Document downloaded from http://www.elsevier.es, day 03/12/2015. This copy is for personal use.

Any transmission of this document by any media or format is strictly prohibited.

Articles

Effect of weekly zinc supplements on incidence of

pneumonia and diarrhoea in children younger than 2 years in
an urban, low-income population in Bangladesh:
randomised controlled trial
W Abdullah Brooks, Mathuram Santosham, Aliya Naheed, Doli Goswami, M Abdul Wahed, Marie Diener-West, Abu S G Faruque, Robert E Black

Summary
Background Pneumonia and diarrhoea cause much morbidity and mortality in children younger than 5 years. Most Lancet 2005; 366: 9991004
deaths occur during infancy and in developing countries. Daily regimens of zinc have been reported to prevent acute Published online August 23, 2005
lower respiratory tract infection and diarrhoea, and to reduce child mortality. We aimed to examine whether giving DOI:10.1016/S0140-6736(05)
67109-7
zinc weekly could prevent clinical pneumonia and diarrhoea in children younger than 2 years.
The Centre for Health and
Population Research,
Methods 1665 poor, urban children aged 60 days to 12 months were randomly assigned zinc (70 mg) or placebo International Centre for
orally once weekly for 12 months. Children were assessed every week by eld research assistants. Our primary Diarrhoea Disease Research,
outcomes were the rate of pneumonia and diarrhoea. The rates of other respiratory tract infections were the Mohakhali Dhaka 1000,
Bangladesh
secondary outcomes. Growth, nal serum copper, and nal haemoglobin were also measured. Analysis was by
(Prof W A Brooks MPH,
intention to treat. A Naheed MPH,
D Goswami MBBS,
Findings 34 children were excluded before random assignment to treatment group because they had tuberculosis. M A Wahed BA,
A S G Faruque MPH); Bloomberg
809 children were assigned zinc, and 812 placebo. After treatment assignment, 103 children in the treatment group
School of Public Health,
and 44 in the control group withdrew. There were signicantly fewer incidents of pneumonia in the zinc group than Department of International
the control group (199 vs 286; relative risk 083, 95% CI 073095), and a small but signicant effect on incidence Health (Prof M Santosham MPH,
of diarrhoea (1881 cases vs 2407; 094, 088099). There were two deaths in the zinc group and 14 in the placebo Prof R E Black MPH), and
Department of Biostatistics
group (p=0013).There were no pneumonia-related deaths in the zinc group, but ten in the placebo group (p=0013).
(Prof M Diener-West PhD), Johns
The zinc group had a small gain in height, but not weight at 10 months compared with the placebo group. Serum Hopkins University, Baltimore,
copper and haemoglobin concentrations were not adversely affected after 10 months of zinc supplementation. MD, USA
Correspondence to:
Interpretation 70 mg of zinc weekly reduces pneumonia and mortality in young children. However, compliance with Prof W Abdullah Brooks
abrooks@icddrb.org
weekly intake might be problematic outside a research programme.

Introduction 2 years,710,13,15 although younger children might be more

Globally, from 2000 to 2003, pneumonia caused about
20 million (19%) of 106 million deaths among
children younger than 5 years,1 more than 90% of which
occurred in less developed countries.2 Pneumonia
prevention strategies typically rely on vaccines and are
complicated by inadequate efcacy of non-conjugate
vaccines for children younger than 2 years,3 the high cost
of effective protein-conjugate vaccines, and incomplete
serotype vaccine coverage for Streptococcus pneumoniae.
An effective alternative is needed for very young children
in high-risk populations.
Mortality from diarrhoea has decreased from
45 million per year in 1979,4 mostly as a result of the
introduction of oral rehydration therapy. However,
diarrhoea causes 19 million (18%) deaths among
children younger than 5 years,1 and about 20% of young
child deaths in Bangladesh.5 Zinc deciency is common
where pneumonia and diarrhoea are common,6 and
daily regimens of zinc prevent and treat pneumonia710
and diarrhoea8,1113 and reduce child mortality.6,13,14
Research into zinc for the prevention of pneumonia
and diarrhoea has mostly included children older than
vulnerable,2 particularly to pneumonia. In most of these
studies zinc has been given daily,710,13,15 which can be
impractical. Pneumonia and diarrhoea are the rst and
second leading causes of patient visits, respectively, to
our urban eld clinic in Kamalapur, Bangladesh.
We studied whether a large weekly dose of zinc
(70 mg) would reduce clinical pneumonia, diarrhoea,
and other morbidity in children younger than 2 years.
We also studied the ponderal and linear-growth effects
of this regimen. Because long-term intake of supple-
mental zinc could affect iron and copper metabolism,16,17
we measured the effects on haemoglobin and serum
copper.
Methods
Participants
We did the study in a poor, urban population:
Kamalapur, southeastern Dhaka, Bangladesh. We
included children aged 60 days to 12 months at the time
of enrolment. We excluded those with known or
suspected tuberculosis, chronic respiratory or congenital
heart disease, or severe malnutrition requiring hospital

