Original Contribution

JAMA. 2003;289(16):2083-2093. doi: 10.1001/jama.289.16.2083

.
.

Effects of Comprehensive Lifestyle Modification on Blood Pressure Control

Main Results of the PREMIER Clinical Trial
Writing Group of the PREMIER Collaborative Research Group
[+] Author Affiliations
Authors/Writing Group of the PREMIER Collaborative Research Group:
Lawrence J. Appel, MD (chair), Departments of Medicine, Epidemiology, and
International Health (Human Nutrition), Johns Hopkins Medical Institutions,
Baltimore, Md; Catherine M. Champagne, PhD, and David W. Harsha, PhD,
Pennington Biomedical Research Center, Baton Rouge, La; Lawton S. Cooper, MD,
and Eva Obarzanek, PhD, National Heart, Lung, and Blood Institute, Bethesda, Md;
Patricia J. Elmer, PhD, Victor J. Stevens, PhD, and William M. Vollmer, PhD, Kaiser
Permanente Center for Health Research, Portland, Ore; Pao-Hwa Lin, PhD, and Laura
P. Svetkey, MD, Duke Hypertension Center and Sarah W. Stedman Center for
Nutritional Studies, Duke University Medical Center, Durham, NC; and Deborah R.
Young, PhD, Department of Kinesiology, University of Maryland, College Park.
Corresponding Author and Reprints: Lawrence J. Appel, MD, MPH, Departments
of Medicine; Epidemiology, and International Health (Human Nutrition), Johns
Hopkins Medical Institutions, 2024 E Monument St, Suite 2-645, Baltimore, MD
21205-2223 (e-mail: [email protected]).
More author information
Next Section
ABSTRACT
Context Weight loss, sodium reduction, increased physical activity, and limited
alcohol intake are established recommendations that reduce blood pressure (BP). The
Dietary Approaches to Stop Hypertension (DASH) diet also lowers BP. To date, no
trial has evaluated the effects of simultaneously implementing these lifestyle
recommendations.
Objective To determine the effect on BP of 2 multicomponent, behavioral
interventions.
Design, Setting, and Participants Randomized trial with enrollment at 4 clinical
centers (January 2000-June 2001) among 810 adults (mean [SD] age, 50 [8.9] years;
62% women; 34% African American) with above-optimal BP, including stage 1
hypertension (120-159 mm Hg systolic and 80-95 mm Hg diastolic), and who were
not taking antihypertensive medications.
Intervention Participants were randomized to one of 3 intervention groups: (1)
"established," a behavioral intervention that implemented established
recommendations (n = 268); (2) "established plus DASH,"which also implemented
the DASH diet (n = 269); and (3) an "advice only" comparison group (n = 273).
Main Outcome Measures Blood pressure measurement and hypertension status at 6
months.
Results Both behavioral interventions significantly reduced weight, improved fitness,

and lowered sodium intake. The established plus DASH intervention also increased
fruit, vegetable, and dairy intake. Across the groups, gradients in BP and hypertensive
status were evident. After subtracting change in advice only, the mean net reduction in
systolic BP was 3.7 mm Hg (P<.001) in the established group and 4.3 mm Hg
(P<.001) in the established plus DASH group; the systolic BP difference between the
established and established plus DASH groups was 0.6 mm Hg (P = .43). Compared
with the baseline hypertension prevalence of 38%, the prevalence at 6 months was
26% in the advice only group, 17% in the established group (P = .01 compared with
the advice only group), and 12% in the established plus DASH group (P<.001
compared with the advice only group; P = .12 compared with the established group).
The prevalence of optimal BP (<120 mm Hg systolic and <80 mm Hg diastolic) was
19% in the advice only group, 30% in the established group (P = .005 compared with
the advice only group), and 35% in the established plus DASH group (P<.001
compared with the advice only group; P = .24 compared with the established group).
Conclusion Individuals with above-optimal BP, including stage 1 hypertension, can
make multiple lifestyle changes that lower BP and reduce their cardiovascular disease
risk.

KEYWORDS: blood pressure, diet, fat-restricted, diet, sodium-restricted, exercise,

hypertension, lifestyle, randomized trials.
High blood pressure (BP) is a common, powerful, and independent risk factor for
cardiovascular disease (CVD). Almost 50 million US adults, or approximately 25% of
the US adult population, have hypertension, defined as BP of 140/90 mm Hg or
higher and/or current use of antihypertensive medication.1 The prevalence of
hypertension increases progressively with age, so that more than half of all
individuals aged 60 years or older in the United States have hypertension.2 The
estimated lifetime risk of developing hypertension is 90%.3
Above-optimal BP that is not in the hypertensive range also confers excess CVD
risk.4 In fact, almost a third of BP-related deaths from coronary heart disease are
estimated to occur in nonhypertensive individuals with a systolic BP of 120 to 139
mm Hg or diastolic BP of 80 to 89 mm Hg.5 Therefore, reduction of BP to optimal
levels, control of hypertension, and prevention of the age-related increase in BP
remain major public health priorities.
Current national recommendations for the prevention and treatment of high BP
emphasize nonpharmacological therapy, also termed "lifestyle modification."6-7
Lifestyle modifications that effectively lower BP are weight loss, reduced sodium
intake, increased physical activity, limited alcohol consumption, and the Dietary
Approaches to Stop Hypertension (DASH) diet.6-8 The DASH diet emphasizes
consumption of fruits, vegetables, and low-fat dairy products; includes whole grains,
poultry, fish, and nuts; and is reduced in fats, red meat, sweets, and sugar-containing
beverages. As such, the DASH diet has reduced levels of total fat, saturated fat, and
cholesterol and increased levels of potassium, calcium, magnesium, fiber, and
protein.9
These lifestyle modifications are recommended in nonhypertensive individuals with
above-optimal BP. Lifestyle modification is also recommended as initial therapy in
stage 1 hypertension (for up to 12 months in those without other risk factors [risk
class A] or for up to 6 months in those with other risk factors [risk class B]).7 For

