HIV Risk and Perceptions of Masculinity among Young Black MSM

Abstract

Purpose

Young Black men who have sex with men (MSM) have among the highest rates of HIV infection in the US. Although reported rates of unprotected anal intercourse are similar to MSM of other racial/ethnic backgrounds, young Black MSM (YBMSM) aged 15–22 are 5 times more likely than comparably aged white MSM to be HIV-infected. We explored contextual social-environmental factors that may influence how YBMSM assess risk, choose partners, and make decisions about condom use.

Methods

We analyzed semi-structured interviews with 35 YBMSM (18–24) in New York City, upstate NY, and Atlanta. We used structured analytic coding based on a theoretical scheme that emerged from the data.

Results

Perception of masculinity was the primary contextual factor influencing partner selection, risk assessment, and condom decision-making. Four primary themes emerged: 1) greater preference for partners perceived as masculine; 2) discomfort with allowing men perceived as feminine to be the insertive partner in anal intercourse; 3) a power dynamic such that partners perceived as more masculine made condom-use decisions within the dyad; and 4) use of potential partners’ perceived masculinity to assess HIV risk.

Conclusions

Perceived masculinity may play a significant role in HIV risk for YBMSM and may be an important concept to consider in prevention strategies directed towards this population.

INTRODUCTION

Black men who have sex with men (MSM) have higher prevalence and incidence rates than MSM of other racial/ethnic groups [1, 2]. A 2005 study of 2261 men in 5 major metropolitan areas, sampled systematically from randomly selected venues where MSM congregated, reported an HIV prevalence of 46% among Black MSM (BMSM), compared with 21% in white MSM and 17% in Hispanic MSM [3, 4]. This disparity in HIV prevalence is similarly pronounced for young Black MSM (YBMSM). A recent study in 7 urban centers reports an HIV prevalence of 16% among YBMSM 15–22-years-old compared with 7% in Latino MSM and 3% in white MSM [5].

Reasons for disparities in HIV prevalence remain unclear, particularly given similar reported rates of unprotected anal intercourse across racial/ethnic groups [5]. Differences in sexual partner characteristics may contribute; for instance, one study found that YBMSM’s anal sex partners were more often older and Black than partners of other young MSM [6]. These different partner characteristics were associated with higher likelihood of HIV infection among YBMSM.

Regarding HIV infection, partner characteristics are downstream determinants, i.e., individually-based risk factors, potentially under personal control [7]. However, myriad upstream determinants, i.e., social-environmental factors, precede and potentially influence both partner selection and the decision to have unprotected sex with that partner. One upstream factor that has received minimal attention is socio-cultural conceptions of gender and masculinity. Masculine expression (i.e., behaviors that portray masculinity, such as one’s manner of dress, the expression of physical strength or aggression, one’s sexual activities and interactions with women ) among Black males has been presented as a response to racial and socioeconomic oppression [8–11]. Limited access to social and economic dominance promotes a masculinity emphasizing attainable hyper-masculine characteristics (i.e., exaggerated versions of sexual prowess, physical dominance, aggression, and other stereotypical masculine characteristics) [11, 12]. Further, attendant cultural perceptions equate male homosexuality with femininity, a construct that contradicts hyper-masculine gender role expectations [13].

Some research suggests that these socio-cultural conceptions of masculinity, sexual identity, and male gender role expectations influence the type of sexual partners BMSM select [13], their sexual roles (as insertive or receptive partner in anal intercourse) in sexual encounters [14, 15], and perceptions of partners’ HIV risk [16, 17]. One qualitative study of sexual risk behavior in BMSM found a preference for partners perceived as masculine [13]. Moreover, being the receptive partner (referred to colloquially as the “bottom”) was seen as feminine, while being the insertive partner (or “top”) was seen as masculine. There were exceptions to this dichotomy particularly for those men who engaged in both receptive and insertive anal intercourse (referred to colloquially as “versatile”). It was acceptable (i.e., their manhood was not threatened) for these men to engage in receptive intercourse with men who were masculine or if there was a strong emotional connection. However, many reported being uncomfortable or threatened bottoming for a man perceived as effeminate [13]. In another study, BMSM described maintaining their manhood and heterosexual identity by being a top, i.e., by only engaging in sexual behaviors they perceived as masculine, while “gay” men assumed the feminine role in same-sex coupling [14]. Wright, in his study of Black male sexual behavior, reports concurrent sex with women as another means of maintaining masculinity used by the men in his study [15]. BMSM have also been found to describe masculine men as men who appeared heterosexual or who were “straight-seeming” men [18].

