Effects of Alcohol Intoxication on Sexual Decision-Making among Men Who Have Sex with Men (MSM): Alcohol's Influences on Self-Control Processes

Abstract

This experiment tested mechanisms linking alcohol intoxication and analogue determinants of condomless anal intercourse (CAI) in a sample of 257 men who have sex with men (MSM). The two mechanisms tested were implicit approach biases toward CAI stimuli and executive working memory. Participants were randomized to 3 conditions (water control, placebo, or alcohol) and following beverage administration completed a working memory task, an Approach Avoidance Task of sexual vs. condom stimuli, and two video role-play vignettes of high-risk sexual scenarios. Sexual arousal and CAI intentions were assessed by self-report, and behavioral skills and risk exposure were derived from participants' role-play behavior. Estimation of four path models showed that the hypothesized mechanisms were supported for the CAI intention outcome, but the findings for the skills and risk exposure outcome were mixed. Implications for development and enhancement of HIV prevention interventions were discussed.

Keywords: MSM; alcohol; condomless sex; executive function; implicit associations.

Figure 1. Hypothesized Conceptual Model and Structural Models of the Relations among Alcohol Consumption, Approach Biases, Working Memory, and Sexual Arousal on Sexual Decision-Making Outcomes

Note. Solid lines are hypothesized direct effects in the model. Dashed lines are moderating effects. Figure A = conceptual model. Figure B = Structural Model of CAI Intentions Analysis ( χ2(22, N = 257) = 16.88, p = .7699, CFI = 1.00). Figure C = Structural Model of Behavioral Skills Prompt 1 Analysis (χ2(22, N = 257)= 16.09, p = .8116, CFI = 1.00. RMSEA =0.00 90% CI[0.00,0.03], SRMR = 0.017). Figure D = Structural Model of Risk Exposure Analysis (χ2(22, N = 257)= 15.41, p = .8440, CFI = 1.00. RMSEA =0.00 90% CI[0.00,0.03], SRMR = 0.017). PrEP use, Site, and Age are included as covariates with paths to each endogenous variable, but omitted from the figures. Effects are standardized. † p <.10, * p < .05, ** p < .01, *** p < .001