Document Type



Health Psychology


Objective: To assess the prevalence of unprotected sex and examine the association between alcohol consumption before sex and unprotected sex among HIV+ individuals in Cape Town, South Africa.

Methods: For 42-days daily phone interviews assessed daily sexual behaviour and alcohol consumption. Logistic and Poisson GEE models were used to examine associations between alcohol consumption before sex and subsequent unprotected sex.

Results: During the study which yielded 3,035 data points, the 58 HIV+ women and 24 HIV+ men drank an average of 6.13 drinks when they drank, and reported 4,927 sex events, of which 80.17% were unprotected. Over half (58%) of unprotected sex events were with HIV-negative or HIV-status-unknown partners. Extrapolating from the data using likelihood of infection per act estimates we calculated that an estimated 2.95 incident HIV-infections occurred during the study. Drinking alcohol before sex by the female, the male, or by both partners, increased the proportion and number of subsequent unprotected sex events. However, these associations held only when the quantity of alcohol consumed corresponded to moderate or higher risk drinking.

Conclusions: Among HIV-positives engaging in moderate or higher risk drinking prior to sex increases the likelihood and rate of unprotected sex. Prevention efforts need to address reducing alcohol-involved unprotected sex among HIV-positive persons.


This is not the final published article. Published in the Journal of Acquired Immune Deficiency Syndrome, Vol.49, No. 2, (Oct. 2008). Publisher URL: