Secondary Logo

Journal Logo

Sex-specific Responses to Zinc Supplementation in Nouna, Burkina Faso

Garenne, Michel*; Becher, Heiko; Ye, Yasome; Kouyate, Bokar; Müller, Olaf

Journal of Pediatric Gastroenterology and Nutrition: May 2007 - Volume 44 - Issue 5 - p 619–628
doi: 10.1097/MPG.0b013e31802c695e
Original Articles: Hepatology & Nutrition

Objectives: To study the different responses by sex to zinc supplementation among young children.

Study Children and Methods: Double-blind randomized controlled trial of zinc supplementation in 686 children aged 6–30 months, conducted in Nouna, a rural area of Burkina Faso. Children received either a 12.5-mg zinc sulfate tablet or a placebo every day for about 6 months. Outcomes were morbidity, nutritional status, and mortality.

Results: Results revealed significant differences between boys and girls in their responses to zinc supplementation. Boys who received the zinc preparation had fewer days with diarrhea than did control boys (RR = 0.88, P = 0.05), especially less nonfebrile diarrhea (RR = 0.72, P < 0.001) and less dysentery (RR = 0.65, P = 0.05), but more ear infections (RR = 4.00, P < 0.001). By contrast, girls who received the zinc supplement had the same prevalence of diarrhea as did control girls, but more dysentery (RR = 3.70, P < 0.001), fewer ear infections (RR = 0.39, P < 0.001), and fewer eye infections (RR = 0.41, P < 0.001). The effect of supplementation on nutritional status was not detectable in boys, but girls who received supplementation experienced a faster growth velocity in height than did control girls (P = 0.004) and a faster growth velocity for weight and height if they were wasted and not stunted at baseline (P = 0.003).

Conclusions: Zinc supplementation had positive, nil, or negative effects depending on pathological condition, and the effects were different for boys than for girls.

*IRD & Institut Pasteur, Unité d'Epidémiologie des Maladies Emergentes, Paris, France

Department of Tropical Hygiene and Public Health, Heidelberg University, Germany

Centre National de Recherche en Santé, Nouna, Burkina Faso

Received 17 March, 2006

Accepted 20 October, 2006

Address correspondence and reprint requests to Michel Garenne, Institut Pasteur, 25 rue du Dr Roux, 75724 Paris Cedex 15, France (e-mail:

Back to Top | Article Outline


Zinc is one of the most important trace elements for the human body. Zinc plays a major role in various aspects of physiology, immunity, and skeletal growth. Its many roles include participation in basic metabolic functions such as cellular respiration, synthesis of many proteins and enzymes, DNA and RNA replication, carbohydrate metabolism, cell division and growth, spermatogenesis and ova formation, pituitary and adrenal gland functions, integrity of membranes and wound healing, sensory responses, and bone metabolism. In addition, it is a key element of the immune system and has been shown to be a major determinant of physical growth (1–7). Zinc deficiency has been shown to be associated with impaired immunity, growth retardation, and a variety of diseases that zinc supplementation has been able to reverse (8–11).

With respect to morbidity in developing countries, zinc supplementation in the community has been shown to have a preventive effect on diarrhea morbidity, reducing incidence by about 15% to 35% depending on the study (12–17). Zinc supplementation has also been shown to have a therapeutic effect on diarrhea episodes, especially persistent diarrhea, in several hospital-based studies in which malnourished children or children with diarrhea had been admitted (18–24). The effect of zinc supplementation could depend on nutritional status, on age, and, of course, on zinc status at admission. The effect of zinc supplementation on acute respiratory infection (ARI) remains controversial because about the same number of studies show a positive effect (22,25) as studies showing no effect or a negative effect (12). With respect to malaria, 2 studies claimed to have shown an effect. In Papua New Guinea the supplemented group seemed to show lower morbidity for Plasmodium falciparum, but not for P vivax, and a lower attendance to clinics for febrile episodes; in Gambia the supplemented group that received zinc only twice weekly also had lower attendance to clinics for febrile episodes (9). However, our study in Burkina Faso did not find any effect of zinc supplementation on malaria morbidity (26,27). Conversely, some diseases seem to have an impact on zinc status. In particular, in Malawi, malaria among women was associated with lower hair zinc but not with plasma zinc (28).

Several studies have noted differences in the impact of zinc supplementation on boys and girls. For instance, in India, zinc supplementation had an effect on persistent diarrhea and dysentery morbidity among boys ages 6 to 36 months but not among girls (14). A similar finding was reported from Peru among children with persistent diarrhea (20). Similarly, a significant effect on linear growth was found in boys but not in girls (19). These studies, however, were based on small numbers, and systematic testing for different responses between boys and girls was not done.

