Background: In South Africa, alcohol contributes to the HIV epidemic, in part, by influencing sexual behaviors. For some, high levels of alcohol consumption may be driven by previous traumatic experiences that result in traumatic stress. The purpose of this study was to quantify the longitudinal association between traumatic stress and unprotected sex among women who attend drinking venues and to assess whether this association was explained by mediation through alcohol use.
Methods: Data were collected in 4 waves over a year from a prospective cohort of 560 women who regularly attended alcohol-serving venues in a Cape Town township. Longitudinal mixed models examined (1) the relationship between traumatic stress and counts of unprotected sex and (2) whether alcohol use mediated the association between traumatic stress and unprotected sex.
Results: Most women reported elevated traumatic stress (80%) and hazardous alcohol use (88%) at least once during the study period. In models adjusted for covariates, traumatic stress was associated with unprotected sex (b = 0.28, SE = 0.06, t = 4.82, P < 0.001). In addition, traumatic stress was associated with alcohol use (b = 0.27, SE = 0.02, t = 14.25, P < 0.001) and was also associated with unprotected sex (b = 0.20, SE = 0.06, t = 3.27, P < 0.01) while controlling for alcohol use (b = 0.28, SE = 0.07, t = 4.25, P < 0.001). The test for the mediated effect established that alcohol use was a significant mediator, accounting for 27% of the total effect of traumatic stress on unprotected sex.
Conclusions: These results highlight the need to address traumatic stress among female venue patrons as an important precursor of HIV risk due to alcohol use.