Among BMSM, socio-cultural conceptions of gender and masculinity may inform HIV prevention heuristics, which are relatively automatic decision-making rules used to determine which HIV prevention strategies are appropriate for a given partner [19]. Across several studies, a commonly reported prevention heuristic employed by MSM is the “known/trusted partner is a safe partner” heuristic, where men feel they can safely engage in unprotected sex with a well-known partner [19–22]. Prior research suggests heuristics are also used with unfamiliar partners to determine their HIV risk and the need for safer sex behavior based on social or physical attractiveness [19, 23]. Given the apparent social desirability of masculinity, coupled with assumptions about sexual behavior and identity of men perceived as masculine, masculinity may be one such characteristic used by BMSM to assess risk and inform condom decision-making.

Studies exploring the potential impact of socio-cultural conceptions of gender and masculinity on HIV risk beyond adult populations, namely among Black adolescent and young adult populations have been limited to heterosexual youth [24, 25]. There has been little exploration of upstream contextual factors among YBMSM who are at arguably greater risk for HIV transmission. Therefore, we sought to understand upstream factors preceding partner selection specifically among young BMSM, with a focus on socio-cultural conceptions of gender and masculinity, as well as the HIV prevention heuristics arising from these upstream factors. Toward this end, we qualitatively explored how YBMSM assess risk, choose partners, and make decisions about condom use; we also explored contextual factors (including racial, sociocultural, and psychosocial factors) that may influence these decision processes.

METHODS

Study Population and Design

We conducted secondary analysis of data collected in three studies of BMSM using semi-structured interviews. The first was conducted in New York City, Buffalo, and Rochester, NY in 2001 with 19 Black men of aged 18–24 who reported same-sex behavior; it covered contextual factors influencing HIV risk behavior. The second and third studies were in Atlanta in 2003 and 2006 with 59 men aged 18 and older with a history of same-sex behavior. The former focused on examining masculinity and its relationship to belief systems, social/sexual networks, sexual behavior and condom use; the latter examined socio-cultural determinants of HIV risk and protective behaviors. For our analysis, we selected the 35 participants 18–24-year-olds (mean=20; SD=7) who identified as Black/African-American and reported same-sex behavior (New York study, n=17; Atlanta studies, n=18). See Table I. The studies were combined to conduct an amplified supplementary analysis – a specific type of secondary analysis where qualitative data are pooled and reanalyzed using research questions that extend the primary studies’ questions [26]. These studies were selected because they were well-suited for our analysis. Each focused on psychosocial factors influencing HIV risk in BMSM and topics of interest to us (e.g. masculinity and sexual risk behavior) emerged as important themes in them.

Table I.

Demographic Characteristics of Study Participants (N = 35)

Characteristic	N	%a
City of Interview/Current Residence
New York Cityb	8	23
Rochester	4	11
Buffalo	5	14
Atlanta	18	51
Race/Ethnicityc
African American (Black American)	16	84
Caribbean	1	5
African	1	5
Sexual Identityd
Gay/Homosexual	17	49
Bisexual	16	46
Heterosexual/Straight	1	3
Other	5	14
HIV Serostatus
Positive	2	6
Negative	30	86
Unknown	3	9
Income
<15,000	15	43
15,000–20,000	5	14
20,000–30,000	3	9
30,000–45,000	2	6
45,000–60,000	-	-
>60,000	1	3
No Answer	9	26
Education
Some High School	4	11
High School Diploma/GED	13	37
Some College	13	37
College Degree	3	9
Graduate Degree	1	3
No Answer	1	3
Employment Status
Employed	18	51
Unemployed	17	49

^aUnless otherwise specified, percentages do not add to 100% because of rounding

^bStudies in New York City were conducted in Manhattan and Brooklyn.