This article attempts to give a comprehensive view of different effects of zinc supplementation on boys and girls living in a rural area of Burkina Faso in west Africa. The study was originally designed to study the effect of daily zinc supplementation on child morbidity, with a focus on malaria (main outcome), on other leading causes of child morbidity (secondary outcome), and on nutritional status (tertiary outcome) for both sexes combined. The in-depth secondary analysis of the effects revealed the pattern of sex differences.

Back to Top | Article Outline


Study Area

The study area was in the Nouna district, in the western part of Burkina Faso, about 300 km from Ouagadougou, the capital city. It included 16 villages, part of a broader area of 39 villages under demographic surveillance since 1992 (Nouna Health Research Center). The area is exclusively rural and is populated by a variety of ethnic groups, the most important being the Marka, Mossi, Bobo, and Poular groups. The study villages were centered on 2 local clinics (CSPS): Bourasso and Koro, which serve about 5000 and 7000 people, respectively. In addition to using the local clinics, people with serious illnesses go to the district hospital in Nouna or to other nearby hospitals. The area is known to have a high prevalence of malaria and of other tropical diseases, in particular schistosomiasis.

The main dish prepared in this area is based on Tho, a paste made from millet cooked in water, to which various vegetables, meats, and spices can be added. Rice and maize can also be used as cereals for preparing the family dish. The zinc content of the average diet has never been measured in this area.

Back to Top | Article Outline

Study Children

The children selected for the zinc supplementation trial were ages 6 to 30 months at baseline in June 1999. The sample size (N = 720) was calculated for testing the impact of zinc supplementation on malaria and was chosen to detect a 20% reduction in malaria morbidity with a power of 90% (26), which is large enough for testing major differences in other morbidities as well. Randomization was organized in 24 blocks of 30 children each (15 zinc and 15 placebos in each block): 1 block for each of the 24 field workers. In large villages, 2 blocks of 30 children each were selected, and in smaller villages 1 block only. In 15 of the 16 villages, the exact number of eligible children was found without difficulty, and in the smallest village the block comprised only 23 children. Except in the last case, children were selected by a lottery in each village from the resident children of the proper age group. Altogether, 713 children were recruited at baseline for the study. There was no evidence of HIV/AIDS in any of the children recruited, and Nouna is in a province with low HIV seroprevalence (2.2% among women ages 15–49 in 2003).

Back to Top | Article Outline

Zinc Supplementation

The zinc supplements were 25-mg tablets of zinc sulfate (Biolectra Zinc, Hermes Arzneimittel GMBH, Munich, Germany). The placebo had the same appearance as the zinc tablets. Tablets were stored in waterproof plastic tubes labeled with a unique identification number for each child. Supplementation was administered daily except on Sundays and major holidays. Tablets were cut in 2 pieces by the field workers to be given to the children in front of the field worker: a daily dose of 12.5 mg per child. This dose was recommended by the WHO at the time of the project, was close to the RDA dose (10 mg/day), and had been used successfully in other trials in developing countries (26). Families, field workers, field physicians, and investigators were blinded to the product given to the children until the code was broken 3 months after the end of the trial. Supplementation started on June 22 and lasted for about 6 months until December 17.

Back to Top | Article Outline

Morbidity Surveillance

The morbidity surveillance was conducted by the 24 field workers during their daily visits for the whole course of the study. Field workers were selected from the local population with at least 4 years of secondary school, and all spoke local languages. They were supervised twice per week by field supervisors who were experienced field workers from other studies. Field supervisors reported to the field physician, who was based in Nouna. The morbidity surveillance covered 24 weeks (169 days on average).

Absences from the village were monitored systematically at each daily visit by the field workers. Children were absent usually for short periods. Children absent for more than 2 weeks in a row were excluded from the study. The mean number of days absent was 6.3 per child during the study period, representing only 4/100 of the accumulated person-days.

The surveillance included a questionnaire wherein main symptoms were prompted (fever, diarrhea, cough), and each child's temperature was measured at every visit with an electronic thermometer (Digital Classic, Hartmann). If the temperature was 37.5°C or above, then a thick and a thin blood film were taken on the spot and sent to the National Center for Malaria Control (CNLP) in Ouagadougou. Blood films were read at CNLP by 2 well-trained specialists and, in case of disagreement, by a third person. A parasitemia of 2500 parasites per milliliter was considered to be indicative of malaria infection. Diarrhea was defined as at least 3 loose stools per day, and dysentery by the presence of blood or mucus. In addition, the field workers noted any other symptom reported by the mother and any treatment received by the child.

