not of sufficient quality to comment on either
of these issues related to safety.

We do not know if the events observed in

the clinical studies were all transient and
would not characterize them as so. A dose
analysis is now included in this revision.

There should be another draft report issued to the TEP to

This is not EPC practice, and there is no
ensure that these issues are addressed to the satisfaction of requirement that the TEP be satisfied
the TEP before a report is finalized.
before the report is finalized. We
intentionally recruit TEP members holding
differing views in order to be made aware
of all viewpoints. Trying to get all such
people to be satisfied with the final report
is an impossibility as demonstrated by the
wildly diverging comments we received
from TEP members regarding the causality
analysis.

A3-8

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

Para 1 I would use the word treatment duration, intervention Change Made
length, or other terms designating the duration for which the
participants were on ephedra/ephedrine instead of using
follow up. I would not use follow-up as it denotes a passive
time post-intervention for which participants were followed
to measure outcomes. e.g. 19 were excluded from pooled
analysis because their intervention periods were less than 8
weeks.
The term follow-up can have a number of meanings in the
context of obesity/weight loss trials. In addition to referring
to the duration of treatment with a test agent during which a
research subject is evaluated, it can also refer to patient
evaluation after treatment has been discontinued. From my
reading of this report, you are using follow-up to only refer
to the time during which treatment is administered. It might
be useful to clarify this in the text as 8 weeks of treatment,
etc so as to avoid any confusion in meaning.

Change Made

I was slightly troubled by the exclusion of studies with less

than eight weeks' follow-up. While I agree that studies with
less than eight weeks' follow-up are undesirable, if a large
number of such studies exist, it does seem unfortunate to
exclude them. I would rather have seen them included and
have separate analyses for a very short-term weight loss
and slightly longer term weight loss. I think that the
exclusion of such studies, if there are many, opens up the
report to allegations from companies who have done such
short term studies that their important data were not
included and that the report is biased. I am not stating that I
believe the report is biased, but only by any exclusion of
such studies if there are many, opens up the report to this
allegation. Moreover, while as I said previously, I do not
favor studies less than eight weeks' duration, I still think that
while such studies exist there is something we can learn
from them.

The exclusion of studies less than 8 weeks

duration was made by the TEP and not
something we can change at this stage.
The key question specified a sustained
period of time for efficacy and this was
judged by the TEP to be at least 8 weeks.

Page 2, para 4: It would be useful to give the reason that

the Technical Expert Panel (TEP) gave for suggesting that
follow-up of less than 8 weeks is insufficient to assess
weight loss. It is because the original charge was to assess
long-term weight loss and the TEP thought 8 weeks could
not be considered long-term?

Yes, and furthermore even short term

weight loss would not be useful below 8
weeks. Explanation made in the Methods.

Follow-up of 8 weeks. This term used on p. 2 and then

many other places is confusing. As a clinical investigator,
follow-up usually means the time after treatment is
complete. You appear to be using it only for the treatment
period. It would confuse me less if you said "duration of
treatment".

Change Made

This review excluded 19-controlled trials that assessed

ephedra or ephedrine for weight loss because there was
follow up of less than 8 weeks in each of these This

These studies were included in the safety

assessment.

A3-9

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

exclusion is rational from the perspective of evaluating the

evidence for efficacy. Information obtained from these trials
about short term adverse effects, or the lack thereof, would
be valuable however in the overall evaluation of safety. We
strongly encourage the inclusion of all such data from these
trials.
The statement, In order to improve health outcomes, long
term weight loss is necessary is not accurate. Usually in
pharmacotheraphy for weight loss, long-term means one
year or more. I am not aware of studies that have used time
in place of percent body weight loss as the important
measure. Because your point is that the studies were short
(<=4 months) I would change loss to maintenance because
Yanovski et al. (2002) states that most nonsurgical obesity
treatments lead to weight loss for the first four to six months
followed by regain.
It is not only that the ephedra interventions did not extend
beyond 4 months but also that there was not sufficient
follow-up to determine if individuals were able to maintain
their loss. See review by Yanovski et al., 2002 New Engl
Journal Med.

Change Made

Rewrite last sentence to say In order to improve health

Change Made
outcomes and reduce the risk of morbidities associated with
being overweight, sufficient weight loss (5 to 10% of body
weight) and long term weight maintenance is necessary.
See comments on page 3 and 5 regarding use of term
follow-up; treatment duration, ephedra intervention, etc.

No Response

Small weight losses (5 to 10%) of body weight reduce the

Reference Added
risk of morbidities associated with being overweight (Clinical
guidelines on the identification, evaluation, and treatment of
overweight and obesity in adults. National Heart, Lung, and
Blood Institute (NHLBI), Clinical guidelines for obesity,
1998.)
This first paragraph seems to blend intervention duration
Change Made
and follow-up post intervention. Please rewrite to reflect
data. Longest intervention: 4 months (this is not follow up).
You do not address whether individuals lost a certain
percent of their pre-ephedra weight. This measure is
important when it comes to defining weight loss success.

Percent of weight loss in the treatment

group is now included in this revision.

Maybe the key points of DSHEA needs to be stated in the

overview or somewhere else to emphasize herbal
supplements versus supplements containing synthetic
alkaloids. Maybe place a sentence after the In addition to
the questions related to ephedra-safety. Because
synthetic ephedrine alkaloids.

We revised the text to try and improving

clarity.

On page 3 the report mentions that an algorithm for

assessing causality was developed by the authors. Was the

This algorithm was deleted from this

revision

A3-10

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

algorithm unique to this study, or is there already significant

scientific agreement to its accuracy and validity? If it is a
new algorithm, who suggested its use? How was
"reasonable certainty" determined?
The Draft identifies question that guided this Report, both in
relation to weight loss and energy enhancement, as "Does
ephedra have additive effects with other agents?" Specific
emphasis was placed on caffeine and caffeine-containing
botanicals, but in Table 1 herbal "agents" were listed as
"Herbs commonly combined with ephedra," presumably
(though not stated) in products marketed for weight loss.

Change Made

It is stated that the majority of ephedrine (up to 97%) is

excreted unchanged by the urine. The 97% seems too
high. The recent paper by Christine Haller et al (Clin
Pharmacol Therap 2002;71:421-32) indicates that about
60% of ephedrine is excreted unchanged in the urine. This
is important because the other 40% can be metabolized to
other pharmacologically active alkaloids.

Change made

In places, particularly the introduction , the report focuses

We present the results stratified by agent.
more on ephedrine than ephedra. Since there were only 5
The efficacy results for ephedrine &
trials assessing ephedra for weight loss (actually 4, since
ephedra were similar.
one is reported twice) and many more synthetic ephedrine,
the ephedrine trials would seem to have greater weight than
the ephedra trials. Not clear how this influences the results.
There are several problems with Table 1. No references
We greatly shortened this Table to include
were given to inform as to how the herbs included in this
just the caffeine-containing herbs, as
Table were identified as "commonly combined with ephedra" suggested by this reviewer.
and in fact it is our belief that several of the listed herbs are
either uncommonly found in products containing ephedra
and marketed for weight loss or are not found in the market.
For example, although the aloe resin is known to be a
cathartic laxative, we are not aware that it exists as an
ingredient in any ephedra product (or in any dietary
supplement product), and if it does it is certainly not
common. Without attempting to be exhaustive, the same is
true for at least the following: cocoas, coffee, scotch broom,
jalap bark, and mayapple root.
In addition, several of the ingredients are at best
questionable for the described categories, and follow-up
should be undertaken to find references to support that
yellow dock root is a cathartic laxatives. These examples
are again not exhaustive; references should be given to
support each herb in its classification.
An additional oversight is that some listings do provide the
part of the plant that is purportedly a commonly combined
with ephedra, though very nearly 50% do not. The federal
law requires that botanical ingredients in dietary
supplements identify the plant part and this Table should do
the same.

A3-11

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

Finally, the Table does not appear to provide any

information that is useful toward answering any of the
questions proposed by the funding agencies or those that
guided the Report. While the question of the additive effect
of other agents was proposed and reportedly guided the
report, there is no attempt in the Report to actually do this,
except in the case of caffeine containing herbs.
In summary, it might be best to eliminate the Table to
reduce it to consist of just the caffeine containing herbs. If
the table is maintained, some effort should be made to
actually find each of the listed ingredients in one or more
products in the market. This is especially true for hers with
significant toxicity potential, such as Scotch broom to
mayapple as the final report should not communicate that
these ingredients are "commonly" sold. Preferably, such
market information would be provided in the form of
references. The part of the plant that is used should be
including for any plant listed in this Table. References
should be provided as to how classifications are made if the
categories in the Table are maintained.
Notwithstanding the above comments the question whether
all of these herbs should be included in the Table, there are
several spelling errors in the botanical names: Coffea is
correct, as in the 1st such listing but Caffea is not; Camellia
is correct as in the 2nd such listing, but Camilla is not; the
correct spelling of the species name for Mate is
paraguariensis; the references species of mayapple is P.
peltatum while Rheum palatum us correctly recorded as
rhubarb, R tanguticum (misspelled in the Table) is
considered to be a variety of R. palmatum (so R. palmatum
var.tanguticum) and R. officinale is misspelled in the Table;
the correct spelling of the botanical name for flax ends in
"m" rather than "n" (so Linum usitatissium);Irish moss is in
the genus Chondrus, not Chrondrus; contemporary
authorities accept the name of the slippery elm to be Ulmus
rubra rather than Ulmus fulva, these corrections may not be
exhaustive.

A3-12

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

We have included some, but not all, of this

The characterization of DSHEA in the Background section
additional material when describing the
of Chapter 1 is inaccurate, biased, unnecessary and badly
DSHEA.
written!! It should either be removed -It has nothing to do
with the assignment-or expanded, to include other elements
of the law. For example: The DSHEA was passed
unanimously in 1994 based in part of Congressional
displeasure with the federal governments 'adhoc, patchwork
regulatory policy on dietary supplements.' Under these
regulations, herbal dietary supplements are not necessarily
required to be tested for safety prior to marketing, although
marketers are required to assure all of their products are
free of significant or unreasonable risks. Also, as with overthe-counter drugs, there is not a requirement to report
health problems that resulted from their use. The federal
regulations that govern this class of goods are different from
this that control either foods or drugs, but as with both of
these classes, FDA and FTC maintain significant authority
to regulate the manufacture, labeling and claims for dietary
supplements and to remove unsafe products."

A3-13

Appendix 3. Reviewer Comments (continued)

Reviewer Comment
The Dietary Supplement Health and Education Act of 1994
(DSHEA) The brief mention of Public Law 103-417 is
inadequate. In 1994, Congress passed the Dietary
Supplement Health and Education Act (DSHEA) amending
the Federal Food, Drug, and Cosmetic Act. In DSHEA, the
term "dietary supplement" is defined as: 1. A product other
than tobacco intended to supplement the diet that bears or
contains one or more of the following dietary ingredients: * a
vitamin; * a mineral; * an herb or other botanical; * an amino
acid; * a dietary substance for use by man to supplement
the diet by increasing the total dietary intake; or * a
concentrate, metabolite, constituent, extract, or combination
of the above listed dietary ingredients. 2. A product that is
intended for ingestion is not represented as food or as a
sole item of a meal or diet, and is labeled as a dietary
supplement. 3. It includes an article that is approved as a
new drug, or licensed as a biologic, and was, prior to such
approval, certification, or license, marketed as a dietary
supplement or as a food unless the Secretary has issued a
regulation, after notice and comment, finding that the article,
when used as or in a dietary supplement under the
conditions of use and dosages set forth in the labeling for
such dietary supplement, is unlawful. 4. It excludes articles
that are approved as a new drug, certified as an antibiotic,
or licensed as a biologic, or an article authorized for
investigation as a new drug, antibiotic, or biological for
which substantial clinical investigations have been instituted
and for which the existence of such investigations has been
made public, which was not before such approval,
certification, licensing, or authorization marketed as a
dietary supplement or as a food, unless the Secretary, in the
Secretary's discretion, has issued a regulation, after notice
and comment, finding that the article would be lawfully
marketed as a dietary supplement. 5. It deems a dietary
supplement to be a food. 6. It excludes a dietary
supplement from the definition of the term "food additive."
Important safety measures were included in DSHEA. A food
could be deemed to be adulterated if it was a dietary
supplement or contained a dietary ingredient that: 1.
presents a significant or unreasonable risk of injury; 2. is a
new dietary ingredient for which there is inadequate
information to provide assurance that such ingredient does
not present such risk; 3. poses an imminent hazard to public
health or safety; or 4. contains an ingredient that renders it
adulterated. Important clarifications were included in the law
regarding labels and labeling. Section 5 of DSHEA provides
that a publication shall not be defined as labeling when used
in connection with the sale of dietary supplements when it:
1. is not false or misleading; 2. does not promote a
particular manufacturer or brand of supplement; 3. is
displayed so as to present a balanced view of the available
scientific information; 4. is displayed physically separate
from such supplements; and 5. does not have appended to
it any information by sticker or other method. 6. places
the
A3-14
burden of proof on the United States in establishing that
such matter is false or misleading. Additionally DSHEA: 1.
Set forth conditions under which nutritional claims may be
made with respect to such supplements. 2. Deemed a
dietary supplement misbranded unless its labeling meets
specified guidelines. 3. Deemed a dietary supplement which

Rand Response

Appendix 3. Reviewer Comments (continued)

Reviewer Comment
The draft sites in it's Chapter 1 the findings of a 1996
meeting of the FDA's Food Advisory Committee (FAC),
stating that "over half of the members recommended
removal of dietary supplements containing ephedra on the
market" and gives as it's reference Dr. Lori Love's testimony
in August 2000 at another meeting. To assure that the
findings of this meeting are most accurately reported it
would be best to add a statement such as "a finding that
was in direct contravention to the recommendation of the
Special Working Group of experts that had been
empanelled to offer guidance to the FAC.
"The transcript of the 1995 meeting of this Special Working
Group can be seen at http://www.cfsan.fda.gov/~dms/dsephe1.html . The more important factor with regard to this
statement, however, is that it is false. The transcript of this
meeting is available on the FDA's website in two PDF files
(see http://www.fda.gov/ohrms/dockets/ac/cfsan96.htm).
Regardless of how Dr. Love characterized the
recommendations of the FAC members, the record shows
that only 4 of the eleven voting members of the FAC stated
that ephedra products should be removed; even when
calculating the opinions of all the meeting's participants, well
under half made statements to that effect.
The statement in the Draft could be corrected either by
changing "over half of" to "a minority of" or by reversing the
two sentences (At this time over half of the members
recommended that the FDA develop rules on use that would
help reduce risk over adverse events, a recommendation
that trade groups had made two years earlier.
Finally, the use of the word "Thus", at the beginning of the
next sentence in this section implies a direct relationship
between the reported advice of the FAC and FDA's imposed
rule. This is a reinvention of the historical facts. FDA stated
in its proposal was based on information that included, but
was not limited to the opinions of the FAC. More detail
should be added to this section if the report is to be an
accurate record of facts.
If the only limitation accessible about the history of the
controversy regarding the use of ephedra in dietary
supplements was from the Background in the Draft's
Chapter 1, one would conclude that federal health officials,
consumer groups and National Football league had been
actively attentive to this issue while industry stood by. This
is not the case. The Background information should be
expanded to include some or all of the facts: that AHPA
adopted labeling guidelines in 1994 that were substantially
familiar to those later proposed by FDA; AHPA adopted
dosage limits (25mg/servind; 100mg/day of ephedra
alkaloids) in 1995; AHPA and others specifically requested
in public hearings in 1995 and 1996, and in a meeting with
FDA in 1999 that the industry policies be adopted by
rulemaking; AHPA and others submitted a Citizen petition in
October 2000 (prior to the Public Citizen petition identified in
A3-15
the background) to make the same request in a more formal
manner.

Rand Response

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

Many scientists would disagree with the statement [page 4,

and elsewhere] that definite causality cannot be determined
for case reports when the adverse event is very serious [or
various iterations of this statement]

We note this comment, and also note that

many scientists would agree with it. At any
rate, we have deleted from this revision the
causality assessment.

p.9 Background states Three billion servings of ephedra

containing products were consumed during 1999 This is a
misstatement , as in the transcript Mr. McGuffin indicates
servings sold rather than servings consumed. As a
separate comment, it is unclear as to whether the data on
the number of servings actually represents servings
manufactured by a particular company or some other
measure.

We have revised this statement to make

clear this is the industrys contention.

D. Finally, the report should not repeat the industry

assertion that three billion servings of ephedra were
consumed in 1999, unless this is based on hard facts. It's a
self-serving statement that has the effect of diminishing the
safety concerns over ephedra by perhaps inflating the
frequency of exposure. Is the three billion estimate based
on quantities sold? Surely not all dosages were consumed.

Change made to reflect this is an industry

assumption.

p.10 FDA concerns about the safety of ephedrine alkaloid

containing products sold as supplements preceded the
passage of DSHEA, which changed how FDA could deal
with safety in the context of supplements.

We revised the text to reflect this.

Change made
Two references in the background section should, in my
opinion, be changed. On page 11, you state that "weight
loss has been associated with decreased morbidity and
mortality" and cite ref. 26, the Williamson et al study.
Actually, the literature on this point is quite controversial,
and despite the Williamson study, much of the literature
shows an increase in mortality with weight loss. All of these
studies are observational, and subject to serious limitation.
This is why NIDDK is undertaking a very large study (Look
AHEAD) to answer questions about morbidity/mortality with
voluntary weight loss. The DPP does suggest that
intentional weight loss in persons at risk can delay or
prevent the onset of type 2 diabetes in persons at high risk.
I suggest that you state instead that "intentional weight loss
in obese persons leads to reductions in risk factors for
disease" and cite the NIH guidelines: Clinical Guidelines on
the Identification, Evaluation, and Treatment of Overweight
and Obesity in Adults--The Evidence Report. National
Institutes of Health. Obes Res 1998; 6 Suppl 2:51S-209S.
Also, on page 14, ref. 69--when discussing the role of
ephedrine in humans, its role in stimulation of beta three
adrenergic receptors in brown fat is noted. There is very
little brown fat in adult humans, and I'm unsure that this
would play any role in ephedrine's thermogenic effect. The
reference cited is an old one (1982). Someone should be
sure that this citation represents current thinking on the role

A3-16

We deleted this comment.

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

(if any) of brown fat in the thermogenic effects of ephedra

compounds.
The definition of overweight is >=25-29.9 (not in excess of
25 but also inclusion of 25) and the definition of obesity is
>=30 (not greater than 30 but inclusion of 30) See NHLBI,
Clinical guidelines for obesity, 1998

Change made

The attempted intentional weight loss data is only for 1996. Change made
The 1998 data you reference is a paper that only includes a
subset, only 5 states. The latest national data on attempts
for weight control is the 2000 data that is in Reference 10.
Therefore, you may want to delete reference to the 1996
data and instead use the 2000 or just edit the sentences to
say The same survey when administered in 2000 showed
that one third (38.5%) of subjects were actively trying to lose
weight and another third (35.9%) were trying to maintain
their weight.ref 10 Furthermore, among those who were
overweight 45.0% of subjects were actively trying to lose
weight and 34.9% were trying to maintain their weight.
Among those who were obese, 65.7% of subjects were
actively trying to lose weight and 20.8% were trying to
maintain their weight ref 10.
I then go on to reference 29 data. The would suggest using
the estimates from Reference 29 to determine a
denominator for use. I would suggest also using the
Michigan data from this paper to support claims that
consumers are not aware of the ingredients in their herbal
supplements.
Ref 29 In a population-based study of 14,679 U.S. adults
in 5-states using the 1998 BRFSS data, 7% reported using
nonprescription weight loss products; 2% reported using
PPA and 1% reported using ephedra products from 1996 to
1998. More women used ephedra products than men; 1.6%
of women and 0.4% of men reported using weight loss
products containing ephedra. Extrapolated nationally, this
study estimated that during 1996-1998, 2.5 million
Americans used weight loss products containing ephedra.
This study also has data to suggest that many individuals
are not aware they are taking weight loss products that
contain ephedra. Of the 183 respondents in Michigan who
responded no to the questions about using ephedra and
reported to have taken other nonprescription weight loss
products, 33% reported using name-brand products that
claim to contain both ephedra products and chromium
picolinate.
I would rewrite the sentence regarding Harnack et al. (2001) This change was made consistent with
to be the following (inclusion of small n and previous of
previous reviewers comments.
ephedra specific for weight loss). It is hard to follow what the
12% of the total is when you dont give the original total
usage (61.2%); and, it is really 5.3% that used ephedra for
weight loss This is larger than Ref 29 but Ref 29 has

A3-17

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

14679 individuals whereas the Harnack study has only 376.

Among 230 (61.2%) of 376 adults in the St. Paul/MSP area
who reported using an herbal products during the past 12
months, 44 (19.1%) used ephedra. Of these 44, 20 (45%)
used ephedra for weight loss. Therefore, 5.3% of adults (20
of 376) reported using ephedra for weight loss. --Taking
these estimates, you find that 20 (5.3%) of 376 individuals
used ephedra for weight loss during the past 12 months
(1998/1999).
Some of the herbs mentioned in the last sentence are not
listed in Table 1. Many of the latest formulations also
contain bitter orange. I would add these to the Table.

This Table has been greatly shortened and

this part has been deleted.

On page 9, the following key information is provided:

Ephedra has been used for over 5,000 years. Three billion
doses have been sold. Even after the FDA's campaign to
advertise the AERs and to have more AERs reported, there
has been a 65% increase in volume of sales over the
previous five years. Even after the FDA's campaign, there
are only 1,500 AERs out of 3 billion servings. That
calculates to about 1 adverse event in every 2 million
servings. By anyone's standards that is very safe.

This is a judgment and not a statement of

evidence, which is what the Evidence
Report presents.

On page 10, the statement, "Still, the controversy over

ephedra continues," and a reference to litigation have no
place in a scientific analysis. It is doubtful that such
information was garnered from a review of the published
scientific literature. Inclusion of this type of information
takes away from the science.

We disagree that these sentences take

away from the science, we think they are
necessary to put the science in context.

Information from the scientific literature on ephedrine (the

purified alkaloid) regarding it mechanism of action. There is
a fair amount of literature (1910 to 1930) about ephedrine.
For example, Chen KK, Schmidt CF. Ephedrine and Related
Substances, Medicine volume 9, number 1, 1930.

This section of the report was not intended

to be exhaustive, but to provide context for
the reader. Many relevant references may
not be included.

Page 10 paragraph 1: Obesity, The definition of obesity had

changed since 1991. A result of this change was that in the
mid-1990's many more people were considered obese than
previously. So, although the incidence of obesity had been
increasing since 1991, the change in definition makes it
seem more dramatic than it actually was. As a result, this
statement may need to be qualified.

This is probably true, but by most

standards the incidence is increasing. At
any rate we did qualify the statement.

Page 12, paragraph 2: It is appropriate to extrapolate

figures from 511 subjects attending a gymnasium to the
general public? Suggest qualifying this statement.

We indicated this is the authors

extrapolation.

The RAND Corporation has drafted a document entitled

"Ephedra: Clinical Efficacy and Side Effects" in order to
assess the efficacy of herbal and synthetic ephedrine on
weight loss and athletic performance and to assess the
safety of herbal ephedrine products through review of
adverse events reported in clinical trials and in reports on

No specific response to these general

comments. Specific responses to specific
comments below.

A3-18

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

file with the U.S. Food and Drug Administration (FDA). This
report will focus on the safety assessment in the RAND
report. Prior to commenting on this assessment, however, it
is important to note that the RAND report includes a formal
meta-analysis that concludes that products containing
herbal ephedra and caffeine produce significant weight loss
over a 4 to 6 month period. This weight loss is similar to
that documented from synthetic ephedrine plus caffeine.
Given the epidemic of obesity in the United States and the
associated morbidity and mortality from obesity, it must be
emphasized that weight loss may play a large role in
reducing morbidity and mortality.
In fact, several studies have shown that weight loss
associated with herbal ephedra and synthetic ephedrine are
associated with significant reductions in parameters
associated with cardiovascular disease among the obese
(e.g., reductions in triglycerides, 1, ApoB, 1 and LDLcholesterol;2 and increases in HDL-cholesterol2).The
evidence for weight loss is quite robust because it is derived
from controlled randomized trials. The best data for safety
would also be derived from randomized trials. RAND
reports on adverse events within randomized trials, noting
that there were "no serious adverse events (e.g., death,
myocardial infarction, stroke) reported in these clinical
trials." Because of the limited numbers of subjects studied
in these trials, these studies could only detect a serious
adverse event rate of one in a thousand. That is the studies
can exclude a rate of serious adverse events of greater than
one in a thousand. This should not be inferred to mean that
the rate is one in a thousand nor that ephedra even causes
adverse events.
In the absence of additional controlled studies, RAND then
turned to adverse event reports (AERs) filed with the FDA
up to September 30, 2001. The limitations of AERs in
proving causality, especially when viewed in isolation of the
totality of evidence, are well known and have been
discussed extensively in the literature and basic textbooks
of pharmacoepidemiology. Causality cannot be proven by
AERs because there is no comparison control group.
Authors of other reviews of AERs in the ephedra database
have noted that a collection of AERs "does not prove
causation, nor does it provide quantitative information with
regard to risk."3 There are several reasons for this which
will be discussed briefly.
To what end was this inquiry directed? Most exogenously
ingested chemicals that are thought to enhance athletic
performance are banned from competitive sports. To what
end will the results of such an inquiry be applied?

This report was commissioned by AHRQ &

ODS to assess the state of the science
regarding ephedra.

Given the most contradictory recommendations of the

We believe that every person on the TEP
CANTOX report regarding the safety of the ephedra
has a bias. Our goal in selecting the TEP
supplements why was a member of CANTOX included as a was to try and get a balanced set of

A3-19

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

part of the Technical Expert Panel? Some bias may have

been imparted from a member of an organization with such
close ties with the ephedra industry.

biases. We judged it important to include in

the TEP a member with close ties to the
industry who was also scientifically
credible.

Of note, no neurologists (and particularly no stroke experts)

were included on the Technical Expert Panel.

While true, a neurologist was included in

the group assessing the case reports and
a neurologist was included in the peer
reviewers.

The group was charged with assembling and evaluating the

evidence that ephedra and its congeners favorably affected
energy enhancement, affected weight loss and improved
athletic performance. Energy enhancement is a vague
term and it is not clear how it can be measured or tested.

Our TEP defined this for us as indicated in

Table 3.

A basic problem in the method adopted for the pooling of

studies rests with the false assumption that these herbal
preparations have been standardized and are similar
enough in constituents, potency and purity that they can be
assumed to have sufficient homogeneity to justify pooling of
results. In fact, there is much evidence that this is not the
case. Yet much of the report rests on the results of the
pooling of many under-powered studies of herbs or ephedra
where potency and constituents are vaguely described or
even unknown. The herbs are mixtures of many chemicals
with various actions so it is doubtful merging or pooling such
studies represents a scientifically legitimate exercise.

We disagree with this opinion, and point

out the chi-square test of heterogeneity did
not reject the null hypothesis of no
difference in the effects reported in the four
ephedra studies.

In sum, this report addresses two questions that are not

relevant to the public health: energy enhancement and
improved athletic performance. The public health question
it does address, obesity or weight loss, is not answered due
to the heterogeneity of the products examined and pooled
and the lack of long-term follow-up studies. The case
reports of adverse events possibly due to ephedra or
ephedrine are not well described and the algorithm adopted
is too rigid and is being applied to a data collection system
that is unable to obtain the data required for causality in the
algorithm.

The key questions were given to us by

Federal Agencies and defined by our TEP.
Causality was removed from this revision.

There is no reason to engage in further research concerning This is an opinion and not a comment
ephedra or ephedrine. Enough is known about its benefits
about evidence to which we can respond.
and risks to remove the drug from the market. More
research is merely a stalling device to delay the removal of
the product from the market.
That ephedra has efficacy for weight loss seems to be true,
though people can quibble over how much. The key
question, as I see it, is: Is there credible evidence that
ephedra poses a significant or unreasonable risk of harm,
even when taken at recommended dosages? This report
so far is inadequate for addressing this question.

No response as there is no specific critique

of methods or analysis.

Page 23, paragraph 3: This paragraph is not clearly written.

Change Made

A3-20

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

What about: To be accepted for pooled analysis, studies

were required to be controlled clinical trials according to the
following definitions [insert definitions].
Page 26, first sentence: Not sure about the implications for
BMI of assuming an average height 5'8". Where did this
number come from? Does it affect the outcome significantly
is this number is off?

We chose this number arbitrarily. The

results do not change across a range of
potential heights.

Because study ref 87 is in adolescents, I question whether

transformation of the data using a height of 5 8 is a good
decision. I would suggest contacting the study authors and
request the actual individual height data.

A sensitivity analysis using 54 made little

difference in the results. Therefore we do
not feel it necessary to contact the original
authors.

Page 27, Paragraph 1: Less than 20% attrition is a

commonly accepted threshold below which concerns about
bias increase due to loss of follow up. Should this read
greater than 20% would be concern for bias.

The reviewer is correct, change made.

Page 28: Meta-Analysis. Will the two Danish trials be

included in the final analysis?

Yes.

Update literature searches past December 2001, if

appropriate.

Done

The questions guiding the evidence report were relevant,

These questions are the same but refer to,
well formulated and easy to understand. The only problem I respectively weight loss, athletic
saw were questions 3, 7 and 12 were the same (Does
performance, and safety.
ephedra have additive effects with other agents?) Most of
the questions were related to the herbal ephedra, but much
of the data reviewed was based on synthetic ephedrine.
Of the 517 articles collected, 56 were controlled clinical
These numbers have been reconciled.
trials of either synthetic ephedrine or herbal ephedra...If the
number of articles are added together
(56+146+84+19+47+4+3+157), there are only 516 articles.
According to page 53 of the Evidence Report, there are 48
controlled trials identified. It is unclear if the two Danish
trials are included or not. Even if it is, there are still
discrepancies with the number.
In reading the objectives, one assume that synthetic
ephedrine was also part of the study objectives. However,
this is not the original intent (see page 19 on Original
Potential Key Questions). Should the changed in objectives
be explained in the evidence report (rather than just a
statement of agreement by TEP on page 20)?

It is explained in the text why this change

was made, and we also changed the title
to reflect this.

It will be more corrected to state that, Forty-eight were

controlled trials assessing ephedra/ephedrine for weight
loss.

Change Made

Most likely the only detailed analysis of all the trials of

ephedrine in the literature. The detailed explanation of the
method with tables and graphs are helpful to the reader.

No Response

A3-21

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

Should the same list of questions applicable to ephedrine as We listed the question as received from
this is listed in the Objectives, or does ephedrine serve only AHRQ, and then describe how these were
as information (if that is the case, the objectives need to be modified for the task order.
reworded)?
Page 20, paragraph 2, last sentence: categories of
patients: children, adolescents, young athletes, and adults.
These are not generally patients, but rather potential
consumers of ephedra or ephedrine products.

Change Made.

Change Made.
Page 22: Additional sources of evidence. Readers may
think it unusual (as did several of our reviewers) that RAND
would place an announcement seeking unpublished studies
in Phytomedicine and Herbalgram. They would wonder why
such announcements were not put into more mainstream
medical journals such as JAMA and Lancet. It might be
useful to mention that the intent in choosing Phytomedicine
and Herbalgram was to reach individuals who might know of
small studies being done on ephedra or ephedrine the TEP
may not have been familiar with.
It is quite evident that a concerted effort was put forth by the
authors to search all relevant databases and literature
sources for clinical studies assessing the efficacy of the
ephedrine/caffeine and ephedra containing dietary
supplements. I was somewhat surprised that
advertisements were placed only in Phytomedicine and
HerbalGram. Phytomedicine is a relatively obscure journal
while HerbalGram is targeted more toward the layperson.
Were other journals considered?
On page 26, the authors state that when a standard
deviation was missing they imputed an average standard
deviation from all other available data. They further state
that they weighted all other standard deviations equally (IF I
understood them correctly). It was unclear to me why they
would weight all the standard deviations equally rather than
weighting them by same size.

We weighted each study equally in the

imputation procedure, i.e. we did not
weight each study by its sample size (we
assume the reviewer meant sample size
not same size). Neither approach
(weighting equally or weighting by sample
size) is entirely consistent with our
assumed random effects model. The
approach we did take is simply applied,
and we have found in practice that the
results are fairly insensitive to weight
choice.

On page 42, the authors indicate that in their reporting form, The form has been correct to read
BMI greater than 27 was defined as obesity. This seems an overweight/obese.
odd choice given that both the NIH and the World Health
Organization have now reached a consensus that a BMI
greater than or equal to 30 should represent obesity and
this information was available prior to the initiation of the
current project.
On page 55 the authors describe some power analyses. It
was not crystal clear to me what null hypothesis was under
consideration in the power analyses the described I think

A3-22

This text was revised to try and increase

clarity.

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

greater clarity in this section could be achieved.

In the weight loss category, of the 24 trials listed in
Evidence Table 2, 4 were excluded, 2 because of study
design, the other 2 because of "insufficient statistics" and
lack of "weight loss outcome" (addressed weight gain).
However, 4, available only as abstracts, were rated using
the Jadad system, scoring 0,1,2, and 2. Of the 20 trials
included in this weight loss panel, only 5 used herbal
ephedra, the other 15 employing the pure alkaloid
ephedrine. Interestingly the highest scores (5) on the Jadad
scale were to the two Boozer studies, which combined
herbal caffeine ( kola and guarana, respectively).

No Response

Regarding the efficacy aspect of ephedra/ ephedrine use,

These were included in the safety analysis.
the rejection of 19 of the identified 48 controlled trials on the The text has been changed to reflect this.
basis of a lack of 2 month follow-up, appears reasonable,
but assessment of the 19 terms of safety indication may add
to the pool of data.
We did include all available trials in the
The literature search seems to be appropriate, with the
safety analysis and have clarified the text
relevant publications being identified. The study selection
to reflect this.
for efficacy analysis seems justified, whereas the selection
of studies for safety is not appropriate. This reviewer finds it
justified including only the controlled trials with a placebo
arm for efficacy analysis. But for safety evaluation is
obvious that all trials should be included. The safety
information collected during a clinical trial has much better
value and validity than the cases received through the FDA.
I suggest therefore that the analysis of safety in terms of
adverse effect dropouts and side effects should be reexamined with inclusion of all the available trials.
Could you estimate the average amount of weight loss per
month in each of the ephedrine groups and in the placebo
weight loss groups. For the lay press and political readers
this may mean more to them than the difference in weight
loss between the active intervention and the placebo.

This revision now contains the percent

weight loss in the treated group.

It is clear there are data gaps with respect to ephedra use

No Response
and effects which are apparent in this study. However, I did
not identify any evidence of bias in the data collection
process.
It appears that the researchers made every effort to reduce No Response
bias in the data collection process. The data collection
process was systematic and thorough. Problems were
identified and explained in the Limitations section of the
report. The researchers did acknowledge that missing
information did exist and also described this in the
Limitations section. It seems as if the researchers did the
best they could have done with the literature captured in the
meta-analysis.
Is there a minimal amount of missing information regarding

A3-23

We conducted sensitivity analysis on

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

outcomes and other variables considered key to the

interpretation of results? The fact that the studies that were
found has 6 month or less treatment duration. That is too
short a period of time to fully analyze safety or efficacy.

attrition rate greater or less than 20%, and

the results are reported in the text. We
acknowledge in the limitations that the
short duration of the identified studies
limits the conclusions that can be drawn.

I believe you did a fine job in synthesizing the data. There

was one inconsistency that I think might have been a
typographical error. In table 15 you state that the pooled
monthly weight loss in pounds is 2.7. In the text on page 55
and 56, you state that the same monthly weight loss is 2.1
pounds. You my want to check this apparent conflict.

This discrepancy has been corrected.

Reasonable decisions were made concerning whether and

how to combine data. Precisions of results were indicated.
Limitations and inconsistencies were also stated.

No Response

All study designs were considered in the synthesis and

reasonable decisions were made as to combining the data.
Precision was reported and limitations described.
Limitations and inconsistencies were stated along with
limitations of the review process. The meta-regression was
used in an attempt to compare treatment across trials.

No Response

At one or more points the authors used the term "cathartic".

I am not certain I know what they mean by that . Do you
mean laxative?

Yes.

I believe that there are at least three major reviews related

In this revision we do not review previous
to this topic to varying degrees that merit mention. I believe reviews, therefore we did not act on this
the authors have mentioned at least two of these three. The comment.
three of these are: the CANTOX Report; Frank Greenway's
recent review, and a review by Allison and colleagues which
appeared in critical reviews in Food Science and Nutrition in
2001. Each of these reports addressed the use of ephedrine
products for weight loss in part or in whole. I do not think
any of them need to be discussed at great length, but it
should be mentioned an the authors of the current report
should briefly mention whether their conclusions largely
agree or do not agree with those prior reports. The authors
are probably also aware that, subsequent to their producing
this draft document, there was a Senate hearings on the
use of dietary supplements for weight loss, at which a
number of experts provided testimony. The written
testimony from several of these experts are available on the
Senate's website. The authors may wish to briefly mention
this in their report and cite any key relevant information that
appeared in that testimony that was not available to them
this report was written.
I think that as it intimated above, any studies excluded must
be carefully accounted for. The exclusion of studies opens
up the reports to potential allegations of bias. Therefore, I
would advocate that the authors include a very detailed
table of all excluded studies giving the reason for their

A3-24

We considered doing so, but felt the report

had so many tables already that this table
was of marginal extra benefit.

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

exclusion and a full reference to the study. If the authors

have already done that and I missed it (the tables were
quite extensive and I confess that I did not go through them
with a fine toothed comb), I apologize.
I was confused on one point. I thought that the authors only
included studies that were randomized, double-blind,
placebo-controlled trials. If I understand the scoring of the
Jadad system correctly, a study that is randomized would
get at least one point, and a study that was placebocontrolled ( and therefore presumably double-blind) would
get a second point. Therefore, all studies should receive a
score of at least 2 on the Jadad scale. However, I thought
that I saw some point that some of the studies received
scores less than 2. Can this be clarified?

Studies were included if they were

controlled clinical trials. Such studies can
score zero on the Jadad scale.

I was somewhat disappointed by the authors discussion and

use of effect sizes. First, the discussion is slightly simplistic
at points. For example, it seems to imply that the particular
effect size metric they use is "the" effect size rather than "a"
specific metric of effect size. Moreover, it is generally wellrecognized that when the outcome measure in a field of
study is something that had intrinsic or accepted meaning it
is perfectly reasonable to use this outcome measure rather
than the particular effect size metric the authors used, which
scales things relative to within group standard deviations.
This is the case with body weight, where most investigators
and people in general understand pounds and kilograms.
There is no reason to standardize by the standard deviation,
which makes the data less interpretable. In fact, several
meta analyses have appeared in the literature on obesity
and simply use pounds or kilograms. I agree that several
meta analyses have as appeared in the obesity field that
have used the standardized effect size that the authors use,
but I personally see it as unnecessary. It is not only
unnecessary but it can create situations in which there is
less clarity.
For example, two studies can achieve the same absolute
weight loss and yet one study because it is much more
tightly controlled may have a smaller standard deviation.
This latter study would achieve a larger effect size and yet I
do not think that most people would see it as more
efficacious if the same number of pound or kilograms were
lost. The authors themselves seem not to accept this metric
of effect size because they later back transform if two
pounds as a way to help the reader interpret the results.
Finally, although a minor point, the authors should use
metric units of weight (kilograms) rather than pounds. The
metric system is the accepted system in current scientific
practice and the authors report will be perceived as less
professional otherwise.

The effect sizes were transformed back to

changes in weight loss in pounds. We also
assessed whether conducting the analysis
entirely in pounds changed the results. It
did not. We also continue to report the
results in pounds because US audiences
are more familiar with this unit of
measurement.

Being not a statistician I do not understand certain parts in

The methods text has been revised to

A3-25

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

the Methodology. For example, effect size is not defined in

the section Weight Loss Effect Size (page 25), the
paragraph in the Safety Assessment, Controlled Trial
Adverse Events, Meta-Analysis section on exact conditional
inference methods versus asymptotic methods (page 29)
was not very clear.

improve clarity.

On p. 2 you describe the "effect size" determination.

Although it becomes clear later that you are comparing
them with placebo, this one sounds like it is only for single
treatments. I am confused.

An effect size is calculated for any

comparison of two groups.

In general I think that this a reasonable objective, and have A dose analysis is included with this
little doubt that this will make a useful contribution to the
revision. We also tried to make sure our
field. That being said, I think there are a number of points
statements were accurate.
that, if carefully addressed, could improve the document. I
detailed specific comments below.
In the main summary (I.e. pages V and VI) the authors
made no mention of dose. I think that this is a marked
oversight. It could inappropriately be taken as an indication
that they statements they make apply to all doses. Clearly
this is not the case as the statements they make can only
apply at best to the doses for which they observed the data.
I believe the authors should consider substantially softening
several of their statements. The first one to catch my eye
was the statement on page VI that "the effects on weight
loss of synthetic ephedrine plus caffeine and Ephedracontaining dietary supplements with herbs containing
caffeine are equivalent..." As I am sure the authors are well
aware, lack of evidence for an effect is not the same as
evidence for lack of an effect. We can never marshal
sufficient evidence to unequivocally prove the null
hypothesis. We can only fail to reject a null hypothesis. If
the authors had access to multiple, very well controlled
studies comparing herbal and non-herbal ephedrine, this
conclusion might be warranted.
However, based on the data they have observed, a far
softer statement such as "We observed no statistically
significant difference between the effects of herbal and nonherbal sources of ephedrine and caffeine" would be much
more appropriate. The authors may perceive me to be a
stickler on this point. I am suggesting that the authors try to
particularly cautious throughout this report in framing their
conclusions because of the highly contentious nature of the
topic they are studying. Even if these authors never enter a
courtroom, it is highly probable that they will "speak" in one
or more courtrooms through this document. That is, lawyers
and expert witnesses representing multiple diverse interests
are likely to cite this document in court cases. For this
reason, it is crucial that the authors say exactly what they
mean and state exactly what can be supported by data and
be vary catious about making statements that could be

A3-26

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

misinterpreted or overextended
On page 11of the report, the authors state some numbers
regarding how many billions of dollars obesity costs.
Although I do not think these numbers are especially
relevant to the report and could easily be eliminated without
any loss, if the authors are going to cite them they should
cite the most accurate information available. My colleagues
and I published a report in the American Journal of Public
Health in 1999 in which we showed that prior estimates of
the costs of obesity were almost certainly inflated by a fact
of approximately 25%. If the authors are going to cite cost
figures they should probably cite our paper and lower costs
showed therein.

We stated that the reported value is only

one estimate. The point, we think, is that
obesity has an enormous cost in terms of
health. We also included this reference
stating that another estimate was 25%
less.

On Page 14 under Pharmacokinetics, I thought the authors We made this modification.

may wish to consider softening their statements about the
lack of difference between herbal and non-herbal sources of
ephedrine in terms of Pharmacokinetics. It seemed to me
that the studies they reviewed did not fully support what
appeared to be their conclusions, namely that there were no
important differences in pharmacokinetics between herbal
and non-herbal ephedrine.
The authors state that two physicians working
independently extracted data in duplicate and resolved
disagreements by consensus. It would be interesting to
know how often such disagreements occurred. That is, can
the authors present any indication of the reliability of their
coding scheme.

We did not assess in this project (or any

similar project) a measure of
disagreements, such as Kappa, and
therefore cannot report this.

With respect to the search strategy, the authors seem to

have been quite thorough. However, there are two sources
they did not mention using that, in my experience can be
extremely useful for this type of work. The first is the United
States Patent and Trade Office which now has all patents
on line. The online data base is searchable. One can often
obtain quite a bit of additional information on this topic by
finding companies' patents. Second, although, in my
experience a less important source, Dissertation Abstracts
International, Which also had on line searchable databases
can occasionally help uncover additional studies. I can
certainly understand the last thing the authors probably wish
to hear at this point is a suggestion they go back and search
for more literature. Whether they ultimately choose to do so
is obviously up to them. However, at minimum they might
want to do a type of "sensitivity search" to see if it seems
likely that they would have missed a great deal of
information by not searching these databases.

We did not go back and search these

databases. No reviewers identified any
missed trials, so while we can never be
sure, we judge it unlikely that there are
significantly large and well done RCTs that
were not included in our analyses.

Regarding herbal ephedra for weight loss: There are

apparently no studies addressing whether weight loss is
maintained after ephedra use is discontinued. This is a very
important gap in our knowledge and should be explicitly

An addition was made to the limitation

section.

A3-27

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

pointed out. With all other weight loss medications, weight is

regained after their use is discontinued, suggesting that lifelong use (whether continuous or intermittent) is likely
needed to maintain weight loss. If this is also true for
ephedra, adverse events must be considered from the
perspective of chronic ephedra use rather than episodic
use. This has important implications for clinical use, public
health, and study design to detect adverse effects.
The HHS requested this analysis to evaluate the safety and The issue of studying select populations
efficacy of ephedra/caffeine products when used for weight was added to the limitations.
loss or exercise enhancement used in the absence of
medical supervision. However, the reports that have been
analyzed have all been performed on subjects that were
screened for pre-existing medical conditions and were
followed during the trials with medical supervision. For
example, in the most recent study by Boozer et al., the
investigators excluded one of every ten subjects they
screened for medical history or for conditions that made
ephedra/caffeine, in their estimation, to be unsafe.
The only trials that could have adequately addressed the
question posed by HHS would be any that enrolled an
unscreened population and followed them with little, if any,
medical supervision. Such studies are not feasible or
ethical because of the general knowledge that ephedrinecontaining products are dangerous. An Institutional Review
Board would not accept this study design. The report
should note that the clinical trials reviewed (at least the ones
with which I am familiar) had strict criteria for medical
exclusion and require careful monitoring for safety during
the study. This is the result of the general understanding of
the medical community that these products are dangerous
and therefore requires medical supervision during their use
These are topic areas that may be worthy
The other major flaw in the analysis is that it failed to
of review but were outside our scope of
adequately consider the pharmacology and clinical
work.
pharmacology of sympathomimetic amines. The
consistency of the evidence across a range of chemically
related substances must be considered. The relative safety
and efficacy of other drugs that have similar pharmacologic
actions is absolutely relevant. Every drug with
sympathomimetic actions that have been studied
adequately has been associated with serious cardiovascular
and neurological adverse events. Likewise, the actions of
drugs that antagonize the effects of ephedrine should be
considered. For example, adrenergic antagonists reduce
the incidence of strokes and heart attacks.
The questions are clearly formulated, but some of the
answers are difficult to find. For example, the answer to
question 7 on page 19 was buried in paragraphs on page
14. I might suggest adding to the summary chapter, brief
answers to the questions you posed that are based on your

A3-28

We reported the questions as we received

them. We tried to reword our conclusions
to better match the questions.

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

analysis. Although the summary does address many of the

questions, it would be nice to see the answers lined up with
the questions. For example, the answer to questions 7 may
be something short like this: " Ephedrine releases
norephedrine from nerve terminals stimulating alpha and
beta adrenergic receptors. Caffeine magnifies this effect by
slowing the breakdown of cyclic-AMP inside the cell through
the inhibition of phosphodiesterase." This is not an attempt
to suggest text, but just to give an idea of how it might be
possible to address the questions you posed in a two or
three sentence answer.
The questions are understandable but not well formulated.
The questions should be specific for the way the
ephedrine/caffeine products are being used. To ask whether
they are safe without specifying how they are used ignores
the potential selective bias.
The selection of 24 hours as a window for exposure to
ephedra is conservative. It is quite possible that ephedra
could cause coronary or cerebral vasospasm that could
persist much longer. This certainly has been described for
other sympathomimetic drugs such as cocaine.

The 24 hour criterion was set by the TEP

and not something we can change.

In evaluating the adverse events, why was documented use This criterion was set by the TEP
of ephedra with 24 hours made a criterion? This timing
interval is far shorter than that used in the PPA
epidemiology study [use ~72 hours]. Furthermore, this
criterion tends to exclude those adverse events that are not
necessarily time or dose dependent or whose effects are
not ascertained until some critical threshold is exceeded
[e.g., immunological reactions; hemorrhagic stroke with
symptoms of an antecedent headache not considered
typical].
Requiring documentation of ephedra exposure within 24
hours of the acute event may be biased against the most
serious cases when a patient cannot provide a history of
recent use because of death, coma, aphasia, or other sever
impairment. In the absence of toxological results a reliance
proxy history by a household or family member should be
adequate.

We did count as satisfying this criterion a

report of the subject consuming ephedra or
ephedrine within 24 hours.

Most likely the only detail analysis of all the trials of

ephedrine in the literature. The detail explanation of the
method with tables and graphs are helpful to the reader.

No response

Is the FDA data on the products that contain ephedra based It could be based on the label or on direct
on label claims that it is ephedra or there are lab analysis
analysis.
confirmation? This should be stated as many of the
products tested claim to contain ephedra only contain
ephedrine and pseudoephedrine and no other ephedra
alkaloids. This is likely due to a non-naturally occurring
source.

A3-29

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

In September 2001, the FDAs Office of Nutritional

Products, Labeling, and Dietary Supplements produced an
excel spreadsheet...for the dates specified What is the
inclusive date of the requested report?

From inception to September 2001.

The literature on herbal supplements/medicines is replete

with reports based on undefined or poorly defined research
materials. This occurs at least one of three levels: (1)
research, (2) reporting research findings in journals, and (3)
abstracting/indexing journal articles for database entry.
Unless serious efforts are made immediately to set criteria
for researchers at all three levels to follow, further research
in the herbal supplements/medicines field will only continue
to generate data that will continue to lead to ambiguous
conclusions and hence, controversy. Ephedra and
ephedrine are no exception.

No response

Ephedra herb (defined as the green herbaceous stem)

sometimes contains up to 30% root material, which has
different types of chemical constituents than those of
ephedra herb. The root has completely different traditional
uses than the stem as well (e.g., antiperspirant vs.
diaphoretic). And the root contains macrocyclic spermine
alkaloids (ephedradines) that are hypotensive, as opposed
to the hypertensive effect of ephedrine in ephedra herb.
Also, ephedra herb from different sources (Ephedra sinica,
E. intermedia, E. equisetina, etc.) contains widely different
levels of ephedrine among the ephedrine alkaloids (30%90%) present in the herb.4 We cant assume the results
from ephedra herb containing ephedrine are equivalent to
those based on the single-chemical drug ephedrine unless
both the following two conditions are met: (1) the efficacy
and safety evaluation is only based on ephedrine and (2)
the concentration of ephedrine in ephedra has been
specifically defined by definitive chemical analyses.
Otherwise this ephedrine could only be 30% ephedrine,
with the rest (70%) being made up of other
phenethylamines (e.g., pseudoephedrine, norephedrine,
etc.) as well as ephedradines (from root material present as
adulterant in the raw material used for extraction); the latter
have different pharmacological activities and toxicities than
ephedrine.

We added to the limitations the lack of

standardized products for ephedra.

Line 3 should read this: "Less than 20 percent attrition is a

commonly accepted threshold above which concerns about
bias increase due to loss to follow up.

Change made.

The cases of seizures (n=70) and fainting/loss of

The seizure cases are now included in this
consciousness (n=63) may represent serious cardiovascular report.
events such as syncope due to cardiac arythmia.
You may want to add to table 1, Bitter orange extract (Citrus This portion of this Table was deleted in
aurantium) and Garcinia Combogia.
this revision.

A3-30

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

In contrast to the Rand draft report, the relation of the

potential of consumption of ephedra to the dosage involved
was the central point of the ephedra risk assessment
contracted by the Council for Responsible Nutrition and
performed by Cantox Health Sciences International,
Mississauga, Ontario
(http://www.crnusa.org/CRNCantoxreportindex.html). The
Cantox report reflects a true risk assessment that includes
(1) evaluation of the evidence for a hazardous effect, (2)
dose response relationship evaluation, (3) uncertainty
assessment, and (4) identification of a dose that does not
carry significant risk under specified conditions of use. The
Rand report includes one important topic not addressed by
Cantoxthe benefits of ephedra.

No response

Given the animal toxicology data that includes wellA dose analysis is now included in this
documented toxicity at high doses, as well as many
revision and this revision no longer assigns
anecdotal cases from the drug and dietary supplement
causality.
literature that point toward ephedrine or ephedrine alkaloid
toxicity, examination of the ephedra adverse event report
(AER) dataset for possibly, likely, or even definite causality
seems to be a moot point unless the dosage that produced
that causal case is identified and put into context with the
recommended dosages. There are abundant examples
among the essential nutrients of the absolute necessity of
applying this principle. For a comparative example, a
conclusion that vitamin A can cause liver damage may be
true but is misleading, and actually harmful, as a generality.
Clearly, scientists should recognize the critical importance
of dose in any evaluation of causality, but not all
policymakers or legislators, much less the general public,
can be expected to do so. Thus, it is critically important to
recognize and evaluate the dosage involved in any possibly
or likely causal cases of adverse effects by ephedra.
The Rand evaluation of risk stops a major and critical step
short of the Cantox risk assessment in that little attention
was paid to the dosage involved in adverse effects that
might be casually related to ephedra ingestion. The
absence of any significant dose-response consideration in
the evaluation and conclusions is very clear in the
Structured Abstract sections Main Results and Conclusions
(page vi). This omission inexplicably occurs even though in
the Methodology section (page 19), the Safety Assessment
list of considerations asks the appropriate dosage question.
This virtual absence of dose-response assessment in the
entire report is reflected in the section on Attribution of
Adverse Events (pages 20-21). In that section, the dose
question is asked mainly in relation to the temporal
relationship, not a dose-response quantitative relationship.
Likewise, in the section on Causality Analysis of Case
Reports, the three key points of the causality algorithm do
not include any evaluation of the dose that produced the

A3-31

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

adverse effect. A complete evaluation requires an answer

to the following question: If the answer is affirmative on all
three key points, what dose was involved? Paracelsus got it
right some 500 years agothe dose makes the poison.
Without consideration of dose, we can justifiably conclude
that anything, indeed everything, is a poison.
The necessity of adequate information to answer the doseresponse question is exemplified by AER 13408, released
by the Food and Drug Administration (FDA). In contrast to
the labeled dosage of up to six capsules per day, the wife of
the 26 year-old male in this case acknowledged to the FDA
investigator that he took a handful at a time, several times
a day. This case is mentioned only to illustrate actual
dosage may bear no resemblance whatever to labeled or
expected dosage. Regardless of oral reports of specific
dosage, the actual dosage should be assumed to be
completely unknown, without confirming pharmacokinetic
information or other objective information.
Pharmacokinetics. There are two published studies of the
pharmacokinetics of ephedra, both from the same
laboratory (Gurley, references 74, 75). Unfortunately, the
results are not consistent, but rather conflicting. It is
strongly recommended that future research include a
carefully designed comparison of the pharmacokinetics of
ephedrine and two ephedra formulations, one comprised of
powdered whole herb and the other powdered extract of the
whole herb, in human volunteers. This should resolve the
issue of a potential difference between purified ephedrine
and the herbal products.

We agree this is an important line of

research, but think this falls somewhat
lower in priority than our first three listed
recommendations.

The report notes the similarity between ephedrine and

phenylpropanolamine (PPA) but fails to consider the
relevance of the data with PPA to ephedrine. The
suggestion that a trial similar to the one with PPA should be
performed ignores the fact that most would consider the
study to be unethical. The only ethical way to do the study
would be to exclude patients at risk for cardiovascular
events but that would make it impossible to accurately
define the safety in an unscreened population of patients.
Again, it would not be ethical to conduct a case control trial
to quantify the magnitude of harm from a drug known to
have the ability to cause strokes and heart attacks.
The only reasonable recommendation from this analysis is
that the drug (ephedrine/caffeine) has modest short-term
efficacy and probable safety when used under medical
supervision. If it is to remain available to the public it should
only be used under medical supervision, i.e. dispensed only
by prescription.

The case control study suggested is an

observational study design that does not
compel subjects to take anything, and in
most situations starts after exposure has
already occurred. We do not think it any
more unethical to conduct this study than
the PPA study.

The review of the safety of a drug with potentially rare

adverse events must include a complete consideration of
the polymorphisms of adrenergic receptors that have been

We judged this beyond the scope of our

report.

A3-32

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

identified that could explain variable response and

idiosyncratic reactions (Am. J. Human Genetics 2002: 70;
935-42). The polymorphisms that result in failure to develop
tolerance are especially important to be considered.
The section on metabolism should include some mention of
the metabolic polymorphisms that result in deficient
metabolism and accumulation of excessive drug levels.
The evidence report questions were easily understandable.

No response

There was a paper published several years ago in the

We could not find this paper so we could
American Statistician, Unfortunately, I do not recall the
not include this.
authors' names. However they presented a particular
method as a way of analyzing MedWatch Report data from
the FDA. In brief, the method entailed creating a
contingency table between types of events on the one hand,
and drugs or substances ingested on the other hand. By
looking for cells with larger than expected frequencies, one
can potentially identify drugs with particular hazards. The
authors might consider adapting this method to their data, or
at least mentioning it.
On page 54, the authors state that "a sensitivity analysis on We do note this prominently in the text but
also note this effect was only observed for
only those studies scoring 3 or greater on the Jadad scale
studies of ephedrine without caffeine.
yielded a pooled estimate of effect size substantially lower
than the main analysisthis differencedid not quite reach
the conventional levels of statistical significance (p=.053)."
In my opinion this is an extremely important finding. The
literature on supplements for weight loss is riddled with a
large number of trials of a very dubious quality. It is often
difficult to know how to interpret such trials. It is easy to
point out the flaws in these trials, but the obvious question is
do these flaws matter? No study is perfect, and defenders of
the claims companies make based upon these flawed trials
are quick to point this out. The finding from the current
authors suggest that such flared trials may be giving
misleading answers. I believe that the authors should much
more carefully describe this result and its implications and
portray it much more prominently in the report.
On page 58 the authors state that there are data from the
pharmaceutical literature that support the contention that
patients taking pharmaceuticals outside of clinical trials may
have a greater risk of certain adverse events than do
patients selected to participate in clinical trials. The authors
should supply one or more references supporting this
statement.

Reference added.

On page 59 the authors state "Thus bias may exist, as the

events we included were different in terms of type vs. those
we had to exclude." It is unclear to me exactly what they
meant by this. I suggest that they describe exactly what
bias they are referring to.

We revised the text to try and clarify this

point.

A3-33

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

Health Canada discourages its citizens from using ephedra

for weight loss. They say they have at least 60 reports of
adverse events. It's not enough to say that you haven't
received them. You must get them and include them in your
analysis.
The US military discourages its people from using ephedra.
A Col Mike Health, identified as an Army pharmacy
consultant, states on the armymedicine.army.mil website:
"There were 25 documented active-duty deaths of soldiers,
sailors, airmen or Marines who had died and were
coincidentally taking ephedra-containing products." You
must get these and include them in your analysis.
The American Association of Poison Control Centers
collects information on human poison exposure cases,
including cases attributed to dietary supplements. In 2000,
2.2 million cases of poisoning were reported to 63 centers.
The Los Angeles Times reported on September 2, 2002,
that the nation's Poison Control Centers collected 9,000
cases of ephedra poisoning since 1993. Where are these?
You must include them in your analysis.
E'Ola, a manufacturer of ephedra products, admitted in a
lawsuit deposition in 1999 that it had received 3,500
complaints about ephedra from its customers that it had not
forwarded to FDA. Where are these? They should be
included in your analysis.
There are at least 25,585 reports of adverse events
associated with ephedra that you have not included

The EPC did request adverse event

reports from most of these sources. We
did not receive any. The EPC does not
have the power to compel organizations to
provide any data. Furthermore, the
adverse events that were assessed leave
us unable to conclude anything about
causation. Therefore, our expectation is
that the inclusion of additional case reports
is unlikely to increase our certainty about a
causal relationship between ephedra use
& serious cardiovascular or neurologic
events.

This report must deal better with the issue of dosages.

Some people dismiss reports of ephedra-induced reactions
as the consequences of over-dosing. Which events among
the likely or possibly associated with ephedra use involved
subjects taking only the recommended dosages?

It is not possible to tell which patients were

taking the recommended doses.

Why were the criteria for high blood pressure set at systolic
BP > 180 or diastolic > 105 mm Hg? More reasonable
measures for serious or clinically significant hypertension
would be to capture all cases of hypertension where
pharmacologically management is indicated [class 2 and 3
hypertension] . This would be consistent with the
definitions of serious adverse event as defined by
MedWatch and in CIOMS [ required intervention to prevent
serious outcome.] Ascertainment of the rate and risk of
clinically significant hypertension would be particularly
critical in any safety assessment of ephedrine alkaloid
containing products for use the general population where a
learned intermediary is not required.

The criteria were set at a level sufficiently

high that treatment would be warranted
that day.

p. 50 Figure 4 brief data collection form for case report:

what criteria were used to establish the categories under
psychiatric [e.g., severe depression, psychosis]

The implicit review of experienced

clinicians.

A3-34

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

On page 59 it is stated that we did not examine the

remaining 251 adverse events because the descriptors in
the master excel spreadsheet were of conditions less
severe.. The descriptors mentioned in the listing were of
the adverse event as reported [usually by a consumer]
rather a diagnosis or precise description of signs and
symptoms. Consequently, this description may be an
unreliable or inadequate characterization of the adverse
event, its severity or seriousness. It may be better to state
that the remaining 251 AERs appeared to fall outside the
focus of serious adverse events [deaths, cardiovascular,
CNS, etc.].

We changed the text to reflect this.

Were other measures of variation included, e.g. confidence

intervals or limits? Could these be used instead of having to
impute standard deviations?

If possible we back calculated the standard

deviation from other information include
din the report. Otherwise, we reported the
standard deviation.

The questions were clearly formed.

No response

The search methods were appropriate and resources were

clearly documented.

No response

Inclusion of Non-Scientific Adverse Event Reports Invalidate Animal and laboratory data were outside
the Integrity of the Study
our scope.
There is a potentially fatal weakness in the report in that
there is no inclusion of a discussion on the peer-reviewed
animal and laboratory research but extensive discussion of
the non-peer-reviewed, non-scientific FDA Adverse Event
Reports (AERs). The inclusion and heavy dependence on
data that the General Accounting Office has already
concluded was flawed is likely to nullify the scientific
integrity of the report. The GAO report stated:
We do not see how this critique of FDA is
While FDA's conclusions regarding the desirability of the
proposed action may be valid, we believe these conclusions applicable to our report.
are open to question because of limitations and
uncertainties associated with the agency's scientific and
economic analyses. The GAO found that the AERs were
poorly documented; that the FDA did not perform a causal
analysis to determine if, in fact, the adverse events reported
in the 13 AERs it used to set dosing levels were caused by
supplements containing ephedrine alkaloids; and that the
FDA indicated in its proposed rule that 10 to 73 percent of
reported adverse events might not be related to
consumption of dietary supplements containing ephedrine
alkaloids.
Have AERs ever been included in an AHRQ or RAND
Evidence-based Center review before? An important
hallmark of the evidence-based review or meta-analysis is
the establishment of strict criteria prior to the review and an
adherence to the established criteria once the review
begins. Any deviation from criteria once the study begins
may result in a flawed analysis and a loss of credibility

A3-35

Case reports have certainly been included

as a course of evidence in other AHRQ
evidence reports, for example our own
report on a Best Case Series for CAM
Treatments of Cancer.

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

The inclusion of AERs as part of this review appears to be a

serious deviation both from what AHRQ requested and from
the standard criteria used in conducting a meta-analysis.
On page 4 of the draft, the authors conclude, "the majority
of FDA case reports are insufficiently documented to make
an informed judgment about the relationship between the
use of ephedra-containing supplements and the adverse
event in question." Devoting approximately 50 pages to
AER reports in the report seems incongruent with the space
devoted to descriptions of the peer-reviewed scientific data.
There is also no explanation in the AER evaluation of
products that were found to be illegally marketed as dietary
supplements, which in fact were misbranded. Some of the
early and most serious adverse events were from products
that were adulterated with high doses of synthetic ephedra.

In fact, as our analysis shows, there were

more deaths as a percentage of total
AERs reported in the more recent data
compared to the older data.

