1020 Vitamin A supplementation to reduce maternal mortality

In South Africa with an estimated maternal mortality rate of around 400/100 000, and with the pressure to achieve reductions in terms of Millennium Development Goals, there has to be an interest in strategies to bring that rate down. Certainly there is a close relationship between the high maternal mortality and prevalence of HIV/AIDS in the country, so any strategy must be linked to HIV/AIDS initiatives. Those interested in the problem will no doubt be noting the result of a recently published study from Ghana that involved >200 000 women of reproductive age who were given placebo or 25000IU retinol in a study aimed at reducing maternal mortality. Whereas a 1999 study from Nepal had found a 44% reduction in maternal mortality following supplementation during pregnancy, a Cochrane review in 2002 concluded that further trials were necessary to establish a beneficial effect. Interestingly causes of death that were reduced in the Nepalese study were in areas such as hanging, drowning and snakebite, so it was not surprising that the results were treated with skepticism. Other Nepalese-based studies by the same highly respected US-based group have produced results showing that offspring of vitamin A supplemented women have improved lung function at 9-13 years, maternal micronutrient supplementation may increase symptoms of birth asphyxia (by 60%) and mortality risk in term infants, and maternal folic acid ± iron supplementation reduces mortality within the first three months in preterm infants. The Ghana study showed no benefit of vitamin A supplementation on maternal mortality, with an odds ratio of 0.92 in the supplemented group (95% CI 0.73-1.17). A recent Bangladeshi study has also not shown any benefit. While apparently accepting that there is no role for vitamin A supplementation during pregnancy, the authors of a Lancet editorial do however ask whether the positive results in the Nepalese study imply differences between the populations (e.g. higher baseline mortality rates or varying patterns of vitamin A deficiency). They then add that for quick answers (within 3 years) to a question that involves a statistic expressed as a rate per 100 000 one must perform multisite trials in order to get the numbers, obviously ensuring that all the right data are collected to establish whether divergent results from preliminary studies were on the basis of regional or population differences.

Lancet 2010; 375: 1640-9 and 1675-7
BMJ 1999; 318; 570-5
N Engl J Med 2010; 362: 1829-31

 

<BACK

HOME