www.thelancet.com Vol 366 September 17, 2005 999

Document downloaded from http://www.elsevier.es, day 03/12/2015. This copy is for personal use. Any transmission of this document by any media or format is strictly prohibited.
Articles
admission. Children with suspected tuberculosis were
referred to Dhakas hospital for tuberculosis.
The research review committee and ethical review
committee of the International Centre for Diarrhoeal
Disease Research approved the study. Parents or carers
gave informed written consent at enrolment.
Procedures
Random assignment to zinc or placebo was done with
permuted blocks of variable length between two and
eight. Zinc was given orally as a syrup (35 mg zinc
acetate per 5 mL). The placebo was non-nutritious and
vitamin-free, designed to be identical to the zinc syrup
in colour, odour, and taste. ACME Laboratories, Ltd
(Dhamrai, Dhaka, Bangladesh) prepared, labelled, and
masked the identity of both preparations.
Field research assistants (FRAs) did active
surveillance, visiting every enrolled child at home once
weekly. They collected information about specic signs
of respiratory disease and diarrhoea for each day since
the last visit by use of a pre-coded calendar
questionnaire standardised during training before
deployment. To avoid bias, FRAs used a timer to count
every childs respiratory rate for 60 s, irrespective of
reported symptoms, at each visit. If the respiratory rate
was above 50 breaths per min, they counted a second
time and calculated the mean. FRAs also inspected any
diarrhoeal-stool samples. The FRA recorded all
reported and observed data on the questionnaire,
referring children to the clinic for any observed or
reported ndings that were suggestive of respiratory
disease or diarrhoea.
At the end of each visit, the FRA observed the child
take the syrup, and waited at least 5 min before leaving.
Compliance required intake of two teaspoons of syrup
(10 mL). If the child vomited within 5 min, the syrup
was given again once.
At the clinic, medical ofcers and nurses involved in
this study assessed patients. Nurses collected vital signs
(respiratory rate, axillary temperature by mercury
thermometer, pulse rate, and blood pressure). Fever
was dened as an axillary temperature of at least 38C.
During the assessment, clothing was removed from the
childs torso to assess chest indrawing and cyanosis.
The medical ofcers auscultated the chest by
stethoscope. Danger signs, such as chest indrawing,
cyanosis, lethargy, or inability to drink, were recorded.
Inability to drink was assessed by having the mother
offer the child oral rehydration solution. Children who
were too lethargic to drink or unable to swallow were
recorded as unable to drink. Dehydration was
assessed with WHO criteria.18 The medical ofcer
diagnosed pneumonia if crepitations were heard on
inspiration with a respiratory rate greater than
50 breaths per minute; severe pneumonia was
diagnosed if there was also chest indrawing, or at least
one other danger sign. Children with wheezing or
1000
rhonchi with crepitations were also diagnosed with
pneumonia. Children who had only wheezing or
rhonchi without crepitations were not diagnosed with
pneumonia, but with reactive airways disease or
bronchiolitis (reversible lower airway obstruction).
Diarrhoea was diagnosed if the child had three or
more watery stools or one bloody stool in a 24-h period.
A stool sample was required for diagnosis.
Patients with purulent ear discharge were diagnosed
with suppurative otitis media. Upper respiratory tract
infection consisted of cough, rhinorrhoea, and fever
without tachypnoea (ie, no sign of lower airway
obstruction).
We treated pneumonia with co-trimoxazole (10 mg/kg
trimethoprim, twice daily for 5 days). Children with
WHO-dened pneumonia18 were also treated, although
their illness was not counted unless it met study
criteria. Children on antibiotics were assessed within
72 h of starting treatment; those who did not improve
(ie, the respiratory rate did not change by more than
5 breaths per min from baseline) were switched to
treatment with amoxicillin (40 mg/kg, three times daily
for 5 days). If oral treatment failed, or if they had severe
pneumonia, we referred them to hospital for parenteral
treatment (ceftriaxone 75 mg/kg intramuscularly per
day). Children with only expiratory wheezes or rhonchi
were managed with salbutamol syrup (03 mg/kg, three
times daily), or referred to hospital for danger signs.
Patients with blood in their diarrhoea were treated with
co-trimoxazole for 5 days, or if ineffective, with nalidixic
acid. Children with some dehydration were treated
according to WHO guidelines with oral rehydration
solution,18 and those with severe dehydration were
admitted to hospital.
FRAs followed up patients who were home and taking
antibiotics within 72 h to document compliance (asking
the caretaker to explain the dosing regimen, and
checking the remaining volume) and the end of the
illness episode. The end of an episode of respiratory
illness was dened as 7 disease-free days after the last
day of illness; for diarrhoea, the disease-free interval
was 3 days. Disease-free for pneumonia meant a
respiratory rate less than 50 breaths per min without
danger signs or fever, for suppurative otitis media no
ear discharge, and for diarrhoea no watery or bloody
stool.
Medical ofcers collected blood (3 mL) at baseline
before zinc or placebo was given and again during the
10th month. The nal sample was taken at 10 months
rather than 12 months to minimise attrition. Trace-
element-free vacutainers and storage tubes were used
to measure serum zinc, haemoglobin, and copper
concentrations. Serum was separated at the
International Centre for Diarrhoeal Disease Research,
Dhaka, within 3 h of collection. Serum samples
(150 mL) in polypropylene tubes were diluted 1/12 with
HNO3 and Brij35, a non-ionic detergent. Flame atomic
www.thelancet.com Vol 366 September 17, 2005
Document downloaded from http://www.elsevier.es, day 03/12/2015. This copy is for personal use. Any transmission of this document by any media or format is strictly prohibited.

