Predictors of reporting bareback sex among a diverse sample of MSM recruited through a Swedish website

This study aimed to understand how person, health and sexual behavior, HIV-risk cognitions, and gay community involvement relate to barebacking among 3634 men who have sex with men (MSM) recruited by way of the Swedish website QX.se. In this sample, 10% reported that they had engaged in barebacking in the past year. Variables found to be statistically significant in the bivariate analyses were incorporated into a logistic regression domain-specific model. Variables that remained significantly associated with barebacking in domain 1 were being HIV-positive and having had a sexually transmitted infection (STI) in the past year. The variable talked with someone in HIV services remained significant in domain 2, and for domain 3, the variable used the Internet to look for a bareback partner remained significant. Two variables, believing taking semen in one's mouth involves no or low risk of HIV transmission and believing that engaging in insertive unprotected anal intercourse (IUAI) involves no or low risk of HIV transmission, remained significant in domain 4. The final multivariate regression analysis included sixvariables (from domains 1 to 4) and had a significant fit (chi(2)(6) = 2.571, p = 0.958). The likelihood of engaging in barebacking was higher for those men who reported being HIV-positive (odds ratio [OR] = 2.77), having had an STI in the past year (OR = 1.67), and having used the Internet to look for a bareback partner (OR = 12.59). This first study to explore the predictors of bareback sex among a Nordic MSM sample suggests that bareback sex among northern European MSM is less common than among other samples. The findings reconfirm that MSM who engage in bareback sex may represent a unique subset of MSM with distinct HIV prevention needs.