Résumé
Abstract only Breastfeeding (BF) determinants were investigated in 1028 HIV‐infected African women participating in a 5 site PMTCT study, which provided antiretroviral prevention during pregnancy and postpartum (PP) (a triple antiretroviral prophylaxis up to 6 mo PP or the standard ZDV‐based regimen until 7 d PP). Women were counselled to either breastfeed exclusively up to 6 mo or formula‐feed exclusively for 6 mo. Some 76% ever breastfed, among whom 73% still breastfed at 3 mo (40% exclusively and 24% predominantly). In multiple logistic regression analysis, significant determinants of formula‐feeding were study site, CD4 <200mm 3 , secondary schooling and emergency C‐section. BF at 3 mo (if breastfed at 1 mo) was independently associated with study site (P<0.001) and neonatal morbidity reported by the mother (OR: 0.60 (0.37–0.985)) but not with report of maternal morbidity (fever or fatigue) during the first mo PP (0.57 (0.31–1.02)), CD4 cell count or schooling (P>0.20 for both). By 3 mo, boys tended to be less breastfed (0.70 (0.46–1.05)) and were less often exclusively breastfed than girls (0.52 (0.34–0.78)). This study suggests that exclusive breastfeeding is shorter among boys of HIV‐positive African women. The importance of neonatal morbidity for breastfeeding duration requires further investigation. Funding: ANRS, DFID, EDCTP, Trasher Research Fund, GlaxoSmithKline Fund, CDC, NIH, WHO