Sexual violence and associated factors among women in HIV discordant and concordant relationships in Uganda

HIV serodiscordance is a sexual partnership in which one partner is infected with HIV while the other is not. Managing emotional and sexual intimacy in HIV serodiscordant unions can be difficult due to concerns about HIV transmission and the challenge of initiating and maintaining safe sex. In situations where couples are jointly aware of their HIV status, women in serodiscordant unions may face increased risk of partner violence. We conducted an investigation to assess risk factors for HIV serodiscordance and determine if HIV serodiscordance is associated with incident sexual violence among a cohort of women attending HIV post-test club services at three AIDS Information Centers (AICs) in Uganda. Using a prospective study of 250 women, we elicited information about sexual violence using structured face-to-face interviews. Sexual violence and risk factors were assessed and compared among HIV positive women in HIV discordant unions, HIV negative women in discordant unions, and HIV negative women in nega tive concordant unions. Multivariable logistic regression was used to assess the association between participants ' serostatus and sexual violence. HIV negative women in serodiscordant relationships (36.1 ± 11.1 years, range: 19 - 65 years) were significantly older than either HIV positive women in serodiscordant relationships (32.2 ± 9.0 years, range: 18 - 56 years), or HIV negative women in concordant relationships (32.3 ± 11.0 years, range: 18 - 62), (p = 0.033). Early age at sexual debut was associated with a 2.4-fold increased risk of experiencing sexual violence (OR 2.4, 95 % CI 1.27 - 4.65). Based on unadjusted analysis, HIV positive women in discordant relationship were at highest risk for sexual violence compared to HIV negative women in discordant unions, and HIV negative women in negative concordant unions. HIV negative women in discordant relationships and those in concordant negative relationships showed no increased risk for sexual violence. However, couples ' HIV serostatus was not significant related to incident sexual violence after controlling for potential confounding covariates. Nevertheless, the results were able to elucidate the sexual violence risk factor profile of participants based on couples ' HIV serostatus. Couple counseling protocols at HIV voluntary counseling and testing centers in Uganda should identify those at risk for sexual violence and develop interventions to reduce its incidence.