The limitations of ‘Black MSM’ as a category: Why gender, sexuality, and desire still matter for social and biomedical HIV prevention methods

Abstract

The USA faces disproportionate and increasing HIV incidence rates among Black men who have sex with men (BMSM). New biomedical technologies such as pre-exposure prophylaxis (PrEP) have been developed to address their HIV risk. Very little consideration, however, has been given to the diversity obscured by ‘BMSM’ as a category, to how this diversity relates to men’s sexual partnering strategies, or to the relevance of these issues for new HIV prevention methods. We conducted a community-based ethnography from June 2013 to May 2014 documenting factors that affect the acceptance of and adherence to PrEP among BMSM. We conducted in-depth interviews with 31 BMSM and 17 community stakeholders, and participant observation. To demonstrate the diversity of social identities, we present a taxonomy of indigenous categories organised along the axes of sexual identity, sexual positioning, and gender performance. We analyse how HIV prevention strategies, such as PrEP, may be more effective if programs consider how gender, sexuality, and sexual desire shape sexual partnering strategies. This article underlines the importance of attending to the diversity of sexual and social subjectivities among BMSM, of bringing the study of sexuality back into HIV prevention, and of integrating biomedical prevention approaches into community-based programs.

Introduction

In the context of rising HIV incidence rates among non-heterosexual and racial-ethnic minorities in the United States, HIV research and intervention efforts frequently employ the epidemiological category Black men who h*ave sex with men (BMSM), without giving due consideration to how this category obscures diversity. This article highlights the importance for rolling out new HIV prevention methods of addressing (1) diversity in gender, sexuality, and desire among BMSM, and (2) power inequalities that operate within sexual networks in this category. We begin by describing the historical creation of the category of BMSM and by discussing its key limitations. We then draw on our community-based ethnographic research to present a taxonomy of indigenous cultural (‘emic’) categories organised along the axes of sexual identity, preferred sexual position and gender performance. The emphasis on social processes that contextualise sexual partnerships provides important insights for understanding the HIV epidemic in many places throughout the world where sexuality and gender performance shape HIV vulnerability at the same time that they are fluid and negotiable.

In the US as in many parts of the world, HIV prevention began through safer sex education efforts that were both initiated and disseminated at the grassroots level (Parker, 1996; Patton, 1996; Watney, 1990); early prevention efforts were heavily grounded in local expertise about community organisation and sexual cultures (Parker, 1996; Patton, 1996; Watney, 1990). In the 1980s, amid concern that non-heterosexual men who did not identify with the gay community were not engaging with prevention efforts, grassroots community groups turned their attention to homosexually-active other men who have sex with men (Aggleton & Parker, 2015; Patton, 1996). In the 1990s, epidemiologists appropriated the category ‘MSM.’ Although the category was originally created to refer to non-gay identified men, its public health appropriation transformed MSM into a catch-all category, lumping together all men who have sex with men, independent of social identities or sociocultural and community ties (Boellstorff, 2011; Kippax, Stephenson, Parker, & Aggleton, 2013; Young & Meyer, 2005). The remaking of this category into ‘MSM’ erased the initial concern with social identities and obscured the social organisation of sexuality (Altman et al., 2012).

The ways public health research and practice have attempted to include MSM who do not identify as gay or bisexual into this category have sometimes failed to address the range of variation in how stigma and discrimination affect those aggregated within it (Altman et al., 2012; Young & Meyer, 2005). Furthermore, the category itself became a source of stigma (Malebranche, Arriola, Jenkins, Dauria, & Patel, 2010), as both popular culture and research literature hold ‘hard-to-reach’ MSM responsible for HIV transmission to heterosexual populations. Researchers have argued that the stigma generated by the discussion about men on the ‘down low’ (Ford, Whetten, Hall, Kaufman, & Thrasher, 2007; Malebranche, 2008; Saleh & Operario, 2009) may make the men to whom it refers even more vulnerable to HIV, by rendering them less willing to participate in research and interventions.