Morbidity episodes were recorded from reports at the daily visits. An episode was defined as starting with first reported symptom and terminating at the last symptom followed by 1 week without symptoms. Prevalence was estimated as the number of days of the condition during daily visits. Incidence was estimated as the number of episodes during the study period. The duration of episodes was calculated as the number of days between the beginning and the end of each episode. This was not precisely equivalent to the number of visits when the condition was observed because the duration of episodes included days with no visit overlapping an episode (in particular, Sundays) and days without reported symptoms in the middle of an overlapping episode. However, differences between the 2 estimates were small because most episodes were short. A malaria episode was defined as an episode in which the child's temperature was 37.5°C or above during which parasitemia was equal to or higher than 2500 parasites per milliliter. An episode of ARI was defined as an episode of cough with at least 1 day of fever. There was no attempt to count the respiratory rate for ARI.

Back to Top | Article Outline

Clinical Examination and Testing

A clinical examination of the study children was conducted 3 times: at baseline (June 8–18), at midterm (September 10–20), and at endpoint (December 6–17). For each visit, the study children were examined by a physician, and basic anthropometric measures were taken: weight, height, and mid–upper arm circumference (MUAC), although height was not taken at the midterm examination. Clinical examination included systematic temperature and spleen size by Hackett grade. Children with severe disease at baseline were excluded from the sample. z-Scores of weight for age, height for age, and weight for height were calculated by use of the US National Center for Health Statistics standards (29). The Waterlow classification (30) was used to tabulate nutritional status: stunting was defined as below −2 z-score in height for age, wasting as below −2 z-score in weight for height, and severe malnutrition as wasting and stunting combined.

In addition, the serum zinc concentration was evaluated at baseline on a subsample of 81 children. Details of the laboratory procedures for malaria parasite screening and for the dosage of zinc at baseline are given elsewhere (26).

Back to Top | Article Outline

Statistical Analysis

Given that the main findings seemed to be the different responses between boys and girls, the differences between zinc-supplemented children and control children were tested systematically for each sex separately and compared. Statistical tests were standard tests for relative risks, and the Fisher exact test was used when the number of cases was small. Similarly, multivariate analysis was conducted to investigate the net effect of zinc supplementation on morbidity (logistic regression) and on nutritional status (linear regression) after age, sex, and other relevant factors were controlled for. In each multivariate model the effect of zinc supplementation was tested separately for boys and for girls.

Back to Top | Article Outline


Sample Size and Randomization

Among the 713 children recruited, 27 children were excluded: 3 because of severe diseases at baseline (2 hydrocephalus, 1 cerebral palsy), 1 who never participated, and the remaining 23 because of absences longer than 14 days. Among the 27 excluded children, 15 belonged to the treatment group and 12 belonged to the control group.

The treatment and control groups were similar for virtually all of the variables available at baseline and for fieldwork variables (Table 1): same number of children, same proportion of boys and girls, same anthropometric indicators at baseline, same number of visits, same number of supplements (zinc or placebo) received, and same number of days absent. In addition, serum zinc concentration was the same in the 2 groups, with an average of 11.7 μmol/L, and 72% of children were considered zinc deficient according to the reference laboratory's threshold of 13.0 μmol/L (26). However, the treatment group was slightly older by 1.2 months than the control group. The difference came only from the male treatment group, which comprised fewer children ages 6 to 18 months and more children ages 24 to 30 months. It seems unlikely that this minor bias could have significantly affected the results, and it will be seen later that controlling for age did not change the results of the multivariate analysis.



Back to Top | Article Outline


Fever and Malaria

A total of 2044 episodes of fever were recorded during the study period, all certified by temperature ≥37.5°C (average 3.0 episodes per child). Among those, 1282 (63%) were considered malaria episodes with parasitemia ≥2500 parasites per milliliter; however, for 321 cases of fever (16%) no data on parasitemia were available. Fever episodes were quite short (3.0 days on average), as were malaria episodes (3.2 days on average). Most malaria episodes occurred between September and November, which is usually the peak of the malaria season in this area. The prevalence of both fever and malaria was low during the first 4 weeks of the study (June 20–July 17); it then increased rapidly peaked in late August and early September (August 29–September 18), decreased and peaked again in late October and early November (October 10–November 6), and decreased again in December to low levels, close to those in June.

