Maternal vitamin A deficiency during pregnancy and lactation continues to be a concern in many regions of the world. In this issue of IJE, Gogia and Sachdev report a systematic review and meta-analysis of the effect of maternal postpartum vitamin A supplementation on infant mortality and morbidity. A total of six trials from developing countries included in the meta-analysis revealed no impact on infant mortality [relative risk (RR)1⁄4 1.05, 95% confidence interval (CI)1⁄4 0.92–1.20]. Only one trial that reported morbidity found no effect on diarrhoea or respiratory infection. Out of the six postpartum vitamin A supplementation trials included in the meta-analysis, two large ones were cluster-randomized controlled trials that also included routine pre-conceptional and/or pre-natal supplementation using small weekly doses of vitamin A, representing a supplementation regimen that was restricted neither to the postpartum period nor to the large bolus dose as per the existing WHO guidelines for postpartum supplementation with 200 000 IU of vitamin A within 6 weeks. Additionally, for one of the trials in the meta-analysis, the authors selected infants from the placebo arm of a newborn vitamin A supplementation trial, which was nested within the maternal supplementation trial. Also, three other trials included in the meta-analysis had sample sizes of 300 or less per group. Although these trials reported mortality, they did not have this as a pre-specified outcome, nor were they designed or powered to assess mortality as an outcome. The one trial with the appropriate sample size and which used the currently recommended dose of 400 000 IU vitamin A was conducted in an HIV-infected population. Thus, it is evident that the meta-analysis was based on diverse data of limited utility for discerning a mortality impact. Additionally, it mixed different forms of maternal vitamin A dosing (single with routine) over varying periods of time (from pre-conceptional through postpartum) and therefore both cumulative and timing of dosages are variable. The primary motivation for the policy related to prophylactic postpartum dosing of vitamin A did not stem from reducing infant mortality; instead, the two primary goals for this approach were to improve maternal vitamin A status and the vitamin A status of infants <6 months of age via enhancing the vitamin A content of breast milk. Several previous trials of postpartum vitamin A supplementation, not included in the meta-analysis, have demonstrated improvements in biochemical indicators of vitamin A status in mothers and their breast-fed infants, including increased breast milk concentrations of vitamin A. Based on these studies, in 1997, the WHO, UNICEF and the International Vitamin A Consultative Group (IVACG) provided guidelines for postpartum vitamin A supplementation recommending a single dose of 200 000 IU soon after delivery. This recommendation was updated in 2002 to two doses of 200 000 IU (or 400 000 IU) based on an analysis of expected change and adequacy of infant vitamin A stores suggested to be suboptimal with a single 200 000 IU dose. The revised recommendation, however, has not been widely adopted in developing countries. Only two of the six trials included in the meta-analysis tested the revised 400 000 IU of vitamin A. In a parallel meta-analysis effort, these authors have also evaluated neonatal/early infancy dosing with 50 000 IU vitamin A using data from six trials and shown no impact on infant mortality. However, combining trials which provided vitamin A within 48 h of birth vs those in which infants were dosed later during the first month of life may have yielded the null findings. Limiting the analysis to the former studies, at least in the Asian context, has shown to have a beneficial impact. In a meta-regression analysis, it was shown that newborn vitamin A supplementation likely reduces infant mortality in regions where vitamin A deficiency is common. In contrast, studies from Africa, where vitamin A deficiency was not common, have shown an overall null effect, and a potential negative effect among girls. These apparently contradictory findings have resulted in three more trials of newborn vitamin A supplementation being commissioned by the WHO, Published by Oxford University Press on behalf of the International Epidemiological Association
Read full abstract