As the epidemic in the US shifted toward socioeconomically disenfranchised racial-ethnic minority men, researchers and policymakers generated further subcategories, such as Black MSM (BMSM), that aimed to capture health disparities associated with race and ethnicity. However, this continued to erase potentially important differences between diverse kinds of Black MSM (gay-identified or not, behaviourally bisexual, gender variant, etc.). There are, to be sure, vital reasons to focus on the undeniable epidemiological trends towards increased and disproportionate infection among non-heterosexual and racial-ethnic minorities. While BMSM represented only 2% of the US population from 2008–2010, they accounted for nearly 75% of new HIV infections (CDC, 2012). A recent study of BMSM in six US cities reported an HIV prevalence of 21% (Koblin, 2012). Our fieldsite, New York City, has one of the highest HIV infection rates in the US. In 2011, 66.1% of all new HIV diagnoses in NYC were among Black men, of whom 56.2% were MSM (NYCDOHMH, 2012). Compared to other racial-ethnic groups, HIV-infected BMSM are less likely to know their HIV status (Centers for Disease Control and Prevention (CDC), 2005; MacKellar et al., 2005; Millett, Peterson, Wolitski, & Stall, 2006).

A significant body of literature underlines contextual factors (i.e. structural, social and cultural) as key drivers of the epidemic (Maulsby et al., 2014; Millett et al., 2012). However, most medical and public health responses to increasing incidence and barriers to linkage to care have de-emphasised the sociocultural and economic drivers of HIV vulnerability, and instead focused on more proximal biomedical prevention approaches (Aggleton & Parker, 2015). In addition to structural, social and cultural issues, some researchers have argued that the response to HIV itself, through its erasure of meaningful social differences—e.g., men’s notions of ‘selfhood,’ community affiliation, and experiences of stigma— has perhaps itself contributed to the rising rates of HIV among BMSM (Altman et al., 2012; Arnold, Rebchook, & Kegeles, 2014; Boellstorff, 2011; Garcia et al., 2015).

Stigma is a critical element of this broader social context that shapes the disproportionate rates of HIV infection within the epidemiological category of BMSM. Stigma heightens HIV vulnerability by creating relationships of unequal power among BMSM (Jones, Wilton, Millett, & Johnson, 2010; Parker & Aggleton, 2003; Wilton, 2009), challenging engagement with healthcare (Eaton et al., 2015) and contributing to sexual risk behaviours (Radcliffe et al., 2010). The consequences of stigma for the health of Black MSM are particularly severe because Black MSM experience multiple forms of stigma along the intersecting social axes of marginalisation, including sexuality, race, gender, and class (Bowleg, 2013; Earnshaw, Bogart, Dovidio, & Williams, 2013) and HIV status (Arnold et al., 2014; Radcliffe et al., 2010). Although ethnographic research has begun to reveal considerable sexual diversity (Miller, Serner, & Wagner, 2005) as well as gendered diversity (Fields et al., 2014) among BMSM, there remains an urgent need for empirical research documenting how this heterogeneity in social subjectivities may affect the acceptance and uptake of emerging HIV prevention methods.

This article (1) identifies meaningful subcategories that we found within the BMSM category to note the diversity in ‘local’ or ‘emic’ organisation of social and sexual experience; (2) problematises local categorisations, noting the fluidity that exists in navigating sexual relationships and gender expectations; and (3) describes how recognising diversity and fluidity can potentially shape and inform the use of the expanding HIV prevention tool kit, especially with the introduction of PrEP and other emerging prevention technologies.

Methods

A community-based ethnography conducted in New York City from June 2013 to May 2014 examined how these intersectional identities and strategies may affect engagement with various methods of HIV prevention, including PrEP. The goal of the study was to describe and analyse the connections between the sociocultural and economic context of BMSM’s lives, including but not limited to their vulnerability to HIV, and their desire for, acceptance of, and adherence to PrEP.