Back to Top | Article Outline

Diarrhea and Dysentery

A total of 697 episodes of diarrhea and 46 episodes of dysentery were reported during the study period (∼1.0 diarrhea episode per child). Most diarrhea episodes were short (3.3 days), and only 11 (1.6%) lasted 14 days or more (persistent diarrhea). The epidemic pattern of diarrhea followed closely that of malaria, which also showed 2 visible peaks in mid-September and throughout October, the first peak being much more pronounced than the second. However, only 18% of diarrhea episodes occurred at the same time as a malaria episode, and only 15% of the days with diarrhea overlapped the days with malaria. Because diarrhea with fever had a different epidemic pattern and seemed in large part to be due to malaria, we separated the nonfebrile diarrhea episodes in the final analysis.

Back to Top | Article Outline

Cough and Acute Respiratory Infection

A total of 470 episodes of cough were reported during the study period (0.7 episode per child on average). Among those children, 266 were febrile (57%), and in 24 cases the mother reported difficulty in breathing. Cough and ARI morbidity remained low during the first part of the study, then peaked suddenly throughout October and decreased again in November.

Back to Top | Article Outline

Other Symptoms

Episodes of other symptoms reported by the mother were as follows: vomiting (244 cases), most of them associated with fever; common cold (172 cases); skin rash (96 cases); eye infections (96 cases); ear infections (41 cases); skin lesions (33 cases); stomachache (20 cases); icterus (7 cases); and convulsions (8 cases). Most of these conditions were highly seasonal: vomiting followed the pattern of fever and malaria; rash was concentrated in 2 months, from mid-August to mid-October; otitis media occurred primarily in November; conjunctivitis followed a pattern similar to that of vomiting; common cold was rare in the first 2 months of the study and remained high in the second part, from late August throughout mid December. Medical diagnosis was available only at time of the 3 clinical examinations. No striking difference was seen between medical diagnosis at clinical examination and symptoms reported by the field workers at the same time. In particular, several cases of reported ear and eye infections were diagnosed by the physician as otitis media and conjunctivitis during the clinical examinations.

Back to Top | Article Outline

Overall Responses to Zinc Supplementation

There were only a few significant differences in the prevalence of the various conditions investigated between the zinc and placebo groups when both sexes were combined: lower prevalence of diarrhea (P = 0.015), higher prevalence of dysentery (P = 0.010), and lower prevalence of conjunctivitis (P = 0.042).

Back to Top | Article Outline

Sex-specific Responses to Zinc Supplementation

More differences were seen when the sexes were considered separately (Table 2). Zinc supplementation reduced the prevalence of diarrhea for boys (−21%) because of the lower incidence and shorter duration of episodes, but not for girls (despite lower incidence, but longer duration). The difference between male and female responses was even more striking when restricted to nonfebrile diarrhea: zinc reduced its prevalence by −33% for boys, whereas no difference was seen for girls. For dysentery, prevalence was much lower for boys (−33%) in the zinc group but much higher for girls (3.61 times higher). For these 3 conditions, the response was significantly different between boys and girls (P = 0.042 for diarrhea, P < 0.001 for nonfebrile diarrhea; P = 0.001 for dysentery). Zinc supplementation did not affect the prevalence of ARI among boys but increased its prevalence by 32% among girls (P = 0.021). The difference between male and female responses was borderline significant (P = 0.071).



Three other symptoms were also significantly different between male and female responses: vomiting was more frequent among girls receiving zinc (+46%), whereas no significant difference was seen among the boys receiving zinc; the difference between the sexes was significant (P = 0.017). Most of the vomiting episodes were associated with fever. Conversely, boys receiving zinc experienced 3.1 times more ear infections (P < 0.001) and about the same number of eye infections, whereas girls receiving zinc experienced much less of both: 57% less otitis (P < 0.001) and 58% less conjunctivitis (P < 0.001). For those 2 symptoms the differences between male and female responses were highly significant (P < 0.001 in both cases).

These differences between male and female responses to zinc supplementation were striking in comparison with the absence of differences for other diseases, such as malaria (RR = 1.048; P = 0.672) and common cold (RR = 1.017; P = 0.905).

Back to Top | Article Outline


Treatments reported by the family were primarily antimalarial drugs (1317 cases), antipyretic drugs (458 cases), antibiotics (133 cases), oral rehydration therapy (73 cases), other modern treatments (154 cases), and traditional treatments (366 cases). Among antimalarial drugs, most were standard chloroquine treatment and quinine (44 cases) and a few treatments with sulfadoxin-pyrimethamin (Fansidar; 11 cases). The most common antipyretic drugs were paracetamol and aspirin. Various antibiotics was administered, mostly tetracycline (62 cases), ampicillin (37 cases), and cotrimoxazol (34 cases). Among other treatments, the most common was oral rehydration therapy. No details were recorded for traditional treatments. These drugs could be given at home or in the villages as part of the primary health care system, in the clinics, or in the hospital. Most treatments (82%) occurred at home or in the village, most of them being antimalarial treatments. In the first CSPS (Bourasso) all antibiotics were given from the clinic, whereas in the second CSPS (Koro) most were distributed in the villages.

There was no difference in treatments by sex according to main pathological conditions (fever, diarrhea, cough), neither in the zinc group nor in the placebo group, with the sole exception of somewhat less antipyretic (P = 0.045) and more other modern treatments (P = 0.033) among girls in case of fever.

There were also no differences in type of treatment received between the zinc group and the placebo group except for antibiotics provided by the local clinics. In both clinics (Bourasso and Koro) more zinc recipients were treated with antibiotics than children from the control group (P = 0.002). These antibiotics were the classic antibiotics administered in the area for ARI and dysentery (primarily ampicillin, amoxycillin, and cotrimoxazol, but also tetracycline). Precise information on the diagnosis made by the nurses was not available. Episodes associated with these antibiotic treatments in clinics (n = 58) were a variety of febrile episodes: malaria (n = 20), diarrhea (n = 15), cough (n = 14), and vomiting (n = 5). Differences between the treatment and control groups did not seem to be associated with diarrhea (n = 8 and 7 in each group) but seemed to be associated with other symptoms (n = 23 vs 9 for fever; n = 15 vs 5 for malaria; n = 9 vs 5 for cough; n = 4 vs 1 for vomiting). More specific diagnosis was not possible from the available information.

When illnesses were broken down by main pathological condition, other differences emerged. For diarrhea episodes, there was no difference in treatments between the zinc and placebo groups. However, for fever episodes and for cough, more modern treatments were received by children in the zinc group than by those in the control group (+22% more modern treatments for fever; P < 0.001; +53% more modern treatments for cough; P = 0.001). These differences were similar for boys and girls. In conclusion, the sex differences in prevalence, incidence, and duration of episodes shown earlier could not be explained by different treatments between boys and girls.

Back to Top | Article Outline

Nutritional Status

The study children had a relatively low nutritional status at baseline, balanced between wasting and stunting: about half (50.9%) were above the −2 z-score threshold for both weight for height and height for age; about one fourth (24.6%) were stunted but not wasted; and about one eighth (13.1%) were wasted not stunted; and a similar proportion (11.3%) were severely malnourished.

Differences in anthropometric measurements (weight, height, and arm circumference) between the zinc and placebo groups were remarkably small at the endpoint (the December clinical investigation), and most of the differences were not statistically significant (Table 3). However, girls who received zinc supplementation had a significantly faster growth velocity for height than did boys (P = 0.004), whereas the growth velocities were identical among the placebo recipients. Furthermore, in the multivariate analysis, the proportion of girls with a MUAC below 12.5 cm at endpoint was significantly lower in the zinc group (P = 0.017) than in the placebo group after sex, age, arm circumference at baseline, and season of birth were controlled for, whereas the difference for boys was not significant.



When the results were analyzed according to nutritional status at baseline, only 1 significant difference emerged. Among children wasted and not stunted at baseline (≤−2 z-score in weight for height and >−2 z-score in height for age), girls receiving zinc supplementation experienced a faster growth velocity for height (P = 0.005) and for weight (P = 0.020), whereas no difference could be seen for arm circumference. In any case, the possible effect of zinc supplementation on anthropometric indicators remained small.

Back to Top | Article Outline


Among eligible children, 17 deaths were reported during the study period: 5 in the treatment group and 12 in the control group, the difference being not significant (P = 0.088). Among boys there was no difference at all (4 deaths in each group), whereas among girls there was only 1 death in the zinc group, compared with 8 deaths in the placebo group, the difference being significant (Fisher test; P = 0.037). However, this difference between boys and girls is to be viewed with caution because it is based on rather small numbers. In particular, the sex ratio of deaths in the zinc group was not statistically significant from the sex ratio of deaths in the placebo group (P = 0.111). Furthermore, if it is assumed that the baseline mortality was 4 deaths per group, as exemplified by the case of boys, then none of the difference would be significant among the girls. There was no obvious correlation with causes of death.

