The Journal of Nutrition

Community and International Nutrition

Daily Zinc but Not Multivitamin

Supplementation Reduces Diarrhea and Upper
Respiratory Infections in Tanzanian Infants: A
Randomized, Double-Blind, Placebo-Controlled
Clinical Trial1,2
Christine M McDonald,3 Karim P Manji,4 Rodrick Kisenge,4 Said Aboud,5 Donna Spiegelman,6,7
Wafaie W Fawzi,6,8,9 and Christopher P Duggan3,8,9*
3
Division of Gastroenterology, Hepatology and Nutrition, Boston ChildrenÕs Hospital, Boston, MA; Departments of 4Pediatrics and
Child Health and 5Microbiology and Immunology, Muhimbili University of Health and Allied Sciences, Dar es Salaam, Tanzania; and
Departments of 6Epidemiology, 7Biostatistics, 8Nutrition, and 9Global Health and Population, Harvard T.H. Chan School of Public Health,
Boston, MA

Abstract
Background: Although various micronutrient regimens have been shown to prevent and treat common infectious
diseases in children, the effects of daily multivitamin (MV) and/or zinc supplementation have not been widely evaluated in
young African infants.
Objective: The objective was to determine whether daily supplementation of HIV-unexposed Tanzanian infants with MVs
or zinc reduces the risk of infectious morbidity compared with placebo.
Methods: In a 2 3 2 factorial, double-blind, randomized controlled trial, 2400 infants who were 6 wk of age and born to
HIV-negative mothers in a low-malaria setting were randomly assigned to receive daily oral supplementation of MVs
(vitamin B complex and vitamins C and E), zinc, zinc + MVs, or placebo for 18 mo. Morbidity was assessed by study nurses
at monthly visits and by physicians every 3 mo and/or when the child was acutely ill.
Results: No significant differences were found in the percentage of nurse visits during which diarrhea, cough, or any other
symptom were reported throughout the previous month when receiving either zinc or MVs. However, physician
diagnoses of all types of diarrhea (RR = 0.88; 95% CI: 0.81, 0.96; P = 0.003), dysentery (RR = 0.84; 95% CI: 0.74, 0.95; P =
0.006), and acute upper respiratory infection (RR = 0.92; 95% CI: 0.88, 0.97; P = 0.0005) were significantly lower for
infants supplemented with zinc than for those who did not receive zinc. Among the 2360 infants for whom vital status was
obtained, there was a nonsignificant increase in all-cause mortality among infants who received zinc (HR = 1.80; 95% CI:
0.98, 3.31; P = 0.06) compared with those who did not receive zinc. MVs did not alter the rates of any recorded physician
diagnoses or mortality. Neither zinc nor MVs reduced hospitalizations or unscheduled outpatient visits.
Conclusions: Daily zinc supplementation of Tanzanian infants beginning at the age of 6 wk may lower the burden of
diarrhea and acute upper respiratory infections, but provision of MVs using the regimen in this trial did not confer additional
benefit. This trial was registered at clinicaltrials.gov as NCT00421668. J Nutr 2015;145:2153–60.

