Published online by Cambridge University Press: 14 December 2007
HIV-1 (subsequently referred to as HIV) infection may be transmitted to the infant during breast-feeding. The risk of transmission varies with stage of maternal infection, duration of breast-feeding, type of breast-feeding (i.e. exclusive or mixed breast-feeding), and breast pathology. Recent studies support the findings of a meta-analysis (Dunn et al. 1992) and indicate that when breast-feeding is practised for over 2 years, the risk of infection is about 14%. Shorter durations of breast-feeding may therefore minimize the risk of transmission. Although the risk of infection appears to be greatest in the first 6 months this risk needs to be weighed against the excess risk of morbidity and mortality if children are not breast-fed in the first 6 months. In resource-poor settings any policy decision to replace breast-feeding with formula feeding in order to prevent postnatal HIV transmission needs to be balanced against the risks to the infant of malnutrition, morbidity and death if not breast-fed. New information suggests that exclusive breast-feeding, instead of the norm of mixed breast-feeding, may not increase risk of HIV transmission at 6 months and suggests that exclusive breast-feeding with early cessation may be a viable option for some women. The present review suggests options for reducing risk of HIV transmission through breast-feeding. Finally, current knowledge points to the dangers of large-scale replacement feeding programmes in contexts where women will have low education, poor access to water and health services, and strong cultural pressures to breast-feed. Emphasising replacement feeding in these contexts may fuel the mixed breast-feeding practice for mothers who will have no other choice than breast-feeding, yet will try to follow the recommendations applied to replacement feed.