Selection of participants

We recruited 31 BMSM for in-depth interviews through recruitment cards that provided brief information about the project and a phone number. The cards described the study as focused on the life stories of Black men ages 15 and older, and had the name of the ethnographic component of the study: ‘Your Life, Your Words,’ but did not mention sexual identity/behaviour or HIV. Cards were available at health centres, community-based organisations, bars, and online. Eligibility criteria included (1) being born as and identifying as biologically male, (2) identifying as Black or African American, (3) having had anal or oral sex with a man in the past 12 months, (4) being 15 years or older. The average age in our sample was 29.0 years (SD, 12.3), with 17 men between ages 15–24 and 16 men ages 25–54. The majority self-reported as HIV-negative (N=23), while a few did not know their status (N=3) or were HIV-positive (N=5). We used participant observation and the assistance of our Community Advisory Board to identify community stakeholders who were involved in service provision for BMSM in New York City. Twenty-one potential key informants were contacted by phone/e-mail and invited to participate. Four did not respond to those messages; the 17 who responded agreed to participate. In accordance with the protocol approved by the Columbia University Medical Center Institutional Review Board, the study was granted a waiver of parental consent to protect the sexual identity of men under 18. All participants were read a verbal informed consent form. Because participants were asked sensitive questions that could potentially reveal information about sex work and drug use, verbal consent provided a higher level of confidentiality than written consent. Participants were made aware that a Certificate of Confidentiality granted by the National Institute of Mental Health protected data collected. Their understanding was assessed with a series of follow-up questions about the content of the consent form. This assessment of understanding was audio-recorded and data collection began following participants’ recorded declaration of verbal consent.

Procedures

Interview guides for community stakeholder and in-depth interviews were developed by the research team and then discussed with and validated by our Community Advisory Board. In-depth interview guides were divided into three sessions as described in Table 1.

Table 1.

Methods.

Data collection method	Specific Data elicited	Sample Description
Participant Observation	Organizational behaviour, group composition, ways people discuss sexuality, race, gender, age, class;/interaction, spoken rules of conduct and implicit cultural norms expressed, enforced, followed and navigated.	11 months of observation in: private spaces (homes, parties); public spaces (parks, streets, events); virtual spaces (chat rooms, blogs, social media); and institutions (community organisations, health centres, religious institutions).
Key Informant Interviews	Organisational mission; role in organisation/group/community; knowledge and attitudes about black MSM, HIV vulnerability, institutions and networks available to Black MSM; views on PrEP and other for HIV/STI prevention	17 informants, including 2 physicians; 3 mental health providers; 4 community organisation program administrators; 5 outreach workers; 3 community mobilisers
In-depth Interviews	Session 1: History of family relations, coming of age, education, housing, making money, friends; community, recreation Session 2: Sexual history, including desire, casual and steady relations, sexual identity and racial identity Session 3: Perceptions of health and risk; practices and attitudes about medications and seeking health services; knowledge and attitudes about HIV prevention	31 participants

In-depth interviews (3 sessions, each roughly 90 minutes) and community stakeholder interviews (1 session, 60 minutes) were conducted face-to-face. Participant observation took place twice per week over the course of 11 months in spaces recommended by our Community Advisory Board or that emerged in interviews. Participant observation consisted of “community observation” and “short term observation” with in-depth interviewees. In community observation, we participated in weekly community events and public forums (e.g., weekly happy hour at Black LGBT groups and mobilisation discussions at community-based organisations). In “short term observation,” the lead ethnographer accompanied five in-depth interviewees to places where the in-depth interviewees agreed to invite the ethnographer (e.g., private house parties, gay family dinners, police precincts, public cruising in parks among other situations). Participant observation was recorded in jottings, fieldnotes, maps, and analytic memos. The lead ethnographer’s (first author) own subjectivities and sensibilities as a gay Latino man may have facilitated ‘hanging out’ with other men who were also racial-ethnic and sexual minorities.

Data analysis

Interviews were digitally recorded, transcribed verbatim and entered into Atlas.ti 7.0, the qualitative analysis software used for coding and organising qualitative and visual data. Data were analysed using an approach that drew on a range of ethnographic and qualitative research strategies, including aspects of extended-case method (Burawoy, 1998) in relation to the interrogation and extension of existing theoretical frameworks in relation to HIV among MSM, as well as aspects of grounded theory (Corbin & Strauss, 2008), which helped to problematise the ways in which existing research has theorised the question of MSM. In particular, we sought to extend this methodological point of departure by emphasising aspects of an interpretive approach to cultural and linguistic meanings (Geertz, 1973), and to highlight the analysis of sexual categories and classifications (Parker, Herdt, & Carballo, 1991) that were found to operate social interaction and sexual practice on the part of the BMSM community where our research was carried out. This emphasis on the cultural and linguistic categories that organise the lived experience of sexuality on the part of different communities has long been used in HIV research, especially in cross-cultural studies of sexual practices in different social contexts, but has less frequently been employed in the investigation of key populations and communities in the USA. This hybrid approach allowed for theory modification, based on narrative and observation data, which problematised the way that existing work has theorised the category of MSM and drew on new insights from empirical data (men’s sexual and gender presentation and lived experience). The use of multiple methods allowed for data triangulation during the analysis; this involved comparing the perspectives of community stakeholders with BMSM, as well as with observed behaviour. We presented data analysis reports to the clients and providers in CABs to validate salient themes and perform member checking. These analyses employed a codebook that was developed based on domains (code families) derived from the interview guides and through open coding. Here we draw on several code families, including ‘identity,’ ‘community,’ ‘stigma,’ ‘masculinity,’ and ‘sex position.’ Two members of the research team (first and third authors) coded the dataset after arriving at an adequate level of inter-coder agreement (over 80%).

Results

Multiple dimensions of identity: Sexual identity, sexual position, and gender performance

Drawing on social interactions revealed through participant observation and described in narrative data, we came upon three interrelated axes that were central to men’s understanding of who they were, as well as their partnering strategies (i.e., ways in which men seek sexual encounters while negotiating between erotic ideals and social presentation): sexual identity, preferred sexual position, and gender performance. Table 2 describes the local categories that constituted these three axes.

Table 2.

Taxonomy of sexual identities, sexual roles, and gender performance.