individuals taking BP medication, lifestyle modification is recommended as

adjunctive therapy to lower BP. Although lifestyle therapies are generally
recommended as a group, no previous trial has evaluated the effects of implementing
these recommendations simultaneously, and no trial has tested the feasibility of
implementing the DASH diet in free-living persons.
Previous Section
Next Section
METHODS
The rationale for the PREMIER clinical trial10 has been published. Participating
institutions included the National Heart, Lung, and Blood Institute Project Office
(Bethesda, Md), the coordinating center (Kaiser Permanente Center for Health
Research in Portland, Ore), and 4 clinical centers (Johns Hopkins University,
Baltimore, Md; Pennington Biomedical Research Center, Baton Rouge, La; Duke
University Medical Center, Durham, NC; and Kaiser Permanente Center for Health
Research, Portland, Ore). Institutional review boards at each center and an external
protocol review committee approved the protocol. Each participant provided written
consent.
Study Participants
The target population consisted of generally healthy adults with above optimal BP
including individuals with stage 1 hypertension who met Joint National Committee on
Detection, Evaluation, and Treatment of High Blood Pressure (JNC-VI) criteria for at
least a 6-month trial of nonpharmacological therapy.7 Persons were eligible if they
were not taking antihypertensive medication and had a systolic BP of 120 to 159 mm
Hg and diastolic BP of 80 to 95 mm Hg, based on the mean BP across 3 screening
visits. Nonhypertensive individuals with above optimal BP (120-139 mm Hg systolic
and/or 80-89 mm Hg diastolic) were included because of the potential for preventing
hypertension and the excess CVD risk associated with BP in this range.4 Individuals
with stage 1 hypertension (140-159 mm Hg systolic and/or 90-95 mm Hg diastolic)
were included because of the potential for nonpharmacological control of
hypertension.
Other inclusion criteria were 25 years of age or older and body mass index (BMI) of
18.5 to 45.0 (measured as weight in kilograms divided by the height in meters
squared). Major exclusion criteria were regular use of drugs that affect BP, JNC-VI
risk category C (target organ damage and/or diabetes), use of weight-loss medications,
prior cardiovascular event, congestive heart failure, angina, cancer diagnosis or
treatment in past 2 years, consumption of more than 21 alcoholic drinks per week, and
pregnancy, planned pregnancy, or lactation. Although individuals with diabetes were
excluded, persons with other cardiovascular risk factors (ie, cigarette smoking and
dyslipidemia) could enroll. Vitamin and mineral supplement use was not an exclusion.
Trial Conduct
Participants were recruited using mass mailings, community-based screening, and
mass-media announcements. Enrollment began in January 2000 and ended in June
2001. For logistical purposes, participants were enrolled in 3 or 4 cohorts at each
center. Baseline data were collected during 3 screening visits and a randomization
visit, each scheduled at least 7 days apart. Follow-up data were collected at 1 visit 3
months after randomization and at 3 visits 6 months after randomization. Participant
flow during the trial is shown in Figure 1.

View larger version:

In this page In a new window
Download as PowerPoint Slide
Figure 1. Participant Flow in the PREMIER Clinical Trial
BP indicates blood pressure; DASH, Dietary Approaches to Stop Hypertension.
Randomization
Randomization assignments were made centrally by a computer program. Clinical
center staff then notified participants of their assigned group. Assignments were
stratified by clinic and hypertension status; the randomization block size was 24.
Eligible participants were randomly assigned to 1 of 3 groups: (1) an "advice only"
comparison group; (2) a behavioral intervention, termed "established" that
implemented traditional lifestyle recommendations,11 ie, weight loss among those
who were overweight, reduced sodium intake, increased physical activity, and limited
alcohol intake among those who drank alcohol; or (3) a behavioral intervention,
termed "established plus DASH" that implemented the same traditional
recommendations plus the DASH diet.7
Advice Only Group
An interventionist, typically a registered dietitian, discussed nonpharmacological
factors that affect BP (weight, sodium intake, physical activity, and the DASH diet)
and provided printed educational materials. This advice was provided in a single 30minute individual session immediately following randomization. Counseling on
behavior change was not provided. No further contact with the interventionist
occurred until after completion of the data collection visits at 6 months.
Behavioral Interventions
Participant goals for both the established and established plus DASH interventions
were as follows: (1) weight loss of at least 15 lb (6.8 kg) at 6 months for those with a
BMI of at least 25, (2) at least 180 min/wk of moderate-intensity physical activity, (3)
daily intake of no more than 100 mEq of dietary sodium, and (4) daily intake of 1 oz
or less of alcohol (2 drinks) for men and oz of alcohol (1 drink) for women.
The established and the established plus DASH interventions differed from each other
with respect to certain dietary goals and the strategies to achieve weight loss. Only the
participants in the established plus DASH intervention received instruction and
counseling on the DASH diet. In this intervention, participant goals designed to
accomplish the DASH diet were increased consumption of fruits and vegetables (9-12
servings/d) and low-fat dairy products (2-3 servings/d), and reduced intake of
saturated fat (7% of energy) and total fat (25% of energy). The established
intervention did not have goals for fruit, vegetable, or dairy intake; the goal for
saturated fat was 10% of energy or less, and the goal for total fat was 30% of energy
or less. To achieve weight loss, both interventions emphasized increased physical
activity and reduced total energy intake; in addition to these strategies, the established
plus DASH intervention also emphasized substitution of fruits and vegetables for
high-fat, high-calorie foods.
The format and contact pattern of the established and established plus DASH
interventions were identical. During the initial 6 months, there were 18 face-to-face
intervention contacts (14 group meetings and 4 individual counseling sessions).

Participants in both interventions kept food diaries, recorded physical activity, and
monitored calorie and sodium intake. Participants in the established plus DASH group
also monitored intake of fruits, vegetables, and dairy products and monitored their
intake of fat.
Measurements
Staff who were masked to randomization assignment collected measurements. Blood
pressure measurements were obtained by trained, certified individuals who used a
random zero sphygmomanometer. The BP measurement protocol was similar to
protocols used in prior studies.12 -13 After the participant sat quietly for 5 minutes,
the observer measured BP in the right arm with an appropriately sized cuff. At each
visit, 2 BP measurements separated by at least 30 seconds were obtained. Systolic BP
was the appearance of the first Korotkoff sound, and diastolic BP was the
disappearance of Korotkoff sounds. At each assessment point, BP was the mean of all
available measurements (baseline [8 BP measurements across 4 visits], 3-month
assessment [2 BP measurements at 1 visit], and 6-month assessment [6 BP
measurements across 3 visits]).
Weight was measured using a calibrated scale, and height was measured using a wallmounted stadiometer. Other data included the Rose Angina questionnaire14 ; a
medication questionnaire; a symptoms/adverse effects questionnaire; 24-hour urine
collections for sodium, potassium, phosphorus, and urea nitrogen; submaximal
treadmill tests; waist circumference; 24-hour dietary recalls; fasting blood analysis;
and 7-day physical activity recalls. Each of these measurements was obtained at
baseline and 6 months after randomization.
Intake of nutrients and food groups was assessed from unannounced 24-hour dietary
recalls conducted by telephone interviewers.15 Two recalls (one obtained on a
weekday and the other on a weekend day) were obtained at baseline and 6 months by
the Diet Assessment Center of Pennsylvania State University. Nutrient and food group
intakes were then calculated using the Nutrition Data System Version NDS-R 1998
(University of Minnesota). Biomarkers of dietary intake were 24-hour urinary
excretion of sodium, potassium (reflecting fruit and vegetable intake), phosphorus
(reflecting dairy intake), and urea nitrogen (reflecting protein intake). Alcohol intake
was obtained from questionnaire.
Cardiorespiratory fitness was assessed using a submaximal treadmill exercise test
developed for the PREMIER trial. This 2-stage, 10-minute protocol was designed to
achieve an age- and sex-specific effort of moderate intensity.16 The first stage
achieved a light-intensity effort (approximately 40% estimated maximal metabolic
equivalents [METs]), followed by a second stage of moderate intensity
(approximately 60% estimated maximal METs). The main fitness outcome was heart
rate at the end of stage 2 or the last available heart rate from stage 1 for participants
who did not complete stage 2. A 7-day physical activity recall was used to assess
physical activity.17 Participants who reported 35 kcal/kg or less daily of physical
activity were classified as sedentary.18
Specific Aims and Outcomes
The specific aims of the trial were to test the effects of the established intervention
compared with the advice only intervention; the effects of the established plus DASH

intervention compared with the advice only intervention; and the effects of the
established plus DASH intervention compared with the established intervention. The
primary outcome was change in systolic BP from baseline to 6 months. Hypertension
status and change in diastolic BP at 6 months were secondary outcomes. Blood
pressure measurements were censored if the participant reported taking any
antihypertensive medication or other medications known to have major BP effects (eg,
oral steroids). Hypertension was defined as a mean BP of 140/90 mm Hg or higher or
use of antihypertensive medication.
Although the intervention programs lasted 18 months, the protocol-specified, primary
outcome assessment occurred at 6 months because national guidelines recommend
that individuals with persistent BP of 140/90 mm Hg or higher after a period of
lifestyle modification be referred for medication treatment.7 During the design of the
trial, we anticipated that approximately 30% of participants would have stage 1
hypertension at baseline.10 Hence, we expected that a large number of individuals
would need to be referred for medication treatment at 6 months, requiring censoring
of their BP data, and that medication treatment would occur differentially across the 3
randomized groups. Defining the primary outcome at 6 months reduced the risk of
bias and ensured that we would have a maximum number of BP measurements for
analysis.
Data Analysis
Blood pressure data were analyzed using a linear regression model in which change in
BP (mean 6-month value mean baseline) was regressed on indicators of the 2
behavioral interventions, indicators of clinical center and cohort, and baseline BP. In
prespecified subgroup analyses (hypertensive and nonhypertensive), the models also
included a main effect for the subgroup indicator and interactions between this
indicator and the 2 treatment group indicators. The effects of the interventions did not
differ by clinical center (P = .62 for center treatment interaction for systolic BP and
P = .54 for diastolic BP).
Primary analyses of BP change are based on intention to treat. For individuals without
BP at the 6-month assessment and for those who had been taking antihypertensive
medication, 3-month BP measurements were carried forward; if a 3-month BP
measurement was unavailable, values were imputed using a "hot deck" procedure that
drew values from participants in the advice only group.19 We also conducted post-hoc
"on treatment" analyses limited to those participants in the established and established
plus DASH groups who attended at least 15 intervention sessions.
For all other variables, including hypertension status, we used available data and did
not impute values for missing data. To analyze continuous indicators of intervention
effects, such as change in body weight, we used a similar analytic model. We used the
Mantel-Haenszel 2 test for 2 2 tables to compare the proportion of individuals
meeting intervention targets at 6 months.20 Because the focus of these analyses was
the proportion actually meeting target at 6 months in each group and not necessarily
the change from baseline status, these analyses did not condition on initial status.

Hypertension status at 6 months was assessed separately for those who were and were
not hypertensive at baseline, reflecting persistent and incident hypertension,
respectively. We also compared the prevalence of hypertension in all participants.

