Vitamin A

Retinol metabolites play essential metabolic roles. They maintain nightvision and the integrity of the cornea. The metabolite, retinoic acid has been shown to play a fundamental role in embryonic development. Retinoic acid receptors have been identified, which activate transcription of genes. In animals, a vitamin A deficient diet induces malformed offspring, mostly affected by microphthalmia and anophthalmia associated with cardiac, lung, and urogenital system defects (Azaïs-Braesco & Pascal 2000). In the light of these findings, a higher incidence of malformed babies would be expected in areas of endemic vitamin A deficiency, but this is not the case.

On the other hand there is now a considerable consensus that vitamin A deficiency, even marginal, can affect survival in children, probably by reducing morbid periods or their effects (Beaton et al . 1993). This has triggered supplementation studies in other population groups. In a large vitamin A supplementation trial in Nepal, 44 646 women were followed (West Jr et al . 1999). They received either a weekly dose of vitamin A, or placebo. During the follow up period 22 189 pregnancies were recorded. Deaths during pregnancy and up to 12 weeks postpartum were recorded. The morality was 704 (n=51), 426 (n=33) and 361(n=36) per 100 000 pregnancies in the placebo, vitamin A and ß carotene group respectively. This study has raised quite an interest but caution is necessary in the interpretation. All types of deaths were used in the comparison, even if there is no functional explanation to do so. Death due to accidents and chronic illness contributed to a large number of deaths in the placebo group. If one excludes them, the number of deaths in the different groups change (placebo 43 deaths with RR 1.0, vitamin A 33 deaths with RR 0.45-1.18 and ß carotene 23 deaths RR 0.31-0.94). Only the carotene group shows significant differences. The period of follow was also much longer than the usually accepted six weeks postpartum. Taking the usual definition of pregnancy related deaths, the results are also no longer significant (Sachdev 1999, Ronsmans et al. 1999, Azaïs-Braesco & Pascal 2000). Overall, the evidence is not conclusive enough to warrant a vitamin A supplementation during pregnancy.

The latest theories on the role of oxidative stress in the pathophysiology of pre-eclampsia and eclampsia have triggered the interest in the direct role of ß-carotene during pregnancy. Free radicals are proposed as the toxic elements that negatively affect maternal vascular function. Reactive radicals start peroxidation of lipids on cell membranes changing the structure of the cell wall and secondarily the normal function of the cell (Halliwell 1994). Markers of lipid peroxidation are increased in plasma of women with pre-eclampsia, and the low concentrations of water-soluble and lipid-soluble antioxidants in plasma and placenta further suggest a state of antioxidant stress (Yanik et al . 1999, Shaarawy et al . 1998, Mikhail et al . 1994). In these studies lower levels of vitamin E, C and ß carotene were also found to be associated with a higher risk of pre-eclampsia. A recent randomised trial seems to confirm this oxidative stress theory as a cause of pre-eclampsia. Participants either received a placebo or a dose of vitamin E and C. In the intention to treat cohort, pre-eclampsia occurred in 24 (17%) of the 142 women in the placebo group and 11 (8%) of the 141 in the vitamin group (adjusted odds ratio 0.39, p=0.02). In the cohort who completed the study, the odds ratio for pre-eclampsia was 0.24 (0.08-0.70, p=0.002). It needs to be remarked that all the participating women were recruited on the basis of an abnormal uterine artery Doppler at 18-22 weeks of gestation and represent a selective population (Chappell et al . 1999). Effects of a supplementation on a population basis will be much smaller.

Vitamin A and ß carotene levels in the third trimester or at birth have also been found to be predictive of low birth weight and prematurity (Ramakrisham et al . 1999). So far no supplementation studies during pregnancy are available to determine a causal relationship. In the Nepal study (West Jr et al . 1999) these parameters were included but they have not yet been published.

Because of its accepted effects on morbidity and mortality, vitamin A has recently been investigated in relation to HIV infections. Some studies documented an association between serum retinol levels of the mothers and the risk of mother to child transmission of HIV(Greenberg et al . 1997). This has triggered controlled supplementation studies in Tanzania (Fawzi et al . 1998) and South Africa (Coutsoudis et al . 1999). In Tanzania 728 HIV pregnant women received either a daily vitamin A (with ß carotene) supplement or a placebo. There was no difference in the risk of HIV infection by 3 months of age between the two groups, nor were there differences in foetal mortality rates. However, vitamin A seemed to protect against pre term deliveries, and the pre term delivered babies were also less likely to be HIV infected. In South Africa, 1075 HIV pregnant women were assigned to either a vitamin A group, a multivitamin group, a multivitamin with vitamin A group or a placebo. In the group who received multivitamins less foetal deaths were recorded giving a relative risk for foetal deaths of 0.61 (0.39-0.94). The multivitamins also decreased the risk of low birth weight by 44%, severe pre term birth ((34 weeks gestation) by 39% and small size for gestational age at birth by 43%. Vitamin A supplementation alone had no significant effects on these variables. Multivitamins but not vitamin A, resulted in a significant increase in CD4, CD8 and CD3 counts.

What the exact effect is of a vitamin A deficiency on pregnancy remains up to now unclear. Although there is a theoretical framework to explain the negative effect of a deficiency no studies have been conducted that show beyond doubt that a supplementation program has a benefit on maternal mortality, birth weight or prematurity. In women at risk of pre-eclampsia a supplement does have a benefit. In situations where women are deficient it is warranted to correct the deficit to protect the newborn in the first months of life.