2.1 Sexual Identity
Term/Phrase	Meaning	Representative Narrative	HIV Prevention Practice
Gay (N=15)	Most believed the gay ‘label’ has a racially White connotation, leading many times to feeling doubly excluded from gay and black communities.	‘Gay is a lifestyle. It definitely is. It’s a way of life. A lot of people don’t understand the terms and things of that nature.’ (24 year-old)	More than half of these men had condomless sex, although they expressed remorse about it. Many practised sex work, were homeless, and many had links to CBOs. A few of them (N=5) had private health insurance and saw staying healthy as a life project.
Same Gender Loving (SGL) (N=3)	A few used the counter-hegemonic term to undercut feelings of exclusion through solidarity based on sexuality and race.	‘Same-gender-loving, me being an African American man, means that I am in fact that of a man who just happens to love the same sex.’ (39 year-old) ‘We kind of use the term “same-gender-loving,” which kind of burst out of the black gay community… We’d have these conversations, anything from politics or what’s happening specifically in the black gay or black homosexual community’ (31year-old)	All three were very critical of HIV campaigns that equated African American with ‘at-risk’ for HIV; some knew about PrEP, and thought it was a good idea. They saw racial mistrust of medication and medical institutions as a major barrier to PrEP uptake.
Bisexual (N=4)	A few self-identified as bisexual, although for the most part they did not disclose homosexual desire or behaviour with friends, family, and other social networks.	‘But there is no denial I have sex with men, but it’s just the culture is the part that I can’t reconcile myself with and so when approaching people and dating and seeing a cavalier approach around sex and health and wellbeing, initially it was hard to understand and I didn’t understand.’ (32 year-old)	Some of these men used condoms with men and women; two men were open to male and female partners about their sexuality to establish honest relationships.
Straight (N=3)	Men who had steady relationships with women as primary partners identified socially as straight	‘So some of the guys who might consider themselves straight, once they get high, they’ll get fucked too or suck some dick. It’s all good. I don’t judge nobody.’ (45 year-old)	For the most part did not use condoms; expressed using condoms with men but not with women. All of these men used drugs to overcome dissonance between sexuality and identity. Participant observation in public cruising spots (e.g. parks) indicated heavy drug and alcohol and sex among men engaged in secretive sexual encounters.
Discreet (N=4)	Those who identified as discreet could be strictly homosexual or bisexual in behaviour; Discreet identity and discretion independent of identity were associated with feelings of shame and social risk.	‘And I kept it discreet, if I went to the studio I had on a hoodie just trying to blend in…which was still uncomfortable because they could still see – because I used to get my eyebrows arched.’ (26 year-old) ‘I let him walk in front of me, you know, because where I used to work at, sometimes it’d be the trucks would be delivering stuff, and I don’t want them to think I’m with him, you know? Yeah, he dressed differently. He don’t wear high heeled shoes or anything, but he likes buying feminine clothes.’ (47 year-old)	These men had more sexual partners than others, met men online more often, had sex with and without condoms, and two were superstitious or afraid of HIV/STI testing; key informants mentioned that discretion was a challenge for enrolling men in prevention programs.
Other (N=2)	A couple of men had either no sexual identity or preferred to be called MSM. Both were used to reject gay stereotypes.	‘I normally don’t like to refer to myself as gay… I think me just being a black male, I like to identify with being a black male. Gay always referred to being white, you know what I mean? When you’re a black man I think for me it’s so important to identify with that first, right?’ (45 year-old)	They displayed greater concerns for the harmful effects of political sexual identities; and HIV prevention was important to both stemming from life experiences and formal education.
2.2 Preferred Sexual Position
Term/Phrase	Meaning	Representative Narrative	HIV Prevention Practice
Bottom (N=10)	Receptive in anal sex, also refers to enjoying giving oral sex. Some bottoms reframed this sexual position as a source of power (see power bottoms and “the cunts”). Some bottom men complained that men on hook-up smart phone apps were too often bottoms, although several discreet tops preferred this venue.	“When I was younger, I wasn’t a top neither. The guy he was topping me I guess because I was young and kind of dumb and whatever. So he was doing that to me so imagine if that would have gotten out to my grandmother that would have been embarrassing… She always raised me to be a boy, she always raised me to be a man, and if she would have found that out it probably would have just crushed her. You’re doing what; she always raised me to be a man, none of that girly, sissy stuff. So those are all the roles that play with that.” (24 year-old)	Men expressed that bottoms were more at-risk for HIV infection; and self-identified bottoms were more open to using PrEP.