^cRace and ethnicity data were not collected in the Atlanta Studies.

^dPercentages do not add to 100% because some participants used more than one term to describe their sexual orientation in the New York study. Additionally, homosexual and gay were listed separately on the questionnaire because of stigma associated with “gay” identity.

Participant recruitment and data collection

All 3 studies used convenience sampling. The Atlanta studies employed internet recruitment via an America On-Line BMSM chat room, intercept recruitment at a local park, and snowball sampling (described in detail elsewhere [13]). The 2006 Atlanta study also employed venue-based sampling from community-based organizations (CBO), bars/clubs and bookstores, and passive recruitment through advertisements in a local Black newspaper. At each venue, project staff distributed postcards with study information. In New York, participants were selected through venue-based sampling at CBOs serving YBMSM. Participants were recruited by CBO staff using their existing client outreach strategies. Combining these datasets from convenience samples collected using different recruiting strategies created a more diverse overall participant sample.

In each study, a BMSM interviewer conducted 1–2 hour semi-structured interviews after participants completed a brief socio-demographic survey. Participants received a $25 incentive and transportation reimbursement. All interviews were audio-recorded and transcribed verbatim.

Primary data collection procedures were approved by the institutional review boards of Columbia University and Emory University. Participants provided informed consent prior to completing interviews. The present secondary data analysis was deemed “not human subjects research” by the Johns Hopkins Bloomberg School of Public Health Committee on Human Research.

Qualitative Interview Protocol

The qualitative interview protocols were developed based on prior literature and investigators’ previous work with the study population [8–13]. Although interview instruments were not identical across studies, each focused on the social context of sexual risk among BMSM and included questions about family, race, religion, masculinity, sexual identity, homosexuality, and sexual risk behavior. Questions most relevant to the present analysis appear in Table II.

Table II.

Relevant Interview Protocol Questions

Concept	2001 New York study	2003 Atlanta study	2006 Atlanta study
Masculinity	What does masculinity mean to you?	Tell me what it was like growing up as a young black man.	What do you think it means to be a man?
	What are your family’s expectations of masculinity for you?	What are your family’s definitions of masculinity? Familial expectations of you? Importance of your masculinity/sexuality growing up?	What does it mean to be masculine? What life experiences told you this? Do you consider yourself masculine? Why or why not?
	How does the Black community define masculinity? Is masculinity different for white men?
	How does definitions/expectations masculinity affect sexual behavior(partner selection? sex roles?)	Do you consider yourself masculine? Why or why not? Effect on self-esteem? Effect on sexual behavior? Effect on choice of partners?	How big a role does masculinity play with potential romantic and sexual partners?
Homosexuality	What are your attitudes about sex between men? Your family’s attitude? Attitudes of the black community?	Tell me what it was like growing up as a young black man. Discussions on sex/sexuality in general? Homosexuality?	What family and community messages about homosexuality did you receive growing up?
Sexual Identity	How do you identify yourself sexually?	How would you describe your sexual identity to a stranger?	What word(s) do you use to describe your sexual identity?
	How open are you and people you know about their attraction to men?	How involved are you with the “gay” community?	What does it mean to you to “come out”? Your experience?
		How did your upbringing influence your sexual identity?	What is the “gay” community to you? “Gay” lifestyle? How involved are you?
Family	Where do you find strength or support? Family? Friends?	What’s your relationship with father? Mother?	Where did you grow up? What was your childhood like?
		What are you family’s expectations of you?	What is your relationship with your family like?
Religion	Where does the Black Church fit in?	Tell me what it was like growing up as a young black man.	What is spirituality (or being spiritual) to you? Religious influences growing up?
	Do the views of the church affect how you see yourself?	What you’re your religious/spiritual influences?
Race	Are black MSM different from other MSM?	What does it mean to you to be a black man?	If you had to explain to your son what it meant to be a Black man in America, what would you tell him?
	What issues arise from being both black and attracted to men?	What’s more important to you – your race or sexuality? Why?	What is the Black community to you?
	Are you a part of the Black community	How has being a black man influenced your sexual identity?	What makes being Black and attracted to other men different from men of other races/cultures? Are there different communities/lifestyles?
Sexual Risk Behavior	What are the reasons that you or people you know have unprotected sex (esp. anal sex) or have unprotected sex with multiple partners?	Tell me about a situation when you had anal sex with another man without a condom	What is your definition of “risky” sexual behavior? With what you know about “risky” sex, what has your risk for HIV been in the past 6 months?
	What does risky sex mean to you?		What qualities of a potential sexual partner make you feel they are more or less at risk for HIV?

Qualitative analysis

Using Atlas.ti [27], a qualitative software package, we conducted categorical analysis to explore factors that may influence sexual risk. We used a 3-stage analytic coding strategy that included open coding, axial coding, and selective coding [28–30]. Transcripts were fractured into discrete segments, which were sorted into categories, thus facilitating a comparative examination across participants of how racial, psychosocial and sociocultural contextual factors may influence HIV risk in the lives of YBMSM.

Codes related to contextual factors were identified through open coding and data immersion. Through axial coding, a 3-tiered coding hierarchy was developed based on themes that emerged during the open coding process. We developed a final codebook with codes organized according to this hierarchy and applied it to the data for the final round of coding. During the final round, one rater (the first author) coded all transcripts, and a second rater double-coded 20% of the transcripts. Cohen’s Kappa [31], used to test inter-rater reliability, indicated excellent consistency between coders according to conventions on the use of Kappa for inter-rater agreement [32]. The Kappa value was 0.86 for codes related to masculinity and codes related to risk perceptions; all p-values were less than 0.05.

For selective coding, the final stage of analytic coding, we focused on developing a theoretical framework through identification of a core category/theme that systematically related to all previously identified codes. This was accomplished through “story-telling memos” [29] written throughout the 2 prior stages of the coding process that contained emerging research questions, speculations about participants’ variations in response to or in expression of phenomena, comparisons to theory, recurring themes, and possible relationships among major categories.

RESULTS

Perceptions of masculinity emerged as a primary contextual factor influencing risk assessment, partner selection, and condom decision-making. Direct quotations appear in Table III; participants are identified with italicized pseudonyms in the text and table.

Table III.