Several of the preliminary questions provided to RAND have A dosage analysis is included in this
not been addressed in the report. We expected a review of
revision.
the literature to be included in the report: Questions about
Dosage: What dosage of ephedra produce risk of CVD or
other life threatening events? This may be because there is
little or no data available. If so, it should be made clear in
the report. The CANTOX report drew conclusions about a
safe upper limit. While these were based on the results of a
single study, they were somewhat corroborated by others.
This is not to say that the CANTOX report is definitive.
Also not addressed: Do ephedra-containing dietary
supplement products alter physiologic markers of
cardiovascular function?

This was addressed to the extent that RCT

data in humans was identified. Blood
pressure and ventricular tachycardia were
two physiologic measures of cardiac
function included in the analysis.

Adding AER analyses of ephedra AERs in the published

literature, ephedrine AERs from the FDA's Adverse Event
Database, and those for seizure and would make the report
more complete and well balanced.

These have been included in this revision.

While it's useful to analyze the controlled trials for evidence

of adverse effects, we're not likely to find significant effects
in them because if adverse events were that common the
studies wouldn't have been permitted in the first place. We
must rely instead on case reports and adverse event reports
for evidence. Therefore, every effort should be made to
assemble all the credible case reports and adverse event
reports associated with ephedra use. That was not done.

We disagree strongly with the contention

that we did not expend every effort to
obtain case reports. We have extensive
documentation of our efforts to identify and
obtain case reports for this analysis. Within
the resources available to this project
every possible effort was made.

The report should also make a better attempt at comparing This has been done in this revision.
the commonly reported adverse symptoms with those
symptoms observed upon exposure to ephedra/ephedrine in
controlled experiments. If the symptoms are consistent, or
inconsistent, that's important to know.
The case reports cited are difficult to evaluate as they

A3-36

We were limited by what was available in

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

contain clinical terms incorrectly used, incomplete

descriptions and use an algorithm for causality that is
impractical and unrealistic when using FDA reports. A
vigorous documentation and search for better records at the
time the case-reports were received would have improved
the utility of the case reports.
We regard the handling of adverse consequences as
incomplete and unrealistic. The review by the Clinical
Research and Review staff of the Center for Food Safety
and Applied Nutrition of the Food and Drug Administration
represents a more comprehensive and scientifically valid
approach to reviewing adverse events associated with
ephedra and ephedrine.

the files sent to us. It is not within the EPC

scope to search for better records at the
time the case reports were received. We
disagree that clinical terms are incorrectly
used; in most circumstances we are
reporting the clinical terms used in the
source documents. Finally, if there was
agreement about the best scientifically
valid approach to reviewing adverse
events then there would exist
standardized methods for so doing and we
would not have received the same level of
peer review comments that we did.

Overall Evaluation. (i)The means used to evaluate the

AERs is not clear. It is difficult to determine what role, if
any, the TEP actually played in the review process. From
the description given in the text, it would appear that most
members of the TEP never even saw the AERs. (ii) It is not
clear why an eight-week exclusion criteria was chosen the
review of earlier safety studies. The exclusion of doubleblind placebo control studies of less than 8 weeks duration
resulted in the loss of valuable information about acute
toxicity (and excluded most of the existing data not
demonstrating toxicity). (iii) Important epidemiologic and
scientific data has been omitted. This omission severely
limits the value of this study.

The trials of less than 8 weeks duration

were not excluded from the safety analysis
and epidemiologic studies were outside
our scope of work.

Question Formulation. Questions are well formulated and

easily understood. All of the defects in the study, and there
are many, stem from the methods used to answer the
questions.

No response

Study Identification. Appropriate search criteria were not

used. Not all episodes of ephedra/ephedrine toxicity are a
consequence of chronic exposure. The exclusion of all
studies of less than eight weeks duration may strengthen
conclusions about effectiveness, but it weakens conclusions
about safety. There are, for example, dozens of double
blind placebo control studies where clinically relevant doses
of ephedrine were found to have no effect on blood
pressure or cause arrhythmias, even in asthmatics with
heart disease. There is no reason to exclude such highly
relevant data. Studies where ephedrine was compared to
placebo should not be excluded just because they were not
about weight loss or athletic performance. The scientific
credibility of the report was weakened by the search
strategy that was chosen. Clearly, the authors of the report
assume that (1) all episodes of ephedra/ephedrine toxicity
are a consequence of chronic exposure, and that (2) clinical
trials of ephedrine have been limited to studies assessing
the effect of ephedrine on weight loss. All these
assumptions are easily shown to be incorrect.

There is no assumption that chronic

exposure is necessary and we did not
assume that trials of ephedrine have been
limited to studies of weight loss. We do not
agree that studies of safety in healthy
adults are necessarily relevant to studies
in obese individuals who are at greater risk
for comorbid conditions.

A3-37

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

Data Synthesis. The analysis of the weight loss achieved by We disagree. We tested whether weight
loss was linear over this time period and
ephedrine versus placebo, and ephedrine plus caffeine
we could not prove that it was not.
versus placebo etc., is very problematic because one has
assumed that the weight loss rate is high initially and
subsequently lowers, so that the weight loss from months 3
to 6 is typically very small. It is therefore invalid to simply
calculate the mean rate of weight loss as pounds weight
loss per month when trials of very different duration are
included. Those who are familiar with placebo controlled
weight loss and weight maintenance trials know that most of
the difference between the active and placebo arms is
achieved during the first 3 to 4 months, and that the
difference is subsequently maintained even up to 2 years.
The way the data are handled in this report has therefore
produced projections that severely underestimate the real
efficacy of ephedrine and ephedrine plus caffeine. This has
been carried over into the conclusions, where it is stated
that ephedrine/caffeine is not as effective as other antiobesity medications currently on the market.
Data Synthesis. This must refer to Orlistat (the pancreatic
lipase inhibitor from Roche) and Sibutramine (the centrally
acting compound from Abbott). If one looks at the long-term
of Orlistat ones sees that the mean weight loss difference
between Orlistat and placebo after 6 months to 2 years are
of the order of between 2-5 kg in all the large trials.
Ephedrine plus caffeine produces at least an equivalent
effect. For example: If the weight loss on an active
compound after 3 months is 10 pounds more than on
placebo, and this result is maintained also after 6 months, it
is clear that rate of weight loss would be calculated as 10
pounds divided by 3 (=3.33) if the trial is stopped at 3
months. Whereas the result from a 6 month trial would give
10 pounds divided by 6 months (=1.67), which is exactly half
of the weight loss. This issue should be addressed and the
efficacy section should be revised accordingly. The way the
panel has calculated the weight loss rate actually assumes
that the weight loss rate is linear and that it continues at the
same rate with prolonged use. Obviously, this is not the
case.

We were careful to state in the text that our

results could not be extrapolated beyond
4-6 months. We added data on other
weight loss products for comparison.

Data Synthesis. I note that Astrup et al. International Journal

of Obesity 1992;16:269-77, listed in the bibliography
(accepted articles) as number 1, is not included in the
analysis! The Danish double publication of this is the
Quaade et al., listed as number 48 in the same bibliography.
It is hard to see why the panel quotes the Quaade et al.
publication in Danish, which a condensed version of the
Astrup et al. paper, which I assume must be the paper the
panel had taken the study information from in English. The
panel has used pounds in the analysis of weight loss, but it
would be more appropriate to use weight loss in percent of
initial body weight because the weight loss in pounds in not

We identified these two trials as reporting

identical data, and the inclusion of either
(but not both) should make no difference in
the results. Our practice is to include the
most informative article. For these
reasons we note, a percent weight loss
analysis has several limitations.

A3-38

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

independent of initial body weight. This may introduce a

bias if the initial body weight and body mass index in the 2
arms were not comparable.
Page vi. , paragraph 3; page vi, paragraph 4; page;4,
paragraph 4; and page 30 last paragraph: Some may not
classify anxiety, change in mood as psychiatric symptoms.
Emotional/ mood adverse effects might be more
appropriate.

Psychiatrists may disagree with the

statement that anxiety and change in
mood are not psychiatric symptoms. No
change made.

On page 55, the authors state "the effects of ephedrine and

caffeine appear to be additive. I do not understand the basis
for the authors statement. Unless there is a 2 x2 design in
which to have the opportunity to observe an interaction
between ephedrine and caffeine and observe that no such
interaction occurs, how can the make a statement of
additivity? I believe what they mean to state is that there is
an effect of the combination of ephedrine and caffeine
combined that is greater of the effect of either alone. This is
not that same thing as stating that the effect is additive.

The reviewer is correct. We clarified the

language so that we do not imply the
effects are additive in the arithmetic
sense.

On page VI, the authors state that there are no data from
studies of herbal ephedra-containing dietary supplement
products without caffeine. This is not correct. There is at
least one study. My colleagues and presented an abstract at
the 2002 Experimental Biology meeting from suck a trial.
Unfortunately, we did not present efficacy data. Moreover, I
had thought the community sponsoring the study had
provided the safety data to the NIH for this review.
Although, it is not within my authority to release the data
themselves, I can certainly provide the authors a copy of the
poster presented if they do not have access to it.

Without efficacy data we cannot include

this in the analysis. We did not receive this
study in response to our requests to
industry for unpublished studies.

There is no published study of the efficacy of ephedra

without caffeine, but a large, industry-sponsored study was
done by Coffey et al. at the 2002 Experimental Biology
Meeting the authors reported that there had been no
adverse events, but did not report on efficacy.
I think you did an excellent job in collating the important
available data.

No Response

The Danish study reports a 100% increase on post-exercise

O2 consumption by 100%. Was that immediately after
exercise and for how long, this seems like a very big
increase are you sure this is correct? If you are unclear
have Mary Hardy send me this paper and I will have a look
at it.

On further review, we determined the

Danish study was not relevant to the report
since it did not report differences in
performance between groups.

A3-39

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

As far as athletic performance goes, no studies were

available on herbal ephedra and only a modest affect on
"very short-term immediate performance" was observed with
ephedrine, only when caffeine was co-administered. The
report states that there was one study that assessed the
effect of "sustained use of ephedrine on performance over
time", an " reported that the addition of caffeine to ephedrine
necessary to produce an effect on athletic performance."
But in the structured abstract it is not made clear what the
extent of the effect was. The report text states on pg. 57,
however, that "a study116 published in Denmark concluded
that aerobic training enhanced the effect of ephedrine on
energy expenditure. After, 8 weeks of aerobic conditioning,
ephedrine increased post-exercise energy expenditure by
100%. N. B. The reference numbers in the section are
incorrect e.g. ref 116 cited above should be 43, 115:52
Regarding the 1986 Denmark study do you mean that
ephedra increased energy expenditure during exercise, or
that it increased energy expenditure after the completion of
exercise? Please clarify.

On re-examination we determined this

study should have been excluded, as it did
not measure the effect of ephedrine on
physical activity but rather on basal
metabolic rate.

Were there a disproportionate number of case reports of

adverse events that occurred during or after the
performance of exercise training or physical activity?

Not assessed, and probably not possible

to assess.

The conclusions regarding herbal ephedra for weight loss

are reasonable and defensible as far as they go, but are too
conservative. Saying that there is no evidence for sustained
weight loss with use of these preparations for more than 3-4
months is true, but does not translate into conclusions that
are useful for the clinician or regulatory agencies. The data
presented and summarized support the use of herbal
ephedra or ephedra + caffeine for weight loss.

Our charge was to present the evidence.

Translating the evidence into clinical
recommendations or regulatory decisions
is specifically beyond the scope of the
EPC.

Important parameters were properly identified and

addressed, such as study population and design.

No Response

In general, the appropriate study parameters were

examined. However, I would have liked more data on doseresponse, especially with the efficacy trials

A dose response analysis is included in

this revision

Most of the important parameters were systematically

addressed.

No Response

Should a descriptive statement be made on possible nonstatistical publication bias. For example, funding source for
the published clinical trials may also bias the quality of the
results (see BMJ 2002;325 (August 3):249).

This was added to the limitations.

There should be an expanded general statement regarding This was added to the limitations.
the medical exclusion criteria that are used in all the
published clinical trials. This needs to be emphasized in the
analysis and should specifically point out that many patients

A3-40

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

with underlying diseases (hypertension, heart failure...etc)

were excluded from trials. The emphasis is needed due to
common argument that safety data from clinical trials do not
support the potential serious adverse reactions collected by
FDA.
None of the information provided addressed analytical
We included this information where it was
methods with the ephedrine or herbal ephedra products
available.
used (e.g., certificate of analysis verification that the product
used in the study met label claim) and other issues of
quality.
I think you did a good job in defining the methods you would No Response
use in appraising the studies. Open label treatment
following a controlled clinical trial is often thought to be a
way to screen for safety, but I understand how incorporating
that into your assessment might inject bias due to lack of a
control group. Although caffeine and ephedrine has been
evaluated in a controlled clinical trial for 6 months followed
by an additional 6 months of open label treatment, I
understand the statement that trials do not last more than 6
months refers to the double blind period.
The studies that were obtained for review were evaluated
No Response
carefully. Objective criteria were established prior for
inclusion into the planned meta-analyses. The evaluation of
the case studies provided by FDA was also performed in an
objective manner. The limited number of studies on exercise
and athletic performance are presented objectively. The
limitations of these studies is accurately noted.
It appears that a thorough search for relevant data was
undertaken. I can find no evidence of bias or intentional
inclusion/omission of data or search strategies.

No Response

Criteria for clinical study inclusion and exclusion were well

defined and adhered to. Little bias seemed to be introduced
by the selection process, at least for the clinical studies. All
important relevant studies were evaluated.

No Response

The inclusion and exclusion criteria for the selection of

We contacted Dr. Waddington but did not
articles is adequate. I am unaware that any crucial data is
receive any data from him.
lacking; however, it would be helpful if Dr. Phil Waddingtons
data from Canada could be included in the final report.
Should the search term adverse reaction be part of the
strategy? If not, why not? Please see my comments on
funding source of the clinical trial. In an ideal situation, a trial
funded by a neutral party will probably yield the most unbias information.

Generally, these terms act as limiters and

would exclude studies if they were not
tagged in this fashion. We prefer not to
limit the search in this way. Since tagging
articles is not always accurate and we did
not want to prematurely exclude potentially
relevant studies. We did include adverse
events as text search terms.

There were appropriate inclusion and exclusion criteria for

This revision includes the results of a few

A3-41

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

the studies selected for the meta-analysis and these were

clearly stated and explained. There may have been some
bias based on the studies selected, even though the
inclusion criteria were clearly spelled out. This was
mentioned in the Limitations section of the report. Efforts
were made to identity unpublished studies and several
studies were missed in the evaluation as they had not been
received. The researchers indicated that they will be
considered for future assessment. When this will occur is
unclear.

additional studies that are relevant. We do

not judge that any of the handful of
requested but unretrieved articles are
RCTs.

The discussion of the results of the weight loss trials makes The mean attrition rate was not higher in
no mention of doses. Doses should be mentioned either in the active treatment group. This has been
the text or in the table. It is stated that all of these studies
added to the results
had an attrition rate of greater than 20% Was the attrition
rate higher in the active vs. placebo treatment group? This
should be clarified. If the rates were higher in the active
group, which I suspect, then a specific analysis should be
done as to the cause of attrition.
Athletic Performance, second paragraph. Please clarify the
duration of the exercise test. This is important because
some drugs (like creatine) may produce benefit with short
duration exercise (a few seconds) but not longer duration
exercise.

The exercise tests varied in duration from

short (weight lifting) to an hour or more
(endurance).

The originally proposed key questions included inquiries

regarding dosage levels if ephedra with respect to weight
loss, athletic performance, and safety. The Report does not
address dosage levels with respect to weight loss, athletic
performance, and safety assessment (except for the
mention of possible future research study). Clearly this is an
extremely important concept, and issue, regarding to these
materials, especially with respect to the review of case
reported obtained from the passive AE reporting system. It
is our view that the general omission of dosage
considerations should be mentioned with regard to the AER
case report reviews and that it receive some attention in
weight loss and athletic performance assessments.

A dose analysis was added to the RCT

portion of the report. We did not judge
analysis to be possible in the case report
portion of the report.

Are all of these studies single dose studies? What was the
interval between taking the dietary supplement and the
performance of the exercise test?

This has been clarified in the results.

The meaning of this sentence is obscure. What is meant by This sentence has been reworded.
enhanced mechanisms of heat loss?
See page 26: none of the weight loss studies were beyond
4 months. Therefore, I would rewrite the first paragraph of
the Main Results to say the longest published weight loss
intervention was 4 months. If there was a study with a
post-intervention follow-up then this should be stated as
well. (See comments for page 5)

A3-42

Three studies had 6 months of treatment,

which was too few to perform metaanalysis on this time period specifically.
For studies that reported only 6 month
data, we included these in the 4 month
time point, and specified in the methods
section in the report that this could include
data out to 6 months Hence the 4 - 6

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response
months statements in the text.

On page 3 there is mention that 19 of the 48 controlled trials

were excluded from pooled analysis because they had
follow up of less than eight weeks. There should be some
description of the findings of these studies in this report.

These studies were excluded as evidence

on the advice of our TEP. To discuss them
as evidence of efficacy would be
inappropriate in our view. We did include
them in our safety analysis.

It may not be standard format for the EPC's to cite patient

numbers at this early point in a report, but to make sure
readers understand what a small number of people have
actually been studied in controlled trials, it would be helpful
to include this near the beginning of the report.

This was added to the report in the safety

assessment, where the possibility of a type
II error is increased due to low numbers of
studied patients.

Figure 2, question 18: Chemical analysis of ephedrine

alkaloids was part of the quality review form, but data for
individual studies are not provided. Given the variability of
herbal ephedra, if chemical analysis was not performed in a
particular study, does that call into question the results of
the study?

We do not think so since the results for the

ephedra studies were remarkably
consistent.

In order to improve health outcomes, long-term weight loss No, we meant maintenance of weight loss,
is necessary. Do you really mean that long-term follow-up
since the relationship to health outcomes
would be necessary to determine health outcomes?
is known. We have clarified this.
Page 53: Results Section. Weight loss. It might be helpful to
provide a table of 5 types of comparison studies indicating
sample size in each trial and the power calculations for
each. It is important to highlight when sample sizes are
small and individual power calculations are insufficient.

Considering that we pooled data, we do

not think the addition of our assessment of
power of individual studies is very useful.
We did include this in specific
circumstances where it seemed warranted.

Use of ephedrine, ephedrine + coffee, or dietary

supplements containing Ephedra and herbs with caffeine is
associated with a statistically significant increase in weight
loss (compared to placebo) over relatively short periods of
time (no more than a few months). Please Clarify "(no more
than a few months)". We assume you are not saying that
the data show loss of effect after a few months, just that the
studies don't extend beyond a few months.

The data cannot be extrapolated beyond a

few months. We have clarified this. We
earlier explained the reason for the 4 6
month designation.

It would be helpful to define what is meant by sufficient

evidence.

This is defined as statistically significant.

Several of the preliminary questions provided to RAND have We now include a dose analysis in this
not been addressed in the report. We expected a review of
revision.
the literature to be included in the report: Questions about
Dosage: I. What dosage levels of ephedra are necessary to
achieve weight loss?
When describing the efficacy studies, it might be helpful to
include a table delineating the key elements of the weight
loss studies: Dietary prescriptions, Description of subject
characteristics, Mean weight or BMI at initiation of study,
and inclusion and exclusion criteria for studies.

We considered this change or addition but

decided there were already a great deal of
tables and therefore we did not add this
table.

Although ephedrine is the chemical drug that has been

We are sympathetic to this comment but

A3-43

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

found to be effective in short-term weight loss and athletic

performance, this result cannot be extrapolated to the herb
ephedra that contains multi chemical components. In order
to study the efficacy of herb ephedra, the amount of
ephedrine present in the herb ephedra being tested must be
precisely defined. This ephedrine must be pure ephedrine
and not, say, 50% ephedrine with 25% pseudoephedrine
and 25% norephedrine or other related or unrelated
alkaloids, as are normally present in herb ephedra.
Ephedra is not ephedrine and vice versa, even though
ephedrine is one of ephedras active components. I
personally dont see how one can generate meaningful
results from a study using a material, such as ephedra,
which is not clearly defined. There are just too many
variables. Good science requires a well-defined test
material. For example, we would never accept a singlechemical drug like cortisone with even 25% impurities when
performing a clinical trial on cortisone.
Why should we accept the chemical drug, ephedrine,
present in herb ephedra, whose concentration can vary by
300% (from 30% to 90%, with the balance composed of
other alkaloids)!? Until this problem (which is not
insurmountable, as product definition criteria have been and
can be set)5 is resolved, any studies on herb ephedra for
weight loss or athletic performance (both based on
ephedrine) will not yield meaningful results. It is possible
that some of the papers the RAND report selected do
clearly define the ephedrine content in the herb ephedra
(though I seriously doubt it), the fact still remains that there
are related and unrelated alkaloids also present in addition
to ephedrine.
For example, if a product containing an ephedra extract has
been analyzed to contain specifically 20mg ephedrine per
tablet/capsule to conform to the required amount of
ephedrine for efficacy, what happens to the other alkaloids
also present, which could easily be twice the ephedrine
amount, or 40mg, making the total alkaloids content 60mg?
This is a natural scenario unless made unnatural by
manufacturers or suppliers who take spent ephedra herb
(from which all alkaloids have been extracted) or a token
amount of ephedra herb and add the prescribed amount of
ephedrine, thus rendering the product basically a singlecomponent drug (ephedrine), formulated with inert spent
ephedra or token ephedra herb as carrier/excipient. In this
unnatural case, the ephedra product is basically an
ephedrine drug dosage form and has nothing to do with the
herb ephedra. I doubt there is any published scientific
information on herb ephedra based on sound scientific
definition of the test products containing ephedra.
Furthermore, whatever reports available most likely have
not clearly identified and characterized their test materials

believe that the consistency of our findings

supports the decision to pool studies of
weight loss and compare ephedra to
ephedrine.

A3-44

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

hence rendering their findings of little value to us. Based on

the current state of published information in this field, I dont
believe we will be able to obtain meaningful conclusions
relating to herb ephedras toxicity through adverse events
analysis alone or based on modern published experimental
data. In case of the former, since no precise standards are
required for commercial ephedra products, few if any of the
reported adverse events can be reasonably traced to
ephedra herb. In case of the latter, there are simply too few
useful published reports whose findings are based on work
that used well characterized and well-define ephedra.
Modern medicine and traditional Chinese medicine are two
parallel and distinctly different healthcare systems, each has
its own merits and defects. While modern medicine is
based on scientific experimentation, TCM is based on
empirical practice and trial and error in humans over time.
The latter has accumulated a vast amount of recorded
information, including cautions and contraindications. This
has been an ongoing process and it continues to
accumulate data as TCM practice continues to generate
them. It would be our loss if this valuable resource was not
somehow utilized.
Since herb ephedra has a long use history in traditional
Chinese medicine with an extensively documented record
(safety, cautions, contraindications, etc.) over a 2000-year
period, this should be taken into consideration. Also,
common TCM traditional practice should be heeded. For
example, some of the adverse events reportedly due to herb
ephedra alone may not be so at all, but rather due to the
concurrent and inappropriate use of other common herbs
such as Asian ginseng which is traditionally cautioned
against use in healthy persons with a vigorous (yang)
constitution and which has been known to cause serious
toxicity, including death when used improperly. 6 If one
combines the indiscriminate use of even such common
herbal tonics as Asian ginseng with a relatively potent
herbal drug like ephedra as dietary supplements, to be used
daily with no prominent warnings or precautions, serious
adverse effects are bound to occur. In order to meaningfully
study or evaluate the safety and efficacy of traditional
medicines such as ephedra (not ephedrine, the chemical),
apart from ensuring that the ephedra has been well
characterized and defined, we should also consider taking
its historical record and its traditional use context into
consideration as well as keeping an eye open to the
simultaneous but inappropriate (outside of tradition) use of
tonics such as Asian ginseng. Furthermore, we should
keep an open mind to the possibility that the efficacy and
safety of herbal medicines simply cannot be determined by
Western hard science alone. Common sense and welldocumented historical use and safety data should constitute

A3-45

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

part of the evaluation protocol.

Regarding safety or adverse effects of ephedrine and herb
ephedra, the two drugs need to be evaluated separately.
With ephedrine, there should not be much of a problem
because there must be copious amounts of data on the drug
ephedrine, which can be accessed in various databases.
However, with ephedra herb, it is quite different. Since
ephedra has not entered the market through the usual drugdevelopment-and-approval route, which would have
generated toxicity data during that process, evaluating its
safety as if it were a standard pharmaceutical (the singlechemical drug ephedrine) is not appropriate. There are few
modern scientific or clinical reports published in the field.

For the RCT data, the numbers of patients

studied with ephedra have been too small
to assess adverse events without a high
probability of a type II error. For the case
report analysis, we did separate ephedrine
from ephedra.

You are to be commended for the comprehensive search

you conducted. Your methods seemed well-defined and
unbiased.

No response

I think you did an excellent job of selecting articles using

specific criteria and limiting bias. I know that one of the
major issues prompting this review was concern regarding
the safety profile of caffeine and ephedrine. In the United
States this combination is sold in an unregulated fashion, so
the only estimate of the denominator for adverse events is
the number of doses manufactured. In Denmark, caffeine
and ephedrine is a prescription preparation for the treatment
of obesity. Orlistat and, before 1997, dexfenfluramine were
approved prescription drugs in Denmark competing with
caffeine and ephedrine. I assume that sibutramine is also
approved in that country, but I do not know that for sure. It
may be too late to at this point to include in this report, but
information must exist for the incidence of reported adverse
events to obesity drugs in Denmark.
Although this is not a perfect way to assess safety, it might
be useful to determine the relative incidence of serious
adverse events reported with various prescription obesity
drugs in Denmark. Based on conversations with individuals
familiar with the Danish experience, I suspect that the safety
of caffeine and ephedrine would compare favorably with
sibutramine. One advantage of such an analysis is that one
would be comparing alternative drugs for treatment of
obesity in the same population. The second advantage
would be a better estimate of the denominator based on
prescriptions written rater than manufactured pills.

It was beyond our resources to obtain

safety data (other than published data)
from Denmark. We added to the future
research that this would be a good study to
undertake.

A3-46

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

Denmark Experience. Ephedrine/caffeine combinations are

used extensively in Denmark for weight control purposes.
There is a long history and experience that should be
considered by RAND. Dr. Astrup has indicated that there
are very few adverse event reports associated with such
products in Denmark. RAND should contact the Denmark
health authority, and/or Dr. Astrup, in order to obtain more
information regarding these reports - and should include this
information and Denmark experience in the final report.
With regard to the adverse event reports, significant
amounts of information were often missing; however that
was not the fault of the authors, but rather a shortcoming of
the MEDWATCH program. If anything the study highlights
the inadequacy of voluntary reporting systems for adverse
health effects and the confusion that results when a
"systematic" analysis is attempted on such data. It is well
known among the legal community and the FDA that
thousands of adverse events have been reported to
ephedra supplement manufacturers. Access to these
reports might have affected the outcome of the present
study. If anything these additional reports would have
magnified the gravity of the public health threat attributable
to ephedra-containing supplements. Moreover, Poison
Control Centers throughout the country also log calls on
ephedra supplements. Were attempts were made to access
these additional resources?

We did not contact Poison Control

Centers. We did include in this revision an
assessment of the reports made to one
manufacturer.

There was a thorough search of relevant articles using 9

electronic databases. Both national and international
journals were included and the searches appeared to
capture most of the relevant studies. The majority of the
accepted articles for the meta-analysis were from the U.S.
It seemed that 3 were from Germany. There were no
studies from Asian journals.

No response

Table 19, I do not understand the point of this table or the

conclusions being drawn. Please clarify for the simple
minded.

We have added text to explain this table.

The point is the later cases, that we did not
have access to, contained proportionately
more deaths.

Timing of last ephedrine (ephedra) dose? If it was > 24

Such cases were not reviewed, so a
hours, because of the relatively short half life, one would not negative toxicology screen would not even
expect to detect much or any in the blood at autopsy. Is
have been assessed.
timing of last dose with a tox screen negative test taken into
consideration when determining causality?
Insert percentage next to the # of adverse events for easier
direct comparison of the placebo and intervention groups

We do not feel this comparison is justified

due to small sample sizes, that is why we
did not perform meta-analysis.

Instead of the 5x4 (nxn) test, it may be better to perform the

chi-square test on event type vs. data type, ie. death (vs.
other) x data type (2x3), stroke x type, etc.

We are not sure what this comment

applies to.

A3-47

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

There were clear criteria used to select studies for inclusion

in the report.

No response

The appropriate criteria were used to assess the studies on

efficacy of weight loss with ephedra or ephedra/caffeine.

No response

The initial draft failed to include a review of the published

These published case reports are now
case reports of adverse events. Given that these case
included in this revision.
reports were published in peer-reviewed medical journals
and are prepared by medical professionals, they are likely to
contain data that is more complete, accurate, and of
scientific merit. It would seem more appropriate for these
reports to have been evaluated and included in this report,
especially since no other entity has conducted a review of
them, than to once again include an evaluation of the FDA's
evaluation.
Currently one must read to page 54 to get an answer to the This information is in the appropriate place
key question, "We interpret this data as indicating the use of for an EPC evidence report.
ephedra is associated with a statistically significant 1.3
pounds of weight loss per month more than is associated
with placebo for up to four months of use" and "We interpret
these data as indicating that the use of ephedrine and
caffeine is associated with a statistically significant 2.2
pound weight loss per month more than is associated with
placebo up to four months duration." This should be in the
very beginning of the report.
The 19 efficacy studies not pooled in the analysis because
they had a duration of less than 8 weeks, and the 9 studies
eliminated for a variety of reasons should be accounted for
in the document, and any serious adverse event reports
described should be included in the report.

These studies were included in the safety

analysis.

We added that these events would be

Even in aggregate the clinical trials only enrolled sufficient
number (how many?) of patients to detect a serious adverse classified as rare.
event rate of one per one thousand. And again on page 58:
For studies of ephedra, there was only sufficient statistical
power in aggregate to detect a rate of serious adverse
events if three in one thousand. For the reader, it would be
helpful to know how these event rates, 1/1000 and 3/1000
compare with those reported in the literature for drugs in the
same usage category as ephedra (three billion servings in
1999), i.e. to HRT (approx, 3.8/1000 women has an MI or
developed breast cancer), or to event rates for aspirin and
GI bleed.
References 104 and 108: The Nasser study (ref.108) is the
same as the first Boozer study (ref.104).

This duplicate study has now been

removed from the pooled analysis.

Under Bibliography Accepted Articles, pp 135 137,

This duplicate study has now been
references 12 and 41 are the same study [12=published
removed from the pooled analysis.
study, 41 = published abstract], leaving 4 studies that
assessed the effect of ephedra + herbal caffeine. Data
reported in Chapter 3 Results including Table 14 will need

A3-48

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

to be corrected and re-analyzed.

Monthly Weight Loss. In your results section you report the
data as "Monthly" weight loss. The rationale for this
escapes me. Weight loss with all medications slows with
time and a plateau is reached between 4 and 6 months.
Thus, the most rapid weight loss occurs in the first month.
In trials that last 6 months, the only weight loss will be
slower than one that lasts 2 months. How do we compare
them with this criterion?

We tested, given the data available in

these trials, whether weight loss differed
across the different months. We could find
no evidence that it did. Therefore, within
the limited time frames of these trials (4
months), we included all relevant data
points, as it increased our statistical power.

Miscellaneous Comments. On page 11, the draft report

acknowledge that the "estimate of use of ephedra
containing products may be low." Despite this, the report
fails to emphasize denominator-related concerns associated
with adverse event report reviews. This major scientific
weakness needs to be acknowledged as part of the adverse
event report analysis in a manner similar to that used in
prior AHRQ studies (such as the Garlic Report). On page
29, the draft report indicates that although certain studies
did not record any data for certain even category or indeed
any adverse events at all, such studies were not included in
the adverse event meta-analysis as RAND did not assume
zero observed events if a study did not mention a particular
type of event. Is this approach consistent with most
scientific reviews?

We did assume zero events for serious

events like death or stroke even if they
were not recorded in the RCT. We did not
do so for other events because we could
ever know whether those events were
sought by the investigators if they were not
recorded. In other words, we did not
assume zero for the entire universe of
adverse events, only for those specifically
mentioned and sought and recorded as
zero. This is consistent with most high
quality scientific reviews.

