Use of oral rehydration solution (ORS) in the treatment of diarrhea reduces the risk of mortality through prevention and treatment of dehydration but does not decrease diarrheal duration or stool output (1). Zinc deficiency is common in children from developing countries because of the lack of intake of animal foods, high dietary phytate that limits zinc absorption, and inadequate food intake (2). There are also increased fecal losses of zinc during diarrhea (3,4). In observational studies, low plasma zinc levels have been associated with increased severity of diarrhea (5,6). Administration of 20 to 40 mg of zinc to children with mild gastroenteritis results in lower diarrheal duration and stool frequency in placebo-controlled trials in developing countries (7–10). Stool output should be the main outcome assessed in clinical studies for new therapeutic interventions in acute diarrhea because it can be measured more objectively than stool frequency and is a useful proxy indicator for risk of dehydration. Most of the previous studies of zinc levels during acute diarrhea measured episode duration and stool frequency as the outcomes (7–10). Two earlier studies (11,12) that reported an effect on stool output in children supplemented with zinc included small numbers and the subjects studied were moderate to severely malnourished.
In this double-blind, randomized, controlled study, we determined the efficacy of zinc in reducing stool output and diarrheal duration when administered as an adjunct to oral rehydration therapy in hospitalized children ages 3 to 36 months with acute diarrhea and dehydration.
PATIENTS AND METHODS
Male children attending the diarrhea treatment units of the All India Institute of Medical Sciences and Kasturba Hospital, New Delhi, between April 1997 and December 2000 were enrolled in the study if they were between 3 and 36 months of age, passed three or more liquid stools daily for ≤ 72 hours and had mild or severe dehydration. Assessment of dehydration was according to World Health Organization (WHO) guidelines (13). Only male patients were enrolled to ensure proper stool collection uncontaminated with urine. Participants with severe dehydration were given 30 mL/kg/h lactated Ringer's solution intravenously for a maximum of 2 hours, after which they were considered mildly dehydrated and could be enrolled in the study. Severe malnutrition (weight for height <65% of National Centre for Health Statistics median [NCHS]), visible blood in stool, and severe systemic illness were exclusion criteria. Informed written consent was obtained from the parent(s). The study was approved by the ethics committees of the study institutions and WHO.
Randomization, Assignment, and Blinding of Intervention
Participants were stratified according to age (≤ or >12 months) and weight for height (65%–80% or >80% of the NCHS median). Within each stratum, they were randomly assigned to receive zinc or placebo. A different randomization list for each strata and each hospital, using permuted blocks of fixed lengths of 10, was prepared at the WHO in Geneva. The sequence of random numbers was generated using a table of random permutations of 16 numbers. Zinc and placebo syrups packaged in similar glass bottles and with identical color, taste and appearance were provided by a pharmaceutical company (American Remedies Ltd., Chennai, Tamil Nadu, India). Each bottle was labeled with a unique serial number that corresponded to the randomization lists before being arranged sequentially and sent to the investigators. The randomization code was kept with the WHO until the trial was completed. Blinding was maintained during data analysis by coding each group with a letter.
Both intervention and control syrups contained vitamin B complex (5 mL contained: B1, 2.5 mg; B2, 2.5 mg; B6, 1.0 mg; B12, 3μg; C, 50 mg, D-panthenol, 12.5 mg; niacinamide, 25 mg), and the one with zinc sulfate had 1 milligram per milliliter of elemental zinc. Participants ≤12 and >12 months were given 15 mL and 30 mL of zinc or placebo solution, respectively, in three divided doses, and this was continued until cessation of diarrhea and after discharge from the hospital for a total period of 14 days. All doses were offered between meals and repeated if vomiting occurred within 30 minutes.
Outcome Measures and Definitions
The main outcomes were stool output and time taken for cessation of diarrhea. Diarrhea was considered to have ceased at the time of the last abnormal stool before a 12-hour period when no stool had been passed or before the passage of two consecutive formed stools. An abnormal stool was a watery stool with no fecal matter or a loose stool with a rim of more than 1 cm of water around it visualized on the diaper.
Monitoring of Participants
Each stool was measured in preweighed disposable diapers; consistency was noted and recorded immediately; urine was separated from stools by use of urine collection bags connected to condoms. Vomitus was weighed on preweighed disposable gauze pads. Monitoring continued for at least 48 hours after enrollment or until the diarrhea stopped, whichever was later. Nude body weights were taken at enrollment, after rehydration at 6 hours, and then every 24 hours until the patient's participation in the study ended.
