Extraspousal partnerships in a community in rural Uganda with high HIV prevalence: a cross-sectional population-based study using linked spousal data.

BACKGROUND: The role of concurrent sexual partnerships in the HIV epidemic in sub-Saharan Africa is not well understood. Although most infections in Africa occur among married individuals, transmission may occur from both spousal and extraspousal partnerships. This article explores extraspousal partnerships as a form of concurrency, examining the association with HIV status, demographic characteristics, and sexual behaviors in a population-based cohort in rural Uganda. METHODS: Prevalence of extraspousal partnerships was estimated using cross-sectional data from 2008, and adjusted odds ratios (aOR) were estimated for factors associated with the prevalence of extraspousal partnerships using logistic regression. Among men who were not in polygynous marriages, we used linked spousal data to investigate the association between extraspousal partnerships and wives' serostatus. RESULTS: Extraspousal partnerships in the past year were reported by 17% of married men and 2% of married women. Among both men and women, extraspousal partnerships were associated with not knowing their partners' HIV status (men: aOR = 1.74; 95% CI: 1.13 to 2.67; women: aOR = 1.76; 95% CI: 1.13 to 2.75), and extraspousal partnerships were also associated with increased condom use for men. There was no evidence that men reporting extraspousal partnerships were at increased risk of HIV (aOR = 0.98; 95% CI: 0.48 to 2.01), or that a woman's risk of HIV was associated with her husband reporting extraspousal partnerships (aOR = 0.68; 95% CI: 0.29 to 1.57). CONCLUSIONS: For both men and women, extraspousal partnerships were associated with not knowing their partners' HIV status. There was no evidence of an association of extraspousal partnerships with HIV serostatus in this cross-sectional analysis.