Themes and Quotes from Participant Interviews

Theme	Speaker	Representative Quote
Risk Assessment
Known/trusted partners are safe partners” heuristic	Jordan	(1) Where it’s somebody that I’ve been having sex with on a regular basis, and I wanna say some kind of corny heat of the moment line but it’s not really that… you kind of build up, you feel comfortable with this person after it’s been so long and you just kinda stop, and that’s how it’s kinda been with me when I haven’t used a condom.
“Masculine partner is a safe partner” heuristic	Stacey	(2) I can go and mess wit’ this boy that’s a boy-boy-boy that got a girlfriend and two babies, you know what I’m sayin’, and he’s not gonna transmit nothin’ to me because he ain’t got no AIDS… he don’t mess around.
	Preston	(3) Researcher: do you think more masculine men are HIV positive or feminine men? Feminine…I just feel like they just do a whole lot.
	Carlton	(4) If I walked up to somebody and they had a small voice, and they was just real flamboyant, that’s not what I’m looking for either because I feel that you’ve been wrapped up in the lifestyle and you’ve been out there exposing yourself to things that you shouldn’t be.
“Tops are low risk, bottoms are high risk” heuristic	Shawn	(5) If you’re performing anal sex and you’re a top, the statistics for catching HIV and AIDS is somewhere like 40–60% and as far as being on the bottom… the statistics are higher, you know - real high. So those are the basic differences but I mean you still can catch it, but there are just differences from being on top or the bottom.
	Anthony	(6) I think bottoms; they just like sex (laughter). They don’t care what you do to them; they just want it … I got one cat…and it’s like we met each other and…we was foolin’ around and I was waiting for the condom word to pop out, and it didn’t…it’s my first time seeing you, it’s your first time seeing me, and he ain’t say…I mean he really didn’t say anything til I had to… and that was it but I think bottoms really wanna get it; they don’t care.
Partner Selection
Social desirability of masculine partner	Adam	(7) I would have to say that I am more attracted to masculine. Feminine would just turn me off. That would completely turn me off…I know what I want, how I want it and when I want it and where I want it. I know exactly what is the definition of the type of person I would like to meet. Anything that is feminine, it is just unacceptable and it is not perfect to me.
Choosing masculine partners to help steady one’s personal sense of masculinity	Shawn	(8) I mean I like to get fucked, too, but it has to be like masculine guys! It cannot be someone who’s really fem (clapping hands)! …It has to be another masculine brother so it’s fine, yeah, I’ll accept that, and I’m sure everybody else will accept that, but they won’t accept a feminine guy (tapping table) getting’ on their back. No, I don’t see that. Researcher: Would you feel less of a man if a feminine guy fucked you? Yes, I would!
	Stephen	(9) I’m not so eager for a more feminine man to penetrate me than I would be for a guy who is masculine…. It just doesn’t seem right to me… when, you know, a guy who’s feminine penetrates you, um, I guess penetration as being a, you know, a manly type of thing, that’s what the man is supposed to do as the man.
Choosing masculine partners to help camouflage same-sex behavior	Stephen	(10) I prefer masculine dudes…cause I want to be able to take ‘em around my mom, [and my] mom not know. You know, as far as me taking me, you know, an effeminate person. You know what I’m saying, my mom, you know, my mom and she like, yeah, you fucking with him, you know what I’m saying.
Condom decision-making
More masculine partner controls the sex act and condom decision- making	Christopher	(11) Well one factor for me is if my partner does not want to use a condom, then that may make me like okay partner, so don't use it, or if my partner was like, really like, yeah, I want you to swallow, yeah, I want this, you know so that's one factor.
	Christopher	(12) If I’m with a guy who I think is more masculine than I am, then I’m gonna let him do whatever he wants. Yo I’m gonna be like yo, do you. You know, I’m gonna let him do anything you know.
Insertive partner determines condom use	Deven	(13) When it came to like guys, before I was so educated on stuff and it was just like, it was like, sort of like somewhat versatile or whatever; it’s like if it was some of my partners or whatever, I’d be like, alright, do whatever you want.
	Kasey	(14) Researcher: Have you ever had unprotected sex when you were on the top? No. Researcher: So it’s always been when you were on the bottom? Uh-huh. Researcher: Why? [I was] afraid to talk about it or to say that we need to use condoms.
Older partners determines condom use	Alexander	(15) He didn’t want to use a condom. That’s what he told me…I just went along wit’ it ‘cause I was young, so and even being young I should have knew, but he was older and I was like, well, if it makes you feel better, I’ll just do it.
	Alexander	(16) It happens like being a bottom…I mean you have some that’s the same age- top and bottom be the same age - but like the majority of it just be the bottoms are youngest and the tops be like the oldest.

