Vitamin A supplementation for very-low-birth-weight infants

The review reports a meta-analysis of seven randomized controlled trials of vitamin A supplementation in very-low-birth-weight (VLBW) infants. The addition of new data from a large multicentre randomized controlled trial of vitamin A supplementation in 807 VLBW infants and (1), another smaller trial (2), involving 154 infants to the meta-analysis shows a small but significant reduction in death or oxygen use at one month after birth (Relative risk [RR]: 0.93; 95% confidence interval [CI]: 0.88 0.99).

While the overall strategy for extraction of information and meta-analysis by the reviewers is sound, there are several limitations to the analysis that should be recognized. Given the close relationship of maternal vitamin A status and fetal lung development, maternal nutritional status should have been included as a variable. In particular, given that data from Black South African newborn VLBW infants are also included, it would also be important to look at maternal HIV positivity, which is an important correlate of maternal vitamin A status in developing countries, (3,4). Although most of the VLBW infants were premature, no information is provided on postnatal feeding. There were significant differences in overall vitamin A intake from parenteral feeds in the two studies from North America and it is likely that such differences in vitamin A intake may have influenced the outcome.

More importantly, as the reviewers themselves indicate, these studies span a long period of time which has witnessed dramatic changes in the management of VLBW infants, with increased use of antenatal and postnatal steroids, and almost universal surfactant administration. Surfactant therapy and postnatal steroids have also led to significant improvement in neonatal survival from respiratory distress syndrome (RDS) with reduced rates of chronic lung disease (CLD), and in any meta-analysis with studies from the pre-surfactant era, these confounding factors must be accounted for. The impact of postnatal steroids on increasing plasma vitamin A concentrations are important to take into account when correlating vitamin A dosage and plasma vitamin A concentrations with outcome.

Finally, given that infections account for a large proportion of deaths among hospitalized VLBW infants, (5) and that the objective of vitamin A supplementation programmes globally is to reduce childhood, (6) and neonatal mortality (7), from infectious diseases, it is surprising that infections do not figure as an important outcome in this systematic review. It is important to note that in the recent large multicentre trial of vitamin A supplementation in VLBW infants, (1), there was a trend towards lower rates of sepsis and necrotizing enterocolitis.

The full RHL commentary also includes sections on: