One-quarter of participants in a survey of San Francisco men who have sex with men were infected with HIV, and an analysis of the survey data suggests that prevention efforts need to target “novel cofactors” associated with HIV infection.1 For example, use of sildenafil (Viagra) was positively associated with HIV-infected men’s likelihood of reporting unprotected insertive anal intercourse with a partner who was not infected or whose HIV status was unknown, and optimism about the efficacy of HIV treatments was positively associated with uninfected men’s saying that they had had unprotected receptive anal sex with a serodiscordant partner.

The survey was undertaken to explore links to high-risk behaviors reported in earlier studies, particularly recently identified predictors of HIV transmission, and to improve on previous work by using a probability design. Researchers used random digit dialing to recruit a sample of adult men who identified themselves as homosexual or bisexual and reported having had sex with a man at least once since age 14; some 1,976 eligible men completed telephone interviews, which were conducted between June 2002 and January 2003. At the end of each interview, the respondent was asked to provide a urine sample for gonorrhea and chlamydia testing; men who had not reported having HIV infection were also asked to consent to HIV testing.

Twenty-five percent of respondents were HIV-infected; most of these men were aware of their status before entering the study, but eight learned of their infection through the test taken after they completed the interview. Chi-square analyses revealed significant variations in the prevalence of infection by a number of socioeconomic and demographic characteristics. Black men, participants who had not completed high school and respondents with the lowest incomes had the highest levels of infection (39–41%); Asians and Pacific Islanders, those with a graduate degree and men with the highest incomes had the lowest infection rates (14–18%). Prevalence rose steadily from 14% among men who had been living in San Francisco for less than a year to 31% among those who had been there for more than 10 years.

In the 12 months preceding the interview, 85% of respondents had been sexually active. Using chi-square tests, the researchers found marked differences between HIV-positive and HIV-negative men in the prevalence of risky behavior. Significantly higher proportions of infected than of uninfected men had had a nonprimary male partner, had had sex with someone they had met on the Internet, had had unprotected anal sex with a primary male partner whose HIV status was different from theirs or was unknown (i.e., a serodiscordant primary partner), and had engaged in a variety of risky behaviors with nonprimary male partners. HIV-positive men also reported higher rates of sildenafil, amyl nitrite (poppers) and crystal methamphetamine use than HIV-negative men.

Remaining analyses focused on two behaviors that carry a high risk of HIV transmission: unprotected insertive anal sex among infected men who had serodiscordant nonprimary partners and unprotected receptive anal sex among HIV-negative men with serodiscordant nonprimary partners. Although bivariate findings suggested that a broad range of background, behavioral and attitudinal characteristics predicted the first of these, only two characteristics were significant in multiple logistic regression analyses. The likelihood of having engaged in this behavior was elevated among users of sildenafil (odds ratio, 2.3) and was reduced among men reporting fewer than six partners in the previous year (0.04 for those reporting one partner and 0.1 for those reporting 2–5).

Uninfected men’s likelihood of having had unprotected receptive anal sex with a serodiscordant nonprimary partner was associated with a greater number of factors. The odds of having engaged in this behavior were higher among respondents who had been living in San Francisco for less than a year than among those who had been residents of the city for more than 10 years (odds ratio, 5.5), and were lower among those who had had five or fewer partners than among those who had six or more (0.05–0.3). HIV-negative men had elevated odds of reporting unprotected receptive anal intercourse with a serodiscordant nonprimary partner if they used crystal methamphetamine or amyl nitrite (2.8 and 2.6, respectively), or if they agreed that “[they] are less careful about being safe with sex or drugs than [they] were several years ago because there are better treatments for HIV now” (3.4).

The researchers acknowledge the potential limitations of a cross-sectional telephone survey, as well as the possibility that some men’s HIV status may have been misclassified. Nevertheless, they conclude, “In the third decade of the HIV epidemic, it is clear that [men who have sex with men] remain severely impacted…and that risk behaviors in this group are resulting in ongoing HIV transmission.” They consider their findings indicative of the need for programs aimed at amyl nitrite and crystal methamphetamine use, as well as for expanded types of counseling services. Additionally, they stress the need for high-quality studies of HIV incidence and prevalence, and related attitudes and behaviors, “to keep prevention efforts relevant.”—D. Hollander


1. Schwarcz S et al., Prevalence of HIV infection and predictors of high-transmission sexual risk behaviors among men who have sex with men, American Journal of Public Health, 2007, 97(6):1067–1075.