Background:  High HIV viral loads during recent infection may result in increased likelihood of HIV transmission during unprotected anal intercourse. Behavioral patterns of people with recent HIV infection over time provide insight into secondary transmission patterns.

Methods:  A cohort of recently HIV-infected men who have sex with men (MSM) in Southern California were followed to 3.5 years. Participants were interviewed at baseline, and approximately every 3 months thereafter:  230 completed a baseline interview and a median of 3 follow-up interviews and a median follow-up time of 354 days. Analysis of behavior over time was conducted using all 860 baseline and follow-up interviews.  Hierarchical univariate and multivariate analyses was conducted using generalized estimating equations to determine predictors of unprotected anal intercourse with an HIV^– or status-unknown partner over time.   

Results:  Mean age was 34.4 years; 68.3% were white, 20% Hispanic, and 11.7% other ethnicity. For all follow-up, 62.5% reported an HIV^– or unknown partner, 64.4% disclosed their HIV status to their partners, and 21.6% reported using methamphetamine at the last sexual encounter with their partners. Partner types reported at follow-up included 22.5% unknown partner, 18.7% an acquaintance partner, 14.5% a friend, 14.5% a repeat partner, 18.7% a 1-time partner, 13.6% a main partner, and 1.7% a trade partner;   27% reported a main partner at any interview. Median number of partners in the past 3 months was 3 and a median of 2 were anonymous, trade, or 1-time partners. Using generalized estimating equations and controlling for time since baseline, the following were significantly associated with unprotected anal intercourse with an HIV^– or status unknown partner at any follow-up:  sex with a main partner (OR 2.38, 95%CI 1.29 to 4.40) and number of sex partners during the last 3 months (OR 1.02, 95%CI 1.00 to 1.04), while ethnicity and methamphetamine use were not associated.  

Conclusions:  Among recently HIV-infected MSM, more were likely to expose their partners, especially main partners to HIV.  As the number of reported partners increased, the more likely these MSM were to report a transmission behavior suggesting that those who take risks do so with many partners. This suggests that targeting interventions to such individuals could substantially reduce secondary transmission from newly infected individuals.