Risk Assessment

Participants described several assumptions and rationalizations about the HIV risk potential of sexual partners that informed condom decision-making and/or partner selection. These included prevention heuristics based on how well they knew or trusted their partner (“known/trusted partners are safe partners” heuristic); assumptions about their partner’s HIV risk in relation to his perceived degree of masculinity (“masculine partners are safe partners” heuristic) and/or his sexual role (“tops are low risk, bottoms are high risk” heuristic); and their partner’s appearance of physical health (“looks healthy, looks clean” heuristic).

Participants often described unprotected sex with people they knew or trusted as low risk (Quote 1), and used a variety of rationalizations and assumptions under this heuristic. For example, participants noted that partners can be trusted to be safe if they are known for a long time (and thus their sexual history and serostatus is presumed known); that a trusted partner would want to keep the participant safe; and that an assumed monogamous partner is safe. Jordan’s comment (Quote 1) underlies a commonly expressed rationalization: if one knows his partner well, and trusts him, then condom need not be a priority.

A potential partner’s perceived degree of masculinity was also used to assess HIV risk. While some made no connection between masculinity/femininity and HIV risk, there were others who assumed that men they perceived as masculine were less likely to be or to become HIV infected - the “masculine partner is a safe partner” heuristic. Men perceived as masculine were typically described as not being openly gay, lacking any feminine characteristics or tendencies, being strong or aggressive, or being the insertive partner in anal intercourse. Participants also described a perception that homosexuality was associated with femininity, so men with the aesthetic (e.g. straight-acting) or interpersonal trappings of heterosexuality (e.g., girlfriend, wife, children) were also seen as masculine. Stacey’s comment (see Quote 2) that “he don’t mess around” reflected a perspective shared by some participants that masculine men had minimal involvement with the Black gay culture and lifestyle and were thought, therefore, to have limited exposure to HIV positive partners. The opposite was assumed of effeminate men (Quotes 3 and 4).

Participants’ perceptions of masculinity and roles in anal sex were described in binary terms, with masculine men thought to be the insertive partner (top), and effeminate men thought to be the receptive partner (bottom). Notions of versatility were rarely discussed but one young man felt “regardless of how manly [they were]” he could always tell if someone were versatile because “they’re a little bit more feminine.” These assumptions informed the “tops are low risk, bottoms are high risk” heuristic; however, it was also based in part on beliefs about the lower HIV risk for insertive versus receptive anal intercourse (Quote 5). Moreover, many respondents felt that bottoms were also high risk because they were less proactive about condom use and engaged in anal intercourse with multiple partners (Quote 6).

Partner Selection

Masculinity was not only used to gauge perceived HIV risk, many participants described as a socially desirable characteristic in one’s partner. Most participants described a greater, often exclusive, preference for men who were perceived as masculine (Quote 7). This was true of both those participants who saw themselves as masculine and those who saw themselves as effeminate. As Adam describes, this preference was sometimes a matter of greater physical attraction toward perceived masculine men; however, several other reasons for the preeminence of a partner’s masculinity emerged, including the need to steady one’s personal sense of masculinity and/or camouflage one’s same-sex behavior.

Several participants described partnering exclusively with men they perceived as masculine as a way to maintain their own sense of masculinity in the context of their homosexuality. These participants thought of themselves as masculine and expressed discomfort with the idea of allowing a feminine man to be the insertive partner in sexual encounters. One participant felt that he would feel like less of a man if he allowed an effeminate man to top him (Quote 8). For some, being the receptive partner (or bottom) was reserved for men whom they perceived to be as masculine as or more masculine than them (Quote 9).

Partnering with a masculine man also helped to maintain nondisclosure of sexual behavior or identity. One participant said he did not date effeminate men because he could not introduce them to his mother as a friend without her suspecting they were sexually involved (Quote 10).

Condom Decision-Making

Many participants commented that condom use was not often discussed with partners, especially when there was a power differential. They described a phenomenon in which partners of lower status (based on a lower degree of masculinity, younger age, or being the receptive partner during anal intercourse) submitted to their partner’s desire to not use a condom and were unable or unwilling to assert their own desires regarding condom use. Christopher described his tendency to let his partner make all the decisions about sex (Quote 11). He explained that he often relinquished control to partners he believed were more masculine than him – allowing them to decide whether they would practice safe sex or what type of sex they would engage in (Quote 12).

In such situations, Christopher was the receptive partner. Allowing the insertive partner to control the sexual act and condom decision-making was a common theme. Several interviewees commented that as receptive partners they were often not involved in condom use decisions (Quote 13). Another individual reported always using condoms when he was the insertive partner, but deferring to his partner when he was the receptive partner (Quote 14).

Many participants also described this power differential when engaging in sex with older partners, where they felt disempowered regarding condom decision-making. Alexander describes his first same-sex experience where he was the receptive partner; the insertive partner who was 10 years older did not want to use a condom (Quote 15). Alexander went on to comment that such an age difference was fairly common (Quote 16). Several other participants also described this phenomenon as well as initiation of same-sex behavior with an older partner. These young men often described receptive anal intercourse while in their teens with a partner in his twenties or thirties. Some used condoms during their first sexual experience whereas others acquiesced to their partner’s decision not to use them. The vulnerability and uncertainty of younger less sexually experienced MSM placed them at a disadvantage relative to older men.

DISCUSSION

Perceptions of masculinity emerged as a core theme influencing partner selection, risk assessment, and condom decision-making. These findings are consistent with findings from studies of adult BMSM, in which men also described preferring partners perceived as masculine [15, 18, 33]. In addition, these data reveal two concepts not found in the adult BMSM literature: the degree of perceived masculinity of a potential partner may be used as a gauge to assess HIV risk, and a power differential based on relative masculinity of partners that may influence condom decision-making. The influence of masculinity on risk assessment and condom decision-making suggests that masculinity is a significant social factor affecting sexual risk behavior. Further, participants’ assumptions about masculinity and homosexuality (e.g., masculine men control the sex act, masculine men are tops, masculine men are less involved with the gay lifestyle, masculine men are less likely to be HIV seropositive) seem to indicate a tension between homosexuality and masculinity that may also be important for understanding sexual risk behavior among YBMSM. Some of the study participants may have experienced masculine socialization where homosexuality conflicted with masculine gender role expectations, resulting in their current efforts to affirm and maintain their masculinity in the context of their homosexuality.

Our findings suggest the importance of examining Black male sexuality and HIV sexual risk behavior in the context of racial and cultural conceptions of masculinity that may influence partner selection and condom decision-making. As in some previous studies [15, 18, 33], most participants described a greater, often exclusive, preference for partners perceived as masculine. Moreover, many participants steadied their own masculinity by partnering exclusively with men they perceived as masculine. This finding was particularly pronounced with regard to sexual roles where many of the participants were only willing to be the receptive partner with men whom they perceived to be as masculine as or more masculine than themselves.

Preference for masculine partners among YBMSM has potential ramifications for HIV risk. The strict attention paid to the perceived masculinity of a potential sexual partner, i.e., the tendency described by participants to only allow those men perceived as masculine to be the insertive partner, seems to reflect an underlying power dynamic involved in anal intercourse. In a study of gay couples, Kippax and Smith [34] found that masculinity was often expressed through domination and submission during anal intercourse. Similar patterns seem to play out for the young men in this study, many of whom were only willing to be receptive in anal intercourse with a man they perceived as masculine, so as to not threaten their masculinity. According to this narrative, allowing an effeminate man to dominate sexually would be emasculating. For some, this power dynamic extended to condom decision-making such that partners perceived to be more masculine could control what sexual activity they would engage in and whether they used condoms. Moreover, the partner perceived to be more masculine was often the insertive partner, a position that held relatively less risk of HIV transmission than a receptive partner during UAI. This creates a situation where the partner who is ostensibly at greatest risk of HIV transmission is the least empowered with respect to condom decision making. As described in the results, this power dynamic also existed between younger and older partners, where the older partner, again often the insertive partner, controlled the sex act and condom use. This may provide some insight as to why having an older partner within one’s sexual networks may contribute to increased risk for HIV transmission among YBMSM [6].

