Vitamin A supplementation fails to reduce mother-to-child transmission of HIV, may even harm infants exposed to HIV via breast milk

Vitamin A supplementation, given as a large single dose to either the mother or infant (or both) shortly after delivery, does not reduce mother-to-child transmission (MTCT) of HIV according to the final results of the Zimbabwe Vitamin A for Mothers and Babies (Zvitambo) Study, which were published in the March 15^th issue of the Journal of Infectious Diseases. In fact, the data suggested that vitamin A supplementation may have increased HIV transmission and/or infant mortality in some study arms and doubled the risk of mortality in the subset of infants exposed to HIV via breastfeeding.

Vitamin A

Vitamin A is an essential nutrient, and its deficiency has been linked with night blindness, weakened immunity, infections and decreased survival in children over six months of age. Periodic vitamin A supplementation to children over the age of six months has been recommended and implemented by over 70 countries and is considered to be highly beneficial and cost-effective. Additionally, several studies studies in Africa have found that, in HIV-positive children over six months old, vitamin A supplementation reduces illness (diarrhoea and cough), and almost halves the death rate.

Vitamin A supplementation is also beneficial in adults, especially when they are nutrient deficient because of an inadequate diet or food insecurity. Studies in resource-limited settings where nutrient deficiency is common have demonstrated that vitamin A supplementation (200,000 - 400,000 IUs in one or two doses) in pregnant women and breastfeeding mothers generally improves both maternal and infant health. The findings on the benefits of direct vitamin A supplementation (50,000 IUs) in young infants have not been as consistent, with two large studies showing a benefit and another demonstrating no effect.

Vitamin A and HIV

However, these latter studies were not performed in populations with a high burden of HIV, and aside from the benefits already demonstrated in HIV-positive children from resource-limited settings, the impact of vitamin A supplementation in people with or at risk of HIV may not always be so clear cut.

For instance, in the early 1990s, a study reported that people with HIV in the United States, where nutrient deficiency is uncommon, who took either too much or too little vitamin A, progressed more rapidly to AIDS than those who took only the recommended daily allowance (RDA) of the vitamin. Even taking merely twice the RDA of vitamin A had deleterious effects in these patients — although these results can not be directly applied to settings where people are nutrient deficient.

In addition, cell culture studies have demonstrated a complex interaction between HIV and vitamin A that varies dichotomously depending on timing of infection and exposure to vitamin A. Some studies reported that treating HIV-infected cells with vitamin A directly inhibits viral transcription and replication, while conversely, others have found that HIV infection spreads much more rapidly in cell cultures that have been pre-treated with vitamin A. Some of these divergent effects could be mediated by the various potent effects vitamin A has on cell differentiation, enzymatic reactions and cellular receptor expression, on one hand suppressing HIV while on the other, priming cells for infection. But how or whether these in vitro findings were relevant to clinical settings was unknown.

Then, also in the early 90s, a few observational studies of HIV-positive women noted that vitamin A deficiency during pregnancy was associated with higher rates of HIV transmission and/or infant mortality and higher concentrations of cell-associated HIV in breast milk.

It was theorised that this relationship might be causal — that lower levels of vitamin A drove replication and increased transmission. Given that vitamin A is also important for the health of mucosal tissue, such as mammary glands, researchers postulated that supplementation with vitamin A might decrease HIV transmission and related mortality in infants born to HIV-infected mothers — particularly in resource-limited settings and mothers who were breastfeeding.

However, three subsequent studies of vitamin A supplementation in pregnant and breastfeeding mothers found no such benefit and one Tanzanian study even reported an increased risk of HIV transmission.

Zvitambo

Zvitambo is probably the final and by far the largest MTCT study with vitamin A, enrolling a total of 14,110 mother-infant pairs at 14 maternity clinics and hospitals in greater Harare. Unlike the three prior studies, which evaluated daily supplementation, Zvitambo measured the effect on breastfeeding–associated MTCT and HIV-free survival using a simpler regimen (consisting of a single large vitamin A dose (400,000 IUs) given to HIV-positive women and/or their infants (50,000 IUs) shortly after delivery). Antiretrovirals to prevent MTCT were not available in Zimbabwe at this time.

Participants were randomised to one of four arms 96 hours after childbirth:

• Aa — both mother and infant were given vitamin A
• Ap — mother received vitamin A, while the baby received placebo
• Pa — mother received placebo, while the baby was given vitamin A
• Pp — both mother and child received placebo

Infants born to HIV-infected mothers were further stratified by time of infection. Infants who tested positive for HIV at birth were categorised as “IU” infants since they must have become infected during the intrauterine period. Those who tested HIV negative at childbirth but positive six weeks later were classified as “IP” infants because they likely to have become infected either during the late intrauterine intrapartum or early postnatal period. The remainder were classified as “6-week-negative” infants.