Total Top/Dominant (N=9)	Many men described being strictly insertive partners in anal sex, sometimes misconstruing that for not having anal sex.	‘I don’t know. I guess that would be it. I haven’t been – I haven’t had anal sex. I’m always the top, never bottom. And – so you’re always on top, never the bottom with the man… At 54 I haven’t tried it… It would put me on the other end, and I don’t want to be on the other end. I enjoy being the dominant one. I don’t want to be submissive one…Power and control. It could be that.’ (54 year-old)	This is associated with both sexual and social domination; tops being better ‘men’ than bottoms. Guys who identified as ‘dominant’ were more often straight or bisexual, although several were gay. The sense of control associated with this sexual position kindled HIV stigma, placing complacent tops who perceive less risk vulnerable to HIV.
Vers Top (N=12)	Mostly Top, prefers being insertive partner in anal and/or oral sex but would bottom for the right person (if they are emotionally into them or if the other person is a Total Top)	‘I’m a vers-top…I just put that on me because I’m a top, but I’m a verse/top. It depends on the boy I really like. Like the one I was talking to, he was a top. The one that I was telling that we kept texting each other and then it just changed after a while… I was compromising because I like him, and that’s what it is when you like somebody when you’re willing to compromise… I just feel like compromising is good.’ (22 year-old)	Perceived less risk than men who were mostly bottoms, used strategic positioning to reduce HIV risk, and sometimes reported not using condoms when they are bottoms because they only bottomed when they love or trust someone.
Power Bottom (N=2)	A receptive partner in anal sex that exhibits control/dominance in intercourse	‘We’re like the fun in the sex game, because we’re in control like when …the guys just are pumping. That’s basically what – that’s all they do. They’re pumping. They’re doing this… But like the ones, basically in controlling. Like I said, we choose the positions, how fast we want to – like we – it’s like we lead them on, really. If the guys – if we didn’t communicate with the guys, it would just be like sex. We’re like, “Oh, yeah, daddy, give it to me harder,” or, “oh, no, we take it slow,” like we control them. Bottoms are like the females. We do the scratching, the sucking, the fucking, the kissing. It’s just – it’s amazing. I love being on the bottom. [Laughter]… I am a power bottom.’ (22 year-old)	Having a sense of power as a receptive partner in anal sex counters stigma and indicates more ability to negotiate condom use.
2.3 Gender Performance
Term/Phrase	Meaning	Representative Narrative	HIV Prevention Practice
Trade (N=10)	Highly masculine men, ‘hood but not ghetto,’ mostly a top in sex; term also refers to the penis (‘his trade’), or to sex work (‘the trade’ depending on context	‘Trade, which is the guys you wouldn’t really expect to be gay, but they look straight. That’s what we call trade…I find trades more attractive. I don’t know…I see a guy that looks straight. I feel like it’s more attractive because if he’s feminine, I feel like the whole effeminate thing though wouldn’t fit some guys…. It’s not really attractive.’ (16 year-old)	Key informants that referred to ‘trade,’ described men who had concurrent relationships with men and women; feeling that dominance was protective increased HIV vulnerability because men perceived less risk.
Butch Queen (N=11)	Refers to masculine presentation, but versatile sexual role performance	‘Then you have the butch queen … the sort of effeminate but still masculine man. But he can still walk on the street and be identified as a normal male, so you have that. And the butch queen trades generally go for trades.’ (18 year-old)	Men described ability to navigate as tops and bottoms, used condoms inconsistently and more often when they were bottoms; had some difficulty negotiating condoms with trade men
Boy (N=2)	Some ‘signs’ of being homosexual, but masculine enough to pass as a ‘regular’ man or not be immediately pegged as gay; attracted to other ‘boys’ or masculine guys	‘We’re way too girly, too effeminate. They just weren’t boy-boy about it….I’m not saying you got to be macho and play sports and all this shit. Let’s just be a boy.’ (46 year-old)	Similar to butch queen in gender performance and sexual partnering strategies; being ‘a boy’ also meant these men showed strength by avoiding health services
Femme Queen (N=2)	A term for gender performance and identity; sometimes referred to as ‘chicks with dicks’ by ‘straight’ men in this sample who prefer them sexually.	‘Femme queens are Trannies…they’re femme queens, like they’re girls, yeah. Or they wear really girly clothes where they look like women. Like they just look like women…Or they look like boys in drag.’ (26 year-old)	Several men and key informants talked about a lack of focus of research and prevention for transwomen, even though they are highly vulnerable to violence and HIV.
The Cunts (N=6)	Several gay-identified men identified as ‘the cunts.’ They were in their early 20 and teens. The term connotes solidarity in androgyny; other young men who were not ‘the cunts’ had heard of the term and meaning.	‘I would be considered a cunt. That’s the gay term. The cunts are basically…Pretty gay boys who dress fashionable, basically. They always cute. Always pretty and always on point. They dress nice and yeah. You have a certain androgyny feel to them because they’re boys, but they’re femme, but they’re not trans. So I think I would like to identify as more with that…But yeah, that’s the general meaning of what the cunts are and I hang around probably two other cunts.’ (18 year-old)	These men tended to be socially incorporated in HIV prevention, CBOs, some but not all in the house ball scene; although they were knowledgeable about HIV prevention, some had occasional unprotected sex.