The Garlic Report. The draft report, as noted above, is in

many ways inconsistent with prior AHRQ reports that
address adverse event case reports. The AHRQ Garlic
Report (Garlic: Effects on Cardiovascular Risks and
Disease, Protective Effects Against Cancer, and Clinical
Adverse Effects), and statements contained in the Garlic
Report, should be reviewed by RAND as a potential model
for the ephedra report - particularly in the manner case
reports are assessed and described. For example, the
Garlic Report provides the following with regard to adverse
event reports and confounding factors:
Adverse effects of oral ingestion of garlic are "smelly" breath
and body odor. Other possible, but not proven, adverse
effects include flatulence, esophageal and abdominal pain,
small intestinal obstruction, contact dermatitis, rhinitis,
asthmas, bleeding, and myocardial infarction...The
frequency of adverse effects with oral ingestion of garlic and
whether they vary by particular preparations are not
established...Furthermore, the causality of the adverse
effects was not clear, except for the breath and body odor,
and the expected frequency of adverse effects was not
determined...
In addition to the RCTs, 73 studies were found that
addressed diverse effects. Most (97 percent) were case
reports or small case series (Evidence Table 9). The
literature reviewed gives a limited picture of adverse effects

The inclusion criteria and reporting of

studies in the Garlic Report were shaped
by their TEP and their Partners. The
inclusion criteria and reporting of studies in
our ephedra report were shaped by our
TEP and our Partners. There is no
requirement that these reports be the
same in inclusion and reporting.

A3-49

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

attributable to garlic for many reasons. First, searching for

studies that report adverse effects is difficult. Many studies
may mention adverse effects in passing, but do not use
adverse effects as a key index word or in their abstracts. If
these studies do not otherwise meet selection criteria in a
review, they will be missed. Second, in most case reports
and case series, adverse effects cannot be directly
attributed to garlic because chance, coincidence, or
confounding factors could have been responsible for the
adverse effect.
For example, alternative causes of reported adverse effects
were possible in 22 percent of the reviewed studies and
could not be excluded definitively in 69 percent. Third, case
reports and case series may miss delayed adverse
reactions because such associations are more difficult to
make than those that occur immediately after garlic is
administered. Fourth, although case reports and case
series can provide qualitative information about the nature
of an adverse effect, incidence cannot be estimated from
such evidence...
The frequency and severity of adverse effects that are
related to garlic should be quantified. Whether adverse
effects are specific to particular preparations, constituents,
and doses of garlic should be elucidated. Whether certain
adverse effects are unique to particular types of garlic
exposure (e.g. inhaled, oral, or topical) should be clarified.
The most serious potential adverse effects of garlic that
have been cited are complications related to bleeding.
Whether particular preparations and constituents of garlic
affect physiological parameters related to bleeding such as
platelet adhesiveness, prothrombin time, and partial
thromboplastin time, as well as whether particular
preparations lead to clinically significant bleeding, warrants
more study. (emphasis added).
The limitations of the review process are not stated. You
have adverse reactions and inconclusive studies on the
effectiveness of ephedra/caffeine. That leaves us with
insufficient information to make an assessment of either
safety or efficacy. In the data synthesis, the impression is
given that there is more precision than can be justified
based on the nature of the data.

We disagree with regard to the reviewers

comments on precision. Precision is
determined mostly by sample size and
number of studies. We believe our pooled
results adequately reflect the degree of
precision the data allow.

Possible bias of the report due to members of the TEP and

literature captured. Limitations and quality of the studies and
short-term studies used in the systematic review. Combining
the systematic review by meta-analysis with the analysis of
AERs. No conclusions for herbal ephedra and weight loss.
The AERs evaluated were limited to those provided by FDA.
AERs from the studies were not reviewed. Report did not
emphasize any potential benefits of ephedra/ephedrine and
weight loss.

The AERs from other sources are now

included in this revision. We disagree that
the report did not emphasize the potential
benefits of ephedra/ ephedrine use and
weight loss. The other limitations are
noted.

A3-50

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

Reference 29 can provide a national estimate of 2.5 million

individuals using ephedra wt loss products (during 19961998), which is probably an underestimate since 33% of
one-states respondents did not know that their
nonprescription weight loss product contained ephedra.

The imprecision noted by the reviewer

even in this one estimate is, we believe,
good reason to avoid its use in trying to
calculate a rate using case reports.

Some of the questions guiding the evidence report were not

answered but are available in the literature search and data
collection. For example, the question regarding the dosage
level of ephedra necessary to achieve weight loss was not
answer. There is no summary statement on the dose of
ephedra or ephedrine other than Evidence Table 1 and 2.

A dose analysis is included in this revision.

Conclusions regarding the efficacy of ephedra-containing

No Response
supplements in promoting weight loss and the enhancement
of exercise performance were supported by the available
data.
The conclusions are clearly and concisely laid out and are
consistent with the evidence presented.

No Response

The conclusions of the efficacy of ephedrine and related

compounds are valid for the short term studies evaluated
and conclude that longer term studies need to be
conducted. The remaining conclusions are valid and
appropriate.

No Response

For the efficacy and minor adverse effect evaluations, the

evidence does support the conclusions.

No Response

Finally, there was no demonstrable effect of sustained

ephedrine supplementation on strength training.

No Response

According to the report, caffeine appears to enhance the

effect of ephedrine yet there is no mention of assessing
caffeine in the diet. This would have to be done in a casecontrol study. It may be appropriate to add this separate
bullet under conclusions.

The need to control for caffeine intake was

added to a bullet in the conclusion.

Bullet 1: Conclusion of "sufficient evidence" should be

Changes made to bullets.
tempered with reference doses used and duration of
treatment in the studies.-Bullet 2: This is just a statement
not a conclusion. It would become a conclusion by adding
that out of 1848 cases known, 1344 were selected for
review and of these, 158 showed serious adverse events.
Within this subset of 158 reports, 11 were identified in which
ephedra was possibly causal. Conclusions section should
be able to stand alone. This would require adding more
detail including doses and duration.
It is generally acknowledged in the field of obesity research
that any study with less than a 2-year follow-up is
misleading and not relevant to the evaluation of the antiobesity modality under investigation. The reason for this is
that most individuals regain lost weight on any diet, drug or
weight loss program after 24 months There were no long-

A3-51

We believe the text is clear that these

results cannot be extrapolated beyond four
months. The data support a linear
relationship for weight loss over 4 months
in these studies.

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

term weight loss studies of ephedra or ephedrine reviewed

in this report, yet a mathematical formula for weight loss
associated with ephedra was presented. This seems highly
speculative and hard to defend. The conclusion that shortterm weight loss can be achieved with herbal ephedra and
caffeine rest on studies of herbal concoctions that are not
adequately characterized, chemically and
pharmacologically, to permit pooling of studies; The attempt
to quantitate the weight loss per month attributed to the use
of ephedra-containing herbs is invalid. Further, the model
assumes constant weight loss over time, an unlikely
outcome.
The need to add caffeine to ephedrine to produce any
measurable degree of enhanced athletic performance
suggests caffeine alone may suffice to achieve this effect,
which, in any event, is evident for only short time periods.

The reviewer is incorrect, since one study

reported in the athletic performance
section compared ephedrine, caffeine, and
their combination, and reported only the
ephedrine/caffeine combination produced
an effect. This result refutes the reviewers
hypothesis that caffeine alone may suffice.

It seemed that the decision to review FDA adverse event

reports produced very little useful information and was
almost a duplicate effort given Haller and Benowitz study. It
may be more useful to use the same strategy on the case
studies reported in the literature.

The literature cases are now included in

this revision.

The bullet point "Scientific studies (not additional case

reports) are necessary..." should be deleted as this is not
the conclusion of the Adverse Consequences but rather
Future Research, which has been stated already.

We think it important to also include this as

a conclusion.

Dr. Leung made the point that there were thousands of

Chinese literature on the Ma Huang, and if it wasn't safe the
literature would say so. The literature comments that some
people should not take it, but there is a lot of information to
show that it is safe. Dr. Leung also made a point on the
credibility of this information by stating that for other herbs
the literature shows they are not safe.

We do not disagree that there may be

extensive Chinese literature on MaHuang
but we did not find controlled trials in our
literature search, nor was this literature
offered to us by any of the many reviewers
of this report. Furthermore, we believe that
there is ample evidence to support that the
most valid conclusions come from properly
designed hypothesis testing studies, not a
collection of anecdotal literature, either
supporting or refuting safety.

I disagree with the safety conclusion of this report for two

We added to the limitations the issue of
reasons:
select patient populations. It was outside
1. The clinical trials excluded patients at risk, thereby
our scope to assess other chemicals.
reducing the studys ability to detect harmful effects of the
drugs.
2. The totality of prior pharmacologic information was
ignored in the analyses of the FDA cases. The consistency
between the type of events reported and the known actions
of ephedrine to increase heart rate and blood pressure must
contribute to the assessment of causality The similarity of

A3-52

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

the reports seen with similar chemicals, e.g.

phenylpropanolamine, must be given consideration. The
cases that the authors classify as possibly caused by
ephedrine/caffeine I would classify as probably caused by
ephedrine/caffeine.
I believe the second bullet under Adverse Consequences
(page 112) is unclear and unqualified. This statement
reads, There have been a great number of adverse event
reports filed with FDA regarding herbal ephedra-containing
dietary supplements. I find this statement unqualified and
unhelpful. A great number of reports compared to what?
Total dietary supplements sales? The number of AERs
reported for other dietary supplements? The statement
should be better qualified, in my judgment.

We believe 2000 is a great number by

most peoples definition and have not
made any changes.

The first bullet on page 113 is a very key issue and I believe
deserves further comments. I agree with the conclusion
that, given the rarity of serious adverse events associated
with ephedra, properly designed case controlled studies
would be appropriate. However, I believe it will be difficult
to develop such a properly designed case controlled study,
as the underlying factors (that appear to be idiosyncratic)
are not well understood. How, then, would a case
controlled study be designed to take such unquantifiable
factors into account? This is precisely the continuing
problem in deciding how best to approach both the
regulation of and further scientific research into ephedra

It is beyond the scope of an EPC evidence

report to go into such details of study
design. We note, however, that others
have made detailed proposals to
governmental agencies for just such a
study.

The conclusions seemed fair and stated appropriately. I

would suggest adding a section answering the questions
you posed at the start of the report in a summary fashion.

We organized our bullets to follow the

order of the questions.

Most of the studies reviewed were on synthetic ephedrine

and weight loss, therefore, a relationship between herbal
ephedra and weight loss cannot be made and this appears
problematic. The clinical data that were examined only
included ephedra in combination with another herbal
stimulant. While it is true that many weight loss products
contain a combination of ingredients, not all do (NNFAs
database of ma hang or ephedra reveals that almost half of
the products do not contain another stimulant). A concern is
that the conclusions drawn in the report may be applied to
all ephedra products, regardless of use and regardless of
whether other ingredients are present.

We now include one RCT of ephedra

without caffeine. We have limited the
conclusions to only those concoctions
studied.

It seems important to mention that apparently healthy

individuals have died from the use of these products,
possibly related to exacerbation of previously undetected
disease. And, because these products are available without
prescription, and not even regulated as OTC drugs, the risk
associated with unsupervised use are potentially greater
than with drug formulations of ephedrine.

We have added to the conclusions and

limitations both of these points.

The conclusions of adverse consequences are internally

We disagree since the finding of serious

A3-53

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

inconsistent. Once the limitations are listed, then the

number of serious adverse events (death, stroke, and
myocardial infarction) should not be enumerated. This is the
same thing as saying that results of a scientific study are not
statistically significant and then still enumerating them in the
conclusion.
The most important conclusion is listed second to last on
Page 113 and should be moved up to the front of the
conclusion section of the report. That is, scientific studies
(not additional case reports) are necessary. I disagree that a
case-control study would be the next step. Rather I would
recommend strongly a prospectively randomized controlled
study design with appropriate data safety and monitoring in
place. A population-based study will have the same
drawbacks as the phenylpropanolamine study and will have
the same risk of spurious associations rather than cause
and effect relationships. An intervention study and not an
epidemiologic study is needed to clarify the situation.
As already stated, it is clear that the evidence does not
support the conclusions. The adverse event reports are that
they are. The attempts to connect them to the use
Ephedra/Caffeine remain unconvincing both in this report
and in the New England Journal of Medicine article.

adverse events in otherwise healthy young

people is a cause for concern. We also
disagree that the phenylpropanolamine
study found a spurious association. We
note that a case control study is the
accepted study design to quickly assess a
possible relationship between an exposure
and rare adverse events.

Specific issues related to ephedra are not addressed

We clarified that the use of ephedrine to
adequately, and notions for which there is no proof are
raise blood pressure intraoperatively is
presented as if they were accepted scientific fact. For
with parenteral use.
example, in the Pharmacology section, the report states
"ephedrine increases peripheral resistance and can lead to
a sustained raise in blood pressure...Elevations in blood
pressure appear to be dose dependent in humans.
However, does under 50 mg do not always result in
increased blood pressure." The report fails to state that the
sustained raises seen in hypotensive patients occur after
the intravenous, not oral administration of ephedrine. The
only citation for the dose dependency of an ephedrinerelated rise in pressure is a review article, and that article
does NOT say that ephedrine causes hypertension! It says,
ephedrine and caffeine cause a greater increase in systolic
pressure than ephedrine alone, that there is no effect on
diastolic pressure, and that hemodynamic effects are
transient. The statement is quoted out of context and is
therefore misleading.
The way the sentence is written, readers would be likely to
assume that, even though "doses under 50 mg do not
always result in increased blood pressure, a series of
double-blind, placebo control trials have shown that at most
the effects of oral ephedrine on blood pressure are
negligible (as opposed to intravenous dosing used b
anesthesiologists). A partial listing of some of these studies
is cited here [10-23]. The lack of effect on blood pressure is
even supported by the list of TEP "accepted articles" cited in

A3-54

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

the bibliography. All of the cited articles were from

controlled clinical trials, and none reported clinically
significant blood pressure elevations. These studies should
be included as part of the RAND review, and the statements
regarding increases in blood pressure should be
substantially revised. Indeed, if the authors of the report
cannot cite published, double-blind placebo control studies
showing that taking oral ephedra/ephedrine significantly
increase blood pressure, that that claim should not be
included in the report.
Suggested directions for research were provided, but they
are narrow in scope and may suggest "gaps" that do not
really exist. To date, more than 2000 ephedra/ephedrine
users have been enrolled in clinical trials. Given the
consistently benign results of all the previous trials, are still
more trials needed before the issue is put to rest? On the
other hand, cutting edge issues in obesity-related research
are completely ignored. Does ephedrine interact with
uncoupling protein? Does use of ephedra supplements
have any effect on the production of inflammatory cytokines
by adipose tissue? Or upon lepton homeostasis? On
Lipotoxicity? If supplement manufacturers are to be
believed, they have thousands of testimonials from satisfied
users reporting weight losses of 50 pounds or more. Why
not study these individuals and compare them with other
product users who were unable to achieve weight loss?
Having identified a population of proven ephedra
responders, and non-responders, comparing the two groups
medically, chemically, or genetically, may provide some
truly useful insights.

We would ask the reviewer whether 2000

successful airplane flights mean that
airplanes never crash. The point is that
2000 studied patients is insufficient to
detect a rate of 1/1000 events, and even
rare events, when multiplied by the millions
of people who may be consuming
ephedra, add up to numerous serious
adverse events, if such an association
exists.

The report draws conclusions about efficacy and safety that

are not sufficiently supported by the data. As I have pointed
out below the efficacy of ephedrine/caffeine is
underestimated due to the incorrect method of analysis. In
addition, in my view a number of shortcomings in the safety
assessment tend to exaggerate the adverse events.
My overall conclusion is that in several aspects the report
needs some important revision. This includes the
identification of studies, selection of studies for efficacy and
safety. The data handling is also inadequate in some
aspects. Consequently the report's overall conclusions are
not supported in the current version and I believe that the
revision suggested below will produce a substantially
changed conclusion.

No response to this general comment.

Specific response made to specific
comments. We disagree that the
identification of studies, selection of
studies, data handling, etc. are
inadequate. We also disagree that our
results underestimate the efficacy of
ephedrine/ caffeine, or that our analysis is
incorrect.

Weight Loss. In the first bullet it is stated that compounds

produce weight loss over relatively short periods of time (no
more than a few months). This is misleading as there are
trials for a duration of 6 months. The same applies for the
3rd bullet where the expression "short-term weight loss" is
used Bullet 6 is outrageous here it is concluded that

Six months is still a few months when

one year data are considered necessary
by FDA to assess pharmaceuticals. The
data about phentermine are taken directly
from the graph in the cited reference. We
have added data about weight loss using

A3-55

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

other pharmaceuticals.
ephedrine and ephedrine plus caffeine produce a weight
loss somewhat less than the effect reported for FDA
approved pharmaceuticals for weight loss. The panel has
used phentermine as an example and state that the effect is
"reported at about 20 pounds of weight loss at 6 months".
This is certainly not the weight loss produced by
phentermine above placebo, the weight loss produced by
phentermine from baseline including a diet. For comparison
one can take the Astrup et al. study from 1992 where the
weight loss in the ephedrine plus caffeine arm was about 16
kg. But of course, the weight loss in the placebo arm must
be subtracted, giving an additional weight loss produced by
the compound of 3.6 kg.
Adverse consequences. Again, this reviewer suggests that Our statement refers to the data included
the open trials should also be included. In the first bullet it is in our review, which was restricted to
stated that it is not possible to separate out how caffeine
RCTs and CCTs.
contributes to the side-effects. This is actually possible. In
the Astrup et al. in International Journal of Obesity in 1992
there was a separate caffeine arm in the 6 months trial.
Side-effects are shown in one of the tables in this paper,
and here it is clear which side-effects can be attributed to
caffeine.
A long-term study of comparing ephedra + caffeine with
ephedrine + caffeine at promoting weight loss and adverse
reactions. Expand the pharmacokinetic study of ephedrine
(pharmaceutical preparation) and ephedra (botanical
preparation) absorption (as part of the dose response
studies).

We think this is already subsumed under

the first bullet point.

The most basic and important aspect of any research in

natural products and in the reporting of findings is the
characterization and clear definition of the products or
materials being studied. Without this, research findings
cannot be reproduced and thus are meaningless. In our
case with ephedra evaluation, we not only need to set
criteria for selecting articles for study, but also be sure to
clearly understand what it is that we want to study
ephedra herb or ephedrine.
The whole field of ephedra in weight loss and athletic
performance is twisted backwards. The herb ephedra has
never been traditionally used for either function, nor has it
been first clinically reported (before ephedrine) to have
these effects. Only the drug ephedrine has. Yet ephedra is
being used for these effects and is touted as natural
ephedrine and thus safer. So far, there has been no
credible clinical evidence that ephedra itself (and not
synthetic ephedrine in an inert ephedra carrier) has these
actions, despite the conclusion reached in this report.
It is worthwhile to reevaluate references 104-108 to
determine whether the ephedra used in those studies was
actually natural ephedra containing the total complements of

The included ephedra studies said they

assessed herbal ephedra. We agree with
the reviewer that a future study of ephedra
should adhere to these recommendations.

A3-56

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

ephedrine alkaloids in their natural proportions. If not, was it

composed mainly of synthetic ephedrine formulated with
carriers (e.g., token ephedra or exhausted ephedra marc)
into an ephedra dietary supplement that contains little or
none of the usual complements of other ephedrine
alkaloids? If it is the latter, then this ephedra herb has no
place as a dietary supplement in weight loss or athletic
performance. Such ephedra-containing products should
then be more appropriately placed under the OTC-drug
category which would eliminate much of the problems
currently associated with its abuse and also would save us
taxpayers much money trying to resolve these problems.
In order to show ephedra herb (not synthetic ephedrine) to
also have efficacy in weight loss and/or athletic
performance, it is necessary to first characterize and
standardize ephedra products to specific amounts of
ephedras alkaloids in their natural proportions, before
subjecting them to clinical trials. This would eliminate the
drug ephedrine being formulated into a dietary supplement
to bypass the OTC-drug regulations.
Unless ephedrine-containing products (whether natural or
synthetic) for weight loss and athletic performance are all
considered OTC drugs, adulterated, poorly characterized,
and undefined dietary supplements containing ephedra herb
will continue to be sold and abused. We need to set
standards for manufacturers to meet and follow in order to
be able to label and market their ephedra-containing
products as dietary supplements.
As I have repeatedly stressed, the most important aspect of
any research in herbal medicines/supplements is
characterization and precise definition of the test materials,
without which no meaningful and reproducible results can
be achieved, no matter how well designed and how well
executed the rest of the research. In order to reduce the
continued accumulation and dissemination of ambiguous,
meaningless, and useless research data in the natural
products field, we urgently need to set criteria for the
characterization and precise definition of test materials at
three levels: (1) research; (2) publication; and (3)
abstracting, indexing, and data input into databases. Such
criteria have been published and are available.
The conclusion that a properly designed case control study
would be the appropriate next step would require a study
so large, lengthy and expensive it is unlikely to ever be
funded or completed. (If one assumes a prevalence of use
of ephedra of 1%, an alpha of 0.05, a power of 80% and if
one seeks to detect a doubling of risk, then 2,400 stroke
cases and 2,400 unaffected controls would be required. See
Schlesselman, Case-Control Studies, Oxford University
Press, 1982). The presumed benefits of ephedrine, should
they exist (improved athletic performance enhanced

A3-57

We agree that a properly designed case

control study would need to be large
(perhaps not as large as this reviewer
conjectures). However, the PPA case
control study was also large and was
successfully completed, and we do not
favor substituting opinion for science when
the scientific study is feasible.

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

energy, short-term weight loss) are likely due to the

sympathomimetic effects of the drug and the adverse
consequences are predictable as they were with
isoproterenol, amphetamines and fenfluramine. Since
sympathomimetic drugs have never been shown to result in
safe and sustained weight loss, it is highly unlikely that this
will be the result of long-term controlled trials of ephedrine
and weight loss. But the known adverse effects of the drug
and its congeners are now well characterized.
First, you need to organize your listing of sources of
adverse event information better so that readers can see
which sources you have included and which you have not.
Right now, it's difficult to follow what you have gathered. A
table would be ideal, with each row specifying the source,
the number of complaints, the number of deaths, serious
injuries, non-serious injuries, and the numbers of each of
these you concluded are likely or possibly related to
ephedra use.

All of the serious adverse events came

from FDA data, so in the draft report such
a table would have no meaning. In the
revised report, such a table is included.

You do not provide clear evidence of the pharmacological

and pharmacokinetic equivalences in the use of the herb or
of ephedrine alone. The complexity of the Phytochemistry of
Ephedra (p 13) reinforces the point that the whole herb
contains other alkaloids that are likely to be active or to
qualify the effect of ephedrine. More should be made of this
deficit at various points in the text.

We agree that there is likely heterogeneity

in the herbal concoctions, but note the
striking consistency of our findings relative
to amount of ephedrine alkaloid and weight
loss.

There is insufficient evidence that dietary supplements

We disagree and believe the data speak
made up of the herb Ephedra spp. have any of the effects or for themselves.
risks identified for the alkaloid ephedrine.
The future research directions proposed are reasonable.
This was added to the future research.
One addition may be to recommend examining the
interaction of ephedrine and exercise training on weight loss
and adverse events. Is there some interaction between
physical activity and ephedrine?
If the majority of ephedra users are seeking long-term
weight loss, it would be very helpful to better understand the
age, gender, race, temporal use patterns, concomitant drug
use and other risk factors associated with ephedra usage.
These points are underdeveloped and are, I believe, central
to understanding safe and appropriate use of ephedra in the
general population.

We agree in principle with this comment,

but think it might wait until there is better
evidence of sustained weight loss in any
population.

I agree with the suggestion to analyze and compare the

adverse events reported for ephedrine to ephedra. I would
also add PPA.

This report now includes an assessment of

ephedrine so this bullet has been
eliminated.

The suggestion to consider a dose response study to

While it may be difficult, it is certainly
determine a minimum effective dose of ephedra would be
feasible and consistent with the way FDA
difficult, at best. Effectiveness criteria should be identified in evaluates pharmaceuticals for weight loss.
these comments.

A3-58

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

There remain open questions whether there is a difference

Agreed.
between synthetic ephedrine and naturally extracted
ephedrine alkaloids. It would be a very useful research
activity to analyze the branded products which are identified
in AERs using both AOAC and USP methods to try and
determine whether synthetic or naturally occurring
ephedrine alkaloids are present.
Future Research. The numerous gaps in the literature
regarding the efficacy and safety of ephedra is a central
point.

No response

FDA has recently taken action against six companies selling No response
synthetic ephedrine as dietary supplements. This is not
permitted under current law but, unfortunately, synthetic
ephedrine dietary supplements are being sold to the general
public.
As a final thought, the inadequacy of FDAs adverse event
This is not a proper role for an EPC
reporting system is clear as it relates to ephedrine. I believe evidence report and we decline to make
it is appropriate for RAND to recommend that, with respect
such recommendations.
to ephedra products, FDA/CFSANs process and systems to
evaluate and capture ephedra-related AERs be thoroughly
reviewed, as it is likely that continued reliance will be placed
on this system, despite its weaknesses.
A chapter was devoted to future research. The researchers No response
addressed the gaps in a variety of areas and suggested
meaningful recommendations for further research. Most
significant is the need for long-term studies of
ephedra/ephedrine and weight loss and athletic
performance including both anaerobic and aerobic exercise.
This was emphasized in this chapter.
It might be beneficial to explain the pathophysiology of how
ephedra/ ephedrine can contribute to an acute
cardiovascular event in the setting of mild-moderate
underlying disease. For instance, in individuals with noncritical coronary artery disease, ephedrine alkaloids can
produce platelet aggregation with resultant thrombus,
increased myocardial oxygen demand, and cause
vasospasm, all of which can result in decreased perfusion
and ischemia. The same contributory actions could be
expected in individuals with congenital cerebral aneurysms,
and other underlying abnormalities in the cardiovascular
system.

While we agree that biologic rationale is an

important criterion when assessing
causality, we think that direct evidence of
an association is most important, and
therefore recommend a hypothesis-testing
study.

The implications for future research are fairly stated. I would This was added to the Future Research
suggest adding the analysis of safety or adverse events
Section.
reports in Denmark comparing available prescription obesity
drugs, since Denmark uses caffeine and ephedrine as one
of it's approved prescription drugs for obesity.
Are implications for research discussed? Not adequately.
The major implication of the research is whether the

A3-59

The role of the EPC is to report the

evidence which we believe we have done

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

analysis will be adequate to advise the FDA and HHS on

whether they should take action to protect the public. This
aspect of the analysis is ignored.

The judgment about the adequacy of the

evidence to make a judgment is not a role
for the EPC.

What directions for future research would you recommend

based on this report that we have not covered? As I discuss
in the following general statement, the missing ingredient in
this (in addition to the issue of how the products are being
used by the public) is the need for an analysis of the
complete pharmacology of ephedrine/caffeine products.
This must include consideration of the modern science of
pharacogenomics and genetic polymorphisms of receptors
for these products. This type of analysis adds relevance and
credibility to adverse events that occur in low frequencies. It
explains how some patients can have little or no change in
blood pressure or heart rate and how some can be placed
at risk of stroke, seizures or heart attacks.

A good suggestion, but one that we feel is

probably some years off, as opposed to
the three studies listed first. A genetic
analysis could conceivably be added to a
case control study and used as an effect
modifier in the analysis.

Implications for future research were discussed. Physicians

and most pharmacologists seem to want to lay the blame for
problems associated with ephedra supplements at the feet
of ephedrine/ caffeine. This narrow view excludes the
pharmacological activity or potential interactions with other
phytochemicals present in these products. In the opinion of
this reviewer, the problem is more complex than simply
ephedrine and caffeine.

This is a good point. An assessment of

ephedra use may be able to take
advantage of the heterogeneity in
concoctions to perform subgroup analyses
looking for ingredients other than
ephedrine and caffeine.

Regarding future research aimed at stroke aspects, it would

seem valuable to pursue case-control studies along the
lines of that by Kernan and colleagues cited above but
considering both hemorrhagic and idiopathic ischemic
stroke in relatively young adults.

Agreed. The proposed case control study

should assess all of the serious outcomes
we assessed.

Page 5, paragraphs 3 and 4 would it be appropriate to

We do not think so. The exposure has
mention ethical considerations for case-controlled studies in already occurred.
the summary?
Page 5, paragraph 4: Pre-clinical studies should also be
considered to determine the use of ephedrine or ephedrine
containing the alkaloids increases the risk of development of
heat related conditions such as heat exhaustion, heat
stroke, and rhabdmyolysis, if an appropriate animal model
can be found. What specifically would be learned from this?
Could it be extrapolated to humans?

It might help establish a biologic rationale,

but in this discussion we have deleted the
animal model and pre-clinical aspect to
this and suggest it be included in a study
of adverse outcomes in humans.

Rewrite to redirect emphasis of sentence by placing, If an

appropriate animal model can be found, pre-clinical studies
should be.and rhabdomyolysis.

We actually eliminated the preclinical

studies part of this and suggest a study
assessing this as a potential adverse
event.

I would suggest come additions to the section of future

research including an interaction study to investigate the
effects of ephedra with not just caffeine-containing herbs,
but also combined with botanicals such as citrus aurantium,
garcinia cambogia and the herbal diuretics and cathartics

These suggestions have been added.

A3-60

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

as listed in Table 1.
Also, I would specifically suggest that studies on athletic
performance be conducted in women and adolescents,
since these populations are known users of these products.
Finally, I would recommend that the association between
ephedra and seizures be formally explored.
Future Research Section. You favor a case-control study.
The case-control trial with phenylpropanolamine was
sufficient to remove the drug from the market, but it was a
pretty poor study. The controls and cases had very different
lifestyle habits. There was a no-dose-response to PPA.
The effect was only detected in women. Because of the
large number of things used in the many products on the
market, and the relatively high rate of deaths and disability
from heart disease and stroke, it is not clear that a useful
answer would emerge.

We disagree. We think such a study would

tell us something useful about ephedra
products. We do not think the
heterogeneity in the components of the
products will be any greater impediment to
the analysis of safety than it was to
efficacy. Our data are consistent with the
hypothesis that the only active
components with respect to efficacy and
safety are ephedrine and caffeine.

Another point that might be useful to make is that from the

available data it is not possible to determine which
populations are at greatest risk for serious adverse events,
and that this could only be determined by additional
research.

This point has been made in our

suggestion for a hypothesis testing study.

