Correlates of Risk Patterns and Race/Ethnicity Among HIV-Positive Men Who Have Sex With Men

Ann O'Leary; Holly H. Fisher; David W. Purcell; Pilgrim S. Spikes; Cynthia A. Gomez


AIDS and Behavior. 2007;11(5):706-715. 

In This Article


These data provide support for the hypothesis that ethnic differences in transmission risk behavior and its correlates are similar for HIV-seropositive MSM as for HIV-negative/unknown men. The higher level of bisexual activity among seropositive African American MSM may be partly responsible for recent increases in HIV prevalence rates for women in the African American community. Further, findings from the present study are suggestive that these bisexually active men's sexual behavior with other men is concealed from others in the community. The African American men in this study were less likely to report their sexual orientation as gay, exhibited higher levels of internalized homophobia, and reported lower levels of self-efficacy for disclosing their HIV status to others, compared with men other race/ethnic groups. African American men expressed less comfort than European American men in discussing their homosexual activity with close friends and acquaintances. Further, African American race remained uniquely associated with having sex with women even when a variety of psychological and demographic measures and Latino ethnicity were included in the model.

As has been found with HIV-negative/unknown MSM, African Americans had sex with fewer men than the other groups and they reported less unprotected oral sex with HIV negative or unknown status men. These results suggest a greater differential transmission risk to females than male partners in the African American community (although it is acknowledged that we do not know the serostatus of the female sex partners reported in this study). No ethnic differences in rates of unprotected sex were observed in the study. The disproportionate rates of HIV prevalence (CDC, 2005 a,b) and incidence (CDC, 2001) in African American MSM compared to other MSM is puzzling in light of their lower rate of MSM risk behavior. An article was recently published that explores several hypotheses regarding this conundrum. We found that our African American respondents were less confident in their abilities to disclose their HIV status to sex partners, indicating the potential utility of interventions to address serostatus disclosure (see, e.g., Kalichman et al., 2001). They also were more likely than the other men to drink alcohol before sex, suggesting that this may be a useful target for interventions.

The fact that internalized homophobia was higher among African American men and that they were more likely to report having had sex with women may be related in the sense that heterosexual activity in this community may be a mechanism for maintaining self-perceived (or other-perceived) masculinity (Millett et al., 2005; O'Leary & Jones, 2006). The influence of the African-American Church in discouraging homosexuality has also been cited as contributing to secrecy of MSM behavior (Ward, 2005; Woodyard, Peterson, & Stokes, 2000). In fact, the African American MSM in this study reported greater importance of religion and spirituality in their lives than men of other races; however, this factor was not related to having sex with women in multivariate analysis.

African American participants were also less confident than European Americans that they could tell potential sex partners about their own HIV infection than men of other racial/ethnic backgrounds. It should be noted that we did not assess self-efficacy for disclosure to female partners specifically. We also did not assess participants' self-efficacy to disclose sexual behavior (i.e., telling female partners about sex with men), although failure to reveal HIV status may be, for some participants, related to fear of behavior disclosure. A significantly greater proportion of African American MSM in this sample also reported drinking alcohol before or during recent sex, compared to all other groups, which may be related to conflictual feelings about having sex with one or both gender sex partners.

The risk pattern for Latino MSM in this sample suggests that they may be slightly more open about their homosexual behavior than African Americans but still more secretive than European Americans. Consistent with previous work (Montgomery et al., 2003), Latino MSM were less likely to report being gay, compared to European Americans, but more likely to report being gay than African Americans. Internalized homophobic attitudes also tended to be higher in the Latino sample than the European American one, but lower than the African American men. Similarly, self-efficacy for HIV status disclosure tended to be the highest in European American MSM, with the next highest level evident among Latino MSM; African Americans reported the lowest level of self-efficacy for status disclosure. These findings suggest the possibility that MSM in minority communities may be less open about their sexual orientation and, perhaps, less forthright with their sexual partners (male or female) about their sexual activities.

This study has several limitations. One is that data regarding the serostatus of female sex partners were not collected. Therefore, we cannot comment on whether the behavior reported here could cause new infections. We also included no attitudinal questions regarding sex with women. The data were cross-sectional, rendering causal inferences inappropriate. The data were also collected several years ago and patterns of transmission risk may not hold today. Furthermore, our sample was not population-based but rather was recruited from gay-identified and public sex venues in two large cities. Men also had to be sexually active to qualify for participation. Thus, these results may not generalize to the general population of HIV-seropositive gay and bisexual men.

These results have implications for HIV risk reduction interventions for HIV-positive African American men who have sex with men who, as a group, are in dire need of intervention. No effective behavior change interventions have been developed even for HIV-negative African American MSM (Herbst et al., 2005). We found that our African American respondents were less confident in their abilities to disclose their HIV status to sex partners, indicating the potential utility of interventions to address serostatus disclosure (see, e.g., Kalichman et al., 2001). They also were more likely than the other men to drink alcohol before sex, suggesting that this may be a useful target for interventions.

It is likely that men vary along a continuum in their degree of "outness" to others regarding their MSM behavior. Men whose MSM activity is highly secretive -- i.e., who withhold this information from wives, families, their faith communities, etc. -- may not be most effectively reached through standard outreach efforts. Such men may be difficult or impossible to recruit from gay-identified venues due to internalized homophobia (particularly those not wanting to be seen or associated with "out men"), disclosure of sexual activity with other men, and fear of excommunication from traditional institutions within African-American communities, family and church. As a result, we do not know the magnitude of these men's contribution to the HIV epidemic, an important although difficult research gap to fill. However, the need to reach these men with effective behavioral interventions that promote abstinence, monogamy and condom use and confer skills in negotiating these strategies with all partners is great. These interventions must also seek to reduce internalized homophobia that serves as a barrier to risk disclosure, and to increase men's skills to cope with perceived homophobia. One promising strategy for reaching these men may be through the use of respondent-driven sampling (RDS) (Heckathorn, 1997; Heckathorn, Semaan, Broadhead, & Hughes, 2002). CDC is supporting a formative feasibility study to use RDS to recruit bisexually active men for interviews, to offer HIV testing, and to assess the feasibility of using social networks to disseminate interventions and prevention messages for this important population. Better understanding and intervention strategies to reach African American and Latino MSM is crucial to combat serious HIV epidemics among men and women in those communities.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.