Vitamin A supplementation and neonatal mortality in the developing world: a meta-regression of cluster-randomized trials.

OBJECTIVE: To assess the relationship between the prevalence of vitamin A deficiency among pregnant women and the effect of neonatal vitamin A supplementation on infant mortality.
METHODS: Studies of neonatal supplementation with vitamin A have yielded contradictory findings with regard to its effect on the risk of infant death, possibly owing to heterogeneity between studies. One source of that heterogeneity is the prevalence of vitamin A deficiency among pregnant women, which we examined using meta-regression techniques on eligible individual and cluster-randomized trials. Adapting standard techniques to control for the inclusion of a cluster-randomized trial, we modelled the logarithm of the relative risk of infant death comparing vitamin A supplementation at birth to a standard treatment, as a linear function of the prevalence of vitamin A deficiency in pregnant women.
FINDINGS: Meta-regression analysis revealed a statistically significant linear relationship between the prevalence of vitamin A deficiency in pregnant women and the observed effectiveness of vitamin A supplementation at birth. In regions where at least 22% of pregnant women have vitamin A deficiency, giving neonates vitamin A supplements will have a protective effect against infant death.
CONCLUSION: A meta-regression analysis is observational in nature and may suffer from confounding bias. Nevertheless, our study suggests that vitamin A supplementation can reduce infant mortality in regions where this micronutrient deficiency is common. Thus, neonatal supplementation programmes may prove most beneficial in regions where the prevalence of vitamin A deficiency among pregnant women is high.