Pairwise differences in the incidence, persistence, and overall prevalence of

hypertension between treatment groups were also assessed using the Mantel-Haenszel
test.20
Based on a planned sample size of 800 (267 per group), the study had 90% power to
detect pairwise between-group differences in systolic BP of 1.6 to 1.8 mm Hg in the
whole sample, 3.2 to 3.6 mm Hg among hypertensive participants (assuming that 30%
of the individuals in the sample were hypertensive), and 1.7 to 1.9 mm Hg among
nonhypertensive participants.
All analyses were performed using SAS version 8 (SAS Institute Inc, Cary, NC).
Nominal P values are presented. For a given outcome, we only considered the
pairwise contrasts vs the advice only intervention to be significantly different if at
least one of them achieved a P value of <.025; in that case, the other contrast with
advice only and the contrast between the established and established plus DASH
interventions were evaluated at the .05 level.21
Previous Section
Next Section
RESULTS
A total of 810 participants were enrolled in the trial (Figure 1). Baseline
characteristics were similar in the randomized groups (Table 1). The mean (SD) age
was 50.0 (8.9) years, 62% were women, and 34% were African Americans. Of the 279
African Americans, 74% were women. The participants were generally overweight
and sedentary. Mean (SD) systolic and diastolic BP were 134.9 (9.6) and 84.8 (4.2)
mm Hg. Among the 38% of participants with hypertension, mean (SD) systolic and
diastolic BP were 143.9 (7.6) and 87.5 (4.3) mm Hg; corresponding BP measurements
in the nonhypertensive participants were 129.5 (5.8) and 83.2 (3.1) mm Hg. Six
months after randomization, 94% of participants had their BP measured at 1 or more
visits; 87% attended all 3 visits.

View this table:

In this window In a new window
Download as PowerPoint Slide
Table 1. Baseline Characteristics by Randomized Group*
Intervention Attendance and Effects
Of the 18 intervention sessions offered during the intial 6 months, 70% of participants
in the established group attended at least 15 sessions; just 8% attended 5 sessions or
less. In the established plus DASH group, corresponding data were 78% and 7%.
Mean (SD) attendance was 14.5 (4.5) and 15.4 (4.4) sessions, respectively.
Differences in weight, physical fitness, and diet among the randomized groups were
achieved. Table 2 displays intervention outcomes, and Table 3 lists the number of
individuals who reached the intervention goals. Weight loss occurred in each group,
including the advice only group. While changes in physical activity did not differ
among the groups, fitness significantly improved in both behavioral interventions.
Alcohol intake was low and did not change in any group.

View this table:

In this window In a new window

Download as PowerPoint Slide
Table 2. Intervention Outcomes at Baseline and at 6 Months by Randomized Group
View this table:
In this window In a new window
Download as PowerPoint Slide
Table 3. Participants Meeting Intervention Goals at Baseline and at 6 Months by
Randomized Group
Mean reductions in urinary sodium excretion occurred in both behavioral
interventions, but only the reduction in the established group differed significantly
from that of advice only group. However, in both behavioral intervention groups, the
percentage of individuals who achieved the trial goal of less than 100 mEq/d differed
significantly from the advice only group (Table 3). Also, based on 24-hour dietary
recall data, both behavioral interventions significantly reduced sodium intake in
comparison with the advice only group (data not shown).
In the established plus DASH group, fruit and vegetable intake increased significantly
compared with the other 2 groups; parallel changes in urinary potassium excretion
occurred. One third of participants in the established plus DASH group, but only 6%
of participants in the other 2 groups, consumed the goal of 9 or more servings of fruits
and vegetables per day at 6 months. Compared with the advice only and established
groups, consumption of dairy products increased significantly in the established plus
DASH group as did dietary calcium intake and net urinary phosphorus excretion. The
percentage of established plus DASH participants who consumed 2 or more dairy
servings was 59%. Saturated and total fat consumption significantly decreased in both
intervention groups.
Blood Pressure Effects
Blood pressure declined progressively over time in each group (Figure 2). From
baseline to 6 months, mean (SD) reductions in systolic BP were 6.6 (9.2) mm Hg in
the advice only group, 10.5 (10.1) mm Hg in the established group, and 11.1 (9.9) mm
Hg in the established plus DASH diet group. Corresponding diastolic BP reductions
were 3.8 (6.3), 5.5 (6.7), and 6.4 (6.8) mm Hg, respectively. In hypertensive
participants, mean (SD) reductions in systolic BP were 7.8 (10.3), 12.5 (11.5), and
14.2 (10.1) mm Hg, and mean (SD) reductions in diastolic BP were 3.8 (7.1), 5.8
(7.0), and 7.4 (7.1) mm Hg, respectively. In nonhypertensive participants, mean (SD)
reductions in systolic BP were 5.8 (8.4), 9.4 (9.1), and 9.2 (9.3) mm Hg, and mean
(SD) reductions in diastolic BP were 3.8 (5.8), 5.3 (6.5), and 5.8 (6.6) mm Hg,
respectively.

View larger version:

In this page In a new window
Download as PowerPoint Slide
Figure 2. Mean Systolic and Diastolic Blood Pressure (BP) Over Time by
Randomized Group
DASH indicates Dietary Approaches to Stop Hypertension.

Table 4 displays pairwise differences in BP. In all participants, nonhypertensive

participants, and hypertensive participants, the established and established plus DASH
interventions significantly reduced systolic and diastolic BP in comparison with the
advice only group. Although BP change in the established plus DASH group was
consistently greater than corresponding BP change in the established group, none of
the pairwise differences was statistically significant.

View this table:

In this window In a new window
Download as PowerPoint Slide
Table 4. Mean Between-Group Differences in Blood Pressure (BP) Change in All
Participants, Nonhypertensive Participants, and Hypertensive Participants
The pattern of results was similar in the "on treatment" analyses, which included those
individuals in the established and established plus DASH groups (70% and 78% of
participants, respectively) who attended more than 15 intervention sessions. For
contrasts with the advice only group, BP reductions in the on treatment analyses were
approximately 20% to 40% greater than corresponding reductions in the intention to
treat analyses.
Figure 3 displays the percentage of nonhypertensive participants who became
hypertensive, the percentage of hypertensive participants who remained hypertensive,
and the percentage of all participants who were hypertensive at 6 months. In each
instance, there was a gradient in hypertensive status across the 3 groups. The lowest
prevalence of hypertension (12%) occurred in the established plus DASH group; this
prevalence corresponds to a 53% risk reduction (ie, 1 relative risk) compared with
the advice only group. By 6 months, antihypertensive medication had been started in
19 participants in the advice only group, 2 participants in the established group, and 5
participants in the established plus DASH group. The established and established plus
DASH groups significantly increased the percentage of individuals who achieved an
optimal BP (<120 mm Hg systolic and <80 mm Hg diastolic; Figure 4).