Back to Top | Article Outline

Multivariate Analysis of Morbidity

Multivariate analysis of the daily prevalence of specific diseases was conducted with a logit model separately for each sex. The controls were as follows: zinc supplementation, age, season of birth, and 2 anthropometric measurements at baseline—weight for height z-score and height for age z-score. The results did not change the sign or the magnitude of different responses between boys and girls (Table 4), which indicates that randomization had no bias and that the univariate analysis provided a valid estimation of differences in the population. Some results that were borderline (diarrhea all forms, ARI) became even less significant. The interaction between sex and zinc supplementation was particularly strong and significant for nonfebrile diarrhea, dysentery, vomiting, ear infection, and eye infection. By contrast, and as a control, the effect of zinc supplementation on malaria was nil for either of both sexes.



Back to Top | Article Outline


The Nouna study confirmed previous findings, in particular that zinc supplementation reduces diarrhea morbidity among boys. The order of magnitude of the reduction of diarrhea morbidity (21% for all diarrhea cases, 33% for nonfebrile diarrhea, 33% for dysentery) was similar to that found in other studies from Asia and Latin America. The effects on nutritional status were small and usually were not significant, given the sample size, which is also consistent with several other findings.

The quality of data varied among field workers, and some reported more episodes than others. Although differences between field workers may have affected the overall prevalence of disease, they should not have affected the different responses between boys and girls because randomization should take care of the differences between field workers, given that the proportions of boys and girls in each group were balanced, and controlling for a variety of factors in the multivariate analysis did not change the main effects.

There was no effect of zinc supplementation on malaria morbidity, measured by any type of definition used that was based on temperature and parasitemia (26). This negative finding contrasted with earlier optimistic reports from Papua New Guinea and from Gambia (also in west Africa) (11). Nor did we find an effect on episodes treated in clinics, as reported in the other 2 studies. This absence of an effect on fever and malaria in Nouna at least ensured that randomization was fair and allocated the children in 2 comparable groups for morbidity.

The main finding of the Nouna study was the strong difference between male and female responses to zinc supplementation, often in opposite directions, in particular for diarrheal diseases. This pattern of differences had already been suggested by other authors (14,19,20) but had never been systematically analyzed, as far as we are aware. In the Nouna study we showed that the difference between male and female responses was statistically significant. An increase in ARI morbidity among recipients of zinc supplementation had been reported already from Guatemala (12); unfortunately, the results of that study were not presented by sex. The impact of zinc supplementation on ear and eye infections has never been reported, to our knowledge, and requires confirmation from other studies. It should be noted that this information comes from the spontaneous report of the mother and is therefore less reliable than the diarrhea and ARI results, which were based on systematic questioning at each daily visit. The fact that the impact of zinc supplementation on growth velocity among wasted children was significant for girls and not for boys (although both went in the same direction) is puzzling because opposite findings have been reported from other studies (larger effect on boys). However, here again more research is needed because most reports are based on samples too small to have enough power to demonstrate the finding. Furthermore, the differences in the impact of zinc on growth velocity could depend upon the pattern of diseases prevalent in the population, and therefore could differ from one situation to another.

Behavioral factors are unlikely to explain sex differences in morbidity or mortality. First, large-scale demographic surveys showed no evidence of such different behaviors on child health in sub-Saharan Africa, particularly in Burkina Faso (31). Second, if there were a systematic bias against 1 sex it would be consistent across pathological conditions and would be unlikely to vary both ways depending on diseases.

With respect to the immune system, zinc seems to have complex effects. It stimulates the immune system in a variety of ways, in particular for the development of cell-mediating nonspecific immunity such as in neutrophils, natural killer cells, and macrophages. Recent work has shown that zinc supplementation seems to affect the balance between Th1 and Th2 responses (4,11). In particular, mild zinc deficiency depresses Th1 responses (lower levels of typical Th1 cytokines such as interleukin-1 [IL-1], IL-2, IL-4, IFN-γ), but not Th2 responses (no effect on cytokines such as IL-4, IL-6, IL-10). Conversely, zinc supplementation in vitro promotes the production of IL-1β, IL-6, IFN-α, and TNF-α. Furthermore, zinc supplementation in patients with sickle cell anemia stimulates Th1 responses: it increases the CD4/CD8 ratio, NK cells, IL-2, and thymulin. These different effects on Th1 and Th2 immune responses may have 2 consequences: first, zinc supplementation may have both positive and negative effects depending on the pathological condition; second, it may have different effects by sex, as shown by previous studies on a variety of infectious diseases (32). Indeed, males and females have different Th1 and Th2 responses to a variety of infections, and this is true in humans and in animal models (32). Males usually have a better Th1 response, but too strong a Th1 response may be deleterious for them; females, by contrast, mount a better Th2 response, but too strong a Th2 response may be deleterious for them. This may have been the case here. For instance, zinc supplementation may change the balance of Th1 and Th2 responses and may be beneficial for boys and men for certain diseases (eg, nonfebrile diarrhea, dysentery) and have no effect, or be even deleterious, for girls and women (dysentery, ARI). The opposite effect may be found for other diseases such as otitis media and conjunctivitis. Little is known about Th1 and Th2 responses for the many microorganisms involved in diarrhea, dysentery, and ARI, and most of the available work on Th1 and Th2 responses deals with germ-specific diseases, such as measles, whooping cough, and tuberculosis, or diseases caused by tropical parasites. Some recent work has recently been published on otitis media and conjunctivitis, which seems to go in the same direction (33–36). However, the evidence is still limited, the results are sometimes controversial, and further research is needed before we can explain the Nouna findings.