Keywords: multivitamins, zinc, child morbidity, diarrhea, respiratory infection

Introduction
for the deaths of 1.9 million children <5 y of age in 2011 (1).
Diarrheal diseases and respiratory infections are among the Low birth weight and early childhood undernutrition are major
leading causes of child mortality globally and were responsible risk factors for morbidity and mortality from these and other
infectious diseases, particularly in resource-limited areas of the
1
Supported by the Eunice Kennedy Shriver National Institute of Child Health and world. Recent estimates indicate that 45% of all child deaths
Human Development (NICHD R01 HD048969-01 and K24HD058795). result from fetal growth restriction, anthropometric deficits,
2
Author disclosures: CM McDonald, KP Manji, R Kisenge, S Aboud, D Spiegelman,
micronutrient deficiencies, or suboptimal breastfeeding (2).
WW Fawzi, and CP Duggan, no conflicts of interest.
* To whom correspondence should be addressed. E-mail: christopher.duggan@childrens. Therefore, interventions to improve nutritional status have
harvard.edu. great potential to reduce child morbidity and mortality.
ã 2015 American Society for Nutrition.
Manuscript received February 13, 2015. Initial review completed April 4, 2015. Revision accepted, June 29, 2015. 2153
First published online July 22, 2015; doi:10.3945/jn.115.212308.
 Micronutrient supplementation is one such strategy and is the Enzygnost HIV Integral II Antibody/Antigen (Siemens). Any dis-
particularly compelling given the important role of numerous crepancy between the first and second ELISA was resolved by a Western
vitamins and minerals in systemic immune function and the blot assay. Consenting mothers who were confirmed to be HIV-negative
maintenance of local defenses, coupled with its relative cost- were enrolled into the study and their infants were randomly assigned to
1 of 4 regimens between 5 and 7 wk of age. Infants of multiple births and
effectiveness (3, 4).
infants with congenital anomalies or other conditions that would
Several studies have evaluated the efficacy of different interfere with the study procedures were excluded. Birth characteristics
micronutrient regimens in preventing and/or treating common were obtained immediately after delivery whenever possible. We used
infectious diseases, including respiratory and gastrointestinal reference data from Oken et al. (11) to calculate the percentile of birth
diseases. However, differences in study design and location, weight for each completed week of gestation and defined small-for-
baseline nutritional status of the study population, composition gestational age as #10th percentile. At the time of randomization,
and duration of supplementation, and disease background have clinical examination was performed by a study physician, history of
led to varying results. There is clear evidence that therapeutic morbidity and infant feeding practices was conducted by a study nurse,
zinc supplementation has beneficial effects on the duration and infant blood was drawn for a complete blood count, and anthropometric
severity of diarrhea and other morbidities (5), and the WHO has measurements were performed.
Institutional approval was granted by the Harvard T.H. Chan School
included zinc supplementation in its guidelines for clinical
of Public Health Human Subjects Committee, the Muhimbili University
management of acute diarrhea for the past decade (6). However, of Health and Allied Science Committee of Research and Publications,
policy recommendations do not yet exist for preventive zinc the Tanzanian National Institute of Medical Research, and the Tanzanian
supplementation. A meta-analysis showed that preventive zinc Food and Drugs Authority. Over the course of the study, a Data Safety and
supplementation in children reduced the incidence of diarrhea Monitoring Board met twice annually.
and respiratory infections by ~20% and 14%, respectively (7). A
notable finding, however, was that this effect was generally not Randomization and masking. Infants were randomly assigned in a
observed in infants <12 mo of age. In addition, few studies factorial design to receive a daily oral dose of 1 of the following 4
included in the analysis were conducted in sub-Saharan Africa regimens for 18 mo from the time of randomization: 1) zinc, 2) MVs, 3)
(7). Given regional differences in maternal nutritional status, zinc + MVs, or 4) placebo. The biostatistician in Boston prepared a
birth outcomes, and morbidity patterns, it is critical that micro- randomization list from 1 to 2400 that used blocks of 20 and was
nutrient supplementation trials be conducted in a diversity of stratified by study clinic. Capsules were packaged in a blister pack of 15
each and numbered boxes containing 6 blister packs were prepared
settings, including sub-Saharan Africa, before global recommen-
containing the corresponding treatments. Each eligible infant was
dations can be developed. assigned the next numbered box of capsules at his/her respective site.
Because zinc insufficiency may coexist with other vitamin and The supplement used was an orange-flavored powder encapsulated in an
mineral deficiencies, simultaneous supplementation with other opaque gelatinous capsule and was manufactured by Nutriset. All 4
micronutrients may be an effective strategy. Two recent trials in regimens were field tested and the taste, smell, and appearance were
Tanzania and Pakistan examined whether the provision of found to be indistinguishable between groups. All study personnel and
multiple micronutrients in addition to zinc confers additional participants were blinded to treatment assignment for the duration of
benefits against diarrhea and other morbidities (8, 9). Both trials, the study.
however, were restricted to children $6 mo of age, both included
iron as part of the micronutrient supplement, and both reported Procedures. From the time of randomization to 6 mo of age, infants
that multiple micronutrients either increased diarrhea morbidity received 1 capsule/d, and from 7 mo of age to the end of follow-up, 2
capsules were provided daily. For infants in the zinc group, the capsule
or conferred no additional benefit in comparison with zinc
contained 5 mg of zinc. For infants in the MV group, the capsule
alone. To our knowledge, there have not been any investigations
contained 60 mg of vitamin C, 8 mg of vitamin E, 0.5 mg of thiamine, 0.6 mg
of preventive zinc and multivitamin (MV) supplementation on of riboflavin, 4 mg of niacin, 0.6 mg of vitamin B-6, 130 mg of folate,
infectious disease morbidity that have included African children and 1 mg of vitamin B-12. Infants in the MV + zinc group received
<6 mo of age. Initiation of supplementation earlier in infancy 1 capsule containing the micronutrients listed in both the MV and the
may be appropriate given the fact that many lactating mothers zinc groups. For children 0–6 mo of age, these doses represented between
are deficient in several micronutrients (10). To evaluate the 150% and 600% of the RDA or Adequate Intake, and for children 7–12 mo
potential role of zinc or MV supplementation in early infancy, of age, the doses were equivalent to 200–400% of the RDA or Adequate
we performed a randomized trial to determine whether daily Intake. Mothers were shown how to push the capsule through the back
supplementation of HIV-unexposed Tanzanian infants with of the blister pack, open the capsule, decant the powder into a small
plastic cup, mix the powder with 5 mL of sterile water, and administer
MVs, zinc, or both reduced the risk of infectious morbidity
the solution to the child orally.
compared with placebo.
Our choice of supplement composition was based on several con-
siderations including: 1) previous research that found deficiencies in
vitamin B-12, folate, zinc, vitamin A, and vitamin E among breastfeeding
Methods
women in South Africa (10) and, therefore, suggests that infant micro-
Study design and participants. The study was a randomized, double- nutrient status may be low in sub-Saharan Africa; 2) a previous clinical
blind, 2 3 2 factorial design trial that took place in periurban Dar es trial involving pregnant Tanzanian women that confirmed improved
Salaam, Tanzania (clinicaltrials.gov NCT 00421668). Mothers of poten- birth outcomes with supplementation of vitamin B complex, vitamin E,
tially eligible infants were recruited into the study in 1 of 2 ways: 1) and vitamin C (12); and 3) findings from our previous trial of MV
pregnant women #34 wk gestation presenting at 1 of 3 prenatal clinics supplementation involving HIV-exposed children, which revealed lower
in Dar es Salaam were informed about the study and consented rates of fever and vomiting among supplemented children (13). Dar es
prenatally or 2) women were recruited from the labor ward of Salaam has been described as a malaria-endemic area, although studies
Muhimbili National Hospital within 12 h of delivering a healthy have suggested that rates of malaria are declining (14). Nonetheless,
singleton baby. In both cases, written informed consent was obtained owing to the potential that iron supplementation of nonanemic children
and mothers were asked to present at a study clinic within 1–2 wk of may have adverse consequences in malaria-endemic regions (15), the MV
delivery for HIV testing. Maternal HIV status was determined using 2 supplement did not include iron.
sequential ELISAs that used the Murex HIV antigen/antibody (Abbott Mothers and children were followed from the time of randomization
Murex) followed by the Enzygnost anti-HIV-1/2 Plus (Dade Behring) or for 18 mo, until the childÕs death, or until loss to follow-up. Mothers who

