RESUMO
BACKGROUND: An association has been demonstrated between herpes simplex type 2 (HSV-2) and HIV infection among men, but prospective studies in women have yielded mixed results. OBJECTIVE: To estimate the effects of prevalent and incident HSV-2 infection on subsequent HIV acquisition among women in two African countries. DESIGN: Prospective cohort study. METHODS: HSV-2 and HIV serostatus were evaluated at enrollment and quarterly for 15-24 months among 4531 sexually active, HIV-uninfected women aged 18-35 years from Uganda and Zimbabwe. The association between prior HSV-2 infection and HIV acquisition was estimated using a marginal structural discrete survival model, adjusted for covariates. RESULTS: HSV-2 seroprevalence at enrollment was 52% in Uganda and 53% in Zimbabwe; seroincidence during follow-up was 9.6 and 8.8/100 person-years in Uganda and Zimbabwe, respectively. In Uganda, the hazard ratio (HR) for HIV was 2.8 [95% confidence interval (CI), 1.5-5.3] among women with seroprevalent HSV-2 and 4.6 (95% CI, 1.6-13.1) among women with seroincident HSV-2, adjusted for confounding. In Zimbabwe, the HR for HIV was 4.4 (95% CI, 2.7-7.2) among women with seroprevalent HSV-2, and 8.6 (95% CI, 4.3-17.1) among women with seroincident HSV-2, adjusted for confounding. The population attributable risk percent for HIV due to prevalent and incident HSV-2 infection was 42% in Uganda and 65% in Zimbabwe. CONCLUSIONS: HSV-2 plays an important role in the acquisition of HIV among women. Efforts to implement known HSV-2 control measures, as well as identify additional measures to control HSV-2, are urgently needed to curb the spread of HIV.