Pregnancy outcomes in HIV-infected and uninfected women in rural and urban South Africa

OBJECTIVE: To describe pregnancy outcomes among clade C HIV-infected and uninfected women in South Africa. DESIGN: A longitudinal cohort study. METHODS: Pregnant women attending 9 rural/urban antenatal clinics were prospectively recruited and followed up. Women were seen at the clinic or at home after delivery on 4 occasions after enrollment: 2 times within the first 2 weeks of the newborn's life at home, and every 2 weeks thereafter until their first health clinic visit when the infant was 6 weeks old. RESULTS: A total of 3465 women were enrolled; 615 withdrew after delivery, moved away, or had a missing or indeterminate HIV status, leaving 2850 women (1449 HIV-infected women). Six women died after delivery and there were 17 spontaneous abortions and 104 stillbirths. An adverse pregnancy outcome was independently associated with HIV infection (adjusted odds ratio AOR] = 1.63; P = 0.015), urban enrollment (AOR = 0.39; P = 0.020), and nonhospital delivery (AOR = 13.63; P < 0.001) as well as with a CD4 count <200 cells/mL among HIV-infected women (AOR = 1.86; P = 0.127). Among 2529 singleton liveborn babies, birth weight was inversely associated with maternal HIV (AOR = 1.45; P = 0.02) and maternal middle upper arm circumference (AOR = 0.93; P < 0.001). Early infant mortality was not significantly associated with maternal HIV (hazard ratio HR] = 1.18; P = 0.52) but was with urban sites (HR = 0.34; P = 0.045). Low birth weight substantially increased mortality (AOR = 8.3; P < 0.001). HIV status of infants by 8 weeks of age (14.6%, 95% confidence interval: 12.5% to 17.0%) was inversely associated with maternal CD4 cell count and birth weight. CONCLUSIONS: HIV-infected women are at a significantly increased risk of adverse pregnancy outcomes. Low-birth-weight infants of HIV-infected and uninfected women are at substantially increased risk of dying.