2154 McDonald et al.

were enrolled during pregnancy received standard prenatal care analysis plan. Descriptive statistics were used to summarize baseline
including anthropometric assessment, intermittent prophylaxis for characteristics of the study population. Frequencies were reported for
malaria, tetanus toxoid immunization, deworming using mebendazole, categorical variables and the mean 6 SD for continuous variables. The
anemia assessment, and iron-folic acid supplementation. During this x2-test and ANOVA were used to detect any differences among treatment
follow-up period, mothers and children were asked to return to the groups. We used generalized estimating equations with the log link,
study clinic every 4 wk for data collection and standard clinical care, binomial variance, and exchangeable correlation matrix to compare the
including growth monitoring, immunizations, routine medical treat- proportion of follow-up visits in which the illness symptom had occurred
ment for illnesses, and periodic vitamin A supplementation (100,000 IU in the previous 4 wk between factors. For physician diagnoses that were
at 9 mo and 200,000 IU at 15 mo). Children who were diagnosed with made during routine visits every 3 mo or during unscheduled visits
anemia were treated with iron supplementation. At ~12 mo of age, during acute illness episodes, the mean number of diagnoses over the
CD4 and CD8 T cell counts and percentages were measured from a follow-up period was compared between factors using Poisson regres-
subset of 428 children using FACSCalibur system (Becton Dickinson) sion. In both sets of analyses we introduced interaction terms to test for
in order to determine any potential immune effect of micronutrient joint effects between the zinc and MV factors. We also tested for effect
supplementation. modification by each factor and sex and low birth weight. When the
At each of the monthly follow-up visits study nurses assessed P-interaction term was <0.10, stratified analyses were performed.
compliance by counting the number of unconsumed capsules, assessed Although the study was not powered to detect differences in mortal-
infant feeding practices, and conducted a morbidity history with the ity, we used Cox proportional hazards modeling to explore differences in
aid of pictorial diaries that mothers were instructed to complete daily all-cause and cause-specific mortality by factor. Values in the text are
in order to document the occurrence of any of the following symptoms: means 6 SDs unless otherwise indicated. All analyses were performed
diarrhea; common cold; cough; difficulty breathing; fever; refusal to using SAS software (version 9.2; SAS Institute).
eat, drink, or breastfeed; pus draining from ears; and vomiting. They
also assessed symptoms that were present on the day of the visit,
recorded vital signs (including measurement of the childÕs temperature
and respiratory rate and detection of chest indrawing), and inquired
Results
about the occurrence of any unscheduled clinic visits or hospitaliza- Figure 1 shows the study profile. Between August 2007 and
tions in the past month. Diarrhea was defined as $3 loose or watery December 2009, 2400 infants were randomly assigned, and
stools within a 24-h period. Rapid respiratory rate was defined as
follow-up ended in May 2011. Median (25th and 75th percen-
>50 breaths/min in infants 2–11 mo of age and >40 breaths/min among
infants $12 mo of age.
tiles) regimen compliance among children was 96 (91, 99) of the
At baseline, every 3 mo, and/or when acute illnesses were noted by allocated regimen. With the exception of length-for-age z score,
the study nurse a study physician conducted a physical examination, there were no significant differences in any maternal, socioeco-
diagnosed illnesses, and provided necessary medical treatment. Physi- nomic, or child characteristics between the 4 groups at baseline
cians underwent regular training so that they used standardized (Table 1). Mean maternal age was ~26 y, about three-quarters of
diagnostic criteria and treatment guidelines consistent with the WHO mothers had #7 y of education, and ~30% had not been
and Tanzanian Ministry of Health and Social Welfare policies. For the pregnant previously. The prevalences of low birth weight and
purposes of the analysis of physician diagnosis, ‘‘any form of diarrhea’’ preterm birth in all groups were ~3% and 13%, respectively.
included persistent diarrhea, acute diarrhea, dysentery, and/or intestinal One child was enrolled outside of the age criteria (at 2.4 wk) but
parasites. Acute upper respiratory infection was defined as pharyngitis or
was kept in the trial after review and approval by the Data Safety
rhinitis (both without fast breathing or chest indrawing). Acute lower
respiratory infection was defined as cough or difficulty breathing, rapid
and Monitoring Board and the Institutional Review Board.
respiratory rate (based on the same definition described previously), Table 2 shows the effect of MV and zinc supplementation on
and either a fever of >38.3°C or chest retractions. ‘‘Any form of res- the occurrence of common infectious morbidities, hospitaliza-
piratory infection’’ included acute upper respiratory infection, acute tions, and unscheduled outpatient visits, as recorded by the
lower respiratory infection, pulmonary tuberculosis, or other causes of study nurses at the monthly clinic visits. When children who
pneumonia. received zinc were compared with children who did not receive
Children who missed their scheduled monthly follow-up appoint- zinc, there were no significant differences in the percentage of
ment were visited at home and their vital status was confirmed through visits during which diarrhea (RR: 0.93; 95% CI: 0.82, 1.05;
contact with immediate family members. In cases of child death, a verbal P = 0.26), cough (RR: 0.95; 95% CI: 0.90, 1.01; P = 0.10), or
autopsy was performed to determine the cause of death. The cause of
any other symptom had been experienced during the previous
death forms were then coded by 2 independent pediatricians (KPM and
CPD), and any differences were resolved by a third pediatrician.
month. In addition, the occurrence of hospitalizations (RR:
1.50; 95% CI: 0.83, 2.70; P = 0.18) and unscheduled outpatient
visits (RR: 1.18; 95% CI: 0.98, 1.43; P = 0.08) did not vary
Data management and analysis. The primary outcomes of the study
significantly in children receiving zinc. With the exception of
were the incidence of clinical symptoms of diarrhea and lower res-
piratory infection. Power was calculated for 1 factor in the factorial pus draining from the ears, MVs did not affect the occurrence
design, which represented either the zinc or MV arm. Based on findings of any of the morbidity symptoms, hospitalizations (RR: 1.18;
from our previous trial, we estimated that the mean (SD) number of 95% CI: 0.67, 2.11; P = 0.56), or unscheduled outpatient visits
diarrheal and lower respiratory illnesses per child per year in the placebo (RR: 1.11; 95% CI: 0.92, 1.34; P = 0.28). There was, however,
group would be 3.4 (4.2) and 2.1 (1.0), respectively (16). After applying a nonsignificant interaction between treatment groups in the
a 2-sided a-value of 0.025 to yield an overall type I error rate of 0.05 for case of diarrhea (P-interaction = 0.06). Subsequent stratified
each primary outcome and allowing for a 15% loss to follow-up and analysis revealed that in comparison with the placebo group,
minimum power of 80%, we calculated that we would require 2400 children in the zinc group had significantly fewer reported cases
subjects to detect a reduction of 18% in the mean number of diarrheal of diarrhea (RR: 0.83; 95% CI: 0.69, 0.99; P = 0.04), whereas
episodes per year. We then calculated that with this sample size we would
children in the MV (RR: 0.91; 95% CI: 0.77, 1.08; P = 0.28)
have 90% power to detect an 8% reduction in the mean number of
episodes of lower respiratory illness per year. and zinc + MV (RR: 0.95; 95% CI: 0.80, 1.13; P = 0.59) groups
Data were double entered using Microsoft Access software (Micro- did not.
soft Corp.) at the central study site and then converted to SAS data sets Examination of effect modification showed that zinc lowered
and uploaded to a secured UNIX-based server for analysis. Intent-to- the burden of diarrhea in boys (RR: 0.84; 95% CI: 0.71, 0.99;
treat analyses were conducted according to a pre-established data P = 0.04) but not in girls (RR = 1.07; 95% CI: 0.89, 1.29; P = 0.49;
Effects of multivitamins and zinc on child health 2155
 average number of physician diagnoses over the course of
follow-up was <1/child across all treatment groups. Rates of any
type of respiratory infection (RR: 0.93; 95% CI: 0.89, 0.97; P =
0.0006), acute upper respiratory infection (RR: 0.92; 95% CI:
0.88, 0.97; P = 0.0005), any form of diarrhea (RR: 0.88; 95%
CI: 0.81, 0.96; P = 0.004), and dysentery (RR: 0.84; 95% CI:
0.74, 0.95; P = 0.006) were significantly lower among children
who received zinc compared with children who did not receive
zinc. MVs did not significantly change the rates of physician
diagnosis of acute upper respiratory infection, acute lower
respiratory infection, pulmonary tuberculosis or other causes of
pneumonia, any form of respiratory infection, acute diarrhea,
dysentery, persistent diarrhea, intestinal parasites, any form of
diarrhea, uncomplicated malaria, severe malaria, or pallor/
anemia. There were no significant interactions between treat-
ment arms for any of the morbidities.
Stratified analyses illustrated that zinc reduced the rate of
dysentery in boys (RR: 0.73; 95% CI: 0.61, 0.84; P = 0.004) but
not girls (RR: 0.98; 95% CI: 0.81, 1.17; P = 0.79; P-interaction =
0.02). The sex of the child also modified the effect of zinc on
uncomplicated malaria. Girls who received zinc had a reduced
rate of uncomplicated malaria compared with girls who did not
receive zinc (RR: 0.86; 95% CI: 0.75, 0.98; P = 0.02); however,
no difference was observed among boys (RR: 1.06; 95% CI:
0.93, 1.20; P = 0.39; P-interaction = 0.02). Birth weight did not
modify the effect of zinc on any physician diagnoses. Likewise,
the sex of the child or the birth weight did not influence the effect
of MVs on any physician diagnoses.
Among the subgroup of study participants with immunologic
measures available at 51.3 6 2.2 wk, CD4 T cell percent was
35.0 6 7.4 in infants who received zinc compared with 34.5 6
8.0 in those not receiving zinc (P = 0.51). However, infants who
received zinc and MVs had a CD4 T cell percentage of 36.7 6
7.4 compared with 34.3 6 8.0 among children who received
placebo (P = 0.03). There were no group differences in mean
CD8 T cell percentages or CD4:CD8 ratios noted (all, P > 0.20).
After a median follow-up of 17.8 mo, there were 45 deaths
among the 2360 study participants with a known vital status. Of
the 40 children who were lost to follow-up, 8 received the
placebo, 11 received zinc only, 15 received MVs only, and 6
received zinc and MVs. We observed a nonsignificant increase in
all-cause mortality among children who received zinc compared
with children who did not receive zinc (HR: 1.80; 95% CI: 0.98,
FIGURE 1 Study profile of a randomized trial of MV and/or zinc 3.31; P = 0.06). MVs did not significantly alter the risk of all-
supplementation to infants in Dar es Salaam, Tanzania. MV, multivitamin. cause mortality (HR: 0.73; 95% CI: 0.40, 1.32; P = 0.30).
Information on the cause of death was available for 39 study
participants. Respiratory illness (n = 13), diarrheal disease (n = 5),
P-interaction = 0.05). Conversely, zinc increased the occurrence of malaria (n = 7), other infectious diseases (n = 8), and neonatal
fever among low birth weight infants (RR: 1.66; 95% CI: 1.09, conditions (n = 2) were the primary causes of deaths. There were
2.53; P = 0.02) but had no effect among infants born with a birth no significant differences in cause-specific mortality between
weight of $2.5 kg (RR: 0.95; 95% CI: 0.87, 1.02; P = 0.17; treatment groups (all, P > 0.05).
P-interaction = 0.02). The occurrence of unscheduled outpatient
visits was also higher among low birth weight infants who received
zinc (RR: 5.00; 95% CI: 1.15, 21.69; P = 0.03); however, no
Discussion
differences were observed among infants with birth weights of
$2.5 kg (RR: 1.17; 95% CI: 0.96, 1.41; P = 0.11; P-interaction = In this randomized, double-blind trial in 2400 HIV-unexposed
0.009). Similarly, low birth weight infants who received MVs had Tanzanian infants, we found that zinc supplementation signif-
a cough more often than those who did not receive MVs (RR: icantly lowered rates of physician diagnoses of diarrhea and
1.48; 95% CI: 1.08, 2.02; P = 0.02), whereas there was no acute upper respiratory infections. Zinc specifically reduced the
difference among infants with a birth weights of $2.5 kg (RR: rates of dysentery, particularly among boys. The occurrence of
1.00; 95% CI: 0.94, 1.06; P = 0.94; P-interaction = 0.02). diarrhea, as reported at monthly clinic visits with study nurses,
The effect of MVs and zinc on infectious morbidities was also significantly reduced among children who received
diagnosed by the study physician is shown in Table 3. The low only zinc in comparison with placebo. MVs did not alter the
burden of malaria in the study setting is worth noting: the occurrence of any recorded morbidities, and neither zinc nor
2156 McDonald et al.
TABLE 1 Baseline characteristics of enrolled Tanzanian infants and their mothers by supplementation group1

