UC San Diego

Human immunodeficiency virus type 1 (HIV) continues to disproportionately affect men who have sex with men (MSM). Recent concern has focused on the use of 'club' drugs among MSM as a risk factor for HIV acquisition and transmission. To determine the extent to which the current literature demonstrates a causal association between use of individual 'club' drugs and risk for HIV among MSM, Hill's criteria for causation was applied to existing studies. A theoretical framework that demonstrates potential pathways through which these drugs could be associated with HIV acquisition is proposed, and definitions of what constitutes a 'club' drug discussed. Building upon this review, and addressing the Centers for Disease Control and Prevention's call for research addressing prevention among HIV positive individuals, substance use as a risk factor for unprotected anal intercourse (UAI) was investigated among MSM with recent HIV infection. All MSM were interviewed using a computer assisted survey instrument (CASI) on average 5 weeks after HIV diagnosis and 13 weeks after estimated date of infection. Associations between substance use and UAI in within-subjects analyses were modeled using conditional logistic regression (CLR). Associations between substance use and UAI in the entire sample were examined using generalized estimating equations (GEE). GEE models that included interactions between timing of sexual activity (before versus after diagnosis) and substance use were used to examine change in associations between UAI and substance use before and after HIV diagnosis. Among participants, 16% reported no UAI with any of their last three partners; 60% (n=116) reported UAI with some, but not all, of their last three partners; and 24% reported UAI with all of their last three partners. In multivariate CLR (n=116) and GEE (n=194) models, UAI was associated with use of methamphetamine (OR: 4.9 and 2.7 respectively), marijuana (OR: 4.0 and 1.9 respectively) and erectile dysfunction medications when used with a main partner (OR: 26.0 and 17.1 respectively). In GEE models (n=207) that examined interactions between UAI and substance use before and after HIV diagnosis, UAI was associated with methamphetamine use alone (OR=4.80, 95% CI: 1.4, 16.1) and a combination of methamphetamine and other substances (OR=4.11, 95% CI: 1.9, 8.8) before diagnosis. After HIV diagnosis, UAI was associated with use of substances other than methamphetamine (OR=1.96, 95% CI: 1.3, 5.3) and a combination of methamphetamine and other substances (OR=5.08, 95% CI: 2.3, 11.2), but not methamphetamine alone. These analyses indicate that a direct association may exist between use of methamphetamine and may be differentially associated with UAI based on knowledge of HIV status. These findings have implications for both prevention of high risk sexual behavior and substance use among recently HIV infected MSM