A3-61

Appendix 3. Reviewer Comments (continued)

CAUSALITY COMMENTS
The remaining reviewer comments from the first review concern an attempt in our draft document to
assess causality for some adverse events. We did so using our own modification of published methods.
These comments varied widely, ranging from critiques of our method for being too conservative
(meaning, in the opinion of some reviewers, we had excluded or assigned too low a level of causality to
certain cases) to critiques for being too liberal (meaning, in the opinion of some reviewers, we assigned
too high a level of causality to certain cases). Often, these conflicting comments concerned the same
cases. We believe these peer review comments demonstrate that case report reviews involve
considerably more subjective interpretation than do reviews of randomized trials. Because our goal in this
evidence report is to report the evidence as objectively as possible, we ceased to assign assessments of
causality to the case reports. Rather, we tried to identify those cases that would be classified medically as
"idiopathic" in etiology, meaning the cause is not known. For such cases, given the known pharmacology
of ephedrine, if use of ephedra or ephedrine was documented, a potential role for ephedra or ephedrine in
causing the event must be considered. We classified such cases as "sentinel events." Other than correct
typographical errors and respond to questions of fact, we do not provide a response to the numerous
criticisms of the causality algorithm or suggestions to change our interpretation of these case reports
based on the reviewers opinion or additional information they posses that we did not have in the
documents available to us to review.
Reviewer Comment

Rand Response

Although the present study was compared to that report by

Haller and Benowitz (New England J Med), why was a
comparison not made to the Samenuk et al. study (Mayo
Clin. Proc.)? Were individual case numbers not available
from Dr. Samenuk? On page 51, in Level 2 of the Causal
Flow Model, what is meant by "in more than minimal dose"?
What constitutes a minimal dose? Are you talking about
ephedrine, ephedrine/caffeine, or ephedra supplements? It
must be emphasized that ephedrine and caffeine in
conjunction potentially hundreds of other pharmalogically
active phytochemicals constitute an ephedra supplement?
Accordingly, the pharmacodynamic effects for ephedra
supplements are not directly comparable to synthetic
ephedrine or ephedrine/caffeine combinations. Furthermore,
given the heterogeneity of ephedra supplement
formulations, the pharmacodynamic effects of individual
ephedra supplements are expected to vary.
It should also be acknowledged in the final report that there
is very little consistency in the results of any expert attempts
at assessing causality with this same set of AERs. The draft
touched on this issue in discussing comparison with other
reports and in presenting information in Table 22. The
language of the draft is not, however, consistent with the
data in the table. The draft states that the current judgments
"are more conservative than those of Drs. Haller and
Benowitz" but that there was agreement that some cases
cannot otherwise be explained. It is difficult to understand
how such statements, with their implication that any
differences are either minimal it immaterial, can possibly be
associated with the actual information in the table. The table
identifies 24 cases 20 of which were evaluated by both this

A3-62

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

group and by Haller and Benowitz. In only 5 of these is

there full agreement. Twelve of the 20 cases are cases
where Haller and Benowitz report possible or probable
causality while the current reviewers reported insufficient
information!
It is curious that the Table does not note that, of the 11
cases reported by the draft as probably and the 26 cases
identified as possibly related to the use of ephedra, Haller
and Benowitz only classified 2 of the first and 7 of the
second as either probably or possibly related. Information is
not provided to assist in understanding whether Haller and
Benowitz had classed these as not associated with ephedra
of whether they had not evaluated these cases. If the first
case, it should be disclosed if the second, there should be
some mention that the current causality assessment is
preliminary and subject to review by other qualified experts.
It is also curious why reviews with different conclusions by
other parties were not acknowledged. For example, in at
least one case where both this group and Haller and
Benowitz agreed that there was a possible causal relation
between use of ephedra and death, the local coroner
ascribed the unfortunate incident to congenital problem. In
addition testimony was given by Theodore Farber, Ph.D. on
August 8, 2000 at the HHS Office of Womens Health to
discuss the issue of inconsistency at length and in detail. As
Dr, Farber noted " There was a sufficient lack of
concordance between the FDA's causality analysis and the
causality analysis performed by it outside experts." Other
presenters at this meeting provided analyses of these AERs
that found quite different conclusion that have been drawn
in the Draft. It must be assumed that the record of this
meeting and possibly more specific information, was
accessible as the draft was being prepared but it does not
appear that any attention was paid to any other commentary
on causality reviews to date. This must be corrected.
In discussing the case reports the Draft states that "events
related to synthetic ephedrine" were removed.
Notwithstanding out earlier attempt to clarify that synthetic
ephedrine probably means ephedrine in isolation (or its
salts, e.g. ephedrine hydrochloride), at least 8 of the cases
reported on were associated with a product that was labeled
to contain ephedrine hydrochloride (the E'OLA product) or
was subsequently found, or at least has come out to be
assumed to have been manufactured with undisclosed
ephedrine salts (Formula One). At lease two cases do not
identify the brand so it is not known how this determination
was made from these cases.
Almost every Probable case and Possible case had
either a preexisting condition that could have contributed to
the adverse event or exhibited unhealthy behaviors

A3-63

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

(excessive drinking, smoking, intense effort to lose weight)

which should be noted.
On p 69 and Table 22 there is a comparison of the results of
Rand evaluation of FDA AER with those of Dr. Benowitz
the specific criteria to meet definite, probably or possible
causality are explicitly stated for the Benowitz evaluation [as
was the case with other expert evaluations of these data in
the FDA docket]. It would be useful to specifically list the
criteria used for the Rand Evaluation for their classification.
It is stated that the Rand evaluation is more conservative
than the analysis by Benowitz, but what about comparisons
with other expert reviews of these data [2 FDA reviews,
Woosley, Benowitz, Ricaurte and Stoll]?
Table 22 Summary of comparision with other reports of
ephedra adverse events there is an error in the table for
FDA case number 12720 and 12722. According to
information elsewhere in the report, the following appear to
be the correct data entries:
Case# Adverse event Benowitz EPC
12720 Death Possible Insufficient information
12722 Death Possible Possible
What was the classification of the CanTox study
commissioned by CRN?
Do you have information as to how soon an autopsy was
performed? Could this have any impact on the toxicological
screen results?
I would assume that the prevalence of pre-existing coronary
artery disease is very high. Therefore, when interpretations
are made as to causality and risk for most Americans, it
may be important to have some reference numbers as to
how many Americans have pre-existing CAD.According to
the Am Heart Association (using NHANES III data), 1 in 5
males and females has come form of cardiovascular
disease, see this website for details about CAD.See
American Heart Association. 2002 Heart and Stroke
Statistical Update. Dallas, Texas: American Heart
Association,
2001.http://www.americanheart.org/downloadable/heart/101
48328094661013190990123HS_State_02.pdf
My prior bias about ephedra and stroke was based on
influential case-control study of the relationship between the
use of phenylpropanolamine, a compound with related
physiological effects, and hemorrhagic stroke (Kernan WN
et al. NEJM 2000;343:1826-32). I believe that it is likely, reenforced by the data in this draft report, that ephedra use
occasionally leads to stroke. However, for the purpose of
this review, I have elected to play the devils advocate in
considering the specific question: how strong is the existing

A3-64

A neurologist was included in the review

process in this revision. Grand mal
seizure was the description of the event in
the original source material.

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

evidence that use of ephedra can cause a stroke?

Because determining the cause of strokes among young
people is not that often straightforward, it would have been
optimal to have the stroke cases reviewed and classified by
a stroke expert with experience in evaluation of young
stroke. The case reports (p.63) suggest a lack of
neurological sophistication (i.e. grand mal seizure in case
11062 is not technically correct; generalized convulsive
seizure is probably what was intended). OK,so this is an
irrelevant elitist comment, but in the absence of hard
evidence, credibility is a subjective issue. Case 10874 is
categorized as "probably causal" : along time intravenous
drug abuser with phenylpropanolamine on toxicology
screen. Case 9335 is classified as "possibly causal": 56
year old woman with hypertension, tabacco use, elevated
cholesterol and triglicerides, an MRI with microvascular
changes and whose event was lacunar infarct. Case 12713
was "possibly causal": a 63 year-old woman with artificial
fibrillation with acute loss of conscious and embolic stroke. It
would be easy to take issue with classification of likelihood
of causality in each case.
In short, I agree with the appropriately cautious conclusion
that "there is sufficient evidence to suggst a possible causal
role of ephedra-containing dietary supplements in rare, but
serious adverse events, particularily cerebral hemorrhage."
(p.vi) Support for this statement would be better served by
have a stroke expert review the case reports and perhaps
tossing our the marginal cases (such as noted above).
Further, since this authoratative report may eventually be
used for medical-legal purposes, it would seem responsible
to include a caveat that it is not sensible to consider all
strokes of idiopathic cause in people taking ephedra as
caused by the agent. These comments are not meant to
disparage the overall quality of this impressive report. As
noted at the outset, I have elected to play the devil's
advocate concerning this specific aspect.
The criteria for determining causality were arbitrary and did
not address the true causality. In fact, the term "causality" is
misleading in this connection. Rather, the term "association"
should be used as in "guilt by association". Lay people will
read this report and "probable causality" will be interpreted
as a cause-effect relationship which is not warranted by the
data available from medical record reviews.
The limitations of the data collection are not emphasized
enough in this report. Adverse reactions reports by definition
have no denominator and are subject to reporting bias. In
the famous Phen-Fen debate, initial reports on which the
FDA took action suggested at 35% incidence of valvular
abnormalities. Subsequently, this was found to be less than
8% and reversible following discontinuation of the

A3-65

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

medications.
It is very important to the integrity of this report that the
basic questions asked in the contract are answered, that the
report is well ordered, and that only scientifically valid
information is included. If the AER information, which is not
scientifically valid is included, it should be included as an
appendix, not in the body of the report, as this takes away
from the science.
Use of the terms probably and possibly causally related may
make the causality assessment sound more objective than it
is. Would be the subjective natures of the assessment be
more effectively conveyed by changing those AERs
currently designated as probably causal to possibly causal,
such as events if uncertain relationship? Instead of specific
designations it might be adequate to describe the results in
narrative form. The narrative could explain that although in
some cases cofounders make it difficult to attribute
causality, there is a subset of cases in which cofounders
make it difficult to attribute causality, there is a subset of
cases on which cofounding factors are minimal or absent as
far as can be determined, and it is these cases that raise
concern over safety. Whatever terms or phrases are used,
defining them early in the document will help even those
unfamiliar with adverse event causality analyses understand
their meaning.
Should make it clear that it is not possible to determine the
actual level of risk for people taking ephedra or ephedrine
because the number of people who actually take it is not
known.
It would be helpful to provide possible reasons for the
differences between the RAND causality assessment and
the one done by Haller and Benowitz, this could be done by
adding text to point out that: I. Each group used different
criteria. II. The same group of experts would come to
different conclusions of they were using different sets of
criteria for evaluating the same set of AERs, and III. The
RAND report use more stringent/restrictive criteria for
assigning causality than were used in the Haller and
Benowitz review, resulting in more conservative
assessment.
Requirement of angiography for assigning causality for M.I.s
to Ephedra (similar comments from two reviewers): Page V
Paragraph 5: for cases of myocardial infarction, we
required coronary angiography to have been performed and
the results available. This seems like a very restrictive set
of myocardial infarction cases. What would be the effect on
the results if angiography had not yet been done? Why was
this restriction used? Results section explains this better,
Assume all such cases would have been classified as

A3-66

This table has been dropped from this

revision since causality is no longer
assessed.

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

possibly causal , so the data is still listed. Pages 3 and 32:

While we understand the importance if documenting the
occurrence of myocardial infarction by restricting the
documentation of the event so that the cardiac
characterization is required to assign causality to ephedra,
are the number of MI events being underestimated? Why
not also use enzyme changes in laboratory specimens and
Q-wave changes on the EKG to assign causality?
Probably not causal was used for events that had clear
other causes discovered on detailed investigation. This
assumes all events had a single cause. But can't someone
with known atherosclerosis die suddenly because of
superimposed effects of a substance.
This sentence is confusing In the 935 reports, there were
data in 968 subjects of which 925 reported taking ephedra.
Not clear how there can be more subjects than reports.

A single FDA MedWatch report can

contain information on more than one
person.

A case presented (#12843, 15-year-old female) without any

reference to ephedra exposure. Absent that information, it
would be hard to make this even a possibly causal
classification.
A couple of reviewers were confused by the mention of
We added text to try and help explain this.
AERs that took place after September 30, 2001. Perhaps a
footnote on the table would be informative to remind readers
if the timeframe for the AERs analyzed.
Table 20 provides a lot of useful information, but it might be
easier for readers to interpret the data if another table were
added. This table would have 5 columns across the top,
labeled Product, #Probably Causal, #Possible Causal, #
Insufficient Information, and Total (terminology may change
based on other comments).

We considered but did not make any

changes to this table. A different kind of
Summary Table is included with this
revision.

Data from the case report of the death of a 28-year-old

female indicates that the MiniThin was one of the products
that she was taking. MiniThins were shown to contain
synthetic ephedrine, not ephedra, and the FDA required the
company stop marketing it for weight loss and change the
name and marketing focus (product name was
subsequently changed to MiniTwoWays and was marketed
for use on people with bronchial asthma). Should this AER
be included there?

This case was included in the ephedra

FDA MedWatch file. The patient was also
taking Yellow Jacket.

Specific cases. Page 60, Deaths, Probably Causal: A 21year-old male collapsed": This patient has been taking
hydroxycut, which I assume is hydroxy citrate. Hydroxy
citrate is probably quite toxic, though it has not been
systematically assessed in clinical trials. Biochemically it
may be assumed to have a substantially liver toxic effect. I
think it is therefore very difficult to attribute the case to
ephedrine. I think that there are too many examples of
patients with many other risk factors such as those included

A3-67

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

under the "probably causal" myocardial infarctions, e.g. a

54-year-old, who has smoked for 30 years and been an
alcoholic.
Specific cases. Page 62, Deaths, Probably Causal (cont'd):
Page 62: Another example is the "Stroke, Probably
Causal": "She was a long-time intravenous drug abuser and
alcohol abuser. She also smoked cigarettes for 10 years."
She tested positive for benzodiazapines and
phenylpropanolamine, whereas there was no positive test
for ephedrine. I strongly disagree with the conclusion that
this case can be classified as probably causal with respect
to ephedra use. It is more likely, with the given history and
the positive test of the patient, that the stroke was caused
by other vaso-active drugs taken by the patient. These
weaknesses apply to several of the other stroke cases, and
I think this is particularly interesting in light of the metaanalysis of adverse events reported from control trials
(Table 17, page 80) where it is found that there is no
statistically significant increased risk of hypertension. This
also quite clear from the control study by Ingerslev et al. on
hypertensive patients treated with ephedrine/caffeine. One
should therefore be cautious about drawing conclusions on
the causality with respect to stroke.
HHS and GAO Statements Regarding the Same AERs that
RAND Reviewed. The draft report attempts to ascribe
degrees of causality to the ephedra AERs, thereby ignoring
recent statements by the Department of Health and Human
Services and the General Accounting Office ("GAO"). HHS
and FDA recently reviewed the same AERs that RAND
reviewed, and issued a response to Public Citizen on June
14 that provides the following: The primary purpose of a
voluntary adverse event reporting system is to generate
'signals' of potentially related events, rather than assessing
product safety. While a 'signal' has been generated by
these reports, FDA has determined that questions remain
on the likelihood and strength of association between
ephedrine alkaloids and the adverse events reported to the
FDA...
There are situations when background rates of the observed
event are so rare or unusual that, in combination with
physiologic responses and biologic plausibility, a significant
relationship between the events is self-evident from the
reports in a voluntary reporting system. However, the FDA
has advised me that the types of observed outcomes
reported in relationship to the ingestion of ephedrine
alkaloids are not uncommon in the general population and
therefore the reports alone do not provide a scientific basis
for assessing the safety of ephedrine alkaloids or establish
a link between the reported adverse events and the
ingestion of ephedrine alkaloids. (emphasis added).
The draft report should mention and cite the above-

A3-68

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

mentioned letter and FDA's and HAS' position with regard to

the AERs and attempts to ascribe causality. The draft report
also does not acknowledge the GAO report on ephedra
AERs, which reviewed the same AERs and determined that
they were "poorly documented," further weakening RAND's
reliance upon reports for causation analysis. GAO also
noted that the AERs have "inherent weaknesses" and
lacked or had inconsistent information...such as the amount
of the product used, how often it was used, or for how long it
was used. These limitations were not prominently identified
in the RAND report - nor were the general limitations
associated with attempts to ascribe causality based upon
review of information obtained from a passive surveillance
system. GAO also noted that, based upon its review of
specific AERs, it was not possible to draw conclusions
regarding the "causal relationship between ingestion of the
implicated product and the adverse event observed."
Potential Product Variation. As noted above, the draft report
does not even acknowledge the possibility that certain
ephedra supplements may not be standardized and/or
manufactured according to GMPs. Accordingly, even
assuming causation (which is a major assumption), it is
conceivable that certain adverse events may have been
caused by problems with a specific product (such as having
more ephedrine alkaloids or caffeine than stated on the
label). Although it is my understanding that most
manufacturers of ephedra employ stringent quality controls,
this is still nevertheless a significant possibility that should
be reflected in the report. Attachment B contains an article
prepared by Dr. Gurley entitled "Content versus label claims
in ephedra-containing dietary supplements." Although this
article reviewed only a small subset of ephedra products,
and only reviewed a small sample-size of bottles, it
nevertheless supports the conclusion that it may be
inappropriate to assume that all ephedra products are
identical with regard to product quality.
Accordingly, even assuming causation, it should be noted
that there is no assurance that any potential adverse health
events were not caused as a result of consumers ingesting
non-standardized products that contain too much ephedrine
or caffeine. It would be inappropriate for RAND to assume
that consumption of ephedrine and caffeine within labeled
amounts is a potential problem if the possibility exists that
certain incidents may have been caused by consuming nonstandardized products. Accordingly, my strong opinion is
that as part of its review, RAND should call for FDA to
finalize dietary supplement GMPs and impose stringent
quality control requirements on ephedra manufacturers to
ensure that such products contain what they are claimed to
contain.
Haller/Benowitz Review of AERs The draft report places

A3-69

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

great importance of the review of ephedra adverse events

reports conducted by Dr. Haller and Benowitz. In fact, the
draft report compares RAND's assessment of AERs with the
Haller/Benowitz assessment. The draft report, however,
fails to mention that Drs. Haller and Benowitz subsequently
wrote a letter to the editor of the New England Journal of
Medicine (Attachment C) indicating that their review did
NOT prove causation or provide quantitative information
with regard to risk. Specifically, their letter provides the
following: Finally, our report describes a series of cases in
which the use of ephedrine-containing dietary supplements
was associated with a diverse cardiovascular events. Our
report does not prove causation, nor does it provide
quantitative information with regard to risk. A large-scale
case-control study similar to the Hemorrhagic Stroke Project
for phenylpropanolamine is needed to determine the risks
associated with these dietary supplements.
Based upon this letter and clarification, it is unclear why the
Haller/Benowitz review of the AERs is used as a baseline
for purposes of comparison. Moreover, it is unclear why
their statement regarding causation and the
recommendation of a case-control study is not highlighted as this would appear to be information that RAND should
consider as a recommendation for further research.
Importance of Background Risk - Kimmel Study. This draft
report fails to cite favorable analyses of FDA AERs including a detailed study conducted by Dr. Steven Kimmel
that was presented before the Office of Women's Health
(Attachment D). Dr. Kimmel reviewed the AERs and
determined that the number of events was consistent with
background rates in the general population. His report
highlights the importance of background risk - an issue that
should be highlighted in the RAND report. He concludes
that the AERs - even assuming significant under-reporting are not suggestive of causation. In this regard, he quoted
FDA" "it is possible that the reported serious adverse events
are reflective of coincidental background spontaneous
occurrences in the population and are not necessarily
causally related..."
OTC Drugs Containing Ephedrine and Caffeine. The draft
report does not prominently refer to the wide usage in the
United States of ephedrine in OTC drug products. The
report should include use-data for OTC drugs, and should
explain that FDA has already determined - under the OTC
Drug Review - that ephedrine is safe and effective (as a
bronchodilator) in does well over 100 mg per day (the
maximum dose level for the vast majority of ephedra
supplements on the market). In addition, the FDA does not
require such products to contraindicate caffeine ingestion
(i.e. consumers routinely ingest such products along with
coffee tea and other beverages that contain caffeine) This

A3-70

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

information must be factored into the final report, as they

have a direct bearing on any safety assessment of
ephedrine and caffeine. OTC drug use-data indicates that
the combination of ephedrine and caffeine is safe.
Scientific Data and the Landmark Six-Month HarvardColumbia Trial. The draft report acknowledges that "there
were no serious adverse events reported in these clinical
trials." Despite this, the report barely addresses this issue.
Rather, the vast majority of the report reviews and
subjectively interprets AERs that GAO, HHS, and FDA have
already reviewed. Even though the number of subjects in
the clinical studies is limited, and therefore it is conceivable
that small subsets of the population may have some
susceptibility, the clinical data is far more reliable than the
anecdotal adverse event reports and should receive greater
prominence than the AERs. In addition, the report does not
place enough significance upon scientific data such as the
Cantox Report and the landmark six-month HarvardColumbia trial (published in the International Journal of
Obesity). The Harvard-Columbia trial addresses the review
of adverse event reports, and makes suggestions regarding
future research.
The RAND report should contain a more detailed discussion
of this landmark trial - including the researchers
assessments regarding product safety and efficacy. For
ease of review, sections of the report addressing adverse
event reports, product safety and efficacy, and future
research are included below: In a FDA-sponsored analysis,
Haller and Benowitz categorized 140 adverse-event reports
based on how likely they believed the reported events to
have resulted from the use of ephedra supplements. The
difficulty in making such judgments is illustrated by the
controversy regarding their conclusions.
With millions of American consuming ephedra-containing
products it is obvious that some number of adverse events
is expected each year regardless of consumption of these
products. The real question is not whether adverse events
occur in a population undergoing treatment, but whether
these occur at a rate that is higher than that of a matched,
untreated group. This is impossible to determine from
adverse event reports alone. The randomized, placebocontrolled trial allows evaluation of cause and effect
relationships vs. coincidental events. Most clinical trials
purposely exclude individuals with pre-existing medical
conditions to avoid confounding of results. It is therefore not
justified to extrapolate results from such trials to individuals
with such exclusionary medical conditions or to extrapolate
results beyond amounts or time periods that have been
studied.
The possibility of unfavorable interactions between herbal
combinations and other medications either prescription or

A3-71

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

illicit, should be recognized and warning labels present on

herbal products should be adhered to. Some have
expressed the theory that adverse event reports may reflect
an unusually high degree of sensitivity in a small fraction of
individuals. Because of the low suspected incidence, this
type of sensitivity might not be revealed in a clinical trial, but
requires a case-control study of a very large number of
individuals. Such a study would be difficult to conduct, but
may be the only way to address the question of rare
hypersensitivity. In total, these [ephedra studies] suggest
that herbal ephedra/caffeine herbal supplements, when
used as directed by healthy overweight men and women in
combination with healthy diet and exercise habits, may be
beneficial for weight reduction without significantly
increased risk of adverse events.
In total, these [ephedra studies] suggest that herbal
ephedra/caffeine herbal supplements, when used as
directed by healthy overweight men and women in
combination with healthy diet and exercise habits, may be
beneficial for weight reduction without significantly
increased risk of adverse events. The current widespread
usage of herbal products and the increasing incidence of
obesity warrant additional clinical trials to confirm and
extend these results. (emphasis added). Finally, it should
be noted that the Harvard-Columbia Trial researchers, and
Drs. Haller and Benowitz appear to agree that in order to
evaluate the safety of ephedra, a long-term control study
would be beneficial.
The draft report in the end recognizes the futility of trying to
reach scientific conclusions from the AER's, recommending
that a case control study be done to assess risk and
recommending against further AER analysis. Nevertheless,
a detailed causality assessment was performed and
included in the draft report, and conclusions of this
assessment are presented without context. Further, the
draft report describes the involvement of the TEP in a way
makes it appear this assessment was done on the
recommendation of the TEP., when I and others thought
that there was a general agreement within the TEP and
RAND further assessing causality based on the AER's was
not recommended and would not be part of the ephedra
review.
The draft report on page 21 states that the "Highest level if
causality that could be ascribed. Was "probably" causal".
This is not a position taken by the TEP at the November
meeting; in fact it is contradictory to the TEPs position as
quoted above. There was a discussion of the
characteristics of the case reports, but not in the context as
stated on Page 21..."that would be necessary in order to
assign a classification of "probably causal." The criteria
quoted were of causality The report not only implies (also

A3-72

The causality analysis was dropped from

this revision. Regarding the involvement of
the TEP, there was a lengthy discussion of
the criteria by which we would assess case
reports, so we are surprised this TEP
member concluded the TEP agreed that
such a review was not warranted. We did
not receive any such comment from any
other TEP member, all of whom reviewed
this report.

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

on page 31) but also states that the TEP agreed with this
causality classification categorization, which was not the
case.
The classification scheme was developed after the
November meeting, and any category such as "possible
causal" or "probably causal" that suggests a causal
relationship is in contrast to the position of the TEP at the
November meeting. Terminology of "possibly" or "probably"
causal is too strong and more than suggestive of causality,
and to suggest that these terms are less than "definitely"
causal is too fine a point for readers of this report and also
an incorrect representation of the TEP's position. If the
causality assessment remains in the final report, I suggest
that all reference to this classification scheme be changed
accordingly, to omit the word "causal". The conclusions
could be characterized as weak evidence or possibly
suggestive evidence, but the words causality and causal are
too strong.
The draft report notes correctly about the AERs that "The
most important limitation is that the study design, that is an
assessment of case reports, is insufficient to warrant
definite conclusions regarding causality." Yet when it came
to assessment of individual case reports, there were definite
conclusions that the AERs prove ephedra to be unsafe. The
result is a misleading presentation of the available
information.
For the reasons explained above, I feel the report requires
major revision and subsequent further review by the TEP.
Because I have been focused on the fundamentals of the
reports as described above, I haven't even considered the
comments on details that are included in the draft report.
The weakness of the FDA AER database must be better
addressed, and the causality analysis should either be
removed from the report or substantial revised to, among
other things, provide the necessary context and to change
the classification of the AERs to avoid using the terms
"causal" and "causality". The fact that the safety section is
dominated buy the AER analysis reduces the credibility of
the section and indeed the whole part of the report.
However, in the opinion of the reviewer, those conclusions
regarding the case reports are limited by a combination of
the conservative causality assessment criteria and the
limited medical records and toxicology data available for
most case reports. For example, hypertension was defined
as a systolic pressure in excess of 180 and a diastolic
pressure in excess of 105. Also, no consideration appeared
to have been given to the contribution multi-component
ephedra-containing dietary supplements might have had in
those individuals with underlying cardiovascular or
cerebrovascular disease. I think it is generally accepted
among the medical and scientific community the presence

A3-73

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

of sympathomimetic agents could potentially exacerbate the

likelihood of adverse events in such populations? I would
think most clinicians would factor such information into their
differential diagnosis rather than dismiss them altogether.
Adverse Events Reports: One limitation with your approach
taken in evaluation of the adverse events reports made to
the FDA is that your causality algorithm does not include an
assessment of whether ephedra played a contributory role
on the adverse event. Because ephedra is available as a
dietary supplement, it is likely that many persons taking
these products are not using them under doctor's
supervision, and may have medical contraindications to
their use. Therefore, the role of underlying disease becomes
a crucial factor in causation assessment, particularly when a
potential risk factor often goes undetected (i.e. essential
hypertension, structural heart defect), or when a condition is
omitted from the ephedra product label warning (i.e. family
history of premature CAD, sickle cell trait).
Two AERs that you assess as no higher than possibly
causal illustrate this point. AER 12485 did indeed have a
moderate degree of coronary artery disease detected at
autopsy. However, he was reportedly in good health without
history of angina, and had been jogging regularly without
adverse effects. Because he collapsed suddenly after
returning from jogging, we felt this was a primary arrhythmic
event due to ephedra. Similarly AER 12843 was a healthy,
adolescent who had participated in competitive sports for
many years. She had appeared to have been wellcompensated for a serious underlying coronary artery
abnormality that was clinically undetected since birth. Only
with use of Ripped Fuel, did she suffer a catastrophic
cardiac event resulting in death. We felt that the cardiac
stimulant effects of ephedra resulted in myocardial ischemia
in this case.
It would be helpful to specify what degree of pre-existing
This case was reviewed by a cardiologist
coronary artery disease would constitute a significant risk
who made this judgment.
factor to result in myocardial infarction or sudden death in
the absence of stimulant use, thereby ruling out ephedra in
the causation assessment. (page 32 of chapter 2
methodology). In the case of AER 14530 (page 63), I would
disagree that 20-30% stenosis would be significant enough
to result in acute M.I. in a 43-year-old female smoker
without a significant contributory effect from the ephedra
alkaloids in Metabolife.
On page 59 I suggest the authors be slightly cautious with
We have revised this language.
their use of language such as "probably causal" and "no
other possible explanation." The latter phrase is particularly
troubling. What they mean is no other explanation that they
could identify. Similarly, on page 69 they state that there are
a certain number of cases of serious adverse events that

A3-74

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

"cannot be explained by causes other than ephedra use."

While I do not deny that it is extremely likely that many
cases of adverse events happen due to ephedra use, simply
because we do not have in our hand an explanation of why
an event occurred other than a particular explanation under
consideration, does not mean that the particular explanation
under consideration is the correct one.
Because of the paucity of large randomized trials, evidence
concerning stroke and ephedra by necessity consists of
analysis of case reports. "An assessment of case reports
is insufficient to warrant definite conclusions regarding
causality." (p.110) Nevertheless, arbitrary criteria are used
to define "probably cause": documentation that a stroke
occurred, that ephedra was used, and that there was
exclusion of other potential causes. The definition may be
too liberal. Of ischemic strokes in relatively young adults
(i.e. those <50 years old), perhaps 20-35% are "idiopathic"
despite thorough evaluation. The definition implies that all
idiopathic strokes would be classified as "probably causal" if
ephedra was used in any dose in proximity to the event.
Given the frequency of idiopathic stroke, many (perhaps
most) neurologists would consider "possibly causal" to be a
better designation in this situation.
Are there specific clinical circumstances in which the
relationship of ephedra use and idiopathic stroke could be
certain? Perhaps if acute, striking elevation of blood
pressure were known to precede the stroke onset of of
angiographic features characteristic of vasospasm were
present in the abscence of migrane? Arbitrary to be sure,
and not very helpful.
The evidence for effectiveness supports the conclusion.
Except for AERs, however, little evidence of toxicity is
actually provided, and evidence of safety has been largely
ignored. No evidence is provided to even suggest "a
possible causal role of ephedrine-containing dietary
supplement in rare, but serious events," let alone extremely
common events such as heart attack and stroke. Even
critics of ephedra have concluded that the clinical effects of
pharmaceutical ephedrine, and the ephedrine contained in
herbal preparations, are indistinguishable. Gurley states
that the increased incidence of ma huang toxicity does not
stem from differences in the absorption of botanical
ephedrine compared with synthetic ephedrine." Haller and
Benowitz, in their most recent publication, conclude,
"Botanical stimulants have disposition characteristics similar
to their pharmaceutical counterparts..."
The Cantox Report, and the Report of the Expert Panel of
the EEC reached similar conclusions. Since there are no
real differences, studies demonstrating the safety of
pharmaceutical ephedrine and pharmaceutical ephedrine in
combination with pharmaceutical caffeine should not be

A3-75

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

excluded when considering the safety of herbal equivalents.

The explanation most frequently offered for alleged cases of
ephedrine-related stroke is drug-induced blood pressure
elevation, this in spite of the fact that no clinical trial, of any
duration, has ever demonstrated that a clinically significant
effect on blood pressure exists. Indeed, the studies that
have addressed this question. including the most recent
paper by Drs Haller and Benowitz, have shown diminishing
cardiovascular effects over time. In other words, if
dangerous blood pressure elevations do not occur with the
first dose ephedrine, they are even less likely to occur with
prolonged dosing. These studies should be included in the
RAND review of ephedra and should be used to address the
question of potential increases in blood pressure and other
safety issues.
Ephedrine has been studied in more than 50 double blind,
placebo-controlled clinical trials, some of long duration. A
far from exhaustive literature search produced the attached
list of peer-reviewed, published, clinical trials. Most have
compared ephedrine to placebo, and to other
sympathomimetic drugs used to treat asthma. However,
others have involved smoking cessation, sexual function
and athletic performance. Nearly a dozen of these trials
involved caffeine/ephedrine combinations using does
exceeding those found in herbal supplements. In total,
more than 2000 individuals have been enrolled in these
trials. In several studies there was even continuous cardiac
monitoring in middle-aged patients with known heart
disease; no effect was observed. No clinically significant
episodes of toxicity were reported. Including these and
other studies on ephedrine that have been excluded from
the RAND review will increase the power of the safety
calculations that can be derived from clinical data.
One of the major limitations of the report was the
composition of the TEP and the reviewers who made
subjective assessments of the AERs. Given the importance
placed on assessment of AERs, it is unfortunate that no
pathologist was included in the view or on the panel. The
lack of expertise is obvious from the comments made about
the individual AERs. The failure to provide information about
any potential conflicts of reviewers also detracts from the
study. Why were the findings of Expert Panel of the
Ephedra Education Counsel not considered? The analysis
of this panel was in some ways unique, as it is the only
consensus opinion on ephedra safety. In addition, this
panel conducted the most comprehensive review of the
ephedra AERs to date, and yet the causality assessment,
which conflicted the findings of the draft report, are not even
mentioned. If RAND believes that the EEC review and
analysis was, in some way scientifically flawed, then the
reasons for that belief should be stated.