Placebo (n = 604) Zinc only (n = 596) MVs only (n = 598) Zinc + MVs (n = 602)

Maternal characteristics
Age, y 26.5 6 5.0 26.8 6 5.1 26.2 6 5.0 26.1 6 5.0
Formal education, n (%)
None 9 (1.5) 8 (1.4) 10 (1.7) 9 (1.5)
1–7 y 453 (75.3) 421 (71.1) 416 (70.2) 441 (73.4)
$8 y 140 (23.3) 163 (27.5) 167 (28.2) 151 (25.1)
Employment, n (%)
Housewife without income 386 (64.4) 365 (62.3) 337 (56.7) 357 (59.8)
Housewife with income 176 (29.4) 175 (29.9) 212 (35.7) 201 (33.7)
Other 37 (6.2) 46 (7.9) 45 (7.6) 39 (6.5)
Married or cohabitating with partner, n (%) 537 (90.0) 534 (90.5) 542 (91.4) 542 (90.5)
Prior pregnancies, n (%)
None 184 (30.6) 169 (28.6) 205 (34.5) 187 (31.2)
1–4 398 (66.2) 410 (69.3) 375 (63.1) 396 (66.1)
$5 19 (3.2) 13 (2.2) 14 (2.4) 16 (2.7)
Midupper arm circumference, cm 27.0 6 3.1 27.1 6 3.1 27.1 6 3.1 26.7 6 3.2
Recruited prenatally, n (%) 86 (14.2) 76 (12.8) 90 (15.1) 92 (15.3)
Socioeconomic characteristics
Daily food expenditure per person in household is ,1000 TSh, n (%) 158 (27.6) 162 (28.6) 164 (28.8) 170 (29.4)
Household possessions,2 n (%)
None 171 (28.4) 192 (32.7) 173 (29.2) 180 (30.0)
1–3 358 (59.5) 308 (52.4) 330 (55.7) 339 (56.5)
.3 73 (12.1) 88 (15.0) 90 (15.2) 81 (13.5)
Child characteristics
Age at randomization, wk 5.9 6 0.4 5.9 6 0.4 5.9 6 0.4 5.9 6 0.4
Male, n (%) 293 (48.5) 296 (49.7) 282 (47.2) 313 (52.0)
Low birth weight (,2500 g), n (%) 18 (3.0) 21 (3.6) 21 (3.6) 22 (3.7)
Born ,37 wk gestational age, n (%) 77 (14.0) 80 (14.6) 66 (12.0) 67 (12.1)
Born ,34 wk gestational age, n (%) 15 (2.7) 14 (2.6) 12 (2.2) 22 (4.0)
Born small for gestational age (,10th percentile), n (%) 45 (8.4) 47 (8.7) 52 (9.7) 47 (8.7)
Apgar score #7 at 5 min after birth, n (%) 9 (1.6) 17 (3.1) 5 (0.9) 10 (1.8)
Hemoglobin, g/dL 10.7 6 1.6 10.6 6 1.6 10.7 6 1.4 10.6 6 1.4
Length-for-age z score3 20.17 6 1.28a 20.33 6 1.92a,b 20.26 6 1.20a,b 20.37 6 1.23b
Weight-for-length z score 0.05 6 1.32 0.16 6 1.33 0.14 6 1.29 0.15 6 1.31
Weight-for-age z score 20.16 6 1.05 20.23 6 0.97 20.17 6 0.99 20.26 6 1.03
1
According to the x2-test or ANOVA, there were no significant differences in baseline characteristics among treatment groups (P . 0.05) with the exception of length-for-age
z score (P = 0.03). Values are means 6 SDs or percentages. MV, multivitamin; TSh, Tanzanian shilling.
2
From a list that includes sofa, television, radio, refrigerator, and fan.
3
Labeled means in a row without a common letter differ, P , 0.05.