The fact that zinc supplementation may sometimes have negative effects has already been hypothesized by several authors (6,37). Furthermore, it may also be argued that the dose of zinc received by the youngest recipients was too high and could have a negative effect, in particular in inducing some vomiting. However, there was no evidence that vomiting, more prevalent among the female recipients, was associated with age or with nutritional status. On the contrary, vomiting seemed to be associated with other infections, inasmuch as only vomiting with fever was significantly higher among girls. Therefore, it seems unlikely that the vomiting effect was due to too high a zinc dose among the youngest or the lightest-weight children.

The fact that zinc supplementation may have different effects for boys and girls, and may even have positive and negative effects, raises various issues about its applicability in the general population. Overall, the effect of zinc supplementation seemed rather positive in Nouna, at least for eye and ear infections in girls and for nonfebrile diarrhea and dysentery in boys. However, more research seems to be needed for fine-tuning the optimal dose, for identifying the optimal way of providing zinc (different salts are available; some soluble, others not), and for limiting the potential negative effects.

Of course, even though systematic daily supplementation with zinc remains controversial, zinc has been shown to be useful for the treatment of diarrhea, even in hospitals, as we are reminded by a recent editorial in this journal (23).

Back to Top | Article Outline


1. Prasad AS. Clinical, biochemical, and pharmacological role of zinc. Annu Rev Pharmacol Toxicol 1979; 19:393–426.
2. Prasad AS, (ed.). Trace Elements in Human Health and Disease. New York: Academic Press, 1976:1–20.
3. Prasad AS, (ed.). Essential and Toxic Elements in Human Health and Disease. New York: Alan R. Liss, 1988:69–82.
4. Prasad AS. Zinc and immunity. Mol Cell Biochem 1998; 188:63–69.
5. Shankar AH, Prasad AS. Zinc and immune function: the biological basis of altered resistance to infection. Am J Clin Nutr 1998; 68:447S–463S.
6. Chan S, Gerson B, Subramaniam S. The role of copper, molybdenum, selenium, and zinc in nutrition and health. Clin Lab Med 1998; 18:673–685.
7. Beisel WR. Single nutrients and immunity. Am J Clin Nutr 1982; 35(Suppl 2):417–468.
8. Allen LH. Zinc and micronutrient supplements for children. Am J Clin Nutr 1998; 68:495S–498S.
9. Black RE. Therapeutic and preventive effects of zinc on serious childhood infectious diseases in developing countries. Am J Clin Nutr 1998; 68:476S–479S.
10. Fuchs GJ. Possibilities for zinc in the treatment of acute diarrhea. Am J Clin Nutr 1998; 68:480S–483S.
11. Zinc for Child Health. Report of a meeting held at the Johns Hopkins University Bloomberg School of Public Health, November 17–19, 1996. Child Health Research Project Special Report, Vol. 1, no. 1, Baltimore, MD: Bloomberg School of Public Health.
12. Ruel MT, Rivera JA, Santiazo MC, et al. Impact of zinc supplementation on morbidity from diarrhea and respiratory infections among Guatemalan children. Pediatrics 1997; 99:808–813.
13. Sazawal S, Black RE, Bhan MK, et al. Zinc supplementation reduces the incidence of persistent diarrhea and dysentery among low socioeconomic children in India. J Nutr 1996; 126:443–450.
14. Sazawal S, Black R, Bhan MK, et al. Zinc supplementation in young children with acute diarrhea in India. N Engl J Med 1995; 333:839–844.
15. Sazawal S, Black R, Bhan MK, et al. Efficacy of zinc supplementation in reducing the incidence and prevalence of acute diarrhea: a community based, double blind, controlled trial. Am J Clin Nutr 1997; 66:413–418.
16. Bahn MK, Bhandari N. The role of zinc and vitamin A in persistent diarrhea among infants and young children. J Pediatr Gastroenterol Nutr 1998; 26:446–453.