A3-76

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

The danger of drawing conclusions from AERs without a

control group can be illustrated by examining data from
randomized trials in which participants are blinded to
whether they are receiving the study treatment or inactive
placebo. In the placebo group of a recent randomized trial,2
there was an increase in ventricular couplets (extra heart
beats) at the 4th week of the study 9from 3% at baseline to
around 14% at week 4). This is, of course, not due to the
placebo, which is inactive, but rather just spontaneous
ventricular couplets that occurred by chance. However, if
these participants had been given ephedrine alkaloids in an
uncontrolled study (without placebo), this change could
have been attributed, incorrectly, to the ephedra. That is,
these could be AERs that were attributed to ephedra. In
fact, in the controlled trial, a similar increase in ventricular
couplets was not seen in the ephedra/caffeine arm.
Another example is the 15-year-old female (case 12843)
with Bland-White-Garland syndrome who died while playing
soccer. This disorder has been associated with sudden
death after physical exertion. In the absence of a unique
pathologic process, it is almost never possible to establish a
causal association on the basis of adverse event reports.
There is nothing pathologically or diagnostically unique
about the adverse events noted in the ephedra database
(e.g., myocardial infarction, stroke) that allow one to
distinguish a spontaneous event from one caused by use of
Ephedra products. In fact, a review of all autopsy data from
ephedra AERs by Dr. Grover Hutchins, a Professor of
Pathology at the John Hopkins University School of
Medicine and member of the Expert Panel of the EEC,
concluded that "The pathology data available do not show
any pattern consistent with ephedrine alkaloid-containing
dietary supplements as a cause of death."5
Similarly, 10 participants in the ephedra/caffeine group
(12% of these participants) withdrew because of
cardiovascular symptoms (palpitations, elevated blood
pressure, arrhythmias). If there were no control group,
these also might have been attributed to the
ephedra/caffeine combination. However, the same
proportion of participants in the placebo group (13%)
withdrew for the same reasons. The withdrawal rate in the
ephedra/caffeine group thus was consistent with the
background rate of these events in a placebo group
unrelated to ephedra use. A second limitation of adverse
event reports is that other potential causes of the event are
often present, making it extremely difficult to determine if an
event truly is related to the exposure.
As an example, it is well established that physical exertion
can trigger myocardial infarctions and cardiac arrest (up to a
74-fold increase in the risk of sudden death, according to a
recent report in the New England Journal of Medicine6)

A3-77

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

Therefore, a case such as the 38-year-old man with three

vessel coronary disease and a dilated heart who died after
jogging could very well have been related only to the
physical exertion and not the ephedra. According to his wife,
his heart problems had been known for at least five years.
In addition, this man had been taking an ephedra product
for one year without apparent ill effect.
The third limitation of AERs is that their interpretation
remains subjective. Even experts in the field will disagree
about the possibility that an event may or may not be due to
an exposure. The RAND report (Table 22) displays a
comparison of the causality assessment between their
review panel and a published report by reviewers for the
FDA.3 In only two out of 20 cases that both groups
reviewed was their agreement about the highest level of
possible causality. For example, with respect to the 38year-old discussed above, the Haller and Benowitz review
recorded this event as "Definitely or probably related" to
ephedra while the RAND report classified it as only
"Possibly causal."
Another review of AERs performed by Dr. Theodore Farber
and Dr. Norbert Page, members of the Expert Panel of the
EEC, reveals similar disagreements. The two "probable"
cases reviewed by EPC were rated as "Low Possible" (case
12980) and "Improbable" (case 13418) by Drs. Farber and
Page. The fact that the etiology of events can be debated
simply illustrates the substantial limitation of case reports
that lack a comparison control group. This echoes reviews
done by FDA and its consultants in which agreement about
causality was poor. For example, two consultants from
FDA, Drs. Ricaurte and Stoll, reviewed 28 AERs related to
neurological events. Dr. Ricaurte classified eleven cases as
"attributable" while Dr. Stoll classified only five as
"attributable."5 Only two of the consultants findings
overlapped - that is, there were only two cases that both Dr.
Ricaurte and Dr. Stoll agreed should be categorized as
"attributable."5 This disagreement is not a flaw of the
reviewers, but rather a flaw of AERs.
A fourth limitation of AERs is that ingestion of the substance
in question cannot always be substantiated. For example,
case 10276 is a 32-year-old with an enlarged heart who was
found dead in his truck. Although a product that contains
herbal ephedra was found in his truck, so were several
bottles of cold medications, including Nyquil. Toxicology
revealed no ephedrine, but did identify pseudoephedrine
and doxylamine, both components on Nyquil. Thus, there is
no evidence that this person even ingested the herbal
ephedra. A similar case (13096) revealed no ephedrine in a
toxicology screen again suggesting that the man had not
ingested ephedrine around the time of death. Equally
importantly this man died of a disease that appeared to run

A3-78

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

in his family (aortic dissection), including an 18-year-old

niece.
In summary, AERs cannot be used to assess causality. As
stated by several authors with experience in interpreting
adverse event reports for the FDA: "It is probably
impossible to do comparative analyses employing ADE
[adverse drug event] reports for drugs that have received
extensive publicity in the mass media for an adverse
event..."7 In fact, the FDA's Center for Drug Evaluation and
Research pointed out, in a February 10, 2000 memorandum
concerning ephedra products, that "it is possible that the
reported serious adverse events are reflective of
coincidental background spontaneous occurrences in the
population and are not necessarily causally related to [the
use of dietary supplements containing ephedrine-type
alkaloids]." The RAND review notes this as well, stating that
"The most important limitation [of their assessment of
adverse events] is that the study design, that is an
assessment of case reports, is insufficient to warrant
definite conclusions regarding causality." They list this
limitation as one of the "most important...gaps" in the current
knowledge-base.
The RAND review also states that "Disentangling the
relative importance of the pre-existing condition and the
ephedra use is not possible." They state further that
"Continued analysis of case reports cannot substitute for a
properly designed study to assess causality. A case control
study would probably be the study design of choice." Their
Technical Expert Panel also "judged that case reports alone
would be insufficient to establish definite causality between
ephedra use and serious adverse events." Because of these
limitations, terms such as "probably causal" and "possibly
causal" in AER reviews are potentially misleading (see, in
particular, the "Conclusions" section of the "Structured
Abstract," page vi, and the "Conclusions" section, page
112). They represent only reviewers' assessment of
causality based on uncontrolled data and subjective
assessments. Although these terms are often used in
scientific publications, their use in the RAND report may
suggest a level of evidence that does not exist from the
current data. These statements, therefore, should not be
taken out of context.
RAND's stated limitation that "Definite causality cannot be
determined from case reports" must be kept in mind when
interpreting this report. It is also critically important to
remember that case reports can produce false signals of
cause and effect.8,9 Most importantly, because of the
limitation of AERs, it is unclear why the RAND report states,
in the Structured Abstract, that "These [sic] is sufficient
evidence to suggest a possible causal role of ephedracontaining dietary supplements in rare but serious adverse

A3-79

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

events, particularly cerebral hemorrhage." With respect to a

possible causal role of ephedra in adverse events, RAND
acknowledges that AERs are not sufficient to draw
conclusions about causality, consistent with the known
limitations of AERs discussed above. In addition, this
comment is puzzling given that their "Conclusions" section
(ages 112-113) does not state this at all.
With respect to cerebral hemorrhage, the only comment in
the body of the report on cerebral hemorrhage refers to a
case-control study of phenylpropanolamine (PPA) and
cerebral hemorrhage.10 However, the RAND report refers
to this study only as an example of case-control
methodology that could be applied in the future to ephedra.
The report does not discuss the PPA study further. In fact,
this study has been heavily criticized. Despite this, there
was not a formal review of this study by RAND (and there
were no members of the technical expert panel listed with
expertise in epidemiology to perform such a review). In
addition, PPA and ephedra have different chemical
structures and different pharmacological activities. Finally,
the RAND report does not mention that the PPA report also
presented data on non-PPA, ephedrine-alkaloid containing
products. These agents included medications that
contained pseudoephedrine hydrochloride, phenylephrine,
ephedrine, and epinephrine. In the report, there was similar
prevalence of use of these products among those with and
without hemorrhagic strokes.
Although this is not a definitive analysis, it suggests that
there was no association between these ephedra-containing
products and hemorrhagic stroke. Therefore, the statement
in the Conclusions section of the Structured Abstract of the
RAND report is inconsistent with currently available
scientific data. In summary, the RAND review supports the
use of herbal ephedra and caffeine for weight loss, an effect
that may have beneficial health consequences. The report
also suggest, from controlled studies, that adverse events
following ephedra use are, at most, rare. (The should not
imply that the events are even causally related to ephedra
us.) Most importantly, because of the reliance on AERs, the
report cannot establish a causal effect of ephedra on
serious adverse events.
Review of Anecdotal Adverse Event Reports - Limitation of
Review. On pages 109-110, the report identifies potential
limitations associated with the review of anecdotal adverse
event reports. These limitations are buried at the end of the
report, rather than being incorporated into the appropriate
sections of the report (as per prior AHRQ reports, such as
the Garlic Report - see Section IX, herein). Moreover, a
number of limitations are not prominently identified,
including but not limited to: a) The poor quality of the data
and information contained in the anecdotal adverse event

A3-80

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

reports; b) Inherent problems associated with voluntary

reporting systems; c) The number of individuals in the
general population who experience the adverse events
identified (i.e. background risk); d) The possibility that
certain products may not have been standardized and/or
manufactured according to Good Manufacturing Practices
("GMPs") - resulting in potential adverse events that have
nothing to do with ephedra or caffeine when consumed in
recommended amounts. Specifically, in the absence of
standardization and quality control, it is conceivable that
certain products may contain far more ephedra or caffeine
(or other constituents) than indicated on the product label.
This possibility must be considered when evaluating
anecdotal adverse event reports. In the absence of
identical product identity, any general conclusions regarding
ephedra and caffeine are inappropriate and highly suspect
based upon adverse event reports (see the AHRQ Garlic
Report for an appropriate way to address this issue). In my
opinion, RAND should strongly support immediate issuance
by the FDA of dietary supplement GMPs and should
endorse stringent quality control measures to ensure that all
ephedra supplements contain what they are claimed to
contain; e) The possibility that the consumer abused or
misused a product by ingesting more than the
recommended amount - or that the consumer ignored
detailed product warnings and contraindications. RAND
should emphasize the detailed warning label contained on
the vast majority of ephedra supplements - and should
acknowledge that there is little way to know from anecdotal
data whether a consumer abused a product (either
intentionally or more likely inadvertently); f) The possibility
that the anecdotal adverse event reflects chance,
coincidence , or confounding factors - including but not
limited to the possibility that ingestion of a different product
or substance led to the stated event.
I I. Review of Anecdotal Adverse Event Reports - B.
Ascribing Causality to Adverse Event Reports - 1. Reliance
on Unpublished Article. In order to establish a framework
for analyzing the adverse event reports, the draft report
relies upon an unpublished article written by Cynthia
Mulrow, M.D. Reliance upon an unpublished, non peerreviewed article to establish the framework for a critical
portion of the report is entirely unacceptable. AHRQ studies
have not, to our knowledge, ever relied upon an unpulished
article to establish the framework for this type of analysis.
In addition, reviewers such as myself have no way of
analyzing the article - thereby defeating one of the primary
reasons for review of the draft report.
II. Review of Anecdotal Adverse Event Reports - B.
Ascribing Causality to Adverse Event Reports - 2. Failure to
Review Factors Critical to the Interpretation of Anecdotal
Adverse Event Reports Table 4 of the draft report

A3-81

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

summarizes the various methods researchers use to assess

causality from adverse event reports. The following nine
factors are identified: a. Temporal relationship. b.
Dechallenge response. c. Rechallenge response. d. Could
there be an alternative explanation? For example,
dehydration or consumption of other toxic substances. e)
Prior reaction to same substance. f) Dose response. g)
Objective evidence of adverse event. h) Previous conclusive
reports. i) Definition of substance. Despite the report's
reference to these nine factors, it is my understanding that
the report concludes that events are "probably causal"
based upon a review of only two factors - a and g. The draft
report does not explain why RAND believes only two factors
out of nine can be used to ascribe degrees of causation to
anecdotal adverse event reports.
II. Review of Anecdotal Adverse Event Reports - B.
Ascribing Causality to Adverse Event Reports - 2. Failure to
Review Factors Critical to the Interpretation of Anecdotal
Adverse Event Reports (cont'd). The draft report also
indicates that three factors are used to determine if an event
is "probably causal": a) Reasonable certainty that the
adverse event occurred. b) Reasonable certainty that the
patient took ephedra in a dose and timing compatible with
the known pharmacology of ephedrine. c) An adequate
evaluation must have been done to rule out other potential
causes for the adverse event. The third factor (factor c,
above) is exceptionally problematic from a scientific
perspective. The report acknowledges that the third factor
is subjective. Specifically, in an effort to rule out other
potential causes, the report indicates that such a
determination was made by determining if the subject had a
pre-existing condition that was identified in the adverse
event report.
Adverse Event Reports. While CRN acknowledges that the
judgments made about the AERs were, overall, much more
conservative than those made by other reviewers, there is
concern that, in some cases, a much more likely
explanation was evident, but still possible causality was
assigned. Some examples follow, although they are not allinclusive:
1. Case 10276. A deceased truck driver was found with
cold tablets, Nyquil and Vicks Formula 44, in addition to
ephedra-containing supplements. The toxicology screen
was negative for ephedrine, but positive for
pseudoephedrine. The much more obvious and likely culprit
here would be one of the pseudoephedrine-containing cold
formulas, since ephedrine is the dominant alkaloid by far in
ephedra products, and the clearance rates for the alkaloids
are roughly the same.
2. Case 12843. A 15 year old died of a congenital
abnormality of the left coronary artery No ephedrine was

A3-82

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

reported in her system. How could this possibly be causal?

3. Case 10874. A woman with considerable substance
abuse problems and some use of ephedra supplements for
weight loss tested positive for phenylpropanolamine (PPA),
but not for ephedrine or pseudoephedrine. The former is
but a minor constituent of ephedra and products derived
there from, but were a common primary component of a
number of OTC weight loss products until recently. Such
products are a much more likely source of the PPA than the
possibly causal ephedra supplements.
We have several concerns about the way the information is
presented in the sections related to safety. It is AHPA's
position that the report's safety assessment section reviews
case reports from a passive event reporting system without
fully and redundantly disclosing what has already been
determined about the nature of the FDA's current AER
system. Appropriate disclosures include, at a minimum, a
reference to the GAO report on the subject and recognition
that the Special Working Group of the office of Special
Nutritional (FDA: Food Products Containing Ephedrine
Alkaloids, Washington D.C., October 11th-12th, 1995)
explicitly stated that such a system cannot, by it's nature,
show causality.

We acknowledge that the case reports

cannot show causality. We do not need to
discuss the findings of other with respect
to the adequacy of the FDA AER system.
We assessed the information we did
receive using explicit criteria, and our
findings are reported.

AHPA recognizes that limitations in the clinical trial data

lead one necessarily to consider case reports for an
assessment of serious adverse events. The fact that there
are no serious adverse events reported in any of the clinical
study should however be stressed, even as it is identified as
of insufficient statistical power to detect a rate of serious
adverse events. This fact should be repeated at the
Structured Abstract and in the Conclusions, for example,
and the total number of patients in these studies (is that
2319 in the intervention groups?) should be identified. In
addition, Table 17 should be expanded to include each of
the serious adverse events that are subject to safety review
in the draft (e.g. death, myocardial infarction, and stroke)
and the number "zero" should be entered in both the
placebo and intervention columns, if that was in fact the
published observation.

We do not favor, as a general rule, adding

rows or columns to a Table when all the
entries in each cell will be the same. Such
information can more expeditiously be
conveyed in the text.

The Draft contains an extensive review of the specific AER

case reports. This inherent emphasis presents an
unbalanced appearance with respect to the intent of the
original key questions. In comparison, assessment of
efficacy is presented in a much more summarized fashion.

We cannot change the amount of space

needed to describe what we did.

In arriving at criteria for judging the causal relationship in

case reports of adverse effects from ephedra, the concept
of biological plausibility is conspicuously absent. All else
bring equal, adverse events that are biologically plausible
(consistent with the mechanism of action of the drug in
question) are more likely to be causally related to drug use

A3-83

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

than those, which are not. Sudden death, myocardial

infarction, and stroke are biologically plausible toxic effects
of ephedra. This should be taken into account when
interpreting the data.
Clearly there is no right or wrong answer with regard to how
much data is needed to judge the causality of an adverse
event case report as "probable" or "possible" but I believe
that the criteria used in this review have resulted in
conclusions that that are understated. There is nothing
wrong with the criteria per se, but using the term "probable"
for a death that has fulfilled every review criteria (except rechallenge which is by definition impossible) understates the
quality of the data and implications of the case. Similarly,
the term "possible" for cases that have satisfied several but
not all criteria makes it sound like these importance of these
cases should be minimized, which I do not think is the intent
of the report. For example, requiring negative angiographies
to support causality for myocardial infarction has a rationale,
but will necessarily exclude many or most cases because
not all patients have this procedure.
The importance of this wording is illustrated by the
comparison of the Benowitz ephedrine data and your
group's reanalysis of it, which would have downgraded so
many case reports as to make the report unpublishable.
Instead, It was published and shows a remarkable similarity
in adverse event profile with the current report. This
congruence of findings is in fact some of the strongest
literature support for the conclusions of the current report
regarding toxicities from ephedrine and ephedra, and these
two reports suggest just that.
The alternative to changing the terminology of the report is
to provide additional commentary on the interpretation of the
findings; that, in a view of 1) biological plausibility, 2) the
considerable number of case reports emanating form a
spontaneous reporting system, 3) similar toxic effects of
pharmaceutical ephedrine, and 4) similar toxic effects of
phenylpropanolamine, the findings of the current review are
highly suggestive of a relationship between herbal ephedra
and serious adverse events such as sudden death,
myocardial infarction, and stroke.
Discussion at the NIH requested that the title "possibly
causal" was misleading and should be retitled to indicate
more accurately the Rand staff interpretation that there is no
proof of causality and while causality is possible, it is not
probable.
Study Selection. Study selection was not appropriate.
Partly for the reasons stated above, but also because of the
reliance placed upon passively collected anecdotal data.
The mere fact that existing clinical trials contained "too few

A3-84

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

(subjects) to allow adequate statistical power to access the

rate of serious adverse events," does not make AER
analysis any more reliable or probative. In fact, case reports
cannot be "relied upon to assess serious adverse events,"
except, perhaps for the occurrence of rare disorders such
as coronary artery dissection.
The decision to include an analysis of AERs is particularly
puzzling, given that the TEP chose to reject the Haller and
Benowitz analyzing the same AERs (see "rejected articles"
#195, record #116)! Heart attack and stroke are common
disorders in our society, and thousands of ephedra product
users would be expected to experience vascular events,
even if ephedra did not exist. Analysis of AERs for common
disorders, which are even more frequent among the
overweight, is virtually guaranteed to show a connection
with ephedra use, even if no such connection exists (for
example, see the August 1 article, "Obesity and the Risk of
Heart Failure" in the New England Journal of Medicine).
Appraisal of Studies. Important parameters that could alter
study results have not been systematically addressed. The
brief discussion of obesity is confined to generalities.
Obesity is a prothrombotic disease. [1]. Overweight people,
presumably the majority of ephedra supplements users, are
at greater risk for sudden cardiac death (SCD), and heart
disease [2]. The report fails to provide any sort of
epidemiologic prospective, leaving the false impression that
the occurrence of these disorders among ephedra product
users is somehow surprising or unanticipated. In fact, when
the GAO wrote its highly critical analysis of the FDA's
proposed rule on ephedra products, one of the issues raised
was FDA failure to account for the reality that "there is
almost always an underlying background rate for any clinical
event in a population, regardless of whether there was
exposure to a particular product..." The RAND report states
that 3 billion servings of ephedra were sold in 1999.
Assuming that 3 servings are used per day for 12 weeks (as
Haller and Benowitz do in the NEJM paper), then there were
12 million users.
The accepted rate for sudden death, heart attack, and
stroke in the U.S. is 0.1, .5, and .2 percent per year
respectively [3]; which means that even if
ephedra/ephedrine has absolutely no relationship to any of
these disorders, 12,000 cases of SCD, 60,000 cases of
myocardial infarction, and 24,000 cases of stroke would still
be expected among ephedra users each year. Not
providing this information to general readers paints a
completely misleading picture and leads to a misinformed, if
not false, impression of relative risk. It also repeats the
same FDA error already criticized by the Government
Accounting Office.
Data Collection No effort is made to reduce bias in the

A3-85

The reviewer is incorrect The autopsy was

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

data collection, or even to assure that the data is valid. This

is immediately apparent in the discussion of the first AER.
The history given for AER #13914 is simply incorrect. The
autopsy was, in fact, included in the docket (copy attached).
The heart was actually examined by a consultant cardiac
pathologist, and that report states "the cause of death may
be attributed to myocarditis in the absence of other
demonstrable cause." The summary in the report misstates
the data available and misrepresents the conclusion in the
report.

not included in the docket we received.

Data synthesis. Limitations of the review process are not

adequately stated. Many, if not most, of the interpretative
problem seem to be the result of the medical experts on
whom RAND relied upon the review the AERs (it is not clear
from the draft report who reviewed which AERs). Medicallegal death investigation is customarily performed by
pathologists with specific training and expertise in sudden
death investigation, yet it appears that not a single
pathologist was included among the reviewers. As a
consequence, many of the AERs were almost certainly
misinterpreted. I do not have access to many of the AER
files that RAND reviewed, so I cannot comment specifically
on RAND's subjective assessments of causality in most of
the cases. However, the errors and misinformation in the
AERs that I can check show clearly this experienced death
investigators were not involved in the project.
For example, Case #14390, classified by the panel as
"probably causal", was said to have had a "shunt" in place.
It follows that the brain could not, as the report states,
possibly have been "normal". For one thing, there would
have been a shunt in place, which is decidedly not normal.
There would have been tissue reaction around the shunt,
both the heart and brain. If the shunt had been placed for
traumatic injury, trauma residuals would have present. If
the shunt had been placed for traumatic injury, trauma
residuals would have been present. If the shunt was for
congenital hydrocephalus, it is quite likely that the
abnormalities associated with the Arnold-Chiari
malformation would have been present. Sudden cardiac
death secondary to acute obstructive hydrocephalus is well
recognized by forensic pathologists [4, 5].
Even if a pathologist was on the AER review panel, the
methods section is so vague that it is not clear whether the
pathologist would have been asked to review this AER.
One can only conclude that the panel of reviewers do not
know the accepted causes of sudden death in young people
with ventriculatrial shunts. Evidence of potential bias in the
AER reviews is provided by the AERs chosen as "probable"
and "possible" as well as by the summaries of the AERs
contained in the draft report. For example, Case #12485
was classified as "possible causal " and the presence of

A3-86

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

"mild cardiomegaly," noted. The autopsy report in the

docket shows a heart weight of 490 grams, two standard
deviations above predicted [6]. No one with training in
vascular pathology and sudden death investigation would
call that degree of enlargement "mild." Cardiac enlargement
is a recognized risk factor for SCD [7], over and above the
severe coronary artery disease that was also present [8].
Referring to such an important pathologic finding as "mild"
can only be explained in two ways: either the reviewers
were unaware of the significance of this abnormality, or they
were trying to minimize the finding because it did not fit a
bias towards finding evidence of ephedrine toxicity.
Preconceived bias is strongly suggested by the inclusions of
AER #12722, a child who died of a type of congenital heart
disease where the left coronary artery arises from the left
pulmonary artery (not vein, as stated in report). No
ephedrine was detected in tissues analyzed by the FDA,
and there was extensive scarring of the myocardium,
reflecting early episodes of healed myocardial infarction. By
their very nature, the morphologic changes detected, which
almost certainly were the cause of death, antedate by
weeks or months the alleged history of ephedrine ingestion.
Classification of this case as "possibly causal", violates the
report's own stated criteria, which specifically reserve the
"possible" category for those cases where "another
condition by itself could have caused the adverse event, but
ephedra use may have helped precipitate the event." No
ephedrine was detected at autopsy, and anatomic changes
were present that had to have occurred weeks or months
before the first use of any ephedra product is even alleged.
Bias is also suggested by the discussion of
phenylpropanolamine. The section on Phytochemistry
correctly states that the phenylpropanolamine (PPA) content
of ephedra is low. But then in the discussion of AER
#10874, the report also states "there is a described
association between PPA and cerebral hemorrhage, PPA is
also a component of some herbal ephedra." Had a
balanced presentation been intended, the report would have
provided the additional information that the most PPA ever
detected in a serving of an ephedra supplement was half a
milligram, and that in studies with volunteers, 50 mg of PPP
is needed just to modestly raise blood pressure [9].
The emphasis of the potential linkage between PPA and
ephedra in the report is distressing for three reasons.
Firstly, it reiterates the same unproven argument
propounded by Drs. Haller and Benowitz in their most
recent publication (Haller, et al., Clin Pharmacol Ther 2002;
71:421 432). The unquestioning repetition of an unproven
hypothesis tends to lend legitimacy to that hypothesis, even
in the absence of evidence, thereby detracting from the
value of the report Secondly and quite improperly the

A3-87

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

report nowhere discloses that this unproven mechanism of

toxicity is, in fact, an unproven theory propounded by one of
the TEP members. Thirdly, the report makes no mention of
the dispute between professional epidemiologists over the
validity of the study that lead to the withdrawal of PPA in the
first place.
Indeed, the entire discussion of PPA in this report, the
failure to mention the PPA content of the average herbal
supplement, the failure to mention the fact that ephedrine is
minimally metabolized to (-) norephedrine, and the failure to
mention existing disagreements among epidemiologists
about PPA, is simply not scientifically supportable. A
consumer would have to simultaneously ingest 100 servings
of an ephedra supplement in order to receive enough PPA
to minimally raise blood pressure, and probably twice that
amount to cause a clinically significant increase. In AER
#14019, death was attributed to "dissection of a left anterior
coronary artery" in a 26-year-old woman.
This is not a supportable conclusion. Coronary artery
dissection has never been reported in an ephedrine user, or
even in amphetamine abusers (a drug to which ephedrine
has frequently been compared). Almost all reported cases
of spontaneous coronary artery dissection involved young
women following childbirth, or in cocaine users. Had the
decedent just delivered? Was she a cocaine user, or both?
Was toxicology testing performed? This case and others
classified as probable and possible are examples where
"crucial information is missing" and they should have been
classified as "insufficient".
Safety assessment. As mentioned above, the panel should
also include the non-placebo controlled and nonrandomized trials in the safety assessment. The information
obtained from such trials is superior to that from case
reports. Page 58, 4th section: Here it is stated that patients
taking pharmaceuticals outside of clinical trials may have a
greater risk of certain adverse events than patients selected
to participate in clinical trials. I strongly disagree. In all the
clinical trials we have conducted, which have been
conducted in Denmark, it is quite normal that the patients
are referred by general practitioners or hospital departments
because they have a high degree of overweight (are
typically obese, with a body mass index of 29-40) and suffer
from complications to the obese state. This may not
necessarily be ischemic heart disease, heart failure or type
3 diabetes, because these subjects will typically be
excluded, but patients with pre-diabetes, dyspnoea,
osteoarthrosis in knee or hip, etc. In addition, one of the
large trials was conducted on the hypertensive obese
patients (Ingers, et al.).
In contrast, individuals in the community taking preparations
containing ephedrine will typically be less overweight and be

A3-88

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

generally healthier. They will be less likely to experience

serious adverse events than subjects in clinical trials I think
the conclusion reached by the panel should therefore be
reversed.
The Danish Experience. It is quite natural that the panel
has received the list of case reports from FDA's office of
nutritional products, labeling and dietary supplements.
However, why did the panel not ask the Danish FDA for
their full report of collected adverse events during the 12
years from 1990 to 2002 where an ephedrine/caffeine
prescription compound has been on the market in
Denmark? This is a substantial body of experience that
could give more valid conclusions than those received the
American FDA alone.
During the last 8 years, the defined day doses have ranged
between 3.6 and 4.6 per 1,000 inhabitant/day in Denmark.
It also means that the Danish Drug Administration has, in its
surveillance program, obtained anecdotal data regarding
reported side effects from General Practitioners and other
Doctors in Denmark. The post market surveillance program
is very effective in Denmark and there are 134 reports of
side effects, but they are all very mild side effects and the all
the well-known side effects we know from the
pharmacological action of ephedrine/caffeine. They include
tremor, insomnia, palpitations; side effects from the use of
ephedrine/caffeine even though Denmark has had a
substantial amount of sales and ten years experience.
First there is no control group. Because there is often an
underlying baseline risk of disease unrelated to exposure to
a product, there will be events reported in people exposed
to that product that are in no way associated with use of the
product. Given the large number of users of ephedra to a
product, there will be events reported in people exposed to
that product that are in no way associated with use of the
product. Given the large number of users of ephedracontaining products,4 there will be events among users that
are coincidental with the use of ephedra (i.e., not causally
related) even in the absence of other explainable causes of
these events. In an analysis performed by the Expert Panel
of the Ephedra Education Council (EEC), an estimate of the
rate of serious events in ephedra users was compared with
the rate of events expected in the general population. 5
Although this analysis was not designed to rule in or out a
possible cause-and-effect relationship between ephedra
and the outcomes evaluated, it did suggest that the adverse
events reported among ephedra users may very well
represent simply the background rate of events expected
among such a large number of users of ephedra-containing
products, unrelated to ephedra itself. This was true even
under assumptions that inflated the risk from ephedra and

A3-89

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

even assuming the events reported to FDA represented a

small percent of the adverse events occurring among
ephedra product consumers.
I. Research Parameters - Overview. Based upon my review
of the draft questions provided to RAND, it is my
understanding that RAND was asked to provide a sciencebased review in order to identify gaps in the data relating to
ephedra safety and efficacy. Identification of these data
gaps was deemed essential to prepare an agenda for future
research. Despite this charter, RAND instead
acknowledged that it engaged in a subjective review of
anecdotal adverse event reports in order to ascribe potential
causality to such reports. As explained below (see Section
II, herein), the inherently subjective nature of such a review
by definition reduces the objectivity of the report and may
lead one to question the proposed research agenda. It is
my strong belief that RAND should issue a report that is
entirely objective and science-based. A subjective review of
anecdotal adverse event reports that attempts to ascribe
potential causality to such reports in my opinion provides no
meaningful benefit with regard to identifying data gaps and
developing a research agenda.
I. Research Parameters - Overview (cont'd). If the goal of
RAND's review is to identify data gaps, RAND should
address this issue by reviewing the scientific studies and the
types of anecdotal adverse events that have been reported which should be used as a signal to identify additional
research projects. Ascribing degrees of causality to poor
data is not helpful in this regard. Moreover, by including a
subjective component to the review, and indeed
emphasizing this component, one does not control for
potential inadvertent reviewer bias. One of the reasons
placebo controlled double-blind trials are considered to be
the gold-standard for scientific research is that researcher
bias must be accounted for and eliminated to the greatest
degree possible. It is understood in the scientific community
that despite the best of intentions, inadvertent bias can
impact researcher conclusions.
I. Research Parameters - Overview (cont'd). RAND
acknowledged that its review of FDA's AERs was
subjective, and even focused on potential publication bias
by certain researchers and organizations, yet failed to even
acknowledge the possibility that inadvertent bias could
potentially impact its own report based upon the subjective
nature of the review of adverse event reports. I am not in
any way alleging that the draft report is actually biased either intentionally or inadvertently. Rather, my point is that
a subjective review is subject to inherent inadvertent bias
and therefore the report should focus on objective
information If the goal of RAND's report is to identify data

A3-90

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

gaps and suggest areas for additional research, ascribing

degrees of potential causality to anecdotal adverse event
reports appears to be entirely counter-productive. Finally,
the General Accounting Office and Department of Health
and Human Services (along with the FDA) have already
acknowledged that it is not possible to determine causation
based upon a review of these anecdotal adverse event
reports.
I. Research Parameters - Overview (cont'd). In light of this
determination, it is difficult to comprehend the benefit that
can result from RAND attempting to ascribe degrees of
causation to such reports. From an objective perspective,
the adverse event reports should not be used to ascribe
degrees of causation - but also should not be ignored.
Rather, as noted, the reported adverse events should be
documented and identified and then used to help target
endpoints for future research.
The draft report's causality assessment is not consistent
with how other Agency for Healthcare Quality and Research
(AHRQ) reports have addressed dietary supplement
adverse events. From our review of other reports, the draft
reports causality assessment is unprecedented.
The central question is why RAND conducted a causality
assessment of the AERs and prominently reported it
admittedly subjective attributions of causality. RANDs
causality assessment obscures the objective findings of the
report to such a degree that the research agenda RAND
recommends will not be achievable. Further. Rends
subjective attributions do little or nothing to answer the
questions that RAND has been asked to address. In a
broader context, the final report, if published with the
causality assessment, as it now exists, will threaten future
support for similar reviews of other dietary supplements.
Despite serious concerns, the draft report contains a very
worthwhile and comprehensive review of the objective data
on ephedra and ephedrine, and the recommendations for
future research are commendable and attainable, and will
serve the valuable function of answering important
questions concerning ephedra products. The final report can
become a monument for how to address controversial
safety issues such as those that exist for ephedra, provided
the final report focuses on objective science rather than
subjective assessments of the AERs.
The purpose of the RAND review is to perform an objective
review of the science pertaining to ephedra in order to
answer specific questions from the current data, and to
identify research gaps and recommended additional
research to answer the questions where the current data
are insufficient or do not exist. All parties to the ephedra
discussion agree that the AERs raise serious questions that