MVs reduced hospitalizations or unscheduled outpatient visits. of age did not reduce morbidity from diarrhea or respiratory
Although our study was not powered to detect differences in infections in Bangladeshi infants. Likewise, a large trial of
mortality, we observed a nonsignificant increase in mortality preventive zinc and/or folic acid supplementation involving
among infants who received zinc. Furthermore, in a subset of Nepalese children 1–35 mo of age saw no differences in the
children with immunologic measures available at 12 mo of age, frequency or duration of diarrhea or acute lower respiratory
mean CD4 T cell percent was slightly but significantly higher infection between treatment groups (18). However, weekly
among children who received zinc and MVs in comparison with zinc supplementation for 12 mo was found effective in re-
placebo. To our knowledge, this is the first study of preventive ducing the incidence of diarrhea and pneumonia among urban
zinc and MV supplementation to be conducted among young Bangladeshi infants aged 2–12 mo (19). Similarly, a study in
infants in sub-Saharan Africa. Given the large global burden of Delhi, India, involving infants 6–11 mo of age, reported that a
diarrhea and acute respiratory infections in young children, we short course of daily zinc supplementation for 2 wk effectively
believe that the 12% reduction in diarrhea, 16% reduction in reduced the subsequent number and duration of diarrhea
dysentery, and 8% reduction in acute upper respiratory infec- episodes (20). In contrast, a multicenter study involving infants
tions among children who received zinc are clinically significant 1–5 mo of age with acute diarrhea found that infants who
and potentially of major public health importance. received therapeutic zinc supplementation for 14 d had more
Although, to our knowledge, no other studies have assessed days of diarrhea and similar prevalence of pneumonia and
the effects of daily zinc supplementation in tandem with MVs respiratory infection compared with the placebo group (21).
from such a young age, our findings can be compared with Differences in the age of study participants, dose and duration of
trials of zinc vs. placebo on infant morbidity. Osendarp et al. supplementation, and length of follow-up make it difficult to
(17) reported that daily zinc supplementation from 4 to 24 wk identify the reasons behind these contrasting results. However, it
Effects of multivitamins and zinc on child health 2157
TABLE 2 Effect of daily MVs and zinc on the occurrence of common infectious morbidities in Tanzanian infants by nurse evaluation1

Received zinc2 Received MV3

Morbidity reported in Yes (n = 1198) No (n = 1202) Yes (n = 1200) No (n = 1200)

the past 4 wk % (n event/visit) 4
% (n event/visit) 4
RR (95% CI) 5
P 5
% (n event/visit) 4
% (n event/visit)4 RR (95% CI)5 P5 Pint6

Diarrhea 3.7 (547/14,703) 4.0 (584/14,725) 0.93 (0.82, 1.05) 0.26 3.9 (566/14,611) 3.8 (565/14,817) 1.02 (0.90, 1.16) 0.74 0.06
Cough 22.1 (3266/14,782) 23.4 (3474/14,823) 0.95 (0.90, 1.01) 0.10 22.8 (3,350/14,697) 22.7 (3,390/14,908) 1.01 (0.95, 1.07) 0.73 0.78
Difficulty breathing 1.1 (159/14,777) 1.0 (141/14,818) 1.15 (0.89, 1.47) 0.29 1.1 (161/14,692) 0.9 (139/14,903) 1.16 (0.90, 1.49) 0.25 0.51
Cough + fever 4.9 (723/14,782) 5.3 (789/14,823) 0.90 (0.81, 1.01) 0.08 5.2 (760/14,697) 5.0 (752/14,908) 1.02 (0.91, 1.13) 0.78 0.53
Cough plus7 1.7 (247/14,782) 1.8 (261/14,823) 0.95 (0.78, 1.15) 0.57 1.8 (268/14,697) 1.6 (240/14,908) 1.14 (0.94, 1.39) 0.17 0.80
Cough with rapid respiratory rate8 0.1 (16/14,782) 0.1 (14/14,823) 0.98 (0.40, 2.43) 0.97 0.1 (15/14,697) 0.1 (15/14,908) 0.85 (0.34, 2.10) 0.72 0.99
Fever 10.0 (1471/14,780) 10.4 (1539/14,821) 0.95 (0.88, 1.03) 0.18 10.3 (1,514/14,694) 10.0 (1,496/14,907) 1.01 (0.94, 1.09) 0.78 0.80
Cold 21.3 (3153/14,782) 22.5 (3329/14,822) 0.95 (0.90, 1.01) 0.09 22.0 (3,226/14,697) 21.8 (3,256/14,907) 1.01 (0.95, 1.07) 0.82 0.82
Vomiting 1.8 (261/14,779) 1.7 (249/14,819) 1.05 (0.87, 1.26) 0.60 1.9 (272/14,691) 1.6 (238/14,907) 1.16 (0.97, 1.39) 0.11 0.79
Refusal to eat, drink, or breastfeed 2.3 (336/14,780) 2.5 (368/14,817) 0.90 (0.77, 1.06) 0.22 2.5 (363/14,691) 2.3 (341/14,906) 1.07 (0.91, 1.26) 0.38 0.75
Pus draining from ears 0.5 (74/14,778) 0.5 (73/14,817) 1.06 (0.74, 1.52) 0.75 0.4 (59/14,691) 0.6 (88/14,904) 0.65 (0.45, 0.93) 0.02 0.91
Hospitalizations 0.2 (34/14,620) 0.1 (21/14,669) 1.50 (0.83, 2.70) 0.18 0.2 (29/14,528) 0.1 (26/14,761) 1.18 (0.67, 2.11) 0.56 0.63
Unscheduled outpatient visits 2.0 (282/14,286) 1.7 (237/14,358) 1.18 (0.98, 1.43) 0.08 1.90 (270/14,194) 1.72 (249/14,450) 1.11 (0.92, 1.34) 0.28 0.46
1
MV, multivitamin.
2
Received zinc ‘‘yes’’ refers to children who received zinc alone as well as those who received zinc and MVs. Received zinc ‘‘no’’ refers to all children who received MVs alone
and those who received the placebo.
3
Received MVs ‘‘yes’’ refers to children who received MVs alone as well as those who received zinc and MVs. Received MVs ‘‘no’’ refers to all children who received zinc alone
and those who received the placebo.
4
Total number of events occurring during follow-up, defined as being reported in the 28 d (4 wk) before visit or being present on the day of the evaluation.
5
RR, 95% CI, and corresponding P values were obtained from generalized estimating equations with the binomial variance, log link, and exchangeable working covariance structure.
6
P-interaction effect.
7
Cough plus defined as cough with one or more of the following events: difficult breathing, chest retractions, and refusal to eat, drink, or breastfeed.
8
Cough with rapid breathing on the day of the evaluation (respiratory rate: .60/min in infants ,2 mo old, .50/min in infants 2–11 mo old, .40/min in infants 12–59 mo old).