Standardization exercises on collection of stool output, vomitus, and measurements of weights and lengths were conducted at regular intervals.
Rehydration therapy for mild dehydration was according to WHO guidelines using oral rehydration solution (ORS) 100 mL/kg body weight for a period of 6 hours (13). Immediately after rehydration was completed, feeding was resumed with traditional milk cereal diet (calorie density: 92.4 cal/100 g; zinc: 0.27 mg/100 g, 0.32 mg/110 kcal) and was offered at the rate of 110 kcal/kg/d. Water (zinc content: 0.06 mg/L) was allowed ad libitum. ORS was given as replacement for ongoing stool losses on a volume-to-weight basis.
Unscheduled intravenous fluids were given during the initial 6 hours of rehydration or during the maintenance period if the participant experienced severe dehydration despite the intake of estimated oral fluids. Participants requiring intensive care treatment for severe systemic infection continued in the study but were given additional treatment by physicians.
Serum Zinc and Copper Levels
A venous blood sample (∼5 mL) was obtained at enrollment. The sample was collected in zinc- and copper-free tubes, spun, and the serum stored at −20°C. A second blood sample was obtained at a follow-up visit after the completion of 14 days of treatment. The zinc and copper concentration in the serum specimens was analyzed using a flame furnace atomic absorption spectrophotometer (GBC Avanta, Dandenong, Victoria, Australia) using standard techniques and with SERONORM (Sero AS, Billingstad, Norway) as the reference (14). Mean corpuscular volume was estimated routinely using standard methods.
Sample Size Calculations
Trial size was calculated for 90% power and 95% confidence to detect at least a 33% lower total stool output and 20% lower duration of diarrhea in the zinc-treated group. The expected values in the control group were based on a previous study in the same hospital (15). A sample size of 130 per group was estimated to be sufficient for total stool output and 106 per group for duration of diarrhea. Thus, we enrolled approximately 145 children per group to account for a 10% loss to follow-up.
Data were entered into Fox Pro for Windows (version 2.6, Microsoft Corporation) with built-in logic, range, and consistency checks. Statistical analysis was undertaken using the Statistical Package for Social Sciences (SPSS, version 9.0, Chicago, Illinois) software. The preadmission duration of fever, vomiting, and diarrhea all suggested somewhat greater severity in the placebo group. All effect sizes were estimated adjusted for these variables and age. Continuous variables with skewed distribution (e.g., stool and vomitus output, ORS intake, and total volume of intravenous fluids received) were normalized by log transformation, and the values were exponentiated and expressed as geometric means. The effect on stool weight, vomitus output, ORS intakes, and total volume of intravenous fluids received was estimated by multiple regression analysis and presented as a ratio of geometric means. The effect on weight gain is presented as medians, with the 95% confidence interval (CI) and differences between the two groups calculated using nonparametric tests. The proportions of participants with diarrhea continuing beyond 4, 5, 6, or 7 days are presented as adjusted odds ratios from logistic regression models. Cox proportional hazards model was used to estimate the relative hazards (RH) of continuation of diarrhea.
Admission Characteristics of Participants
The flow of participants through each stage is described in Figure 1. Although not statistically significant, preadmission duration of fever, vomiting, and diarrhea indicated somewhat greater severity in the placebo group. All other baseline host characteristics were comparable between the two groups(Table 1). Rotavirus was detected in the stools of 41.6% and 46%, respectively, of patients in the zinc and the placebo groups. The mean (SD) serum zinc level at baseline was comparable between the intervention and placebo groups (zinc: 76.9 μg/dL [24.6 μg/dL]v placebo 75.4 μg/dL [22.6 μg/dL];Table 2).
Effect of Zinc Administration on Serum Zinc Level
Serum for zinc estimation could be collected from only 118 participants (zinc, 47%; placebo, 42%) 14 days after completion of the intervention. The mean serum zinc level was significantly higher in the zinc group than the placebo group (difference in means, 11.5 μg/dL; 95% CI, 2.7, 20.3) (Table 2).
Effect of Zinc Administration on Key Study Outcomes
The total stool output (g/kg) was lower in the zinc-treated group (ratio of geometric means [GM], 0.69; 95% CI, 0. 48, 0.99). In addition, zinc treatment resulted in lower stool output per day of diarrhea (ratio of GM, 0.76; 95% CI, 0.59, 0.98) (Table 3).
The Kaplan-Meier survival curves (Fig. 2) for postenrollment duration of diarrhea showed a significantly faster recovery from diarrhea in the zinc group (log rank test:P = 0.043), and this difference became evident 72 hours after enrollment. The risk of continued diarrhea on a given day in the Cox proportional hazards model was lower (RH for continuation of diarrhea, 0.76; 95% CI, 0.59, 0.97) in participants receiving zinc than in those receiving placebo (Table 3). The proportion of diarrheal episodes lasting ≥5 days (odds ratio [OR], 0.49; 95% CI, 0.25, 0.97), or ≤ 7 days (OR, 0.09; 95% CI, 0.01, 0.73) was lower in the intervention group.
Effect of Zinc Administration on Total Intakes of ORS and Water
Consistent with a decreased stool output, the total intake of ORS and water (mL/kg) was 18% less in the zinc group (ratio of GM, 0.82; 95% CI, 0.68, 0.98). The water intake by itself was significantly lower (difference in means, −43.7; 95% CI, −75.5, −11.8) and that of ORS was somewhat less (ratio of GM, 0.89, 95% CI, 0.72, 1.09) in the patients treated with zinc.
Effect of Zinc Administration on Use of Unscheduled Intravenous Fluids
Overall, 23% and 19% participants received intravenous fluids in the zinc and placebo groups, respectively, for any reason; 20.2% (zinc) and 17.4% (placebo) required intravenous fluids for rehydration after the initial correction of dehydration (OR, 1.2; 95% CI, 0.66, 2.2). The total GM (95% CI) volume of IV fluids received for rehydration was 57 mL/kg (24, 134) in the zinc group and 89 mL/kg (32, 252) in the placebo group (ratio of GM, 0.65; 95% CI, 0.17, 2.5). Other reasons for administering intravenous fluids were hypokalemia (zinc, 3 patients; placebo, 1 patient) and suspected sepsis (zinc, 2 patients; placebo, 2 patients).
Effect of Zinc Administration on Total Food Intake and Weight Gain
The total intake of calories (/kg/day) from semisolid food was similar in the two groups (ratio of GM, 0.95; 95% CI, 0.80, 1.13). The median and 95% CI for weight gain at the end of the rehydration phase expressed as a percentage of the nude weight at admission was not different in the zinc (3.2 [2.4, 4.0]) or the placebo group (2.7 [1.9, 3.4]) (P = 0.24). Similarly there were no statistically significant differences in the weight gain (median [95% CI]) at the end of 14 days of therapy, expressed as a percentage of the weight at the end of the rehydration phase, between the two groups (zinc, 3.5 [2.0, 4.6]); placebo 2.9 [2.1, 3.9]; P = 0.94).
Effect of Zinc Administration on Vomiting
There was no significant increase in the number of children who vomited during the first 24 hours (zinc 29%v placebo 32%; OR 1.1, 95% CI, 0.69, 1.9) or at any time during the study (zinc 65%v placebo 59%; OR, 0.74; 95% CI, 0.45, 1.2). The amount of vomitus (g/kg) was the same in the initial 24 hours (ratio of GM, 1.3; 95% CI, 0.98, 1.74) and for the period of study from enrollment to cessation of diarrhea (ratio of GM, 1.2; 95% CI, 0.84, 1.6) in both the groups.
Effect of Zinc Administration on Serum Copper Level and Mean Corpuscular Volume of Red Blood Cells
The mean serum copper level and mean corpuscular volume were not significantly different in the zinc and placebo groups at baseline. After 14 days of therapy with zinc or placebo, there was no significant difference in serum copper level (difference in means, −6.5, 95% CI, −20.4, 7.5) or mean corpuscular volume (difference in means, 2.2; 95% CI, −1.0, 5.3) (Table 2).
Effect of Zinc Administration on Severity of Diarrhea in Subgroups
The impact of zinc supplementation examined in subgroups based on age (≤12 months and >12 months), serum zinc level (<55 μg/dL and ≥55 μg/dL), anthropometric status (weight for height and height for age <−2z and ≥−2z), and presence of rotavirus infection is given in Figure 3. There was no statistically significant interaction for any of these variables with the impact of zinc supplementation on stool output per day of diarrhea and duration of diarrhea. The effect of zinc appears not to be restricted to any particular subgroup. However, there is a trend toward greater effect of zinc on stool output and diarrheal duration on those with serum zinc level ≥55 μg/dL.