View larger version:

In this page In a new window
Download as PowerPoint Slide
Figure 3. Percentage of Participants With Hypertension at 6 Months by Randomized
Group Among Nonhypertensive, Hypertensive, and All Participants at Baseline
DASH indicates Dietary Approaches to Stop Hypertension.

View larger version:

In this page In a new window
Download as PowerPoint Slide
Figure 4. Percentage of Participants With Optimal Blood Pressure at 6 Months by
Randomized Group Among Nonhypertensive Participants at Baseline, Hypertensive
Participants at Baseline, and All Participants
DASH indicates Dietary Approaches to Stop Hypertension.
Other Effects
A serious musculoskeletal injury occurred in 20 participants in the advice only group,

17 in the established group, and 16 in the established plus DASH group. One stroke, 1
transient ischemic attack, and 1 myocardial infarction occurred in the advice only
group. No cardiovascular event occurred in the established group, and 1 myocardial
infarction occurred in the established plus DASH group.
Previous Section
Next Section
COMMENT
The PREMIER trial documented that individuals with above-optimal BP, including
stage 1 hypertension, can make multiple lifestyle changes that lower BP and control
hypertension. Both of the PREMIER behavioral interventions accomplished
substantial weight loss, reduced sodium intake, and increased physical fitness.
Individuals assigned to the established plus DASH intervention also made dietary
changes consistent with the DASH diet, ie, increased their intake of fruits, vegetables,
and dairy products. In aggregate, these lifestyle changes should substantially lower
the risk of CVD as well as the risk of other chronic diseases, including diabetes,
osteoporosis, and perhaps cancer.
Trial participants were demographically heterogeneous. More than 50% were women,
and more than 30% were African American. The distributions of educational
attainment and household income were broad, but slightly skewed toward persons
with higher education and income. The BP inclusion criteria of PREMIER would
include approximately 50% of US adults. These aspects of the study suggest that trial
results should be applicable to a large portion of the US population.
Markers of adherence, including several objective measurements, confirmed that
participants in the behavioral interventions made lifestyle changes. Each behavioral
intervention led to substantial weight loss as well as increased fitness. The reduced
heart rate on the treadmill test, an objective measure of improved fitness, suggests that
participants increased their physical activity, even though self-reported physical
activity as measured by 7-day physical activity recalls did not change significantly.
Both behavioral interventions significantly reduced sodium intake, although not to the
same extent as behavioral interventions that focused exclusively on this factor.12-13 ,
22 The established plus DASH intervention also increased consumption of fruits,
vegetables, and dairy products; analyses of urinary potassium and phosphorus were
consistent with data from the 24-hour dietary recalls.
Across the 3 groups, gradients in BP and hypertensive status were evident. The
smallest BP reduction occurred in the advice only group, while the greatest BP
reduction occurred in the established plus DASH group. Hypertension control was
most successful in the established plus DASH group, in which 77% of individuals
with stage 1 hypertension at baseline had a systolic BP of less than 140 mm Hg and a
diastolic BP of less than 90 mm Hg at 6 months. In the established group, the
corresponding figure was 66%. These rates compare favorably with survey data2 and
trial data,23 in which drug therapy controls BP in approximately half of hypertensive
individuals. Hence, these behavioral interventions should be viable treatment options,
at least among those hypertensive individuals who are motivated to make lifestyle
changes.
The established and established plus DASH interventions also reduced BP in
nonhypertensive individuals with above-optimal BP. Specifically, an optimal BP level

(<120 mm Hg systolic and <80 mm Hg diastolic) was achieved in 40% and 48% of
participants assigned to these 2 interventions, respectively. Overall, this pattern of
findings suggests that even in the context of other effective lifestyle modifications,
adoption of the DASH diet further improves BP control.
Still, the BP effects attributed to the DASH diet, specifically the BP differences
between the established plus DASH and the established interventions, were less than
previously found in the DASH feeding studies,8 , 24 and none of the contrasts in
hypertension status was statistically significant. One potential reason is that
participants received an inadequate "dose" of the DASH diet. For instance, even
though mean fruit and vegetable intake increased from 4.8 to 7.8 servings per day in
the established plus DASH intervention, the latter is below what was provided in the
DASH feeding studies, namely, 9.6 servings per day.
Another plausible reason is subadditivity of intervention effects. Specifically, the net
BP effect of the DASH diet in the PREMIER trial likely underestimates the BP effects
of the DASH diet if it were implemented alone.8 , 25 It has been well documented
that the combined effect of an intervention that implements 2 or more BP-reducing
components is less than the sum of BP reductions from interventions that implement
each component alone. Subadditivity can occur from reduced adherence13 because of
the effort and complexity of making more than 1 lifestyle change. For example,
participants in the established plus DASH intervention were advised to increase their
intake of dairy products while reducing total caloric intake. Still, even in the setting of
high adherence, such as feeding studies, subadditivity occurs.24
The advice only comparison group in the PREMIER trial likely accomplished some
lifestyle modifications. Such behavior changes might have resulted from recruitment
of motivated participants; secular trends (eg, growing awareness of the obesity
epidemic); the 30-minute intervention session postrandomization; and the multiple,
regular data collection visits and contacts (4 visits and 2 dietary recalls at baseline; 1
visit at 3 months and 3 visits at 6 months, along with telephone calls and reminders).
Although before-after changes must be interpreted cautiously, the weight loss of 1.1
kg at 6 months in this group exceeded what has occurred in other trials, which have
often reported weight gain in the control group.13 Also, the 20-mEq reduction in
sodium excretion may reflect efforts to lower salt intake. Such modest changes in
behavior might have reduced BP in this group, thereby attenuating the pairwise
differences in BP between the advice only group and the 2 behavioral interventions. In
fact, the within-group BP reductions observed in the advice only group (systolic BP
reduction of 6.6 mm Hg and diastolic BP reduction of 3.8 mm Hg) greatly exceed
what was observed in the control groups of other studies.8, 12 -13,22
In the PREMIER trial, the primary outcome variables were collected at 6 months
postrandomization. The inclusion of hypertensive individuals in the PREMIER trial
precluded use of BP measured at a later follow-up visit as the primary outcome
because national guidelines recommend initiation of drug therapy for individuals with
Class B hypertension who remain hypertensive after a 6-month period of nondrug
therapy.7 Still, evidence from clinical trials suggests that as long as adherence is
sustained, BP effects persist.26 -27 Also, in longitudinal observational studies, healthy
dietary patterns indicative of long-term habits are associated with reduced CVD28 -29
and mortality.30