At study entry, 4495 infants were born to women who tested HIV-positive. In general, baseline characteristics were similar the treatment groups (although there were small but statistically significant but imbalances in maternal education and birth weight. Around 60% of mothers in all treatment groups had serum retinol levels

Mortality:

However, the authors conclude vitamin A supplementation did have an effect on survival depending upon the timing of HIV infection. A total of 381, 504, and 2876 infants were categorised as IU, IP, and 6-week-negative infants, respectively. Of these, 339 (89%), 478 (94.8%), and 2644 (91.9%), respectively, were followed to twelve months.

Vitamin A supplementation in either the mother or infant or both had no effect on mortality in IU infants, however, in IP infants, it significantly reduced mortality at 24 months, by 28%. Hazard ratios for death were similar in both the Aa, and Pa arms, but maternal supplementation alone (Ap) had no effect.

Conversely, for the 6-week-negative infants, all three vitamin A supplementation regimens were associated with around a two-fold increase in mortality. Since the researchers believe that vitamin A supplementation did not increase transmission (although over longer follow-up many HIV test results were unavailable), they conclude that "our finding...could be explained if priming with vitamin A increased viral load in those who became infected during breast-feeding, thereby hastening their progression to death."

Results

Transmission: An estimated 8.6% (95% confidence interval 7.2%–10.0% ), 26.6% (25.1%–27.9%), and 37.5% (35.7%–40.1%) were infected at baseline, six weeks, and 24 months, respectively. Proportionately, 22.9%, 48%, and 29.1% of all HIV transmission occurred during the intrauterine, late intrauterine/intrapartum/early postnatal, and late postnatal periods, respectively. The cumulative proportions of infection or death were 29.1% (26.2%–32.5%) and 43.2% (38.2%–48.8%) at six weeks and 24 months, respectively.

As for the individual study arms' effect on transmission or transmission or death together (HIV-free survival), the authors wrote that “neither maternal nor neonatal vitamin A supplementation significantly affected postnatal MTCT....between baseline and 24 months.” However, infection rates and infection-or-death rates were actually higher in the Ap and Pa groups, compared with those in the Aa and Pp groups — and the difference reached statistical significance at twelve months.

But the differences were not statistically significant when looking at HIV-free survival in only the 6-week negative infants. Thus the difference was already apparent by six weeks, so the authors surmised that the vitamin A dosing could not have had an impact on transmission at that time. But while it is true that postnatal vitamin A supplementation would have no effect on late intrauterine or intrapartum transmission, it could impact postnatal transmission occurring after dosing, during in the first few weeks of life. HIV PCR testing detect many infections within two or three weeks of exposure so it would pick up many, though not all, of early postnatal cases by six weeks. And one would presume that if the vitamin A was to have a negative effect on transmission — it would be most apparent proximal to dosing rather than later.

The authors discount their own findings because a consistent increase in transmission or death was not seen in the Aa group, writing that the results may have been due to "chance rather than the implausible interpretation that giving vitamin A only to the mother or only to the infant increases MTCT, whereas giving vitamin A to both the mother and the infant has no effect." It is curious though, that the researchers choose to believe that the fault lies in the two arms that showed an increased risk of transmission, rather than in the one arm that didn't — especially in light of the results of the earlier study from Tanzania.

Furthermore, in an accompanying editorial, another prominent nutition researcher, Dr Wafaie W. Fawzi of Harvard University, suggests that the researchers are dismissing that 'implausible interpretation' too quickly: "Little is known about the complex relationships between vitamin A supplementation and the mix of possible adverse and beneficial effects on immunological and virological parameters at the systemic and mucosal levels, and it would be difficult to ignore the increased risks that Humphrey et al. observed when vitamin A was given only to either the mothers or the infants."

Public policy implications

Regardless of how these anomalies are explained away, it seems clear now that the low vitamin A concentrations in HIV-infected women reported in the earlier observational studies were not necessarily the direct cause of increased transmission or higher levels of viral replication in the breast milk. In fact, the vitamin deficiency may have just been a marker or the result of advanced disease — which itself was the cause for higher transmission. Changing public policy on the basis of those early observations could have harmed some children.

"The findings of the study in Zimbabwe highlight the importance of conducting rigorous research to assess the importance of low-cost interventions that are often presumed to be beneficial... For now, the evidence available—including these latest findings from Zimbabwe—raise concerns about the safety of maternal vitamin A supplementation programs as recommended by the World Health Organization (WHO)," Dr Fawzi wrote in the accompanying editorial.