Determinants of HIV seroconversion among men who have sex with men living in a low HIV incidence population in the era of highly active antiretroviral therapies.

OBJECTIVE: To estimate human immunodeficiency virus (HIV) incidence and associated risk factors among men who have sex with men (MSM) participating in the Omega Cohort Study in Montreal, 1996-2003. METHODS: Longitudinal study of 1587 MSM seronegative at baseline with > or =1 six-month follow-up visit. Multivariate Cox regression with time-dependent variables was used for data analysis. RESULTS: HIV incidence was 0.62 per 100 person-years (95% confidence interval: 0.41-0.84). In multivariate analyses compared with subjects not reporting any anal sex with serodiscordant or casual partners, those reporting anal sex with such partners (all P values <0.05), whether consistently protected [hazard ratio (HR) = 3.4], or unprotected exclusively receptive (HR = 12.0), exclusively insertive (HR = 4.7), or both receptive and insertive (HR = 8.3), were at increased risk of seroconversion. Sexual behaviors with seroconcordant regular partners were not associated with seroconversion. CONCLUSION: These results observed in a cohort of MSM with low HIV incidence provide new insights regarding the debate about harm-reduction strategies to prevent sexual HIV transmission.