Alcohol use and the pre-exposure prophylaxis continuum of care among men in rural South Africa

Abstract

Introduction Despite freely available HIV pre-exposure prophylaxis (PrEP), HIV incidence among young men in South Africa is high. There is conflicting evidence around the association between alcohol use behaviors and PrEP utilization. We explore the impact of alcohol use on PrEP initiation and continuation among South African men. Methods We performed a secondary analysis of data from a trial that included men aged 16-29 randomly selected from a demographic surveillance site in KwaZulu-Natal. All participants were referred to HIV and sexual health services, where those who were HIV negative and sexually active were offered oral PrEP. Alcohol consumption was assessed at monthly visits and categorized as: non-drinking (0), low/moderate risk drinking (1-5), and high/very high-risk drinking (6-12) based on AUDIT-C criteria. Primary outcomes were PrEP initiation and PrEP continuation defined as refilling prescriptions for >3 months. We fitted logistic regression models, adjusted for potential clinical and demographic confounders, to estimate relationships between PrEP initiation/continuation and reported alcohol use. Results Of the 325 men in the analytic cohort, the average age was 22.9 years (SD 3.6) and 131 (40%) had high/very high-risk alcohol consumption (AUDIT-C score ≥6). Men with the highest risk alcohol use also reported more frequent condomless sex (89%, vs 68% in no alcohol group). We found the greatest uptake of PrEP among the high/very high-risk alcohol group (46/131, 35%), followed by the low/moderate-risk group (17/53, 32%) and the no alcohol group (25/141, 18%). Those with high-risk alcohol use remained more likely to initiate PrEP compared to the no alcohol group in multivariable models adjusted for confounders (aOR 2.44 95% CI 1.29-4.60; p-value 0.006). Overall, only 30% (26/88) of men remained on PrEP at 3 months. Men with high/very high-risk drinking had similar PrEP continuation at 3 months compared to men who reported no alcohol use (aOR 1.02 95% CI: 0.28-3.86, p=0.98). Conclusions High-risk alcohol use is common among men in rural South Africa and associated with increased PrEP initiation. However, PrEP continuation was low overall, and similar across all levels of alcohol use. Hazardous alcohol use should not discourage PrEP implementation efforts to engage and retain young men.