Articles

censored (ie, data were included in analysis up to the

1655 children aged 212 months
enrolled in the community
point of withdrawal, even though they did not complete
the observation period) after the last visit date, and
34 had tuberculosis
earlier data were kept for analysis. We did exit interviews
for all withdrawals to identify possible biases, and for
those who died we did verbal autopsies between 2 and
1621 randomised
6 weeks after death.
Disease incidence was the number of episodes divided
by child-years of observation. Because outcomes were
809 assigned zinc 812 assigned placebo infrequent, we compared incidence rates by Poisson
regression for relative risk with 95% CIs, controlling for
103 withdrew 44 withdrew clustering (multiple observations) within groups.
from study from study Analysis of variance was used to compare mean serum
zinc, copper, haemoglobin, and white-blood-cell
706 completed (427 768 completed (511 concentrations, and growth before and after
child-years of child-years of
supplementation. For dichotomous variables, p values
observation) observation)
were calculated with a two-tailed Fishers exact test.

Figure: Trial prole Role of the funding source

The funding sources had no role in the study design,
absorption spectrophotometry was used to assess zinc data collection, data analysis, interpretation of results, or
and copper concentration, as previously described.19 decision to publish this research. The corresponding
Similarly, measurements taken at baseline and at author had full access to the data and the nal
10 months assessed growth; the mean of two responsibility for the decision to submit for publication.
consecutive measures of nude bodyweight and length
were recorded. These measurements were analysed for Results
only a small sample of the study population (329 in the Between April 21, 1999, and August 29, 2000,
zinc group and 309 in the placebo group) primarily 1655 children age 212 months were enrolled; 34 of
because of logistical constraints of getting well children these had tuberculosis and were excluded, leaving 1621
to clinic. We estimated that if there was a 10% difference children to be randomised. 809 children were randomly
in growth we would need 190 children per group if the assigned zinc, and 812 placebo (gure). 147 (91%)
SD were about 03 (a third of the growth velocity). Thus, children withdrew from the study, and their
we recruited 300 per group plus 20% for attrition. Power demographic and illness characteristics were not
is thus 948%. Measurements were compared with different from those who completed it. Withdrawal from
population data at the National Center for Health both groups was most commonly attributed to the
Statistics.20 childs reaction to the taste of the syrup, which
sometimes resulted in regurgitation. Most of those who
Statistical analysis withdrew were young, primarily breastfed infants. The
We used the clinical pneumonia incidence from Matlab, highest proportion (371%) of withdrawals for both
a rural surveillance site, to calculate sample size because
urban data were not available. Among children age Zinc group, n=809 Placebo group, n=812
259 months the rate was 02 episodes per child per Number aged 25 months (%) 439 (543%) 405 (499%)
year. To detect a 30% decrease in the incidence of pneu- Number aged 611 months (%) 370 (457%) 407 (501%)
Number of boys (%) 445 (550%) 397 (489%)
monia, with a type 1 error of 5%, a type 2 error of 20%,
Number with smoker in family (%) 629 (777%) 609 (750%)
and 20% attrition, we needed 776 child-years of Number who use outdoor gas cooker* (%) 561 (693%) 592 (729%)
observation per group. Diarrhoea rates of 46 episodes Number with a full-term gestation (%) 739 (913%) 748 (921%)
per child per year were used. Allowing for 20% attrition, Number ever admitted to hospital (%) 101 (125%) 105 (129%)
Number of mothers who reported vaccinations up to date (%) 512 (633%) 499 (614%)
a type 1 error of 5%, and a type 2 error of 10%, we
Mean age in months (SD) 54 (31) 52 (29)
needed 163 child-years per group. Mean age of mothers (SD) 238 (51) 242 (55)
Statistical analysis was done with StataSE Release Mean income in US$ per month (SD) 617 (410) 615 (389)
82.21 Each childs observation period started from the Mean number of years mother had been educated (SD) 26 (32) 25 (34)
Mean household size (SD) 54 (21) 55 (23)
rst FRA visit at which syrup was given and continued
Number of children 5 years age (SD) 14 (06) 14 (06)
for 12 months. If a child was absent when the FRA
visited, that week was removed from analysis owing to *This is a marker for exposure to air pollution that can exacerbate respiratory disease and is important to determine common
risk between groups.
the absence of morbidity data. If the child was present,
irrespective of zinc compliance, the week was included. Table 1: Demographic characteristics of children at enrolment
Children who withdrew or emigrated were right-

www.thelancet.com Vol 366 September 17, 2005 1001

Document downloaded from http://www.elsevier.es, day 03/12/2015. This copy is for personal use. Any transmission of this document by any media or format is strictly prohibited.

Articles

Baseline 10 months Number in Number in

Zinc group Placebo group p Zinc group Placebo group p zinc group placebo group
n=329 (SD) n=309 (SD) n=329 (SD) n=309 (SD) Diarrhoea/malnutrition 2 1
Serum zinc (mol/L) 99 (22) 97 (18) 0324 110 (26) 99 (15) 00004 Oral infection/malnutrition 0 1
Haemoglobulin (g/L) 98 (11) 100 (118) 0123 96 (91) 97 (94) 0299 Pneumonia 0 10
Serum copper (mol/L) 019 (004) 019 (004) 0442 022 (003) 022 (004) 0863 Poisoning 0 1
White blood cells (109/L) 118 (25) 118 (20) 0732 112 (24) 112 (25) 0711 Sepsis 0 1
Height (cm) 625 (58) 628 (50) 0374 752 (35) 743 (40) 0022
Table 4: Cause of death
Weight (kg) 64 (15) 63 (13) 0496 87 (11) 85 (12) 0137

Table 2: Clinical and laboratory results before treatment and after 10 months of treatment
groups occurred at age 2 months, with 771% younger
than 6 months. The mean length of time in the study
before withdrawal was 55 days, with no differences
between groups (p=0392). There was no difference
between the number of children under age 6 months in
the zinc (511%) and the placebo (470%) groups who
completed the study (p = 012).
There were no signicant demographic differences
between groups at baseline (table 1), except for a slightly
higher proportion of boys in the zinc group. 16 children
(21%) in the placebo group and 29 (41%) in the zinc
group received less than 80% of the doses; mean
proportion of doses received by each group was not
different (943% in the placebo group compared with
932% in the zinc group; p=010). Before treatment,
there were no differences between the groups in
concentrations of serum zinc, haemoglobin, copper, or
white blood cells, or in anthropometry; however, after
10 months of treatment, the zinc group had signicantly
greater serum-zinc concentrations than the placebo
group (table 2). By contrast there were no differences
between the groups at 10 months for haemoglobin,
copper, white blood cells, or weight gain. In addition, the
groups white-blood-cell differential count did not differ
at baseline and 10 months (data not shown).
Children given zinc grew a mean of 09 cm taller
during 10 months than those given placebo (table 2).
Mean monthly linear-growth velocity was higher
(p=0003) in the zinc group (14 cm, SD 04) than in the
placebo group (13 cm, SD 03).
Follow-up per child was for a mean duration of
324 days (SD 97) for the zinc group and 320 days (78) for
the placebo group (p=0332). The overall incidence of
pneumonia that met our study denitions in the placebo
Zinc Placebo Relative risk
(child years=427) (child years=511) (95% CI)
Diarrhoea 1881 2407 094 (088099)
Upper respiratory infection 4834 6294 092 (088097)
Reactive airways disease or bronchiolitis 232 314 088 079099)
Suppurative otitis media 394 572 058 (041082)
Pneumonia 199 286 083 (073095)
Severe pneumonia 18 42 051 (030088)
Death 2 14 015 (003067)
Table 3: Number of medical ofcer diagnoses in the clinic
group was 056 episodes per child-year. The power for
the observed sample size was thus 849% for
pneumonia. There were signicantly fewer incidents of
pneumonia in the zinc group than in the placebo group
(table 3). Fewer children in the treatment group (274,
388%) than in the control group (394, 513%)
contracted pneumonia more than once (p00001).
There was a small but signicant difference between
groups in the incidence of diarrhoea; 681 (964%) in
the zinc group and 743 (967%) in the placebo group
had more than one episode (p=0775). All other
illnesses diagnosed by medical ofcers were
signicantly less frequent in the zinc group than the
placebo group (table 3). The preventive effects of zinc
increased with disease severity: zinc reduced upper
respiratory infections by 8%, reactive airways disease or
bronchiolitis by 12%, pneumonia by 17%, severe
pneumonia by 49%, and pneumonia mortality by
100%. Adjustment of the Poisson model for the
proportion of doses received (compliance) had no effect
on outcome.
Mortality was signicantly lower for the zinc group
than the placebo group. Although ten of 14 deaths in the
placebo group were pneumonia related (table 4), there
were no pneumonia-related deaths in the zinc group.
Children younger than 6 months who took zinc had
less pneumonia (relative risk 076, 022229), severe
pneumonia (063, 006144), and diarrhoea (090,
063128) than those who took the placebo. Children
younger than 12 months also had less pneumonia (085,
062117) and diarrhoea (0.95, 086106) but not
severe pneumonia (101, 039256).
Discussion
Zinc substantially reduced the incidence of pneumonia
and other upper and lower respiratory tract disease, and
modestly reduced that of diarrhoea. However, the effect
p value of zinc on mortality was strong. This nding in a study
of young children in urban Bangladesh extends those of
0030 earlier studies, which found a 68% decrease in mortality
0001 in full-term infants who were small for their gestational
0042
0002
age and treated with daily zinc in India,14 and a
0004 50% reduction in mortality in children younger than
0016 5 years in rural Bangladesh for daily zinc used for
0013
14 days to treat diarrhoea.13 The effect on diarrhoea
is modest compared with other reports,8,13 but
supports the preventive benet described in manage-