is important to note that, to our knowledge, our study provided Our findings build on previous investigations into the pos-
supplementation for the longest period of time and followed sible effects associated with the addition of multiple micronu-
infants well into their second year of life, capturing the period trients to zinc supplementation regimens among older infants
during which the incidence of infectious morbidities typically and children. We previously reported that the same MV regi-
rises. men did not affect the risk of mortality among HIV-exposed

TABLE 3 Effect of daily MVs and zinc on the incidence of common infectious morbidities in Tanzanian infants by physician diagnosis1

Received zinc2 Received MV3

Yes No Yes No
Physician diagnosis n Diagnosis, n 4
n Diagnosis, n 4
RR (95% CI) 5
P 5
n Diagnosis, n 4
n Diagnosis, n4 RR (95% CI)5 P 5 Pint6

Acute upper respiratory infection 1094 3.39 6 2.60 1072 3.70 6 2.80 0.92 (0.88, 0.97) 0.0005 1082 3.52 6 2.68 1084 3.57 6 2.73 0.98 (0.94, 1.03) 0.43 0.35
Acute lower respiratory infection 1040 0.63 6 0.95 1034 0.63 6 0.95 1.01 (0.91, 1.13) 0.86 1039 0.64 6 0.96 1035 0.62 6 0.94 1.02 (0.92, 1.14) 0.69 0.56
Pulmonary tuberculosis or other 1027 0.003 6 0.07 1022 0.001 6 0.03 3.01 (0.31, 28.86) 0.34 1027 0.002 6 0.06 1023 0.002 6 0.04 0.99 (0.14, 7.05) 0.99 —
causes of pneumonia
Diagnosis of any form of 1096 3.94 6 2.91 1074 4.26 6 3.06 0.93 (0.89, 0.97) 0.0006 1083 4.09 6 2.97 1087 4.11 6 3.01 0.99 (0.95, 1.03) 0.62 0.26
respiratory infection
Acute diarrhea 1033 0.46 6 0.77 1028 0.50 6 0.85 0.91 (0.81, 1.03) 0.15 1032 0.48 6 0.79 1029 0.47 6 0.83 1.02 (0.90, 1.16) 0.75 0.29
Dysentery 1048 0.43 6 0.79 1040 0.52 6 0.87 0.84 (0.74, 0.95) 0.006 1045 0.49 6 0.85 1043 0.46 6 0.80 1.05 (0.93, 1.19) 0.42 0.25
Persistent diarrhea 1028 0.003 6 0.05 1022 0.009 6 0.10 0.33 (0.09, 1.23) 0.10 1027 0.009 6 0.10 1023 0.003 6 0.05 1.99 (0.81,11.01) 0.10 0.71
Intestinal parasites 1030 0.14 6 0.38 1023 0.15 6 0.40 0.94 (0.74, 1.17) 0.56 1029 0.14 6 0.39 1024 0.14 6 0.39 0.97 (0.77, 1.21) 0.76 0.33
Diagnosis of any form of 1054 1.00 6 1.18 1043 1.14 6 1.33 0.88 (0.81, 0.96) 0.003 1049 1.09 6 1.29 1048 1.05 6 1.23 1.03 (0.95, 1.12) 0.43 0.24
diarrhea
Uncomplicated malaria 1052 0.87 6 1.08 1037 0.91 6 1.08 0.96 (0.87, 1.05) 0.33 1048 0.88 6 1.07 1041 0.90 6 1.09 0.98 (0.89, 1.07) 0.63 0.92
Severe malaria 1032 0.06 6 0.24 1023 0.06 6 0.24 1.05 (0.73, 1.52) 0.79 1028 0.05 6 0.22 1027 0.06 6 0.25 0.79 (0.55, 1.15) 0.22 0.34
Pallor/anemia 1029 0.16 6 0.45 1026 0.15 6 0.42 1.10 (0.88, 1.37) 0.41 1030 0.15 6 0.42 1025 0.16 6 0.45 0.88 (0.70, 1.09) 0.25 0.42
1
MV, multivitamin.
2
Received zinc ‘‘yes’’ refers to children who received zinc alone as well as those who received zinc and MVs. Received zinc ‘‘no’’ refers to all children who received MVs alone
and those who received the placebo.
3
Received MVs ‘‘yes’’ refers to children who received MVs alone as well as those who received zinc and MVs. Received MVs ‘‘no’’ refers to all children who received zinc alone
and those who received the placebo.
4
Mean 6 SD diagnoses over the course of follow-up.
5
RRs, 95% CIs, and corresponding P values were obtained from generalized estimating equations with the Poisson distribution and log link and by using the log of the follow-up
time as the offset variable.
6
P-interaction term.

