Background: Changes in sexual risk behavior may occur following HIV-1 infection, and these changes could have an important influence on the risk of secondary HIV-1 transmission. This investigation tested the hypothesis that HIV-1 seroconversion and disease progression would be associated with changes in the rate of unprotected intercourse in a cohort of Kenyan female sex workers (FSWs).

Methods: We conducted a prospective study of risk factors for HIV-1 acquisition among FSWs in Mombasa. At monthly visits, we administered a standardized interview to assess sexual risk behavior and performed HIV-1 serological testing. HIV-1 seroconverters were invited to continue with follow-up. Between 1993 and 2004 (when antiretroviral therapy was introduced), 1,597 HIV-1 seronegative women enrolled in the cohort, of whom 265 seroconverted for HIV-1 (7.7/100 woman-years). Generalized estimating equations were used to evaluate differences in the frequency of risk behaviors before versus after seroconversion in this group of 265 women.

Results: Participants accrued 541 years of pre-seroconversion follow-up and 840 years of post-seroconversion follow-up.  Overall, unprotected intercourse was reported at 574/2,128 (27%) pre-seroconversion visits versus 557/3,732 (15%) post-seroconversion visits (p <0.001). These findings remained significant after adjustment for potential confounding factors including time since enrollment, age, and use of contraception (adjusted odds ratio [aOR] 0.70, 95% confidence interval [CI] 0.55-0.88). Compared to HIV-1 seronegative women, there was a progressive step-wise decrease in unprotected intercourse among women with CD4 counts >500 cells/μL (aOR 0.83, 95% CI 0.46-1.51), 200-499 cells/μL (aOR 0.57, 95% CI 0.35-0.92), and <200 cells/μL (aOR 0.49, 95% CI 0.23-1.02). These decreases in unprotected intercourse reflected step-wise increases in both abstinence and 100% condom use.  Microscopic identification of sperm in genital specimens (a surrogate marker for recent unprotected intercourse) was significantly less frequent during HIV-1 seropositive versus seronegative follow-up (aOR 0.68, 95% CI 0.46-1.00).

Conclusions: Among Kenyan FSWs, HIV-1 infection and progressive immunosuppression were associated with a reduction in the frequency of unprotected intercourse. Additional research will be critical to determine how the introduction of antiretroviral therapy affects these underlying trends in sexual risk behavior.