1002 www.thelancet.com Vol 366 September 17, 2005

Document downloaded from http://www.elsevier.es, day 03/12/2015. This copy is for personal use. Any transmission of this document by any media or format is strictly prohibited.
ment recommendations for acute diarrhoea.4 Daily zinc
doses might have had more effect on diarrhoea.
We found a small but signicant effect on linear
growth, but not ponderal growth, which supports the
results of meta-analyses.22,23
Five factors are relevant to the use of zinc for
respiratory infection control: age, disease severity,
disease aetiology, safety of long-term intake of zinc, and
dosing interval. First, regarding age, the study showed
that zinc substantially reduces illness among children
aged 224 months; these children are vulnerable and
ineligible for the non-conjugate vaccine. Rates of
reduction for age-specic subgroups were probably not
signicant because of inadequate numbers studied.
Second, the incidence of clinically conrmed
pneumonia, severe pneumonia, and other respiratory
illnesses was lower in the treatment group. Zinc might
be progressively protective against more invasive and
severe disease, leading to an 85% reduction in overall
mortality, primarily owing to pneumonia. Zinc
supplementation was associated with a reduction in
otitis-media incidence. Otitis media is a common
paediatric infection, particularly in less developed
countries,24,25 with important sequelae.2629 Our ndings
suggest that zinc could result in substantial treatment-
cost savings, reduction in hearing-related disabilities,
and improved quality of life.
Third, although we have no aetiological data,
bronchiolitis and upper respiratory tract infection are
likely to be viral, whereas most cases of severe
pneumonia and pneumonia-related deaths are more
likely to be bacterial; the incidence of all these disorders
was reduced by zinc. Zinc could have a substantial effect,
independent of aetiology, but may have larger effects on
more severe (possibly bacterial) disease.
Fourth, zinc was found to be safe in young children at
a weekly dose of 70 mg. Apart from vomiting in a small
subpopulation, which may have caused early dropouts,
there has been concern that long-term zinc
supplementation may interfere with iron and copper
absorption. If this were the case we would expect to see
decreased serum haemoglobin, copper, and white blood
cells, none of which we observed.
Fifth, our ndings with a weekly dose are an important
step forward in our thinking about, and programmatic
use of, zinc. Daily regimens, though promising, might
not be practical for long-term use. This study suggests
that weekly dosing is benecial and cost effective, even
with reduced compliance. A previous study found a 40%
reduction in the rate of acute lower respiratory infection
in malnourished children receiving both iron and zinc
(20 mg once weekly);30 these results could be more
modest than ours because of the dose. Dosing intervals
that are even longer than weekly might be protective
against certain diseases, because 20 mg zinc for 2 weeks
signicantly reduced subsequent diarrhoea up to
3 months in children younger than 59 months.13
www.thelancet.com Vol 366 September 17, 2005
Articles
Exclusive breastfeeding is encouraged for the rst
46 months of life,18 but is uncommon in this age-
group.31 Thus, we did not attempt to address the effect of
breastfeeding on zinc.
A limitation of our study is that more children in the
zinc group than in the control group withdrew. We do
not believe that these withdrawals affect our results for
several reasons. First, they occurred early, typically
within 2 months of randomisation. By withdrawing early
their risk would be less likely affected by treatment
assignment. Early dropouts in both groups should have
had similar risk proles. Second, they did not
signicantly change the proportion of young children in
each group. Third, there was less than 15% attrition
from either group and less than 10% attrition overall,
well below the projected rate of 20%. Finally, blinding of
FRAs was not affected because a proportion of children
in both zinc and placebo groups reacted to the taste such
that the treatment could not be distinguished. The
astringent taste of the concentrated zinc syrup, and
attempts to duplicate the taste in the placebo, probably
contributed to regurgitation among young children, and
their withdrawal. Present strategies for giving zinc orally
are dispersible tablets, which seem to be well tolerated.
Our ndings suggest that among young children zinc
has a substantial protective effect against pneumonia,
severe pneumonia, suppurative otitis media, and most
importantly, mortality secondary to pneumonia.
Treatment with zinc had a modest effect on diarrhoea
and growth, and no measurable adverse effects on
copper or haemoglobin. Future studies should assess
the optimum dose and length of protection after periodic
zinc supplementation.
Conict of interest statement
We declare that we have no conict of interest.
Contributors
WAB was the principal investigator and provided primary conception,
design, execution, data analysis, and preparation of the paper. MS was
the co-principal investigator, and was involved in the design, execution,
data analysis, and paper preparation. AN oversaw the daily execution of
the study, data collection and entry, and the eld staff. DG was the chief
medical ofcer responsible for clinical protocol adherence. MAW did
the assays of zinc and copper. MD-W assisted with data analysis and
paper preparation. ASGF assisted with study conception, design,
training, and completion, and helped to prepare the paper. REB was the
senior team member and contributed to the design, interim discussions
of the research progress, data analysis, and paper preparation.
Acknowledgments
The research was funded by Johns Hopkins Family Health and Child
Survival Cooperative Agreement with the US Agency for International
Development, the Swiss Development Corporation, and a cooperative
agreement between the US Agency for International Development
(HRN-A-00-96-90005-00) and core donors to the Centre for Health and
Population Research. We are grateful to ACME Laboratories Ltd for
preparation of the zinc and placebo syrups.
References
1 Bryce J, Boschi-Pinto C, Shibuya K, Black RE. WHO estimates of
the causes of death in children. Lancet 2005; 365: 114752.
2 Williams BG, Gouws E, Boschi-Pinto C, Bryce J, Dye C. Estimates
of world-wide distribution of child deaths from acute respiratory
infections. Lancet Infect Dis 2002; 2: 2532.
1003
Document downloaded from http://www.elsevier.es, day 03/12/2015. This copy is for personal use. Any transmission of this document by any media or format is strictly prohibited.
Articles
3 Ledwith M. Pneumococcal conjugate vaccine. Curr Opin Pediatr
2001; 13: 7074.
4 United Nations Childrens Fund/World Health Organization.
WHO/UNICEF joint statement: clinical management of acute
diarrhoea. New York/Geneva: UNICEF/WHO, 2004: 18.
5 Baqui AH, Black RE, Arifeen SE, Hill K, Mitra SN, al Sabir A.
Causes of childhood deaths in Bangladesh: results of a nationwide
verbal autopsy study. Bull World Health Organ 1998; 76: 16171.
6 Black RE, Sazawal S. Zinc and childhood infectious disease
morbidity and mortality. Br J Nutr 2001; 85 (suppl 2): S12529.
7 Bhandari N, Bahl R, Taneja S, et al. Effect of routine zinc
supplementation on pneumonia in children aged 6 months to
3 years: randomised controlled trial in an urban slum. BMJ 2002;
324: 1358.
8 Bhutta ZA, Black RE, Brown KH, et al. Prevention of diarrhea and
pneumonia by zinc supplementation in children in developing
countries: pooled analysis of randomized controlled trials. J Pediatr
1999; 135: 68997.
9 Ninh NX, Thissen JP, Collette L, Gerard G, Khoi HH,
Ketelslegers JM. Zinc supplementation increases growth and
circulating insulin-like growth factor I (IGF-I) in growth-retarded
Vietnamese children. Am J Clin Nutr 1996; 63: 51419.
10 Sazawal S, Black RE, Jalla S, Mazumdar S, Sinha A, Bhan MK.
Zinc supplementation reduces the incidence of acute lower
respiratory infections in infants and preschool children: a double-
blind, controlled trial. Pediatrics 1998; 102: 15.
11 Roy SK, Tomkins AM, Akramuzzaman SM, et al. Randomised
controlled trial of zinc supplementation in malnourished
Bangladeshi children with acute diarrhoea. Arch Dis Child 1997; 77:
196200.
12 Bhutta ZA, Bird SM, Black RE, et al. Therapeutic effects of oral
zinc in acute and persistent diarrhea in children in developing
countries: pooled analysis of randomized controlled trials.
Am J Clin Nutr 2000; 72: 151622.
13 Baqui AH, Black RE, Arifeen SE, et al. Effect of zinc
supplementation started during diarrhoea on morbidity and
mortality in Bangladeshi children: community randomised trial.
BMJ 2002; 325: 1059.
14 Sazawal S, Black RE, Menon VP, et al. Zinc supplementation in
infants born small for gestational age reduces mortality: a
prospective, randomized, controlled trial. Pediatrics 2001; 108:
128086.
15 Rosado JL, Allen LH, Lopez P. The effect of zinc and/or iron
supplementation on morbidity: a double blind, randomised
community trial in preschoolers. FASEB J 1995; 9: 157 (abstr).
1004
16 Festa MD, Anderson HL, Dowdy RP, Ellersieck MR. Effect of zinc
intake on copper excretion and retention in men. Am J Clin Nutr
1985; 41: 28592.
17 Fosmire GJ. Zinc toxicity. Am J Clin Nutr 1990; 51: 22527.
18 WHO/UNICEF. Management of the child with a serious infection
or severe malnutrition: guidelines for care at the rst-referral level
in developing countries. Geneva: Department of Child and
Adolescent Health and Development UNICEF, 2000: 162.
19 Brooks WA, Yunus M, Santosham M, et al. Zinc for severe
pneumonia in very young children: double-blind placebo-controlled
trial. Lancet 2004; 363: 168388.
20 Hamil PV, Drizd TA, Johnson CL. Physical growth: national center
for health statistic percentages. Am J Clin Nutr 1979; 32: 60729.
21 StataCorp. Stata statistical software: release 80. College Station,
TX: Stata Corporation, 2003.
22 Brown KH, Peerson JM, Allen LH. Effect of zinc supplementation
on childrens growth: a meta-analysis of intervention trials.
Bibl Nutr Dieta 1998; 54: 7683.
23 Brown KH, Peerson JM, Rivera J, Allen LH. Effect of supplemental
zinc on the growth and serum zinc concentrations of prepubertal
children: a meta-analysis of randomized controlled trials.
Am J Clin Nutr 2002; 75: 106271.
24 Berman S. Otitis media in developing countries. Pediatrics 1995;
96: 12631.
25 Hatcher J, Smith A, Mackenzie I, et al. A prevalence study of ear
problems in school children in Kiambu district, Kenya, May 1992.
Int J Pediatr Otorhinolaryngol 1995; 33: 197205.
26 Attallah MS, al-Essa A. Hearing results in tympanoplasty in
Riyadh. Otolaryngol Pol 1996; 50: 14551.
27 Rutka J, Lekagul S. No therapy: use, abuse, efcacy, and
morbiditythe European versus the third-world experience.
J Otolaryngol 1998; 27(suppl 2): 4348.
28 Smith A, Hatcher J. Preventing deafness in Africas children.
Afr Health 1992; 15: 3335.
29 Smith AW, Hatcher J, Mackenzie IJ, et al. Randomised controlled
trial of treatment of chronic suppurative otitis media in Kenyan
schoolchildren. Lancet 1996; 348: 112833.
30 Baqui AH, Zaman K, Persson LA, et al. Simultaneous weekly
supplementation of iron and zinc is associated with lower
morbidity due to diarrhea and acute lower respiratory infection in
Bangladeshi infants. J Nutr 2003; 133: 415057.
31 Brooks WA, Santosham M, Roy SK, et al. Efcacy of zinc in young
infants with acute watery diarrhoea. Am J Clin Nutr (in press).
www.thelancet.com Vol 366 September 17, 2005