2158 McDonald et al.

Tanzanian infants, but it did reduce episodes of vomiting and Several strengths and limitations of our study deserve
fever (13). A different study of daily zinc and/or multiple comment. As previously noted, to our knowledge, our study is
micronutrient supplementation involving rural Tanzanian chil- the first to examine the effects of preventive zinc and MV
dren 6–60 mo of age showed that zinc supplementation supplementation in African infants 6 wk of age. We enrolled
significantly decreased the rate of diarrhea and protected against a large number of participants and conducted intensive,
fever without localizing signs; however, multiple micronutrient monthly follow-up activities over an extended period of
supplementation actually increased the rate of diarrhea by time. This allowed us to evaluate the effects of supplementa-
approximately one-quarter (9). Similarly, in a trial of micro- tion during the period when rates of morbidity typically
nutrient powders with or without zinc among Pakistani children increase. We also assessed morbidity in a comprehensive
6–18 mo of age, Soofi et al. (8) observed an increase in the manner, using symptom diaries, monthly nurse evaluations,
proportion of days with diarrhea as well as higher rates of and regular physician examinations. The discordance in
bloody diarrhea and chest indrawing among children in the 2 results between the nurse evaluations and the physician diag-
micronutrient groups compared with a nonsupplemented con- noses is notable. One possible explanation is that mild illnesses
trol group. An earlier, relatively small study that provided zinc were only detected by study nurses at the monthly follow-up
with or without multiple micronutrients to Peruvian children visits, whereas more severe symptoms prompted a visit with
6–35 mo of age with persistent diarrhea showed similar results: study physicians (30). Other possible limitations include lack
zinc supplementation tended to reduce morbidity, whereas the of vitamin D in the supplement regimen and what might be
addition of MVs increased morbidity (22). However, it is worth considered a low dose of zinc given the possibility of impaired
noting that iron was included in the multiple micronutrient zinc absorption in African children (31).
supplement that was evaluated in these 3 trials (8, 9, 22), In summary, daily zinc supplementation of HIV-unexposed
whereas our MV regimen did not include iron. When taken as a Tanzanian infants at 6 wk of age significantly reduced physician-
whole, these findings do not support routine MV supplementa- diagnosed cases of diarrhea and acute upper respiratory tract
tion as an effective means of reducing infant and child morbidity, infections. Simultaneous provision of MVs did not appear to
even in resource-limited settings where diets are likely poor. confer additional benefit. Further investigation into the optimal
Our findings of increased CD4 T cell percentage with zinc timing, frequency, and duration of supplementation is needed
and MV supplementation suggest one possible immune effect of before policy recommendations can be formulated.
supplementation. Although evidence on the topic is particularly
limited in HIV-negative children, we previously reported an Acknowledgments
increase in CD4 counts among HIV-infected pregnant women in We are grateful to the field and study staff for their tireless ef-
Tanzania who received a similar MV regimen (23). Zinc and forts: Esther Kibona (deceased), Frank Killa, Michel Alexander,
multiple micronutrients have also been shown to increase CD4 Phares Zawadi, Susie Welty, Rachel Steinfeld, Anne Marie
counts in HIV-infected adults receiving highly active antiretro- Darling, James Okuma, Angela Jardin, Elizabeth Long, Jenna
viral therapy (24, 25). However, in a safety and efficacy study Golan, and Emily Dantzer. We thank Roland Kupka for ex-
involving 96 HIV-infected children in South Africa, daily supple- pert advice. We thank the members of the Data Safety and
mentation with 10 mg of elemental zinc for 6 mo did not Monitoring Board: Paul Jacques, Davidson Hamer, Roger
significantly change the percentage of CD4+ T lymphocytes (26). Mbise, and Zul Premji.
In an HIV-negative elderly population, Fortes et al. (27) reported CMM analyzed the data and wrote the manuscript; KPM
that zinc increased the number of CD4 cells. Although the designed the study, supervised data collection, and provided
compilation of these findings is encouraging, more research is input to the manuscript; RK provided input to the study design,
needed to better understand the potential mechanisms by which oversaw data collection, and reviewed the manuscript; SA
certain nutrients could be affecting this marker of immunologic oversaw all laboratory aspects of the study and reviewed the
function. manuscript; DS supervised the statistical analysis and reviewed
Although not statistically significant, our finding of a possible the manuscript; WWF designed the study, provided input to the
increase in the risk of all-cause mortality among children who statistical analysis, and reviewed the manuscript; CPD designed
received zinc is generally inconsistent with the bulk of the the study, oversaw study implementation, contributed to the
published evidence on this topic. Our results were unexpected, statistical analysis, and provided input to the manuscript. All
especially considering the fact that zinc reduced rates of diarrhea authors read and approved the final manuscript.
and acute upper respiratory infections, and we saw no differences
in rates of hospitalizations or unscheduled clinic visits between
treatment groups. Our trial did not have sufficient power to assess References
cause-specific mortality, which could have helped to elucidate 1. WHO [Internet]. Geneva (Switzerland): WHO; [cited 2013 Sep 10].
possible causal pathways driving this apparent effect. A meta- Available from: http://www.who.int/gho/child_health/mortality/mortality_
analysis concluded that therapeutic zinc supplementation for causes_region_text/en/index.html.
the treatment of diarrhea significantly reduces child mortality 2. Black RE, Victora CG, Walker SP, Bhutta ZA, Christian P, de Onis M,
Ezzati M, Grantham-McGregor S, Katz J, Martorell R, et al. Maternal
(5). Although the effect of preventive zinc supplementation on
and child undernutrition and overweight in low-income and middle-
mortality appears to be weaker, Brooks et al. (19) reported a income countries. Lancet 2013;382:427–51.
17% reduction in all-cause mortality among children who 3. Bhutta ZA, Das JK, Rizvi A, Gaffey MF, Walker N, Horton S, Webb P,
received weekly zinc supplementation for 12 mo, and most other Lartey A, Black RE, Lancet Nutrition Interventions Review Group;
studies, including a large study from Pemba Island, Tanzania, Maternal and Child Nutrition Study Group. Evidence-based interven-
have reported a nonsignificant reduction in mortality (18, 28, tions for improvement of maternal and child nutrition: what can be
done and at what cost? Lancet 2013;382:452–77.
29). Clearly, more research is needed to better understand the
4. Brown KH, Hess SY, Vosti SA, Baker SK. Comparison of the estimated
mechanisms through which zinc may affect certain causes of cost-effectiveness of preventive and therapeutic zinc supplementation
mortality and to resolve differences that could be attributed to strategies for reducing child morbidity and mortality in sub-Saharan
variations in study design and subject characteristics. Africa. Food Nutr Bull 2013;34:199–214.

