A randomized controlled clinical trial of zinc, vitamin A or both in undernourished children with persistent diarrhea in Bangladesh

In a double-blind randomized controlled clinical trial, moderately malnouri shed Bangladeshi children (61-75% of the median weight/age) were studied fo r the effect of zinc and/or vitamin A supplementation on the clinical outco me of persistent diarrhea. Children 6 mo to 2 y of age with diarrhea for mo re than 14 d were randomly allocated into 4 groups of 24 receiving a multiv itamin syrup and (i) zinc (20 mg elemental), (ii) vitamin A, (iii) both zin c and vitamin A, or (iv) neither, in 2 doses daily for 7 d. Clinical data o n recovery and on stool output, consistency and frequency were recorded for 7 d, and weight change from day 1 to day 7 was assessed. The baseline char acteristics of the four study groups were comparable. The mean daily stool outputs from days 2 to 7 of therapy were significantly less in the zinc and zinc plus vitamin A groups, but not in the vitamin A group, in comparison with the control group. In children receiving zinc, the cumulative stool we ight in the 7 d was 39% less than in the control group (p < 0.001) and 32% less than in the vitamin A group (P = 0.006). The cumulative stool weight i n the zinc plus vitamin A group was 24% less than in the control group (p < 0.001), but the 14% lower output than in the vitamin A group was not stati stically different. The change in body weight over the 7 d study period was significantly different between the group receiving zinc and the control g roup (+ 111 g vs -90 g, p = 0.045). The rate of clinical recovery of childr en within 7 d was significantly greater in the zinc group (88%) compared wi th the control group (46%, p = 0.002) or vitamin A group (50%, p = 0.005), but not statistically different from the zinc plus vitamin A group (67%, p = 0.086). Conclusion: The results indicate that zinc, but not vitamin A, supplementat ion in persistent diarrhea reduces stool output, prevents weight loss and p romotes earlier recovery.