In addition to hypertension, high-normal BP is also associated with excess CVD risk.4

Of the general population, about 34% have BP in the nonhypertensive yet above
optimal BP range and another 14% have stage 1 hypertension.2 Furthermore, the
majority of BP-related events occur in the range of BP studied in the PREMIER trial.5
In this setting, it is reasonable to speculate that widespread implementation of the
PREMIER behavioral interventions, particularly the established plus DASH
intervention, should decrease CVD risk through reduced BP in nonhypertensive
individuals and increased BP control in hypertensive individuals.
Yet most health care insurers do not cover behavioral interventions for the prevention
and treatment of hypertension.31 Given the substantial health benefits of these
programs in improving BP control, preventing diabetes,32-33 and controlling
dyslipidemia,34 it is time to consider how such programs might be implemented,
particularly for those patients at elevated CVD risk. The costs of such programs
should be balanced against the benefits of preventing hypertension, diabetes, heart
disease, and other conditions, thereby preventing the need for medical treatments that
may be costly.
In summary, our trial results demonstrate the feasibility of comprehensive behavioral
interventions and their beneficial effects on BP and hypertension control. Benefits
extend to both nonhypertensive individuals at risk for developing hypertension and
hypertensive individuals who are not receiving medication therapy. Although we did
not study individuals receiving drug therapy, available data indicate that
nonpharmacological interventions also reduce BP in these individuals.35 -36
Ultimately, population-wide adoption of healthy lifestyles as promoted in the
PREMIER interventions should substantially reduce the societal burden of CVD and
other chronic diseases.

Previous Section
Next Section
AUTHOR INFORMATION
Authors/Writing Group of the PREMIER Collaborative Research Group:
Lawrence J. Appel, MD (chair), Departments of Medicine, Epidemiology, and
International Health (Human Nutrition), Johns Hopkins Medical Institutions,
Baltimore, Md; Catherine M. Champagne, PhD, and David W. Harsha, PhD,
Pennington Biomedical Research Center, Baton Rouge, La; Lawton S. Cooper, MD,
and Eva Obarzanek, PhD, National Heart, Lung, and Blood Institute, Bethesda, Md;
Patricia J. Elmer, PhD, Victor J. Stevens, PhD, and William M. Vollmer, PhD, Kaiser
Permanente Center for Health Research, Portland, Ore; Pao-Hwa Lin, PhD, and Laura
P. Svetkey, MD, Duke Hypertension Center and Sarah W. Stedman Center for
Nutritional Studies, Duke University Medical Center, Durham, NC; and Deborah R.
Young, PhD, Department of Kinesiology, University of Maryland, College Park.
Corresponding Author and Reprints: Lawrence J. Appel, MD, MPH, Departments
of Medicine; Epidemiology, and International Health (Human Nutrition), Johns
Hopkins Medical Institutions, 2024 E Monument St, Suite 2-645, Baltimore, MD
21205-2223 (e-mail: [email protected]).

Author Contributions: Study concept and design: Appel, Elmer, Harsha, Obarzanek,
Stevens, Svetkey, Vollmer, Champagne, Lin, Young.

Acquisition of data: Appel, Elmer, Harsha, Svetkey, Champagne, Lin, Young.

Analysis and interpretation of data: Appel, Cooper, Elmer, Harsha, Obarzanek,
Stevens, Svetkey, Vollmer.
Drafting of the manuscript: Appel, Elmer, Harsha, Stevens, Svetkey, Vollmer.
Critical revision of the manuscript for important intellectual content: Appel, Cooper,
Elmer, Harsha, Obarzanek, Stevens, Svetkey, Vollmer, Champagne, Lin, Young.
Statistical expertise: Vollmer.
Obtained funding: Appel, Elmer, Harsha, Stevens, Svetkey, Vollmer, Young.
Administrative, technical, or material support: Appel, Cooper, Elmer, Harsha,
Obarzanek, Stevens, Champagne, Lin.
Study supervision: Appel, Elmer, Harsha, Stevens, Svetkey, Vollmer, Champagne,
Young.
PREMIER Collaborative Research Group, Participating Sites: Coordinating
Center, Center for Health Research, Portland, Ore: Mikel Aickin, PhD, Jack Hollis,
PhD, Njeri Karanja, PhD, Fran Heinith, BSN, Kristy Funk, MS, RD, Michael Allison,
BS, Chuhe Chen, PhD, Clifton Hindmarsh, MS, Terry Kimes, MS, Wan-Ru Li, MBA,
Gayle Meltesen, MS, Carrie Meeks, Nadia Redmond, MSPH, and Rina Smith, BA,
PgDipLGA; Clinical Center, Center for Health Research, Portland, Ore: Adrianne C.
Feldstein, MD, MS, Daniel S. Laferriere, RN, MSN, Shirley R. Craddick, RD, MHA,
Dana R. Larson, RD, MS, Diane J. Cook, RD, MPH, Carol L. Young, Susan D.
Arnold, RN, BSN, Donna L. Clark, Stanley B. Postlethwaite, Titza SuvalcuConstantin, Carol M. Maul, Donna M. Gleason, Cheryl A. Johnson, EdM, Pamela G.
McNeal, and Debra D. Burch; Duke University Medical Center, Durham, NC:
Colleen McBride, PhD, Jamy Ard, MD, Kathleen Aicher, Blondeaner Brown, Denise
Ernst, Jeanne Gresko, Madhuri Kesari, Femke Lamers, LaTonya Nealon, Tori Phelps,
LaVerne Pruden, LaChanda Reams, Patrice Reams, Benjamin Reese, PsyD, Fran
Rukenbrod, Sonia Steele, Natalie Thorpe, Olaunda Williams, and Chenghua (Cherry)
Yang; Pennington Biomedical Research Center, Baton Rouge, La: Philip Brantley,
Allison Worthen, Betty Kennedy, Emily Griffin, Erma Levy, Terri Keller, Shantell
Jones, and Katherine Lastor; Johns Hopkins Medical Center, Baltimore, Md: Barbara
Bailey, MS, RD, Jeanne Charleston, RN, MSN, Sharrone Cypress, Arlene Dalcin, MS,
RD, Maura Deeley, Charalett Diggs, RN, Thomas P. Erlinger, MD, MPH, Ann Fouts,
RN, Angela Hall, Charles Harris, Tara Harrison, Megan Jehn, Shirley Kritt, Estelle
Levitas, Phyllis McCarron, MS, RD, Edgar R. Miller III, MD, PhD, Pauline Patrick,
LD, Joy Peterson, Charles Powell, Thomas Shields, LeeLana Thomas, MS, RD,
Letitia Thomas, Bobbie Weiss, and Deborah Young, PhD; National Heart, Lung, and
Blood Institute, Bethesda, Md: Jeffrey A. Cutler, MD, Michael Proschan, PhD, and
Denise Simons-Morton, MD, PhD; National Health and Nutrition Examination
Survey and Global Micronutrient Laboratory, Centers for Disease Control and
Prevention, Atlanta, Ga: Christine Pfeiffer, PhD; Division of Laboratory Sciences,
Centers for Disease Control and Prevention: Elaine W. Gunter, MT; Core Laboratory
for Clinical Studies, Washington University, St Louis, Mo: Thomas G. Cole, PhD;
Pennsylvania State University Diet Assessment Center, University Park: Helen S.
Wright, PhD, and Diane C. Mitchell; Center for Clinical Epidemiology and

Biostatistics, University of Pennsylvania School of Medicine, Philadelphia: Shiriki

Kumanyika, PhD, MPH; Howard University, Washington, DC: Jerome Williams,
PhD; Stanford Center for Research in Disease Prevention, Stanford University School
of Medicine, Palo Alto, Calif: Leslie Pruitt, PhD, and Abby King, PhD.
Data and Safety Monitoring Board: Jerome D. Cohen (chair), Nancy R. Cook, ScD,
Patricia Dubbert, PhD, Keith C. Ferdinand, MD, Jim Raczynski, PhD, and Linda Van
Horn, PhD.
Funding/Support: This work was supported by National Institutes of Health grants
UO1 HL60570, UO1 HL60571, UO1 HL60573, UO1 HL60574, UO1 HL62828, and
MO1 RR00052.
Acknowledgment: We are especially grateful to our participants for their sustained
commitment to the trial. We also thank members of the trial's data and safety
monitoring board.
Previous Section
.
.
.
.
.
.
.

.
.
.

REFERENCES
1.Muntner P, He J, Roccella EJ, Whelton PW. The impact of JNC-VI guidelines on
treatment recommendations in the US population. Hypertension. 2002;39:897-902.
Free Full Text
2.Burt VL, Whelton P, Roccella EJ, et al. Prevalence of hypertension in the US adult
population: results from the Third National Health and Nutrition Examination Survey,
1988-1991. Hypertension. 1995;25:305-313. Free Full Text
3.Vasan RS, Beiser A, Seshadri S, Larson MG, Kannel WB, D'Agostino RB, Levy D.
Residual life-time risk for developing hypertension in middle-aged women and men:
the Framingham Heart Study. JAMA. 2002;287:1003-1010. Free Full Text
4.Vasan RS, Larson MG, Leip EP, et al. Impact of high-normal blood pressure on the
risk of cardiovascular disease. N Engl J Med. 2001;345:1291-1297. CrossRefMedline
5.Stamler J, Stamler R, Neaton JD. Blood pressure, systolic and diastolic, and
cardiovascular risks: US population data. Arch Intern Med. 1993;153:598-615.
Free Full Text
6.Whelton PK, He J, Appel LJ, et al. Primary prevention of hypertension: clinical and
public health advisory from the National High Blood Pressure Education Program.
JAMA. 2002;288:1882-1888. Free Full Text
7.Joint National Committee on Prevention, Detection, Evaluation, and Treatment of
High Blood Pressure. The Sixth Report of the Joint National Committee on the
Prevention, Detection, Evaluation, and Treatment of High Blood Pressure (JNC VI).
Arch Intern Med. 1997;157:2413-2446. Free Full Text
8.Appel LJ, Moore TJ, Obarzanek E, et al. A clinical trial of the effects of dietary
patterns on blood pressure. N Engl J Med. 1997;336:1117-1124. CrossRefMedline
9.Karanja NM, Obarzanek E, Lin PH, et al. Descriptive characteristics of the dietary
patterns used in the Dietary Approaches to Stop Hypertension trial. J Am Diet Assoc.
1999;99(suppl):S19-S27. CrossRefMedline
10.Svetkey LP, Harsha DW, Vollmer WM, et al. PREMIER: a clinical trial of
comprehensive lifestyle modification for blood pressure control: rationale, design and
baseline
characteristics.
Ann
Epidemiol.
In
press.
Available
at:
http://www.sciencedirect.com/science/journal/10472797. Accessibility verified March
31, 2003.

