HIV-related communication between partners is an important behavioral target in interventions to decrease the spread of HIV ^1-3. Communication includes ascertaining sexual histories, risk assessment ⁴, disclosure of HIV status ⁵, and condom use negotiation ^6,7. A number of studies have found a positive association between communication and increased condom use and HIV testing ^8-13. Correlates of partner-specific HIV prevention communication include higher rated relationship quality and positive outcome expectancies of the conversation ^14,15. Barriers include concerns about negative consequences such as rejection or potential for violence ^16,17.

Social networks provide an avenue for communication about HIV-related topics which can influence the formation or alteration of social norms around testing and condom use ^13,14. Social networks also serve as a source of health information and resources ^18,19. HIV prevention communication among network members may be a critical mechanism for diffusing information and messages about testing for HIV and other sexually transmitted infections.

Peer-based interventions focused on training individuals to promote HIV risk reduction within personal networks have shown efficacy ^20-22. Little attention has been given to characteristics of social network members being targeted with prevention messages. This line of research can inform the development of intervention content to build skills for increasing conversations specifically with higher-risk social network members and those with whom conversations are not occurring. African American men who have sex with men (AA MSM) are disproportionately affected by HIV. However, stigma associated with MSM behavior or HIV, particularly in African American communities, may be a barrier to communication. While recent studies have described the structure and function of social networks of AA MSM ^23-26, a gap exists in the literature concerning the social network characteristics and quality of communication with social network members about HIV.

The purpose of this study was to 1) examine attitudes of AA MSM on talking with peers about HIV testing, 2) to examine associations between having conversations about HIV/STI testing and characteristics of the social network members, and 3) to examine associations between having conversations about HIV status and testing together and characteristics of sexual partners.

Data for this study came from the baseline survey from the Unity in Diversity Study (2007-2008), a pilot behavioral HIV intervention conducted in Baltimore, Maryland. Details about recruitment are described elsewhere ²⁷. In brief, two types of participants were enrolled: primary and secondary. Primary participants were AA MSM recruited using a variety of methods including street-based outreach, advertising in area newspapers, and word-of-mouth referrals. Inclusion criteria were being at least 18 years of age, self-report of two or more sex partners and at least one male sex partner in the prior 90 days, unprotected anal sex with a male partner in the prior 90 days, and willingness to take an HIV test. Primary participants who met criteria and provided written informed consent completed a baseline study visit which lasted an average of 90 minutes and entailed an audio computer assisted self-interview (ACASI) assessment of HIV risks and an interviewer administered social network inventory. The social network inventory consisted of two sections: 1) name generator and 2) network characteristics. During the name generator section a participant is asked to list the first names and/or initials of individuals who provide emotional, material or health related support; those with whom the individual socializes; and drug or sexual network members. Once the total network was listed, information was collected about each network member such as their age, sex, type of relationship (i.e., kin, non-kin, sexual partner), sexual risks, and communication quality. From these responses, the total size of the network was calculated as well as the size of the sexual network, defined as individuals with whom the participant had sex in the prior 90 days. Participants received $50 for completing the baseline visit. At the end of the visit, primary participants were asked to recruit both male and female individuals from their social network (i.e., secondary participants) into the study and received $10 for each member recruited. Recruited secondary participants (n=73 total) provided written informed consent and completed the same survey on HIV risk behaviors and social network inventory as the primary participants. The sample for the present study consists of n=187 primary and n=39 secondary participants who were AA MSM .

The sample included 226 AA MSM whose mean age was 37.9 years (SD=10.6 years) (Table 1). More than half of the sample identified as gay (59%), 34% as bisexual or not sure, and 8% as heterosexual. Nearly half were HIV-seropositive (49%) and not working (43%). The majority of the sample reported that it was important/very important to talk to male friends about HIV (85%) and that they were comfortable/very comfortable talking with their friends about sexual behaviors (84%). Over half (51%) reported beliefs that they could change the risky behaviors of their friends.

There is scant research on the communication patterns of AA MSM with their social and sexual networks. This study first sought to describe attitudes of AA MSM towards HIV communication and then identify characteristics of members of the social and sexual network associated with talking about HIV-related topics. Our findings highlight that despite high levels of positive attitudes towards talking about HIV topics with peers, a very small proportion of social network members were actually talked to by the participant about HIV/STI testing (14%). This suggests that positive attitudes about communication may be necessary but not sufficient for actual conversations to occur. Consideration of the social context of the communication, namely the social network, can offer insight about factors that should be included in intervention design.

We report that the characteristics of network members associated with talking about HIV testing were younger age, transgender sex, more frequent contact, greater trust of the network member, and network member behavioral risk (e.g. using the internet to seek sex partners). Conversations with these network members may be associated with their being perceived as being more open and receptive to discussing HIV/STI and testing. Talking about HIV is a sensitive topic. As a result, conversations pertaining to the topic may be limited to individuals with whom there is a certain level of comfort, frequent contact, and opportunities to discuss the issue. This is consistent with other studies that have shown that positive communication outcome expectancies are associated with having actual conversations ¹⁵. This also underscores the value of activities, such as role-plays, that focus on increasing positive outcome expectancies. In developing interventions that focus on increasing conversations about HIV prevention with peers, these results suggests that optimal communication would be with individuals with whom participants trust and see often. One approach is to incorporate activities in programs that enable a participant to assess the social network members with whom they do trust and see often and then personalize communication practice with these network members in mind. Moreover, further research is warranted to determine how to increase communication with other network members with whom there is less trust.

There may also be a potential for generational differences in willingness to engage in HIV-related communication. This may in part be due to issues related to disclosure of HIV status, where older AA MSM have been reported to be less likely to disclose HIV status²⁵ or high levels of MSM-related stigma in older generations.

We report that a majority of the participants (58%) discussed partner HIV status with their sexual network members. Our finding contrasts with a study of white, Asian Pacific Islanders, AA and Latino MSM in San Francisco where authors report that one-third of men discussed status with partners ²⁹ and that AA MSM were less likely to know the HIV-status of their sexual partners than other racial/ethnic groups of MSM. In the current study, asking HIV status was positively associated with main and casual partner type compared to partners with whom money or drugs were exchanged for sex. One explanation for this may be that inquiring about HIV-status is related to the level of commitment in a relationship. Alternatively, the context of exchange sex does not facilitate communication of personal information such as status. Interventions that include opportunities to practice skills of asking about HIV status should provide scenarios that vary based on partner/relationship type. It is not clear from this study whether there are norms about asking status in any of these relationships. Specifically, in cases of exchange partnerships, there may be perceived sanctions or risks associated with asking status, such as risks to housing, food or money. Therefore, developing interventions that provide alternative strategies to asking about HIV status is important for risk reduction.

Our results demonstrate that good/excellent quality of communication in the relationship remained significantly associated with having conversations with sex partners. This may be due to the cross-sectional nature of the survey. We cannot know whether the perceptions of the communication quality caused or resulted from the conversations about HIV status and testing. Programs can offer communication strategies to use based on the perceived quality of communication. For example, in cases where an individual perceives the communication as excellent or good, programs can emphasize choosing the appropriate time for raising the conversation and using direct statements about testing as a couple.

There are some limitations of this study. First, this was a convenience sample of AA MSM and not necessarily generalizable to all AA MSM in Baltimore, especially to younger AA MSM, or to AA MSM in other urban settings, especially where the HIV prevalence may not be as high. Data were based on self-report; we did not confirm or validate with the network members whether these conversations had actually occurred. The cross-sectional design of the study limits our ability to draw causal inferences about network characteristics and communication. We did not include questions to assess contextual factors, such as the social or physical setting where conversations occurred or the exact nature of the conversations. Additionally, while we report statistically significant differences in the trust scores between social networks with whom the Index had talked about testing, it is not clear whether this is a meaningful difference in terms of trust in the relationship.

These limitations aside, this study provides insights into specific characteristics of members of the social network associated with greater odds of having conversations about HIV testing among AA MSM. Promoting communication about HIV/STI testing and condom use between sexual partners and peers is a central component in HIV prevention programs. Designing interventions that increase communication while utilizing natural social processes to alter social norms about testing can serve as one component of HIV risk reduction to address the epidemic among AA MSM.