17. Sachdev HPS, Mittal SK, Yadav HS. A controlled trial on utility of oral zinc supplementation in acute dehydrating diarrhea in infants. J Pediatr Gastroenterol Nutr 1988; 7:877–881.
18. Sachdev HP, Mittal SK, Yadav HS. Oral zinc supplementation in persistent diarrhea in infants. Ann Trop Pediatr 1990; 10:63–69.
19. Castillo-Duran C, Rodriguez A, Venegas G, et al. Zinc supplementation and growth of infants born small for gestational age. J Pediatr 1995; 127:206–211.
20. Penny ME, Peerson JM, Marin M, et al. Randomized, community based trial of the effect of zinc supplementation, with and without other micronutrients, on the duration of persistent childhood diarrhea in Lima, Peru. J Pediatr 1999; 135:208–217.
21. Faruque AS, Mahalanabis D, Haque SS, et al. Double-blind, randomized, controlled trial of zinc or vitamin A supplementation in young children with acute diarrhea. Acta Paediatr 1999; 88:154–160.
22. Roy SK, Tomkins AM, Haider R, et al. Impact of zinc supplementation on subsequent growth and morbidity in Bangladeshi children with acute diarrhea. Eur J Clin Nutr 1999; 53:529–534.
23. Canani RB, Ruotolo S. The dawning of the “zinc era” in the treatment of pediatric acute gastroenteritis worldwide? J Pediatr Gastroenterol Nutr 2006; 42:253–255.
24. Bijleveld C, Buller H, Taminiau J, et al. Zinc in the treatment of diarrhea. J Pediatr Gastroenterol Nutr 1997; 25:363–365.
25. Lira PI, Ashworth A, Morris SS. Effect of zinc supplementation on the morbidity, immune function, and growth of low-birth-weight, full-term infants in northeast Brazil. Am J Clin Nutr 1998; 68(Suppl 2):418S–424S.
26. Müller O, Becher H, Baltussen van Zweeden A, et al. Effect of zinc supplementation on malaria and other causes of morbidity in west African children: randomised double blind placebo controlled trial. BMJ 2001; 322:1567.
27. Müller O, Becher H, Baltussen A, et al. The associations between protein-energy malnutrition, malaria morbidity and all-cause mortality in West African children. Trop Med Int Health 2003; 8:507–511.
28. Gibson RS, Huddle JM. Suboptimal zinc status in pregnant Malawian women: its association with low intakes of poorly available zinc, frequent reproductive cycling, and malaria. Am J Clin Nutr 1998; 67:702–709.
29. Basic Data on Anthropometric Measurements and Angular Measurements of the Hip and Knee Joints for Selected Age Groups, 1–74 Years of Age, United States 1971–75. Bethesda, MD: National Center for Health Statistics; 1981.
30. Waterlow JC, Buzina R, Keller W, et al. The presentation and use of height and weight data for comparing the nutritional status of groups of children under the age of 10 years. Bull WHO 1977; 55:489–498.
31. Garenne M. Sex differences in health indicators among children in African DHS surveys. J Biosoc Sci 2003; 35:601–614.
32. Garenne M, Lafon M. Sexist diseases. Perspect Biol Med 1998; 41:176–189.
33. Beck FW, Prasad AS, Kaplan J, et al. Changes in the cytokine production and T-cell subpopulations in experimentally induced zinc-deficient humans. Am J Physiol 1997; 272:E1002–E1007.
34. Solomons NW. Mild human zinc deficiency produces an imbalance between cell-mediated and humoral immunity. Nutr Rev 1998; 56:27–28.
35. Bernstein JM, Ballow M, Xiang S, et al. Th1/Th2 cytokine profiles in the nasopharyngeal lymphoid tissues of children with recurrent otitis. Ann Otol Rhinol Laryngol 1998; 107:22–27.
36. Calder VL, Jolly G, Hingorami M, et al. Cytokine production and mRNA expression by conjunctival T-cells lines in chronic allergic eye disease. Clin Exp Allergy 1999; 29:1214–1222.
37. Darmon N, Briend A, Desjeux JF. Zinc in the treatment of diarrhea: comments. J Pediatr Gastroenterol Nutr 1997; 25:363–365.

Zinc supplementation; Morbidity; Diarrhea; Dysentery; Malaria; Nutritional status; Immunity; Sex differences; Burkina Faso

© 2007 Lippincott Williams & Wilkins, Inc.