Potential HIV risks were often compounded by the prevention heuristics assigned to men perceived as masculine versus those perceived as effeminate. Similar to the “implicit personality theories” and “characteristic based theories” described in prior research [19], the “masculine partner is a safe partner” heuristic was predicated on the assumption that someone who presented a socially desirable image was less likely to be HIV positive. The data suggest that masculinity was revered and desirable. However social desirability was not the only factor that informed participants’ association of a masculine aesthetic with low HIV risk. This heuristic was based on the presumption that masculine men were inherently involved in less risk behavior. Femininity in a male partner was presumptively associated with receptive anal intercourse, promiscuity, various other gay stereotypes, and HIV. Men who were perceived to be feminine were seen as high risk, whereas masculine men – at times presumptively associated with heterosexual identity, insertive anal intercourse, or limited exposure to the gay community – were perceived as low risk.

The high social desirability of masculine partners, prevention heuristics that assign low risk potential to masculine “tops,” the power dynamics that influence condom decision-making, and the greater ease of HIV transmission from an insertive partner to a receptive partner – may all coalesce to create considerable potential risk for HIV transmission. Although there is no evidence that masculine, “straight acting” “tops” are more or less likely to be infected than any other subgroup of BMSM, the assumption that these men are uniformly low risk can be dangerous in the context of partner selection and prevention heuristics.

Limitations

The study has several limitations. Sampling strategies varied across sites, and may have involved different selection biases such that the samples between the two locales are not entirely comparable. However, drawing on data from three studies potentially provided a richer and more diverse group of respondents than any single study. Although the sample was fairly diverse, we were limited by the size and composition of the sample. For instance, the sample was not large enough to compare themes across different subgroups of the population; nor were various subgroups adequately represented (e.g. younger adolescents, more non-gay identified youth, etc.). This is also a limitation of semi-structured interviews. Focus groups of different subpopulations would have made comparisons more readily available, but the depth of the data, a significant strength of this study would have been compromised.

The general limitations of qualitative research also apply. Our findings provide important insights into this population, but they are not generalizable to all YBMSM. The goal here was to help elucidate quantitative findings of similar populations of YBMSM, particularly those findings that suggest partner characteristics influence HIV risk [6, 35]. Our findings may also help generate additional hypotheses about HIV risk in this population that can be tested quantitatively and generalized to the appropriate YBMSM population.

CONCLUSIONS

Many current HIV prevention research efforts focus on the individual-level risk behaviors that expose individuals to infection. This study identifies the relevance of social determinants and the social context of these risk behaviors. The pressure that YBMSM feel to prove their own masculinity and to measure up to community and personal expectations of manliness, regardless of – or perhaps in spite of –their sexual desires may increase their risk for HIV infection and is, therefore, worthy of further investigation. Additional qualitative studies are needed to further characterize the conflict between masculinity and homosexuality and how this conflict might influence partner selection, risk assessment and overall HIV risk. Future research might also employ focus groups and theoretic sampling of specific subgroups of YBMSM as well as other racial/ethnic groups of young MSM to determine if the findings from this study vary across these groups.

The construction and composition of sex networks among YBMSM would also benefit from further exploration. Analysis of sexual networks may further elucidate masculinity perceptions that affect partner selection and condom decision-making. Similarly, in-depth sexual network analyses are needed to consider the prevalence of age discordance between sex partners, how such discordance might affect HIV risk, and motivations for coupling with widely age discordant partners.

In addition to directing future research, these findings can also begin to inform HIV prevention strategies among YBMSM. Individual level interventions should target beliefs about masculinity, power dynamics in sexual coupling, and the development of HIV prevention heuristics. Community level interventions are also needed to target the aspects of the masculine socialization process that may make these young men more vulnerable to HIV.