.
.

.
.
.
.
.
.
.
.
.

.
.

11.Joint National Committee on Detection, Evaluation, and Treatment of High Blood

Pressure. The Fifth Report of the Joint National Committee on the Detection,
Evaluation, and Treatment of High Blood Pressure (JNC V). Arch Intern Med.
1993;153:154-183. Free Full Text
12.The Trials of Hypertension Prevention Collaborative Research Group. The effects
of nonpharmacologic interventions on blood pressure of persons with high normal
levels. JAMA. 1992;267:1213-1220. Free Full Text
13.The Trials of Hypertension Prevention Collaborative Research Group. Effects of
weight loss and sodium reduction intervention on blood pressure and hypertension
incidence in overweight people with high-normal blood pressure: the trials of
hypertension prevention, phase II. Arch Intern Med. 1997;157:657-667. Free Full
Text
14.Rose G, McCartney P, Reid DD. Self-administration of a questionnaire on chest
pain and intermittent claudication. Br J Prev Soc Med. 1977;31:42-48. Medline
15.Fox TA, Heimendinger J, Block G. Telephone surveys as a method for obtaining
dietary information: a review. J Am Diet Assoc. 1992;92:729-732. Medline
16.American College of Sports Medicine. ACM's Guidelines for Exercise Testing and
Prescription. 5th ed. Media, Pa: Williams & Wilkins; 1995.
17.Blair SN, Haskell WL, Po H, et al. Assessment of habitual physical activity
methodology by a seven-day recall in a community survey and controlled
experiments. Am J Epidemiol. 1985;122:794-804. Free Full Text
18.Writing Group for the Activity Counseling Trial (ACT) Research Group. Effects of
physical activity counseling in primary care: the Activity Counseling Trial: a
randomized controlled trial. JAMA. 2001;286:677-687. Free Full Text
19.Schafer JL. Multiple imputation: a primer. Stat Methods Med Res. 1999;8:3-15.
Free Full Text
20.Agresti A. An Introduction to Categorical Data Analysis. New York, NY: John
Wiley & Sons Inc; 1996.
21.Aickin M, Gensler H. Adjusting for multiple testing when reporting research
results: Bonferroni vs Holm methods. Am J Public Health. 1996;86:726-728.
Free Full Text
22.Whelton PK, Appel LJ, Espeland MA, et al, for the TONE Collaborative Research
Group. Efficacy of sodium reduction and weight loss in the treatment of hypertension
in older persons: main results of the randomized, controlled trial of nonpharmacologic
interventions in the elderly (TONE). JAMA. 1998;279:839-846. Free Full Text
23.Cushman WC, Ford CE, Cutler JA, et al. Success and predictors of blood pressure
control in diverse North American settings: the Antihypertensive and Lipid-Lowering
Treatment to Prevent Heart Attack Trial (ALLHAT). J Clin Hypertens. 2002;4:393405. Medline
24.Sacks FM, Svetkey LP, Vollmer WM, et al, for the DASH-Sodium Collaborative
Research Group. A clinical trial of the effects on blood pressure of reduced dietary
sodium and the DASH dietary pattern (the DASH-Sodium Trial). N Engl J Med.
2001;344:3-10. CrossRefMedline
25.John JH, Ziebland S, Yudkin P, Roe LS, Neil HA. Effects of fruit and vegetable
consumption on plasma anti-oxidant concentrations and blood pressure: a randomised
controlled trial. Lancet. 2002;359:1969-1974. CrossRefMedline
26.Kumanyika SK, Hebert PR, Cutler JA, et al. Feasibility and efficacy of sodium
reduction in the Trials of Hypertension Prevention, phase I. Hypertension.
1993;22:502-512. Free Full Text

.
.
.
.
.
.
.
.
.

27.Stevens VJ, Obarzanek E, Cook NR, et al. Long-term weight loss and changes in
blood pressure: results of the Trials of Hypertension Prevention, phase II. Ann Intern
Med. 2001;134:1-11. Free Full Text
28.McCullough ML, Feskanich D, Rimm EB, et al. Adherence to the Dietary
Guidelines for Americans and risk of major chronic disease in men. Am J Clin Nutr.
2000;72:1223-1231. Free Full Text
29.Fung TT, Willett WC, Stampfer MJ, Manson JE, Hu FB. Dietary patterns and risk
of coronary heart disease in women. Arch Intern Med. 2001;161:1857-1862. Free Full
Text
30.Kant AK, Schatzkin A, Graubard BI, Schairer C. A prospective study of diet
quality and mortality in women. JAMA. 2000;283:2109-2115. Free Full Text
31.The Role of Nutrition in Maintaining Health in the Nation's Elderly: Evaluating
Coverage of Nutrition Services for the Medicare Population. Washington, DC:
Institute of Medicine, National Academy of Sciences; 2000.
32.Tuomilehto J, Lindstrom J, Eriksson JG, et al. Prevention of type 2 diabetes
mellitus by changes in lifestyle among subjects with impaired glucose tolerance. N
Engl J Med. 2001;344:1343-1350. CrossRefMedline
33.Knowler WC, Barrett-Connor E, Fowler SE, et al. Reduction in the incidence of
type 2 diabetes with lifestyle intervention or metformin. N Engl J Med. 2002;346:393403. CrossRefMedline
34.Dattilo AM, Kris-Etherton PM. Effects of weight reduction on blood lipids and
lipoproteins: a meta-analysis. Am J Clin Nutr. 1992;56:320-328. Free Full Text
35.Langford HG, Davis BR, Blaufox D, et al, for the TAIM Research Group. Effect of
drug and diet treatment of mild hypertension on diastolic blood pressure.
Hypertension. 1991;17:210-217. Free Full Text
36.Miller ER, Erlinger EP, Young DR, et al. Results of the Diet, Exercise and Weightloss Intervention Trial (DEW-IT). Hypertension. 2002;40:612-618.
Free Full Text