Alcohol use and HIV serostatus of partner predict high-risk sexual behavior among patients receiving antiretroviral therapy in South Western Uganda

BACKGROUND Antiretroviral treatment restores the physical and immunological function for patients with HIV/AIDS and the return of sexual desire. The frequency and correlates of sexual activity among patients receiving ART have not been widely studied. There is concern that widespread availability of ART may result in sexual disinhibition including practice of high-risk sexual behavior. We determined the correlates of sexual activity and high-risk sexual behavior in an ART-treated population in rural and urban Uganda. METHODS We conducted a cross-sectional study among 329 ART-treated adult patients at two hospitals, one located in rural and another in urban western Uganda. We collected data on sexual activity, frequency of condom use, pregnancy, viral load (VL) and CD4. Patients were considered sexually active if they had had sexual intercourse in the last 6 months. Any unprotected sex was considered high-risk sex. A two-stage logistic regression was performed to determine factors associated with sexual activity and high-risk sex among those sexually active. RESULTS Overall, 222 (67%) patients were women, 138 (41.2%) had been on ART for at least one year, and 168 (51.4%) were sexually active of whom 127 (75.6%) used condoms at the last intercourse. Younger age (<=30 years) (Odds ratio; OR=2.3, 95% CI 1.2, 4.2), higher monthly income (OR=4.1, 95% CI 2.4, 7.4), and being married (OR=22.7, 95% CI 8.2, 62.9) were associated with being sexually active. Undetectable VL, CD4 count and treatment duration were not significantly associated with sexual activity. Among the sexually active, alcohol consumption (OR=3.3, 95% CI 1.2, 9.1) and unknown serostatus of partner (OR=5.8, 95% CI 1.5, 21.4) were significant predictors of high-risk sexual behavior. The frequency of unprotected sex at the last intercourse was 25.9% and 22.1% among the men and women respectively and was not significantly different (p value for chi square test =0.59). CONCLUSION Younger persons receiving ART are more likely to be sexually active. ART clients are more likely to engage in unprotected sex when sero-status of partner is unknown or report use of alcohol. Counseling on alcohol use and disclosure of sero-status may be useful in reducing high risk sexual behavior.