Effects of multivitamins and zinc on child health 2159

5. Walker CL, Black RE. Zinc for the treatment of diarrhoea: effect on 19. Brooks WA, Santosham M, Naheed A, Goswami D, Wahed MA,
diarrhoea morbidity, mortality and incidence of future episodes. Int J Diener-West M, Faruque AS, Black RE. Effect of weekly zinc supple-
Epidemiol 2010;39 Suppl 1:i63–9. ments on incidence of pneumonia and diarrhoea in children younger
6. WHO/UNICEF. Clinical management of acute diarrhoea. Geneva than 2 years in an urban, low-income population in Bangladesh:
(Switzerland): WHO; 2004. randomised controlled trial. Lancet 2005;366:999–1004.
7. Brown KH, Peerson JM, Baker SK, Hess SY. Preventive zinc supple- 20. Malik A, Taneja DK, Devasenapathy N, Rajeshwari K. Short-course
mentation among infants, preschoolers, and older prepubertal children. prophylactic zinc supplementation for diarrhea morbidity in infants of 6
Food Nutr Bull 2009; 30 1 Suppl:S12–40. to 11 months. Pediatrics 2013;132:e46–52.
8. Soofi S, Cousens S, Iqbal SP, Akhund T, Khan J, Ahmed I, Zaidi AK, 21. Walker CLF, Bhutta ZA, Bhandari N, Teka T, Shahid F, Taneja S, Black
Bhutta ZA. Effect of provision of daily zinc and iron with several RE. Zinc during and in convalescence from diarrhea has no demon-
micronutrients on growth and morbidity among young children in strable effect on subsequent morbidity and anthropometric status
Pakistan: a cluster-randomised trial. Lancet 2013;382:29–40. among infants <6 mo of age. Am J Clin Nutr 2007;85:887–94.
9. Veenemans J, Schouten LR, Ottenhof MJ, Mank TG, Uges DR, Mbugi 22. Penny ME, Marin RM, Duran A, Peerson JM, Lanata CF, Lonnerdal B,
EV, Demir AY, Kraaijenhagen RJ, Savelkoul HF, Verhoef H. Effect of Black RE, Brown KH. Randomized controlled trial of the effect of daily
preventive supplementation with zinc and other micronutrients on non- supplementation with zinc or multiple micronutrients on the morbidity,
malarial morbidity in Tanzanian pre-school children: a randomized growth, and micronutrient status of young Peruvian children. Am J Clin
trial. PLoS One 2012;7:e41630. Nutr 2004;79:457–65.
10. Papathakis PC, Rollins NC, Chantry CJ, Bennish ML, Brown KH. 23. Fawzi WW, Msamanga GI, Spiegelman D, Urassa EJ, McGrath N,
Micronutrient status during lactation in HIV-infected and HIV- Mwakagile D, Antelman G, Mbise R, Herrera G, Kapiga S, et al.
uninfected South African women during the first 6 mo after delivery. Randomised trial of effects of vitamin supplements on pregnancy
Am J Clin Nutr 2007;85:182–92. outcomes and T cell counts in HIV-1-infected women in Tanzania.
11. Oken E, Kleinman KP, Rich-Edwards J, Gillman MW. A nearly Lancet 1998;351:1477–82.
continuous measure of birth weight for gestational age using a United 24. Kaiser JD, Campa AM, Ondercin JP, Leoung GS, Pless RF, Baum MK.
States national reference. BMC Pediatr 2003;3:6. Micronutrient supplementation increases CD4 count in HIV-infected indi-
12. Fawzi WW, Msamanga GI, Urassa W, Hertzmark E, Petraro P, Willett viduals on highly active antiretroviral therapy: a prospective, double-blinded,
WC, Spiegelman D. Vitamins and perinatal outcomes among HIV- placebo-controlled trial. J Acquir Immune Defic Syndr 2006;42:523–8.
negative women in Tanzania. N Engl J Med 2007;356:1423–31. 25. Baum MK, Lai S, Sales S, Page JB, Campa A. Randomized, controlled
13. Duggan C, Manji KP, Kupka R, Bosch RJ, Aboud S, Kisenge R, Okuma clinical trial of zinc supplementation to prevent immunological failure in
J, Fawzi WW. Multiple micronutrient supplementation in Tanzanian HIV-infected adults. Clinical Infectious Diseases 2010;50 (12):1653–60.
infants born to HIV-infected mothers: a randomized, double-blind, 26. Bobat R, Coovadia H, Stephen C, Naidoo KL, McKerrow N, Black RE,
placebo-controlled clinical trial. Am J Clin Nutr 2012;96:1437–46. Moss WJ. Safety and efficacy of zinc supplementation for children with
14. Wang SJ, Lengeler C, Mtasiwa D, Mshana T, Manane L, Maro G, HIV-1 infection in South Africa: a randomised double-blind placebo-
Tanner M. Rapid Urban Malaria Appraisal (RUMA) II: epidemiology of controlled trial. Lancet 2005;366:1862–7.
urban malaria in Dar es Salaam (Tanzania). Malar J 2006;5:28. 27. Fortes C, Forastiere F, Agabiti N, Fano V, Pacifici R, Virgili F, Piras G, Guidi L,
15. Sazawal S, Black RE, Ramsan M, Chwaya HM, Stoltzfus RJ, Dutta A, Bartoloni C, Tricerri A, et al. The effect of zinc and vitamin A supplementation
Dhingra U, Kabole I, Deb S, Othman MK, et al. Effects of routine on immune response in an older population. J Am Geriatr Soc 1998;46:19–26.
prophylactic supplementation with iron and folic acid on admission to 28. Sazawal S, Black RE, Ramsan M, Chwaya HM, Dutta A, Dhingra U,
hospital and mortality in preschool children in a high malaria trans- Stoltzfus RJ, Othman MK, Kabole FM. Effect of zinc supplementation
mission setting: community-based, randomised, placebo-controlled trial. on mortality in children aged 1–48 months: a community-based
Lancet 2006;367:133–43. randomised placebo-controlled trial. Lancet 2007;369:927–34.
16. Fawzi WW, Mbise R, Spiegelman D, Fataki M, Hertzmark E, Ndossi G. 29. Yakoob MY, Theodoratou E, Jabeen A, Imdad A, Eisele TP, Ferguson J, Jhass
Vitamin A supplements and diarrheal and respiratory infections among A, Rudan I, Campbell H, Black RE, et al. Preventive zinc supplementation in
children in Dar es Salaam, Tanzania. J Pediatr 2000;137:660–7. developing countries: impact on mortality and morbidity due to diarrhea,
17. Osendarp SJ, Santosham M, Black RE, Wahed M, van Raaij JM, Fuchs pneumonia and malaria. BMC Public Health 2011;11 Suppl 3:S23.
GJ. Effect of zinc supplementation between 1 and 6 mo of life on 30. Schmidt WP, Arnold BF, Boisson S, Genser B, Luby SP, Barreto ML,
growth and morbidity of Bangladeshi infants in urban slums. Am J Clin Clasen T, Cairncross S. Epidemiological methods in diarrhoea studies–an
Nutr 2002;76:1401–8. update. Int J Epidemiol 2011;40:1678–92.
18. Tielsch JM, Khatry SK, Stoltzfus RJ, Katz J, LeClerq SC, Adhikari R, 31. Manary MJ, Abrams SA, Griffin IJ, Quimper MM, Shulman RJ, Hamzo
Mullany LC, Black R, Shresta S. Effect of daily zinc supplementation on MG, Chen Z, Maleta K, Manary MJ. Perturbed zinc homeostasis in rural
child mortality in southern Nepal: a community-based, cluster 3–5-y-old Malawian children is associated with abnormalities in intestinal
randomised, placebo-controlled trial. Lancet 2007;370:1230–9. permeability attributed to tropical enteropathy. Pediatr Res 2010;67:671–5.

2160 McDonald et al.