In rural Nepal, a substantial difference in mortality between females and males aged five years or younger was virtually eliminated with community-wide vitamin A supplementation, according to results of a randomized, controlled field trial involving more than 30,000 children.1 The high-dose supplements also narrowed survival inequalities between caste groups but not those between socioeconomic levels.

For this study, the researchers performed a secondary analysis of data from a cluster-randomized, double-blind, placebo-controlled trial conducted in 1989-1991. During the trial, all children aged 6-60 months in a selected rural district in southeastern Nepal received four home visits, separated by four-month intervals. At each visit, children in communities randomly assigned to receive high-dose vitamin A supplements were administered a capsule with 200,000 IU; the others received capsules containing 1,000 IU. For the current, intention-to-treat analysis, investigators assessed the number of deaths among participants from the first visit to the fourth, and then compared the death rates by gender; caste (the two highest castes - Brahmin and Chettri - combined vs. all other children); and socioeconomic level (the four highest quintiles in the sample vs. the poorest quintile), measured primarily by household asset ownership.

The trial involved 30,059 children, 49% of whom were female. One-tenth of children had a literate mother. The average household had 1.9 rooms and was a nine-minute walk from the nearest water supply. Fifty-one percent of the children received the high-dose vitamin A supplements.

Overall, 2% of the children died during the study. Among placebo recipients, mortality was substantially higher for females than for males (27 vs. 19 deaths per 1,000 person-years); however, the rates of death among females and males who received high-dose vitamin A supplements were nearly equal (18 vs. 17 per 1,000 person-years). According to logistic regression analysis, death was significantly more common among girls given placebos than among boys given high-dose supplements (odds ratio, 1.6).

Differences in mortality between caste groups were also attenuated with vitamin A supplementation: Without supplementation, the mortality rates were 25 per 1,000 person-years in lower-caste and noncaste families and 11 per 1,000 person-years in high-caste families; with high-dose supplementation, those rates were 19 and 11 deaths per 1,000 person-years, respectively. Compared with high-caste children who received high-dose supplements, children in lower-caste and noncaste families who received a placebo had significantly higher odds of death (odds ratio, 2.2), although this was also the case for lower-caste and noncaste children who received the high-dose supplements (1.6).

In contrast to the results regarding gender and caste, the difference in mortality between the poorest socioeconomic quintile and the four higher quintiles was unaffected by the nutritional intervention: Compared with wealthier children, the poorest children had 13 additional deaths per 1,000 person-years without the high-dose supplementation and had 15 additional deaths per 1,000 person-years with the supplementation. An analysis of treated versus untreated children by quintile showed that although mortality in each quintile was lower among treated children than among nontreated children, mortality in the lowest quintile and in the two highest quintiles differed only nominally—by four or fewer deaths per 1,000 person-years. The benefit of vitamin A supplementation was greatest in the second-poorest quintile and in the middle quintile, where there were 10 fewer deaths per 1,000 person-years among treated children than among untreated children.

The authors believe their finding of a supplementation-related reduction in the mortality difference between castes but not between socioeconomic categories suggests that "dietary vitamin A intake (or the risk of vitamin A responsive disease) in the population was less similar across the lines of caste (more sufficient in Brahmin, Chettri) and more similar across the gradient of asset ownership." In discussing the differing degrees of response to vitamin A supplementation across socioeconomic quintiles, the authors speculate that "the overall mortality hazards of children in the lowest [socioeconomic] quintile include several causes of death that may not be as responsive to vi-tamin A as diseases among children in the middle of the [socioeconomic] distribution." They also note that there may be "factors among the very poorest children that preclude them from deriving benefit from vitamin A."

Finally, in commenting on the intervention's success in virtually eliminating the mortality differential between young girls and young boys, the authors note that "public health interventions that rely on door-to-door distribution of health services essentially overcome intrahousehold gender biases."

-C. Coren


1. Bishai D et al., The impact of vitamin A supplementation on mortality inequalities among children in Nepal, Health Policy and Planning, 2005, 20(1):60-66.