This study has demonstrated that zinc treatment has a genuine antidiarrheal effect because it significantly reduces stool output in severe acute gastroenteritis associated with dehydration. The group of patients studied had relatively more severe diarrhea than those studied earlier because they had dehydration and required hospitalization. The study also showed that zinc treatment significantly reduced average diarrheal duration and the proportion of prolonged episodes in hospitalized patients, which confirms earlier findings in children with mild gastroenteritis (8–10,16). The significant reduction in stool output per diarrheal day indicates that the benefit on total stool output is not only due to shortened diarrheal duration in the zinc group. Consistent with the reduced stool output, the total fluid intake also was significantly less in patients treated with zinc.
The reduction in prolonged episodes after zinc treatment is important because persistent episodes are associated with increased case fatality and growth faltering (17). In addition, approximately 25% of patients do not recover within 5 to 6 days when receiving the treatment recommended by WHO for acute diarrhea, which consists of ORS and continued feeding (1,17). Caregivers of children with these prolonged episodes feel frustrated. Physicians who have no additional specific treatment to offer and may resort to prescribing undesirable dietary changes or antimicrobial agents.
Earlier studies of zinc in acute diarrhea have found beneficial effects with daily doses two times the recommended daily allowance (RDA) (8–11). We chose a 3RDA dose to compensate for excessive losses of zinc during an episode of diarrhea and for possible impairment in absorption of ingested zinc among these pediatric patients (3,18). It appears from the findings of the current study that the effects on episode duration with the higher dose are similar to those seen with use of 2RDA of zinc in earlier studies. Therefore, currently there is no indication to use more than 2RDA of zinc for treating acute diarrhea.
Decreased stool output and duration of illness during acute diarrhea by zinc administration is biologically plausible. Zinc is said to improve absorption of water and electrolytes by helping in early regeneration of intestinal mucosa, restoration of enteric enzymes, and enhancement of humoral and cellular immunity (19–24). In malnourished guinea pigs, zinc has also been shown to have antisecretory properties (25). It is unclear from the findings of this and other studies whether the efficacy of zinc is mediated through correction of deficiency alone or if there is also a pharmacologic effect. One would have expected a greater impact of zinc supplementation in children with low serum zinc levels, but no such trend is seen. In fact, the findings were in the opposite direction. It is likely that often, low serum zinc levels in children with gastroenteritis may be a reflection of infection.
Concerns have been raised about a possible adverse effect of oral zinc on copper absorption, which may have important consequences in malnourished children (26,27). In this trial, short-term use of zinc for as long as 14 days did not appear to affect serum copper levels.
The need for unscheduled intravenous fluids for rehydration after the initial correction of dehydration was similar in the two groups. We cannot exclude the possibility of zinc causing a modest, but important, reduction in this outcome, but final determination of such an effect which would require a larger study.
The findings of the current study, together with previous reports, establish the efficacy of zinc in acute diarrhea in developing countries. The effects are substantial and clinically important. Thus, we recommend the use of zinc during acute diarrhea. In addition, the use of zinc during acute diarrhea for preventing prolonged episodes may avoid inappropriate use of antimicrobial and antidiarrheal drugs. Zinc can be administered as syrup or mixed with ORS. The former approach achieves better standardization of dose and duration. The latter approach has not been well evaluated, and its efficacy deserves to be assessed in future studies.
The authors thank Dr. Olivier Fontaine, MD, Medical Officer, from the Division of Child Health and Development, World Health Organization, Geneva, who aided in the design and monitoring of the study; and the Norwegian Council for Higher Education's program for Development Research and Education, which provided core support to our unit. The authors thank the medical and nursing staff of the unit for their help, and the patients and their parents. The authors also thank the pharmaceutical company American Remedies Ltd. for providing the drug and placebo.
1. Victora CG, Bryce J, Fontaine O, et al. Reducing deaths from diarrhoea through oral rehydration therapy. Bull World Health Organization 2000; 78:1246–55.
2. World Health Organization. Complementary Feeding of Young Children in Developing Countries: A Review of Current Scientific Knowledge.
Document ref. WHO/NUT/98.1. Geneva: World Health Organization; 1998.
3. Castillo-Duran C, Vial P, Uauy R. Trace mineral balance during acute diarrhoea in infants. J Pediatr 1988; 113:452–7.
4. Ruz M, Solomons NW. Fecal zinc
of endogenous zinc
during oral rehydration therapy for acute diarrhea
. J Trace Elem Exp Med 1995; 7:89–100.
