Infant feeding and HIV

Abstract

Avoiding transmission is not enough

Recently, the World Health Organization updated its recommendations of 2000¹ on infant feeding in the context of HIV.² At that time, data had just been published quantifying the risk of infection through breast feeding so avoiding breast feeding was acknowledged as the only effective way of avoiding transmission.³ WHO had also just published a meta-analysis of the mortality risks of not breast feeding, but in non-HIV infected populations.⁴ Considerations of these data resulted in the statement that “When replacement feeding is acceptable, feasible, affordable, sustainable and safe, avoidance of all breastfeeding by HIV-infected mothers is recommended.”¹ Since the 2000 recommendations, the main emphasis of most national programmes aimed at preventing mother to child transmission of HIV has been to avert transmission of HIV in young infants.

The most difficult challenge has been how to make breast feeding safer in communities with a high prevalence of HIV where breast feeding is the traditional mode of feeding. Remarkably, the dilemma of infant feeding and HIV has split scientific communities and programme managers into opposing camps. Even with the risk of HIV transmission, some maintain that breast feeding may still be the best option for many mothers infected with HIV because of its anti-infective and nutritional advantages.⁵,⁶,⁷ Others promote commercial infant formula, arguing that the risks of diarrhoea and malnutrition associated with formula feeding are lower in most urban communities, or that the risks of not breast feeding may not be as great for infants born to mothers infected with HIV who, to prevent transmission, choose to give formula milk from birth; it has been suggested that this active decision making and motivation may result in safer preparation and use of formula milk.⁸

In South Africa, formula feeding was reported as a successful intervention with minimal morbidity in an urban transmission prevention programme.⁹ However, this cross sectional study of just 113 infants recorded outcomes only in infants who were still being brought back to the clinics, and data from infants who had already died or had experienced serious morbidity would not have been reported. Recently, two reports have been published from Botswana¹⁰,¹¹ that, together with a compelling body of evidence from the literature in children without HIV,⁴,¹² emphasise that child survival must be seen as the greater goal; greater even than “just” avoiding HIV infection.

The first is a randomised trial evaluating whether giving zidovudine to infants exposed to HIV while breast feeding resulted in more infants surviving and not being infected with HIV than giving formula milk to infants from birth.¹⁰ It found no significant difference in HIV-free survival rates of infants at 18 months; any gains from reducing transmission by giving formula milk were lost through increased mortality from diarrhoea and malnutrition. At 7 months the infants who were formula fed from birth had almost twice the mortality risk of those who were breast fed (9.3% v 4.9%; P=0.003). Importantly, the data suggested that giving antiretroviral drugs to the mother may reduce transmission of HIV through breast feeding, whereas giving zidovudine to the infant alone seems to have little impact on transmission.

The second report, an investigation conducted by Centers for Disease Control, followed an outbreak of severe diarrhoea and malnutrition across the entire country that, in the first three months of 2006, accounted for 22 470 cases and 470 related deaths in children under 5 years.¹¹ In the same months in 2004 and 2005, there were 8478 and 9166 cases of diarrhoea and 24 and 21 related deaths, respectively. In children admitted with severe diarrhoea, not breast feeding was the biggest risk factor (adjusted odds ratio 50.0; 95% confidence interval 4.5 to 100); this was in spite of 97% mothers having access to piped water sources and most having been part of the HIV transmission prevention programme and therefore counselled on safe replacement feeding practices.

Should we be surprised by these data? In the original randomised trial in Kenya when the risks of postnatal HIV transmission in an African setting were first quantified,³ the risk of mortality associated with formula feeding in the first 6 months of life were similar to those described in the randomised trial in Botswana. By 18 months, however, and in contrast to the recent Botswana study, infants who had been randomised to formula feeding had better HIV-free survival. The mothers in this study also had access to safe water, electricity, and medical care and formula milk was supplied by the study (that is, all the effective conditions deemed necessary in the 2000 WHO recommendations). The irony of these reports is that Botswana arguably has the best primary health infrastructure and provision of water and sanitation in southern Africa. The national HIV transmission prevention programme was the first to promote formula milk as the default feeding option for mothers infected with HIV and it seemed to be successful. As alarming as the report of epidemic diarrhoea and ensuing deaths might be, we should be more concerned about possible longstanding unrecognised mortality related to mothers making inappropriate choices of infant feeding practices when their household circumstances cannot safely support those practices and infants are thereby at high risk of diarrhoeal illness, malnutrition, and death.

The updated WHO report and recommendations,² while generally supportive of the 2000 recommendations,¹ provide greater insight and nuance to the complexity of implementing any theoretical approach in programmes. They acknowledge that exclusive breast feeding has a twofold to fourfold lower risk of transmission than mixed feeding.¹³,¹⁴ They recommend that in the absence of safe water, hygiene, and community acceptance exclusive breast feeding should be promoted until the infant is 6 months old rather than just during the first few months of life. The continued risks of serious morbidity and mortality in infants if breast feeding is stopped early at around six months was highlighted stressing the need for continued contact and counselling with mothers during the first months of the infant's life to help guide her decisions about feeding practices after 6 months. Most important, perhaps, was the acknowledgement that the health systems in which transmission prevention programmes are implemented must be able to cope with the diarrhoeal illnesses and malnutrition that are so frequently the consequence of incorrect use of commercial infant formula.²

There is no doubt that breast milk can transmit HIV or that an infant's chances of survival when living in a poor or rural community are greatly decreased by not breast feeding. The challenge is how health systems can, at scale, help individual women, whether infected with HIV or not, appreciate the inherent risks and opportunities of their environment and make good decisions about how to feed their infants. For scientists, the challenge is to study and understand the evidence and not simply follow their hearts and possible prejudices in this crucial component of child survival.