A3-91

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

deserve serious attention. Industry has enthusiastically

supported the concept of the RAND review for that reason.
However, as the draft report recognizes, the AERs do not
represent the objective data. RAND's review of ephedra is
necessary in large part to resolve the controversy that has
been created by widespread reporting of the subjective
assessments of the AERs on ephedra and the conflicting
opinions as to what the AERs mean. Informed critics and
supporters of the ephedra agree that the AERs cannot be
used to assess safety, to establish whether there is in fact a
risk of serious adverse events, or to quantify that risk if the
risk exists.
These limitations on the use of the AERs for ephedra were
the focus of the HHS and the Food and Drug
Administration's (FDA's) recent statements on ephedra on
June 14, 2002 1) and have also been noted by critics such
as Drs. Haller and Benowitz, as well as industry-supported
panels such as the Expert Panel of the Ephedra Education
Panel. 1)"The primary purpose of a voluntary adverse event
reporting system is to generate 'signals' of potentially related
events, rather than assessing product safety...There are
situations when background rated of the observed event are
so rare or unusual that, in combination with physiologic
responses and biologic plausibility, A significant relationship
between the events is self-evident from the reports in a
voluntary reporting system. However, the FDA has advised
me that the types of observed outcomes reported in
relationship to the ingestion of the ephedrine alkaloids are
not uncommon in the general population and there for the
reports alone do not provide a scientific basis for assessing
the safety of ephedrine alkaloids or establish a link between
the reported adverse events and the ingestion of ephedrine
alkaloids
#2 "[O]ur report describes a series of cases which the use of
ephedrine-containing dietary supplements was associated
with adverse cardiovascular events. Our report does not
prove causation, not does it prove quantitative information
with regard to risk. A large-scale case-control study similar
to the Hemorrhagic Stroke Project for phenylpropanolamine
is needed to determine the risks associated with these
dietary supplements" Christine A, Haller & Neal L. Benowitz,
Dietary Supplements Containing Ephedra Alkaloids: Letter
to Editor, 344 New Eng. J. Med 1095,1096-1097 (2001).
#3 the consensus conclusions of the EEC Expert Panel, as
well as extensive reviews of the published literature and the
most comprehensive review of the AERs that has been
conducted to date, were submitted to the FDA's published
docket on ephedra in October 2000 and were made
available to RAND. Since the Expert Panel Report is
nowhere mentioned is referenced in RAND's draft report, a
copy is included with these comments The Expert Panel

A3-92

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

concluded that the available data and information including

the AERs, "do [] not demonstrate an association between
the use of dietary supplements containing ephedrine
alkaloids and serious adverse events when used according
to the American Herbal Products Association (AHPA) trade
recommendation for ephedra products." Expert Panel,
Ephedra Educ. Council, Comments of the Ephedra
Education Council on the Safety of Dietary Supplements
Containing Ephedrine Alkaloids and on the AERs and
Health Assessment Released by the FDA on April3, 2000
on 6 (2000) (on file in FDA docket 00N-1200 as C30)
RAND has also concluded that, because of the subjectivity
of assessing causality from AERs, further analysis is
additional case reports will not lead to any objective
scientific conclusions and would not be useful to establish
whether there is any causal connection between ephedra
and the type of serious adverse events. Continued analysis
of case reports cannot substitute for a properly designed
study to assess causality." Draft Report at 5.
Nonetheless, RAND has conducted a causality assessment
of some of the AERs deemed to report serious adverse
events and has described the results of this review in terms
of "probable" and "possible" causal relationships between
ephedra, death and other serious events. RAND's findings
are presented in a way that will expand the controversy
surrounding the AERs on ephedra rather than resolve it
through objective analysis and recommendations for
additional research. Further, the draft report is open to the
inappropriate interpretation than RAND has concluded,
based on the AERs, that ephedra causes serious adverse
events, and that is exactly how the ephedra critics will
interpret the draft report. This interpretation is a result of the
wording of RAND's findings as well as the presentation of
the causality findings without necessary context.
The headlines that will result concerning RAND's findings on
causality will make is difficult if not impossible to justify the
research agendas that RAND recommends at the very end
of the draft report. Again, given the lack of any scientific
value to RAND's analysis, the ability to discuss the "signal
that the reports raise without reporting subjective
attributions of causality, and the importance of the report as
a means to resolve the controversy that the AERs have
created, the best solution is to remove the causality from the
assessment report.
The problems created by the RAND causality assessment
are aggravated by a number of factors: 1) The vast majority
of the draft reports text and tables relating to safety address
AERs (10 pages of text, 20 pages of tables) rather than
clinical data (one page of text, three pages of tables), even
though RAND acknowledges that clinical data are if much
greater value and even though in the end of the report

A3-93

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

RAND concludes that causality assessments do not provide

objective, science-based answers to the questions that it
was asked to address.
Although the draft report discusses a review of selected
AERs that was conducted under the contract with FDA by
Drs. Haller and Benowitz, and the draft report compares the
results of this review to the RAND review, the draft report
excludes any mention of an extensive review by Cantox
Health Sciences International, and the most comprehensive
review of the AERs to date by the Expert Panel of the EEC.
A copy of the Expert Panels Report is enclosed with a hard
copy of these comments.
While the issues of publication bias and other coursed of
bias were carefully analyzes in the draft report for published
studies, other than a brief mention of the subjectivity of
RAND's causality assessment, the draft report does not
adequately address the potential for reviewer bias in
RAND's causality findings.
Because the AER files that RAND reviewed not the same as
those that FDA had provided to the public, and RAND has
not made its AER files available to the technical expert
panel or the peer reviewers, none of these reviewers will
have the ability to do an in-depth analysis of RAND's
causality assessments. Some reviewers may have their own
files on some of the AERs, but this will not permit the type of
analysis that would be needed for a thorough peer review.
There is a simple solution to the problems that the RAND
causality assessment, and the manner in which the
assessment is presented in the draft report, have created.
The causality assessment should be removed from the
report. The final report will then be focused on the objective
assessment of the data and answering the questions that
RAND was asked with addressing. In addition, dropping the
causality assessment will permit the recommended
research agenda to proceed as intended. To do otherwise
places the whole RAND ephedra project in jeopardy, as well
as future support for similar projects.
If RAND does not remove the causality assessment from
the report, RAND should drop the attributions assigned to
specific reported as either probably causally" or "possibly
causally" related to ephedra consumption. Instead, RAND
should state the certain reports were reviewed for causality
and that those reports raise a "signal" that additional
research is needed to address that signal. This would be a
neutral and objective statement that would be accepted by
parties, and would be consistent with the recent statements
of HHS and FDA on the inability to make safety
determinations or regulatory decisions based on the AERs
for ephedra.
However, in the opinion of the reviewer, those conclusions

A3-94

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

regarding the case reports are limited by a combination of

the conservative causality assessment criteria and the
limited medical records and toxicology data available for
most case reports. For example, hypertension was defined
as a systolic pressure in excess of 180 and a diastolic
pressure in excess of 105. Also, no consideration appeared
to have been given to the contribution multi-component
ephedra-containing dietary supplements might have had in
those individuals with underlying cardiovascular or
cerebrovascular disease. I think it is generally accepted
among the medical and scientific community the presence
of sympathomimetic agents could potentially exacerbate the
likelihood of adverse events in such populations? I would
think most clinicians would factor such information into their
differential diagnosis rather than dismiss them altogether.
Adverse Events Reports: One limitation with your approach
taken in evaluation of the adverse events reports made to
the FDA is that your causality algorithm does not include an
assessment of whether ephedra played a contributory role
on the adverse event. Because ephedra is available as a
dietary supplement, it is likely that many persons taking
these products are not using them under doctor's
supervision, and may have medical contraindications to
their use. Therefore, the role of underlying disease becomes
a crucial factor in causation assessment, particularly when a
potential risk factor often goes undetected (i.e. essential
hypertension, structural heart defect), or when a condition is
omitted from the ephedra product label warning (i.e. family
history of premature CAD, sickle cell trait).
Two AERs that you assess as no higher than possibly
causal illustrate this point. AER 12485 did indeed have a
moderate degree of coronary artery disease detected at
autopsy. However, he was reportedly in good health without
history of angina, and had been jogging regularly without
adverse effects. Because he collapsed suddenly after
returning from jogging, we felt this was a primary arrhythmic
event due to ephedra. Similarly AER 12843 was a healthy,
adolescent who had participated in competitive sports for
many years. She had appeared to have been wellcompensated for a serious underlying coronary artery
abnormality that was clinically undetected since birth. Only
with use of Ripped Fuel, did she suffer a catastrophic
cardiac event resulting in death. We felt that the cardiac
stimulant effects of ephedra resulted in myocardial ischemia
in this case.
It would be helpful to specify what degree of pre-existing
coronary artery disease would constitute a significant risk
factor to result in myocardial infarction or sudden death in
the absence of stimulant use, thereby ruling out ephedra in
the causation assessment (page 32 of chapter 2

A3-95

This case was reviewed by a cardiologist

who made this judgment.

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

methodology). In the case of AER 14530 (page 63), I would

disagree that 20-30% stenosis would be significant enough
to result in acute M.I. in a 43-year-old female smoker
without a significant contributory effect from the ephedra
alkaloids in Metabolife.
We have revised this language.
On page 59 I suggest the authors be slightly cautious with
their use of language such as "probably causal" and "no
other possible explanation." The latter phrase is particularly
troubling. What the mean is no other explanation that they
could identify. Similarly, on page 69 they state that there are
a certain number of cases of serious adverse events that
"cannot be explained by causes other than ephedra use."
While I do not deny that it is extremely likely that many
cases of adverse events happen due to ephedra use, simply
because we do not have in our hand an explanation of why
an event occurred other than a particular explanation under
consideration, does not mean that the particular explanation
under consideration is the correct one.
Because of the paucity of large randomized trials, evidence
concerning stroke and ephedra by necessity consists of
analysis of case reports. "An assessment of case reports
is insufficient to warrant definite conclusions regarding
causality." (p.110) Nevertheless, arbitrary criteria are used
to define "probably cause": documentation that a stroke
occured, that ephedra was used, and that there was
exclusion of other potential causes. The definition may be
too liberal. Of ischemic strokes in relatively young adults
(i.e. those <50 years old), perhaps 20-35% are "idiopathic"
despite thorough evaluation. The definition implies that all
idiopathic strokes would be classified as "probably causal" if
ephedra was used in any dose in proximity to the event.
Given the frequency of idiopathic stroke, many (perhaps
most) neurologists would consider "possibly causal" to be a
better designation in this situation.
Are there specific clinical circumstances in which the
relationship of ephedra use and idiopathic stroke could be
certain? Perhaps if acute, striking elevation of blood
pressure were known to precede the stroke onset of of
angiographic features characteristic of vasospasm were
present in the abscence of migrane? Arbitrary to be sure,
and not very helpful.
The evidence for effectiveness supports the conclusion.
Except for AERs, however, little evidence of toxicity is
actually provided, and evidence of safety has been largely
ignored. No evidence is provided to even suggest "a
possible causal role of ephedrine-containing dietary
supplement in rare, but serious events," let alone extremely
common events such as heart attack and stroke. Even
critics of ephedra have concluded that the clinical effects of
pharmaceutical ephedrine and the ephedrine contained in

A3-96

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

herbal preparations, are indistinguishable. Gurley states

that the increased incidence of ma huang toxicity does not
stem from differences in the absorption of botanical
ephedrine compared with synthetic ephedrine." Haller and
Benowitz, in their most recent publication, conclude,
"Botanical stimulants have disposition characteristics similar
to their pharmaceutical counterparts..."
The Cantox Report, and the Report of the Expert Panel of
the EEC reached similar conclusions. Since there are no
real differences, studies demonstrating the safety of
pharmaceutical ephedrine and pharmaceutical ephedrine in
combination with pharmaceutical caffeine, should not be
excluded when considering the safety of herbal equivalents.
The explanation most frequently offered for alleged cases of
ephedrine-related stroke is drug-induced blood pressure
elevation, this in spite of the fact that no clinical trial, of any
duration, has ever demonstrated that a clinically significant
effect on blood pressure exists. Indeed, the studies that
have addressed this question. including the most recent
paper by Drs Haller and Benowitz, have shown diminishing
cardiovascular effects over time. In other words, if
dangerous blood pressure elevations do not occur with the
first dose ephedrine, they are even less likely to occur with
prolonged dosing. These studies should be included in the
RAND review of ephedra and should be used to address the
question of potential increases in blood pressure and other
safety issues.
Ephedrine has been studied in more than 50 double blind,
placebo-controlled clinical trials, some of long duration. A
far from exhaustive literature search produced the attached
list of peer-reviewed, published, clinical trials. Most have
compared ephedrine to placebo, and to other
sympathomimetic drugs used to treat asthma. However,
others have involved smoking cessation, sexual function
and athletic performance. Nearly a dozen of these trials
involved caffeine/ephedrine combinations using does
exceeding those found in herbal supplements. In total,
more than 2000 individuals have been enrolled in these
trials. In several studies there was even continuous cardiac
monitoring in middle-aged patients with known heart
disease; no effect was observed. No clinically significant
episodes of toxicity were reported. Including these and
other studies on ephedrine that have been excluded from
the RAND review will increase the power of the safety
calculations that can be derived from clinical data.
One of the major limitations of the report was the
composition of the TEP and the reviewers who made
subjective assessments of the AERs. Given the importance
placed on assessment of AERs, it is unfortunate that no
pathologist was included in the view or on the panel. The
lack of expertise is obvious from the comments made about

A3-97

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

the individual AERs. The failure to provide information about

any potential conflicts of reviewers also detracts from the
study. Why were the findings of Expert Panel of the
Ephedra Education Counsel not considered? The analysis
of this panel was in some ways unique, as it is the only
consensus opinion on ephedra safety. In addition, this
panel conducted the most comprehensive review of the
ephedra AERs to date, and yet the causality assessment,
which conflicted the findings of the draft report, are not even
mentioned. If RAND believes that the EEC review and
analysis was, in some way scientifically flawed, then the
reasons for that belief should be stated.
The danger of drawing conclusions from AERs without a
control group can be illustrated by examining data from
randomized trials in which participants are blinded to
whether they are receiving the study treatment or inactive
placebo. In the placebo group of a recent randomized trial,2
there was an increase in ventricular couplets (extra heart
beats) at the 4th week of the study 9from 3% at baseline to
around 14% at week 4). This is, of course, not due to the
placebo, which is inactive, but rather just spontaneous
ventricular couplets that occurred by chance. However, if
these participants had been given ephedrine alkaloids in an
uncontrolled study (without placebo), this change could
have been attributed, incorrectly, to the ephedra. That is,
these could be AERs that were attributed to ephedra. In
fact, in the controlled trial, a similar increase in ventricular
couplets was not seen in the ephedra/caffeine arm.
Another example is the 15-year-old female (case 12843)
with Bland-White-Garland syndrome who died while playing
soccer. This disorder has been associated with sudden
death after physical exertion. In the absence of a unique
pathologic process, it is almost never possible to establish a
causal association on the basis of adverse event reports.
There is nothing pathologically or diagnostically unique
about the adverse events noted in the ephedra database
(e.g., myocardial infarction, stroke) that allow one to
distinguish a spontaneous event from one caused by use of
Ephedra products. In fact, a review of all autopsy data from
ephedra AERs by Dr. Grover Hutchins, a Professor of
Pathology at the John Hopkins University School of
Medicine and member of the Expert Panel of the EEC,
concluded that "The pathology data available do not show
any pattern consistent with ephedrine alkaloid-containing
dietary supplements as a cause of death."5
Similarly, 10 participants in the ephedra/caffeine group
(12% of these participants) withdrew because of
cardiovascular symptoms (palpitations, elevated blood
pressure, arrhythmias). If there were no control group,
these also might have been attributed to the
ephedra/caffeine combination However the same

A3-98

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

proportion of participants in the placebo group (13%)

withdrew for the same reasons. The withdrawal rate in the
ephedra/caffeine group thus was consistent with the
background rate of these events in a placebo group
unrelated to ephedra use. A second limitation of adverse
event reports is that other potential causes of the event are
often present, making it extremely difficult to determine if an
event truly is related to the exposure.
As an example, it is well established that physical exertion
can trigger myocardial infarctions and cardiac arrest (up to a
74-fold increase in the risk of sudden death, according to a
recent report in the New England Journal of Medicine6).
Therefore, a case such as the 38-year-old man with three
vessel coronary disease and a dilated heart who died after
jogging could very well have been related only to the
physical exertion and not the ephedra. According to his wife,
his heart problems had been known for at least five years.
In addition, this man had been taking an ephedra product
for one year without apparent ill effect.
The third limitation of AERs is that their interpretation
remains subjective. Even experts in the field will disagree
about the possibility that an event may or may not be due to
an exposure. The RAND report (Table 22) displays a
comparison of the causality assessment between their
review panel and a published report by reviewers for the
FDA.3 In only two out of 20 cases that both groups
reviewed was their agreement about the highest level of
possible causality. For example, with respect to the 38year-old discussed above, the Haller and Benowitz review
recorded this event as "Definitely or probably related" to
ephedra while the RAND report classified it as only
"Possibly causal."
Another review of AERs performed by Dr. Theodore Farber
and Dr. Norbert Page, members of the Expert Panel of the
EEC, reveals similar disagreements. The two "probable"
cases reviewed by EPC were rated as "Low Possible" (case
12980) and "Improbable" (case 13418) by Drs. Farber and
Page. The fact that the etiology of events can be debated
simply illustrates the substantial limitation of case reports
that lack a comparison control group. This echoes reviews
done by FDA and its consultants in which agreement about
causality was poor. For example, two consultants from
FDA, Drs. Ricaurte and Stoll, reviewed 28 AERs related to
neurological events. Dr. Ricaurte classified eleven cases as
"attributable" while Dr. Stoll classified only five as
"attributable."5 Only two of the consultants findings
overlapped - that is, there were only two cases that both Dr.
Ricaurte and Dr. Stoll agreed should be categorized as
"attributable."5 This disagreement is not a flaw of the
reviewers, but rather a flaw of AERs.
A fourth limitation of AERs is that ingestion of the substance

A3-99

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

in question cannot always be substantiated. For example,

case 10276 is a 32-year-old with an enlarged heart who was
found dead in his truck. Although a product that contains
herbal ephedra was found in his truck, so were several
bottles of cold medications, including Nyquil. Toxicology
revealed no ephedrine, but did identify pseudoephedrine
and doxylamine, both components on Nyquil. Thus, there is
no evidence that this person even ingested the herbal
ephedra. A similar case (13096) revealed no ephedrine in a
toxicology screen again suggesting that the man had not
ingested ephedrine around the time of death. Equally
importantly, this man died of a disease that appeared to run
in his family (aortic dissection), including an 18-year-old
niece.
In summary, AERs cannot be used to assess causality. As
stated by several authors with experience in interpreting
adverse event reports for the FDA: "It is probably
impossible to do comparative analyses employing ADE
[adverse drug event] reports for drugs that have received
extensive publicity in the mass media for an adverse
event..."7 In fact, the FDA's Center for Drug Evaluation and
Research pointed out, in a February 10, 2000 memorandum
concerning ephedra products, that "it is possible that the
reported serious adverse events are reflective of
coincidental background spontaneous occurrences in the
population and are not necessarily causally related to [the
use of dietary supplements containing ephedrine-type
alkaloids]." The RAND review notes this as well, stating that
"The most important limitation [of their assessment of
adverse events] is that the study design, that is an
assessment of case reports, is insufficient to warrant
definite conclusions regarding causality." They list this
limitation as one of the "most important...gaps" in the current
knowledge-base.
The RAND review also states that "Disentangling the
relative importance of the pre-existing condition and the
ephedra use is not possible." They state further that
"Continued analysis of case reports cannot substitute for a
properly designed study to assess causality. A case control
study would probably be the study design of choice." Their
Technical Expert Panel also "judged that case reports alone
would be insufficient to establish definite causality between
ephedra use and serious adverse events." Because of these
limitations, terms such as "probably causal" and "possibly
causal" in AER reviews are potentially misleading (see, in
particular, the "Conclusions" section of the "Structured
Abstract," page vi, and the "Conclusions" section, page
112). They represent only reviewers' assessment of
causality based on uncontrolled data and subjective
assessments. Although these terms are often used in
scientific publications, their use in the RAND report may
suggest a level of evidence that does not exist from the

A3-100

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

current data. These statements, therefore, should not be

taken out of context.
RAND's stated limitation that "Definite causality cannot be
determined from case reports" must be kept in mind when
interpreting this report. It is also critically important to
remember that case reports can produce false signals of
cause and effect.8,9 Most importantly, because of the
limitation of AERs, it is unclear why the RAND report states,
in the Structured Abstract, that "These [sic] is sufficient
evidence to suggest a possible causal role of ephedracontaining dietary supplements in rare but serious adverse
events, particularly cerebral hemorrhage." With respect to a
possible causal role of ephedra in adverse events, RAND
acknowledges that AERs are not sufficient to draw
conclusions about causality, consistent with the known
limitations of AERs discussed above. In addition, this
comment is puzzling given that their "Conclusions" section
(ages 112-113) does not state this at all.
With respect to cerebral hemorrhage, the only comment in
the body of the report on cerebral hemorrhage refers to a
case-control study of phenylpropanolamine (PPA) and
cerebral hemorrhage.10 However, the RAND report refers
to this study only as an example of case-control
methodology that could be applied in the future to ephedra.
The report does not discuss the PPA study further. In fact,
this study has been heavily criticized. Despite this, there
was not a formal review of this study by RAND (and there
were no members of the technical expert panel listed with
expertise in epidemiology to perform such a review). In
addition, PPA and ephedra have different chemical
structures and different pharmacological activities. Finally,
the RAND report does not mention that the PPA report also
presented data on non-PPA, ephedrine-alkaloid containing
products. These agents included medications that
contained pseudoephedrine hydrochloride, phenylephrine,
ephedrine, and epinephrine. In the report, there was similar
prevalence of use of these products among those with and
without hemorrhagic strokes.
Although this is not a definitive analysis, it suggests that
there was no association between these ephedra-containing
products and hemorrhagic stroke. Therefore, the statement
in the Conclusions section of the Structured Abstract of the
RAND report is inconsistent with currently available
scientific data. In summary, the RAND review supports the
use of herbal ephedra and caffeine for weight loss, an effect
that may have beneficial health consequences. The report
also suggest, from controlled studies, that adverse events
following ephedra use are, at most, rare. (The should not
imply that the events are even causally related to ephedra
us.) Most importantly, because of the reliance on AERs, the
report cannot establish a causal effect of ephedra on

A3-101

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

serious adverse events.

Review of Anecdotal Adverse Event Reports - Limitation of
Review. On pages 109-110, the report identifies potential
limitations associated with the review of anecdotal adverse
event reports. These limitations are buried at the end of the
report, rather than being incorporated into the appropriate
sections of the report (as per prior AHRQ reports, such as
the Garlic Report - see Section IX, herein). Moreover, a
number of limitations are not prominently identified,
including but not limited to: a) The poor quality of the data
and information contained in the anecdotal adverse event
reports; b) Inherent problems associated with voluntary
reporting systems; c) The number of individuals in the
general population who experience the adverse events
identified (i.e. background risk); d) The possibility that
certain products may not have been standardized and/or
manufactured according to Good Manufacturing Practices
("GMPs") - resulting in potential adverse events that have
nothing to do with ephedra or caffeine when consumed in
recommended amounts. Specifically, in the absence of
standardization and quality control, it is conceivable that
certain products may contain far more ephedra or caffeine
(or other constituents) than indicated on the product label.
This possibility must be considered when evaluating
anecdotal adverse event reports. In the absence of
identical product identity, any general conclusions regarding
ephedra and caffeine are inappropriate and highly suspect
based upon adverse event reports (see the AHRQ Garlic
Report for an appropriate way to address this issue). In my
opinion, RAND should strongly support immediate issuance
by the FDA of dietary supplement GMPs and should
endorse stringent quality control measures to ensure that all
ephedra supplements contain what they are claimed to
contain; e) The possibility that the consumer abused or
misused a product by ingesting more than the
recommended amount - or that the consumer ignored
detailed product warnings and contraindications. RAND
should emphasize the detailed warning label contained on
the vast majority of ephedra supplements - and should
acknowledge that there is little way to know from anecdotal
data whether a consumer abused a product (either
intentionally or more likely inadvertently); f) The possibility
that the anecdotal adverse event reflects chance,
coincidence , or confounding factors - including but not
limited to the possibility that ingestion of a different product
or substance led to the stated event.
I I. Review of Anecdotal Adverse Event Reports - B.
Ascribing Causality to Adverse Event Reports - 1. Reliance
on Unpublished Article. In order to establish a framework
for analyzing the adverse event reports, the draft report
relies upon an unpublished article written by Cynthia

A3-102

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

Mulrow, M.D. Reliance upon an unpublished, non peerreviewed article to establish the framework for a critical
portion of the report is entirely unacceptable. AHRQ studies
have not, to our knowledge, ever relied upon an unpulished
article to establish the framework for this type of analysis.
In addition, reviewers such as myself have no way of
analyzing the article - thereby defeating one of the primary
reasons for review of the draft report.
II. Review of Anecdotal Adverse Event Reports - B.
Ascribing Causality to Adverse Event Reports - 2. Failure to
Review Factors Critical to the Interpretation of Anecdotal
Adverse Event Reports. Table 4 of the draft report
summarizes the various methods researchers use to assess
causality from adverse event reports. The following nine
factors are identified: a. Temporal relationship. b.
Dechallenge response. c. Rechallenge response. d. Could
there be an alternative explanation? For example,
dehydration or consumption of other toxic substances. e)
Prior reaction to same substance. f) Dose response. g)
Objective evidence of adverse event. h) Previous conclusive
reports. i) Definition of substance. Despite the report's
reference to these nine factors, it is my understanding that
the report concludes that events are "probably causal"
based upon a review of only two factors - a and g. The draft
report does not explain why RAND believes only two factors
out of nine can be used to ascribe degrees of causation to
anecdotal adverse event reports.
II. Review of Anecdotal Adverse Event Reports - B.
Ascribing Causality to Adverse Event Reports - 2. Failure to
Review Factors Critical to the Interpretation of Anecdotal
Adverse Event Reports (cont'd). The draft report also
indicates that three factors are used to determine if an event
is "probably causal": a) Reasonable certainty that the
adverse event occurred. b) Reasonable certainty that the
patient took ephedra in a dose and timing compatible with
the known pharmacology of ephedrine. c) An adequate
evaluation must have been done to rule out other potential
causes for the adverse event. The third factor (factor c,
above) is exceptionally problematic from a scientific
perspective. The report acknowledges that the third factor
is subjective. Specifically, in an effort to rule out other
potential causes, the report indicates that such a
determination was made by determining if the subject had a
pre-existing condition that was identified in the adverse
event report.
It is stated, "With regard to adverse events, it was
recognized by EPC staff and the TEP that, even in
aggregate, the number of patients included in randomized
trials was likely to be few. Because of this, it was
recognized that case reports would have to be relied upon
to assess serious adverse events " It was Dr Shekelle who

A3-103

The causality analysis has been removed

from this revision. We revised the
sentence to indicate EPC staff recognized
assessing case reports was going to be
required to meet the terms of the contract.
Our notes from the TEP meeting are clear

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

advanced his position to this question, but it is not correct to that the majority of the TEP agreed.
state that this was recognized by the TEP. In my remarks I
made it clear that the clinical trials were the most useful
clinical data available, and that the FDA AER database was
to flawed and incomplete to be able to draw any
conclusions. Dr. Shekelle introduces Cindy Mulrow's criteria
as Rands position in the assessment of adverse medical
events. My point was that these criteria would be useful in
assessing adverse events in the course of controlled,
conventional, pharmaceutical trials, not the poor quality of
the voluntary AERs in the FDA database was shocking in
contrast.
We did the best analysis possible in the circumstances for
the CRN report and found that even the most complete
subset was not sufficient quality to draw any conclusions.
Dr. Benowitz agreed that the AERs did not have all the
elements of an AER analysis. Therefore, the statement on
page 21of the draft report is a reasonable summary of the
discussion, i.e., "Consequently our TEP judged that case
reports alone would be insufficient to establish definite
causality" Dr. Shekelle also agreed that the FDA AER
could not be used to assess causality. He said that the
adverse event issue would be the hardest to deal with,
because it is front page and gets wide attention, but he was
not worried about disagreeing with the FDA. Stating that not
all the AERs were true or false, he acknowledged that the
gold standard was lacking to link exposure and outcome,
there is no basis for a conclusion.
The assessment of probability/ possibility respecting
Causality has been removed from this
causality between use of ephedra-containing dietary
revision.
supplements and adverse health events seems shaky
based as noted mainly on FDA case reports, "insufficiency
documented". It is not clear to me where the margin is
between probably and possibly. and whether there is a clear
basis for location on either side of the margin.
It is not clear why you require coronary angiography for
cases of myocardial infarction. Clearly MI can be diagnosed
on the basis of EKG and enzyme changes. Coronary
angiography does address the severity of underlying
coronary artery disease, but that does not address whether
or not ephedra played a causative role. It is well known that
coronary spasm is most likely to occur at the site where
there is some underlying coronary artery disease. If
ephedra can cause coronary spasm, a person with
underlying coronary artery disease would be most
vulnerable to this occurring.

We clarified that coronary angiography

was required in cases of myocardial
infarction in order to evaluate other
causes, such as coronary artery disease,
not to make diagnosis. In the presence of
coronary artery disease, the occurrence of
an MI could be classified no higher than a
possible sentinel event.

The criteria in our causality algorithm may be too

conservative. Requiring coronary angiography in cases of
M.I. would exclude cases diagnosed by cardiac enzymes
and ECG changes alone.

Angiography was required to assess the

possibility of alternative explanations, not
the presence of a myocardial infarction.
The text has been revised to reflect this.

A3-104

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

The interpretation of co-existing cardiovascular disease with In this revision we no longer deal with
respect to causation is an important issue. It is quite likely
causality.
that underlying cardiovascular disease would predispose to
ephedra causing a serious cardiovascular event. This has
been well shown to be the case for cocaine. I think this
issue needs to be made quite clear, especially since this is
the reason why many of the adverse events are classified
as possible in your evaluation, while they were judged to be
probable in the evaluation done by Christine Haller and
myself.
What is meant by more than the minimal dose? How do
you know what a minimal dose is? Were any cases
excluded because of this criterion? In the same figure on
level 3, the question comes up again about the difference
between probably causal and possibly causal. The box
above possibly causal says interaction with ephedrine
likely. If you say that the interaction is likely, then why do
you say its possibly causal?

We no longer use this criterion or assign

causality in this revision.

A 41-year-old female has four stroke events over a 2month period between December and February. This case
seemed to have incomplete description as there was no
mention of the product (Diet Phen) that the patient was
taking and when and for how long (14-60 days).

This text has been revised.

According to Table 20, Ripped Fuel was also involved but

there is no mention of the product in the description (which
is probably important information for the reader).

This change was made.

If the date of death is May 1994, she cannot be admitted to

the Emergency Room in December of 1994. Should it be
1993?

This typo has been corrected.

I found that Table 20, column titled as Key Determinants of This table has been revised to improve
clarity.
Causality rather confusing, incomplete and unclear.
Delete "Timing<24 hours" from all cases as this does not
contribute any additional information but rather add
confusion to the interpretation (reader may interpret a "no"
to Timing <24 hours means ingestion did not occur within 24
hours, which is not the case as many times tox screen is
"yes".)
Change Tox screen**: to Tox screen was done: and
eliminate "**Ephedrine/amphetamines found in toxicology
screen" as this is not true in all cases.
Add Ephedrine alkaloid detected: Yes or No to the column
since tox screen may not include detection of ephedrine or
its alkaloid.
Does Natures Nutrition Formula Three contain ephedrine
(see p. 60 case description)? Should this be included in the
table? In the description on p.60, it stated that Toxicology
screen was negative for ephedrine..., however, table 20
indicated that ephedrine was found The footnote may be

A3-105

Change made.

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

misleading and it may be more appropriate to footnote tox

screen withTox screen was done rather than
Ephedrine/amphetamines found in toxicology screen.
The case description stated that Toxicology screen for
cocaine, ephedrine and amphetamines was negative.
However, the Table indicated that ephedrine /
amphetamines were found (same problem as #15).

Change made.

The age of the patient is different (37 y.o. or 36 y.o as

described on page 62)?

Change made.

Should the product be Metabolife (as described on page 63) Change made.
or Metab-O-Lite as indicated in the table.
Does Accelerator also contain ephedrine? Should this be
Change made.
included in Table 20? Footnote of tox. screen is inconsistent
with the description on p.63 (same problem as #15).
The ages are different between the table and the description Change made.
on page 64.What is the tox. screen results for this patient?
The ages are different between the table and the description Change made.
on page 64-66.
There is no description of the case on page 66 under
The grouping of the cases has been
Stroke, possibly causal section but it is found under
changed to better improve clarity.
Subarachnoid Hemorrhage on page 68. Yet, it is grouped
together with all the other cases of Stroke (CVA) in table 20.
Should this be separately described or should this be
described under Stroke section?
For AER 13672 described as "probably not causal" on page
62, There are toxicology results that showed 280 ng/ml
ephedrine in the blood. These results were reported by the
Medical Examiner on 2/12/02, which may be after the FDA
report was finalized.
Metabolife recently admitted (after this draft was issued)
that it has received 13,000 complaints about its ephedra
products. These should be included
in your analysis.

These are now included.

p 47, question 18: what was the rationale for dividing the
durations of use into the listed classes? Because
tachyphylaxis generally occurs after about 14 days of
continuous use of ephedrine, the evaluation of acute use or
for acute effects is commonly limited to days 1-14, with
durations longer than 2 weeks considered chronic use.

Because we wanted to distinguish dosing

within 24 hours, we divided the categories
in the Table in this fashion.. In the actual
data, we recorded the exact time.

Regarding adverse event adjudication -- you make your

reasoning clear for exclusion of individual events due to
insufficient information or downgrading attribution due to
pre-existing conditions; however, the point could be made
more clearly that this may tend to underestimate the number
of serious adverse events

We have emphasized in the text that our

methods of case report analysis are
conservative and may underestimate the
number of events.

A3-106

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

A concern is that in the case studies there was no analysis

and no sufficient emphasis place on the evaluation of
dosage amount, which in many cases appeared to be
excessive.

In the majority of case reports the dose

was not even reported, preventing any
dose analysis of the case reports.

To demonstrate how additional data could affect the

outcome of the present study, consider case 13672, which
was designated "probably not causal. This reviewer has
access to additional data on this particular case, specifically
that the decedent did not have a toxicological examination
that revealed a postmortem ephedrine blood concentration
of 280 ng/mL and a pseudophedrine level of 100ng/mL.
Given this additional information, is the causality level
assigned to this specific case still valid? More thorough
investigations like those pursued during the discovery
process of specific lawsuits almost always yield additional
information that would likely modify the causality
assessment of specific cases submitted via the
MEDWATCH program. Also, were those case reports
described in the medical literature used in the case report
assessment? They do not appear to have been utilized.

We did include the medical literature case

reports in this revision. The additional
information about a specific case, as
provided by this reviewer for this case and
other reviewers for other cases, we
unfortunately cannot include or assess in
our report, as we have no access to the
original information.

I wasn't clear as to why only case reports documenting

death, myocardial infarction, and/or cerebrovascular
accidents were evaluated. To me this made the comparison
to the Haller and Benowitz study less meaningful.

We have included additional case reports

in this revision. We have deleted the
comparison to the Haller and Benowitz
study since we no longer assess causality.

A3-107

Appendix 3. Reviewer Comments (continued)

Second Review of Safety Analysis including Metabolife Data
Reviewer Comment

Rand Response

Meta-analysis. This study was sent to me for information

purposes only. The analysis is well done and meets the
highest standard. The efficacy of Ephedra is very modest in
terms of weight loss. Treatment causes significant adverse
effects with RRs ranging from 2 to 3. Statistical power is
inadequate to rule out severe adverse events occurring at a
rate of 1.1/1000. The severity of the adverse effects can not
be determined (my assumption). Likewise, the doseresponse relationship can't be estimated.

No response

No response
Metabolife analysis. The database is extremely messy and
does not allow many meaningful analyses and conclusions.
Your approach in terms of coding rules, data extraction and
event classification is good. If a pharmaceutical company
had kept records in this sloppy way my assumption is that it
would be in deep trouble with the FDA. You conclusion is
weak in my view. I would say that the "Findings are
consistent with an increase in rare serious adverse events.
What troubles me is that the population is so young. I would
not expect serious cardiovascular events occurring so often.
I realize that we don't have a denominator so any attempt at
even guessing what the event rates might be are probably
too speculative. In summary, you have from an analytic
point of view done what can be done.
We added to the limitations that MedWatch
Evidence Report. This is another well-done study. My
may underestimate the number of events.
interpretation is colored by two facts. When people
complete a MedWatch form they suspect an association.
There is a marked underreporting ranging from 90 to 99 %.
This means that what appears to be rare may not be very
rare. The temporal relationship between use of Ephedra and
the occurrence of an event may exist even if it isn't
documented. Again, I think your conclusion is too mild. I am
fairly convinced that Ephedra causes serious events but I
can't give a rate. Moreover the benefit-risk ratio is
unfavorable (minor benefit for sure), so I would question the
wisdom of leaving the compounds) on the market. My
position is also influenced by the age of the victims.
The AERs presented in this report appear to be consistent
with the known pharmacologic actions of ephedrine. Would
it be appropriate to include a statement to this effect in the
report?

We included such a statement.

When the summary text under Results and Conclusions is

This was done.
updated to include the analysis of the ephedrine AERs, care
should be taken to present results for ephedra and
ephedrine separately.
For the case descriptions of the cerebrovascular
We included this information to the extent
accident/stroke events, it would be helpful to include the
that we identified it in the source
individuals functional status in the text (it is already included documents.
in the Table 20).

A3-108

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

Weve seen a comment from NCCAM regarding the tables.

We agree with that comment.

These tables are now incorporated into

this revision.

Page 5, paragraph 3, last line: Subject should be defined.

Page 7, paragraph 2: Second sentence: Text would read
better as follows, seizure, only those cases described as
generalized toxic-clonic seizures underwent further review.
Sentence beginning at end of line 5: Text notes requirement
that there be documentation that the individual had
consumed ephedra or ephedrine within 24 hours of the
event, but that this was not a requirement for psychiatric
events. It would be helpful to explain why this decision was
made. Sentinel case was not defined previously or used
subsequently in the report. If it means sentinel event,
should change wording.
Page 24, paragraph: Should doses be changed to
dosage?
Page 27 Paragraph 1, under FDA Cases Ephedra: The
dates arent correct. It appears as if the patient was taken to
the hospital in December 1994 where she signed out AMA
even though she had died in May 1994. What is
chlophoramine? Paragraph 2, line 3: Change toxicology
to toxicology screen.
Page 28, Paragraph 3 (case# 12722): Text doesnt mention
ephedra exposure what product was used? Paragraph 4
(case# 12843): Text doesnt mention ephedra exposure
what product was used? Paragraph 7 (case# 14638): Text
notes that individual had been taking Hydroxycut for seven
days, but Table 20 (page 52) says 2-13 days.
Page 29 Paragraphs 2 and 3 (case# 44): Text doesnt
mention ephedra exposure what products were used?
Paragraph 5: Case# 258 is not included in Table 20.
Paragraph 6 (case# 13672): Change rain to run.
Soldier does not indicate gender. Although from the text
the individual is apparently a male, wording should be
changed. Should indicate whether the toxicology screen
looked for ephedrine or that ephedrine was not mentioned in
the report. Paragraph 7 (case# 1859087): Text says this
individual was taking Max Alert, but Table 20 says the
product is unknown
Page 30: Paragraphs 5 and 6 (case# 13806 and case#
14465) are not included in Table 20. Paragraph 6 (case#
14465), last line: Change not conclude anymore to come
to no other conclusions.
Page 31: Paragraph 1: Product names in text and Table 20
do not match. Paragraph 2: Product names in text and
Table 20 do not match. Indicate whether the toxicology
screen looked for ephedrine or that ephedrine was not
mentioned in the report. Paragraph 6: Text notes that this
individual was taking Eola Amp II Pro drops for 12 days but
Table 20 indicates that he took them for 2-13 days.

We corrected typographical errors. We

made suggested changes in language. We
stated whether ephedrine was looked for in
the toxicology screen. We made the
product names match, e.g. Ripped Fuel
(Twin Labs) was changed to Ripped
Fuel. The reviewer is incorrect about
cases being in the text but not in Table 20
(now table 22); all were present.

A3-109

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

Page 32: Paragraph 2 (case# 9504): Product names in text

and Table 20 do not match. Paragraph 3 (case# 10009): Is
(was) there such a product as Metabolift? Paragraph 4
(case# 13009): Text indicates that the individual described
is a male, but Table 20 notes that it was a female. Indicate
whether the toxicology screen looked for ephedrine or that
ephedrine was not mentioned in the report. Paragraphs 5
and 6 (case# 14114 and case# 14530): Product names in
text and Table 20 do not match. After case# 14530, the last
one in the text under Myocardial Infarction, Table 20
continues with many more case reports of MIs (pages 6466) and then lists other cardiac (pages 67-71) starting
with three possible sentinel events. Why arent the
descriptions in the same order in both text and table?
Page 33: Paragraph 1: Indicate whether the toxicology
screen looked for ephedrine or that ephedrine was not
mentioned in the report. Paragraph 2, line 1 (case# 11062):
Paragraph 2: Text indicates individual was 44 years old, but
Table 20 says she was 42. Insert was between and and
a in was taking Power Trim and a cigarette smoker.
Paragraph 4: Indicate whether the toxicology screen looked
for ephedrine or that ephedrine was not mentioned in the
report.
Page 34: Paragraph 1: Product names in text and Table 20
do not match. Paragraph 5: Sedimentation is misspelled.
Page 35: Paragraph 1: Product names in text and Table 20
do not match. Line 2: editorial - change here to her. Line
7: editorial change with embolus to with an embolus
Paragraphs 3 and 4 (case# 10094 and case# 12713):
Product name in text and Table 20 do not match. Paragraph
6 (case# 515): Text doesnt mention ephedra exposure
what product was used?
Page 36: Paragraph 2: Text indicates individual is 25 years
old while Table 20 indicates she is 26. Paragraphs 2, 3, and
4 (case# 14378, case# 14434, and case# 14553): Product
names in text and Table 20 do not match. Paragraph 5:
Thoracic is misspelled
Page 37: Paragraph 2: Product name in text and Table 20
do not match. Paragraph 3: Delete either other or
additional. Paragraphs 3 and 4 (case# 13829 and case#
13905): These cases are not listed in the same order in the
text and in Table 20 making them difficult to find (they are
located on page 78). Paragraph 4 (case# 13905): Text
notes individual is a female of unknown age while Table 20
indicates she is 36 years old. Paragraph 6: Text notes that
individual was taking 40-60 mg of ephedrine for 10 years.
Was this 40-60 mg per day? Some indication of amount per
unit time should be provided, or a note should be made that
the information is not available.
Page 38, paragraphs 1, 2, 3, and 4 (case # 12851, case#
13031 case# 13643 and case# 13793): These cases are

A3-110

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

not included in Table 20.

Page 39: Paragraphs 1, 2, 3, and 4 (case# 110, case# 297,
case# 260, and case# 13945): These cases are not
included in Table 20. Paragraph 1 (case# 110): This case
was classified as a possible sentinel event but is listed in
the middle of a list of cases with insufficient information.
Paragraph 2 (case# 297): Change taken to taking.
Should indicate whether or not there was any information on
how long he had been taking Herbalife supplements? This
case was classified as a possible sentinel event but is
listed in the middle of a list of cases with insufficient
information. Paragraph 3, last line (case# 260): Delete
intake. Paragraph 5 (case# 13062): Product name,
duration, and dose are not included in the text, but are given
in Table 20 and should be included here. This case was
classified as a possible sentinel event but is listed in the
middle of a list of cases with insufficient information.
Page 40: Paragraph 3: 5am in the morning is redundant.
Paragraphs 3 and 4: Product names in text and in Table 20
do not match.
Page 41: Paragraph 2 (case# 10432): Product name in text
and Table 20 do not match. What is encepholophy?
Change focality to focal nature. Paragraph 3 (case#
11062): Product name in text and table do not match.
Change taking to taken. The last line notes that because
of the possible structural abnormality, this event was
classified as a possible sentinel event. However, it is not
clear from the text what the possible structural abnormality
was. Paragraph 4 (case# 11649): Product name in text and
Table 20 do not match. Indicate whether or not there was
information regarding how long this individual had been
taking Metabolife prior to the event.
Page 42: Paragraph 1 (case# 13408): Product name in text
and Table 20 do not match. Paragraphs 1 and 4 (case#
10874 and case# 11675): Indicate whether the toxicology
screens looked for ephedrine or that ephedrine was not
mentioned in the report. Paragraph 2: Text indicates
individual in case# 14275 was 38 years old, but Table 20
says she was 30. Paragraph 3: Text indicates individual in
case# 11105 was 31 years old, but Table 20 says she was
30. Paragraphs 3 and 4 (case# 11105 and case# 11675):
Product names in text and Table 20 do not match.
Pages 43 Paragraph 2 (case# 9747): This case is not
included in Table 20. Paragraph 3 (case# 9509): Product
name in text and in Table 20 do not match. Paragraph 5,
last sentence (case# 13809): Text indicates the individual
described was intent on doing harm to others, but Table 20
describes her as suicidal. Suggest changing alleviated to
subsided.
Page 44: Paragraph 2 (case# 1855921): Text notes that
this individual was taking Minithin but this information is not

A3-111

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

included in Table 20. Text notes that there was no history of

other drug use, but Table 20 says this individual had a
history of substance abuse. Paragraphs 3 and 4 (case# 48
and case# 238): Change psyc to psychiatric. Paragraph
4 (case# 238): Provide composition of Tedral as is done for
Bronchipax in paragraph 5. Also note that this drug is listed
as Bronchi Pax in Table 20. Paragraph 6 (case# 9751):
Product name in text and in Table 20 do not match. Should
note in Table 20 that problem resulted from discontinuation
of product use as described in the text.
Page 45: Paragraph 3 (case# 12372): Product name in text
and Table 20 do not match. Last line: Change classified
as to classified it as. Paragraph 4, first line Case# 13005):
Change also to used. Paragraph 4 (case# 14436): Delete
(tid). Paragraphs 5, 6 and 8 (case# 14436, case# 14528,
and case# 79): Product names in text and in Table 20 do
not match. Paragraph 6 (case# 14528): Clarify very soon
vs. approximately 1 week after. Paragraph 7 (case#
1682426): End of line 3: Change in residential to in a
residential. Last line: Note is misspelled.
Page 46: Paragraph 1 (case# 79): Text notes that product
name is not given, but that investigators contacted the
manufacturer is the name of the manufacturer known?
Paragraph 2: Should term causality be used here after the
discussion about not trying to determine causality on page
6?
Page 63, Table 20, row 6, and Page 73, Table 20, rows 5
and 6: Care should be taken to provide full product name.
Eola is the manufacturer name and Eola makes some diet
products that are laxative-based which would be
inappropriate for inclusion in this report.
Page 72 and elsewhere in Table 20: What does implicit
review mean?

This was defined in the report.

Page 91 and elsewhere: Replace psyc with psychiatric.

This change was made.

Page 103: Paragraph 1: Use of term causality should be

Causality has been removed from this
reconciled with discussion on page 6 regarding the intent of revision.
the report. Last two sentences would be more accurate if
changed to: Definite causality for adverse events cannot be
determined from case reports. When an adverse event is
very serious it may be infeasible or unethical to conduct a
de-challenge/re-challenge test for causality.
Given the short time frame RAND has to fulfill its contractual We indicate in the report that this section
obligations, the peer review process also has necessarily
did not receive the same level of peer
been severely time-constrained, not to mention coincidental review as the other portion of the report.
with the year-end holiday period. This is regrettable but I
have had the opportunity to review these drafts and reflect
upon what they say in general and how it is said. I have not
had sufficient time to review the details of the reports and
tables for accuracy which is almost certainly true for the

A3-112

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

other peer reviewers. Any claim that the Metabolife Report

in particular has been "peer-reviewed" should be qualified,
as a true peer review was not possible under the
circumstances.
The Metabolife Report spends a large proportion of its
content describing the poor quality of the reports, limitations
of the records and methods, the short time frame, the many
thousands of files, etc., making it difficult to see a scientific
value to this review under these circumstances.

We were requested to do this review as

part of the contract.

A major criticism of the Metabolife Report is the disconnect

between the final limitation listed on p7, which clearly states
that ". . . case reports are in general not considered
sufficient evidence to draw conclusions about causality",
and the overall tone of the report. Any reader will be led to
believe that this report links the occurrence of adverse
events to the consumption of ephedra, despite the
limitations listed at the end. Even though the words
"possible" and "may" do appear in the report, the terms and
phrases used in the methodology, the detailed and
repetitious descriptions of the case reports, and the results,
are all written with such a factual tone that there can be no
doubt that this report will be interpreted to mean that these
events were caused by ephedra. The report should be
rewritten to state the study's major limitations (p7) at the
beginning, i.e., that these case reports cannot be ". . .
considered sufficient evidence to draw conclusions about
causality". Then the report should state at the outset in clear
terms that the purpose of the analysis was not to establish
or prove that there is a risk of serious adverse events, since
AERs are not suitable to that task. The purpose was to
determine whether or not this database might be useful to
"generate a" (rather than "support the") "hypothesis that
ephedra may cause rare serious adverse events", to quote
conclusions on p7. The introduction should also state the
fact, which is not a conclusion of the study but was included
as the final statement of conclusions (p7), that "A
hypothesis-testing study, such as a case-control study, is
necessary to prove or disprove this hypothesis". Each
statement in all sections should be carefully examined and
rewritten if found to be interpretable as drawing a link
between ephedra and effects. Phrases such as "instances
of serious adverse events such as death, heart attack, or
stroke" are repeated several times which undoubtedly will
lead to the impression that these are caused by ephedra. It
should be made more clear that the Metabolife Report, as
well as the review of the FDA AERs and published case
reports, are part of the effort to explore the hypothesis that
ephedra may cause rare serious adverse events. Common
sense and a rudimentary understanding of pharmacology
will lead to the conclusion that this clearly is possible,
depending principally upon the dose Anything at a

RAND did not generate the hypothesis that

ephedra causes serious adverse events,
that hypothesis was already generated and
one reason why our report was
commissioned. Certainly the existence of
serious adverse events in otherwise
healthy young adults must be considered
support for this hypothesis, just as the
lack of such events would be considered a
lack of support. It is not proof of a causal
relationship, and we say so, repeatedly.
We also note that the concern about out
report being overly suggestive that the
case reports imply a cause and effect
relationship is not shared by numerous
other reviewers, who believe just the
opposite, that our report downplays the
possibility of a causal relationship.

A3-113

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

sufficiently high exposure can cause adverse events.

Continued analysis of deficient and flawed adverse event
databases cannot and will not lead to any conclusions about
causality, but hurried evaluation, suggestive language, and
imprecise wording can lead to perceptions of cause and
effect that are not scientifically supported.
Turning to the revised RAND Report, it is difficult to
comment on the "Safety Assessment Excerpt" without
seeing and understanding how it is used and referenced in
the rest of the report. Most of the comments that I made
previously still apply to this draft because 99% of the safety
assessment deals with adverse event reports which are
flawed and inconclusive. Some limited peer review of the
introduction and conclusion sections of the completed report
should be permitted to assure that the wording of these
sections avoids the continuing problem of implications that a
cause and effect relationship can be established from the
number of AERs, or the exhaustive treatment given to the
AERs, in the report. The language in the previous draft has
been changed to reflect the fact that the "causality scale"
leads to erroneous and exaggerated conclusions.
Nevertheless, the new scheme of classification, using the
terminology "sentinel" instead of causal, is still suggestive
that these reports can be used for interpreting cause and
effect. To help avoid this problem, there should be added to
the explanation of the term "sentinel" on p7 of the Metabolife
Report and on p6 of the Safety Assessment Excerpt that
adverse events, even serious events, are commonly
idiopathic in etiology, and that therefore the lack of any
known cause combined with known consumption of ephedra
is not meant to imply that ephedra was the cause -- the
intent is simply to show which events could potentially have
been caused by ephedra, given the lack of a known cause,
with the understanding that a cause and effect relationship
for ephedra cannot be established from such reports.
I have not been able to review in any detail the descriptions
of events categorized in the Metabolife Report or in the
Safety Assessment Excerpt. Nonetheless, a cursory review
indicates that the criteria established for "sentinel" events in
particular has not been met in a significant number of these
cases, and these reports should be reviewed with this in
mind. For example., RAND did not have access to the
results of the autopsy in the first death listed as a sentinel
event on p26, and availability of autopsy results is
appropriately listed as a criterion for qualification as a
sentinel event on p7. Also, there are a number of cases
described as sentinel events where the individuals are also
described as long-term smokers, alcoholics, or drug
abusers. These and other cases do not appear to meet the
criterion that sentinel events be idiopathic when these
conditions are known to be risk factors for events at issue.

A3-114

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

The revised report on p7 states that the Office of Dietary

Supplements had given RAND a key question concerning
"the relationship between dose and the likelihood of serious
adverse events". The authors, however, ". . . do not believe
such an analysis is justifiable on the case report evidence. .
. ." It seems to me, if evidence is insufficient to be evaluated
for dose relationship, then any such evaluation is
unjustifiable, which speaks to the point I made in my
previous review that reliance on flawed case reports can
only lead to flawed analysis and conclusions.
Similar to previous comments on the Metabolife report
above, the limitations and lack of ability to draw conclusions
from AERs should be stated clearly up front in the Safety
Assessment Excerpt. The language used to describe the
large number of reports clearly suggests causality, even if
not intended. The preponderance of the description in the
safety section leads the reader to conclude early on that
ephedra must be responsible for these effects. The very
brief description of controlled trials is dismissive of strong
evidence for ephedra safety, and the extensive toxicology
database is completely ignored. Therefore, the safety
section continues to be unbalanced by the absence of
objective evidence in contrast to the voluminous treatment
given to the case reports.
I agree with the statements in the revised report (p6)
concerning the variability and subjectivity of interpretation of
case reports. This is a principal reason for my objection to
their consideration being the centerpiece of the report's
safety assessment.
It is an important exercise, and RAND has done as thorough
a review as could be expected. It is extremely important,
therefore, that readers of this report not be led to an
impression that the repetitive descriptions of large numbers
of case reports can be interpreted as evidence for cause
and effect. Clearly this is and will be the message unless
the introduction, methods and language throughout are
consistent with the messages about limitations, insufficiency
for causality, and the need for a conclusive study of a
different kind, i.e., a case-controlled study to add to the
existing objective clinical evidence.
As a final point in this regard, the "Conclusions" section on
p103 of the Safety Assessment Excerpt should be revised
to remove the implication that RAND has concluded that the
case reports are useful to establish causation or that the
case reports establish that there is in fact a risk of serious
adverse events. The reports generate a hypothesis, and
whether a risk of serious events exists as well as the
estimate of the level of any risk needs to be determined
through scientific studies, not review of additional case
reports. In particular, the sentence beginning with "For rare
outcomes" in the first paragraph should be revised to make

A3-115

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

it clear that the review of case reports was to assess

whether these reports generate a hypothesis that ephedra
might cause rare outcomes.
In addition, the third bullet should be revised to avoid the
implication that the lack of other identifiable causes
combined with ephedra consumption establishes causation.

We added this important qualifier.

RAND's conclusion that further analysis of case reports is

pointless is the key to moving forward with a scientific
evaluation of ephedra and the resolution of a controversy
that has been created by over-focusing on case reports.
This point should, therefore, be made in clear terms at the
very beginning of the completed Ephedra Report.

The Conclusions is the appropriate place

for this conclusion.

We would like to see summary tables of the sentinel and

These tables are now added.
possible sentinel events by ephedra use, by ephedrine use;
by gender; by broad age groups; by category of AER. The
long tables listing each event are not sufficient.
As to the adverse consequence conclusions it would seem
appropriate to summarize the events for ephedrine as they
are 'bulleted' for ephedra. Right now, it looks as if there is
no conclusion on the sentinel and possible such events for
ephedrine.

This change was made.

A recommendation is made for scientific studies of ephedra

risk. No comment is made on whether it would be
appropriate to also do this for ephedrine. For the present
data, one could argue that PPA-like case-control studies
should be generated for other ephedra and ephedrine
products.

These changes were made.

No response
Adverse Event Data from Randomized Trials. Methods.
RAND identified 44 randomized, controlled studies, and a
pooled meta-analysis was conducted of the risk of adverse
events in treated vs. placebo groups for the most commonly
reported adverse events. Risk was significantly elevated for
psychiatric symptoms (OR 3.24, 95% CI 1.67-6.58),
autonomic hyperactivity (OR 2.91, 95% CI 1.84-.70),
nausea and vomiting (OR 2.37, 95% CI 1.51-3.78), and
palpitations (OR 2.11, 95% CI 1.16-4.02). The risk was
elevated, but not significantly, for hypertension (OR 1.86,
95% CI 0.39-11.74). Tachycardia was reported in only one
study. The methods used are standard; the analyses appear
appropriate.
Adverse Event Data from Randomized Trials. Methods. The
subgroup analyses of adverse events of ephedrine (+)
caffeine were said to be similar to the main analysis. This
data may be important and should be presented in greater
detail. The reason for this is that caffeine can potentiate the
CNS stimulant effects of this class of drugs
(sympathomimetic aminies).

A3-116

The results are the same because the

ephedrine plus caffeine studies contribute
the vast majority of the data to this
analysis. So we have said as much as we
can about this.

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

Adverse Event Data from Randomized Trials. Methods. A

We did this analysis and included it in the
second recommendation is to conduct the pooled analysis
results.
combining similar adverse event groups in an attempt to
reanalyze for a dose-response effect from ephedra or
ephedrine (+) caffeine. For example, it would make sense to
combine palpitations, tachycardia, and hypertension in such
an analysis, since all are cardiovascular, sympathomimetic
events.
Adverse Event Data from Randomized Trials. Methods. I
would also like to see a pooled analysis of headache, and
add this to Table 17. The reason for this is that headache
may be a prodrome to more serious neurologic events, and
was present in all three cases I reported at the 1996
meeting of the American Academy of Neurology.

This adverse event analysis was added.

Adverse Event Data from Randomized Trials. Potential for

Bias. Since this data is a meta-analysis, there is little
opportunity for bias.

No response

Adverse Event Data from Randomized Trials .Clarity of

Reporting. The writing is clear. The report (I) would flow
better if the meta-analysis section was stand-alone and
separate from the case report analysis.

It is separate in the final version of the

report.

Adverse Event Data from Randomized Trials. Conclusions. We reworded this to try and improve
clarity.
The conclusions are to the point. However, I find the metaanalysis conclusion somewhat lacking in methodologic
content and discussion. To be more useful, expansion of the
authors critical point in 1, page 25, should be added to the
Conclusion section (p.103). The conclusions would more
properly read: 1.There is sufficient evidence. (same). 2.
Safety data from relatively small clinical trials of
ephedra/ephedrine are unlikely to reveal rare but serious
adverse events, those that may occur at a rate of less than
1/1000. Thus, such data cannot be used to conclude that
ephedra/ephedrine does not cause such serious adverse
events. In addition, it is likely that, in some of these trials,
differential drop out of treated patients related to a higher
rate of milder adverse events could have removed subjects
at higher risk for more serious events.
Adverse Even Data from Reported Cases. Methods. It is
important to point out that the authors utilized a very much
more conservative method to identify the likelihood of
association with ephedra/ephedrine than that reported by
both Haller and Benowitz5 for cardiovascular and central
nervous system events, and by Samenuk et al for
cardiovascular events. The authors should point out the
differences with these studies, and how these differences
may have led to different counts of adverse events in the
main categories in Reports I and II. A table highlighting the
differences with Haller and Benowitz would allow clearer
comparison of categories One important difference is that

A3-117

Since we dropped a causality

assessment from this revision, we dont
think such a comparison is valid. We do
acknowledge in the limitations that our
methods are more conservative than those
used by some other groups.

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

RAND dropped all cases with any alternative explanation or

competing cause to probably not related. Haller and
Benowitz, however, considered events at least possibly
related, even in the face of co-existing or pre-existing
condition, if those conditions themselves could be severely
exacerbated by ephedra alkaloids (e.g., hypertension, some
psychiatric conditions).
Adverse Even Data from Reported Cases. Methods. One
table should summarize all of the reviewed cases by
adverse event type (e.g., death, seizure) and by authors
conclusions regarding category (sentinel, possible sentinel,
probably not related, and insufficient information). For
example, of 41 reported seizure cases, only two were
deemed sentinel cases. This may highlight the
insufficiency of available data with which one may judge
likelihood of association. For example, I have detailed
knowledge of seizure case 13408. Even with the authors
criteria, this case should be classified as sentinel.

These tables are now included. We

acknowledge that limitations of the source
documents limit our ability to draw
conclusions.

Adverse Even Data from Reported Cases. Potential for

This limitation was acknowledged in the
Bias. With the conservative approach described under
appropriate section.
Methods, there is potential for serious misclassification of
cases, primarily in the direction of probably not related or
insufficient data. It is much less likely that misclassification
substantially went in the other direction.
Adverse Even Data from Reported Cases. Clarity of
Reporting. As mentioned above, at least one or two other
summary tables would be helpful to the reader.

These tables are now included.

Adverse Even Data from Reported Cases. Conclusions. The

conclusions reached on p.103 regarding the case report
assessment are not very helpful in moving things forward on
this issue. There is an underlying assumption that, if
causality cannot be proven from passively reported cases
with poor documentation, then it may take a case control
study to do so. There are several problems with these
conclusions: (1) If ephedra were an FDA-approved drug, its
use would have likely been banned related only to the sheer
number of serious adverse events reported. Even if one
accepts only the sentinel and possible sentinel events
reported here, or those reported by Haller and Benowitz, or
the cases from Texas or Rochester6, the likelihood of
association, to most clinicians, would be overwhelming. (2)
There should be substantial discussion added to the report
related to other converging lines of evidence one would
normally wish to include in an assessment of causal
relations. These would include: (a) Expected actions of
sympathomimetic amines, including effects on the
peripheral vascular system, and the biologic plausibility of
association with milder and severe adverse events. (b) A
summary of the extensive literature on the potential for the
look alike drugs such as PPA to cause similar serious

We note there is a great deal of

controversy among experts about whether
case reports are sufficient to conclude
cause and effect relationship with serious
adverse events. The other kinds of
evidence cited by this reviewer were
outside our scope. We do think a case
control study is possible, and that the
controversy is likely to continue to rage
until such a hypothesis-testing study is
performed.

A3-118

Appendix 3. Reviewer Comments (continued)

Reviewer Comment

Rand Response

adverse events. For example, both PPA and ephedrine are

known to be associated with angiitis. (c) Evidence from
animal studies or basic neuroscience studies related to
adverse events of ephedrine. (3) If one of the problems
relates to poor reporting to the FDA or from the
manufacturers, it would seem that, at a minimum, clearer
reporting standards should be established. (4) It would be
extremely difficult to conduct the type of case control study
recommended. The serious events are rare, and among the
major event categories (e.g., seizures), ephedra is not likely
a frequent cause. I have thought about how to conduct such
a study, either via emergency departments or poison control
centers. However, there would be serious methodologic
issues in proper case and control specification. Can we
really afford to wait for such an imperfect study to be
conducted? Is there really any justification whatsoever not
to ban unfettered use and marketing of these
sympathomimetic amines in pharmacologic doses?

A3-119