5. Bhandari N, Bahl R, Hambidge KM, et al. Increased diarrhoeal and respiratory morbidity in association with zinc
deficiency: a preliminary report. Acta Pediatr 1996; 85:146–50.
6. Bahl R, Bhandari N, Hambidge KM, et al. Plasma zinc
as a predictor of diarrhoeal and respiratory morbidity in children in an urban slum setting. Am J Clin Nutr 1998; 68(Suppl 41):4–7.
7. Sachdev HPS, Mittal NK, Mittal SK, et al. A controlled trial on utility of oral zinc
supplementation in acute dehydrating diarrhoea in infants. J Pediatr Gastroenterol Nutr 1988; 7:877–81.
8. Sazawal S, Black RE, Bhan MK, et al. Zinc
supplementation in young children with acute diarrhoea in India. N Engl J Med 1995; 333:839–44.
9. Hidayat A, Achadi A, Sunoto, et al. The effect of zinc
supplementation in children under three years of age with acute diarrhoea in Indonesia. Med J Indonesia 1998; 7:237–41.
10. Faruque ASG, Mahalanabis D, Haque SS, et al. Double blind, randomised controlled trial of zinc
or vitamin A supplementation in young children with acute diarrhoea. Acta Pediatr 1999; 88:154–60.
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; 77:196–200.
12. Dutta P, Mitra U, Datta A, et al. Impact of zinc
supplementation in malnourished children with acute watery diarrhoea. J Trop Pediatr 2000; 46:259–63.
13. World Health Organization, Division of Child Health and Development. Integrated Management of Childhood Illness.
Document ref. WHO/CHD/97.3E. Geneva: World Health Organization; 1997.
14. Hambidge KM, King JC, Kern DL, et al. Pre-breakfast plasma zinc
concentrations: the effect of previous meals. J Trace Elem Electrolytes Health Dis 1990; 4:229–31.
15. Bhan MK, Sazawal S, Bhatnagar S, et al. Glycine, glycyl-glycine and maltodextrin based oral rehydration solution. Acta Paediatr Scand 1990; 79:518–26.
16. The Zinc
Investigators' Collaborative Group: Bhutta ZA, Bird SM, Black RE, et al. Therapeutic effects of oral zinc
in acute and persistent diarrhoea in children in developing countries: pooled analysis of randomised controlled trials. Am J Clin Nutr 2000; 72:1516–22.
17. Bhandari N, Bhan MK, Sazawal S. Mortality associated with acute watery diarrhea, dysentery and persistent diarrhea in rural north India. Acta Pediatr 1992; 381(Suppl):3–6.
18. Wapnir R. Zinc
deficiency, malnutrition and the gastrointestinal tract. J Nutr 2000; 130(5S Suppl):1388–92.
19. Ghisham FK. Transport of electrolytes, water and glucose in zinc
deficiency. J Pediatr Gastroenterol Nutr 1984; 3:608–12.
20. Gebhard RL, Karouani R, Prigge WF, et al. Effect of severe zinc
deficiency on activity of intestinal disaccharidases and 3-hydroxy-3-methylglutaryl coenzyme A reductase in the rat. J Nutr 1983; 113:855–9.
21. Roy SK, Behrens RH, Haider R, et al. Impact of zinc
supplementation on intestinal permeability in Bangladeshi children with acute diarrhoea and persistent diarrhoea syndrome. J Pediatr Gastroenterol Nutr 1992; 15:289–96.
22. Bettger WJ, O'Dell BL. A critical physiological role of zinc
in the structure and function of biomembranes. Life Sci 1981; 28:1425–38.
23. Shankar AH, Prasad AS. Zinc
and immune function: the biologic basis of altered resistance to infection. Am J Clin Nutr 1998; 68(Suppl 2):447–63.
24. Sazawal S, Black RE, Jalla S, et al. Effect of zinc
supplementation on cell mediated immunity and lymphocyte subsets in preschool children. Indian Pediatr 1997; 34:689–97.
25. Darmon N, Pelissier M-A, Candalh C, et al. Zinc
and intestinal anaphylaxis to cow's milk proteins in malnourished guinea pigs. Pediatr Res 1997; 42:208–13.
26. Bhutta ZA, Nizami SQ, Isani Z. Zinc
supplementation in malnourished children with persistent diarrhea in Pakistan. Pediatrics 1999; 103:e42.
27. Sandstrom B. Micronutrient interactions: effects on absorption and bioavailability. Br J Nutr 2001; 85(Suppl 2):181–5.