As shown in Table 2, men in our sample identified their sexuality as gay (N=15), straight (N=3), bisexual (N=4), discreet (N=4), same-gender-loving (N=3), or preferred no sexual identity (N=2). Although approximately half of the sample identified as gay, the meaning of this identity varied. Some men felt more comfortable than others in feeling ‘part of’ the gay community. In fact, several gay men were consciously resistant of the gay community because they felt discriminated because they were Black. A few men identified as same-gender-loving, an identity that politicised the inherent need for more attention to racial inequalities within the gay community. In addition, nearly half of the sample identified as straight, discreet, bisexual, or preferred no sexual identity. Most straight and bisexual men reported currently having long-term (more than one year in a relationship) girlfriends, while they had sex with men in casual relationships. Men who chose to identify as discreet focused on ‘not being out of the closet,’ as the main characteristic of their sexuality. Similar to a preference of no sexual identity, this sexual description allowed for men seeking sex with men to be ‘up front’ about not wanting to be sexually labelled as gay.

Even though sexual identity labels began to tell a story about men’s social self-representation, Table 2 also shows how men preferred to label themselves when negotiating sexual intercourse. Some had very strict notions of what sexual position they preferred (e.g., total tops, total bottoms, power bottoms), while others were willing to be both insertive and receptive in anal sex to varying extents (e.g., vers-tops, vers-bottoms, vers). In many ways, men’s preferred sexual roles set the boundaries for their expected gender performance: men seeking partners who were solely receptive in anal sex often presented themselves as more masculine, and vice versa. In short, other aspects of their sexual and erotic preferences – what might perhaps better be described as their gendered sexual ‘subjectivity’ – appeared to be far more important than identity in terms of shaping specific sexual practices (Dowsett, 1996; Lorway, Reza-Paul, & Pasha, 2009; Weeks, 1998)

For this reason, the third axis of the taxonomy in Table 2 depicts how men qualify their gender performance or expression using a range of labels. Men who were perceived as hyper-masculine were described as ‘trade’ (Bowleg, 2013; Ford et al., 2007). Approximately a third of the sample was versatile during anal sex and gay-identified, many of whom identified as ‘butch queens’ to denote self-acceptance of the feminine and masculine aspects of their social presentation. In addition, there were those who were proud of their femininity, enjoyed being bottoms (sometimes finding power in being anally receptive). Among these men there was a meaningful play on normative notions of gender and sexuality.

Many of them identified as ‘the cunts.’ They used intentionally vulgar or transgressive language to represent their ‘sassy’ and ‘stylish’ feminine gender performance and to undercut ways that they felt discriminated or undermined due to their femininity. In categorising themselves as ‘feminine,’ these men were not seeking to reproduce symbolic patterns associated with ‘normative femininity,’ but rather to emphasise a transgressive version of femininity – i.e., what is viewed as a transgressive performance of prohibited female sexuality – being sexually aggressive as ‘power,’ etc. At the same time, their sexual objects of desire were normatively masculine, often ‘straight’ or ‘trade’ men. Gender transgression is thus multiple (or at least double): biological males acting like females, and normative female sexual behavior transformed into transgressive female sexuality – the ‘play on’ gender seems to be on male fantasies (or at least symbolic representations) about female sexuality. The kinds of local identities men adopted along these three axes reveals how much notions of gender and sexuality were intertwined.

Thus, beyond looking at these local identities as static or independent of each other, we found that the ways in which these axes were interrelated and sometimes fluid revealed critical insights for culturally nuanced HIV prevention strategies. As discussed below, this fluidity and interrelation were particularly evident in relation to partnering strategies and community formation.

How sexual identity, sexual performance, and gender performance are fluid and intertwine

There was a real fluidity to these categories of identity over time and space; the way men presented and understood themselves shifted depending on where they were and with whom they were socialising. Nonetheless, despite the ways in which gender and sexuality were performative and intentional, there was some consistent social patterning regarding sexual identity, preferred sexual position, and gender performance, as follows.

Group 1 (N=8): Men in this group identified as straight, discreet, or bisexual, had a masculine gender performance and most were ‘total tops’ (only insertive) in anal sex

These men always sought to keep their homosexual behaviour a secret from family, friends, and religious communities. Nearly all of these men had girlfriends and children, and prized discretion as a strategy that preserved their access to heterosexual privilege. This group of men understood competent gender performance to be constituted in part by sexuality, and as a result, these men often perceived their homosexual behaviour as shameful or experienced guilt. Their understanding of what was expected of a man extended beyond appearing heterosexual, however, to include having children, resolving problems through anger and violence, providing for the family, and being (or at least presenting themselves) as self-sufficient.

For men in this group, belonging to what they considered important aspects of the Black community and complying with what was expected of Black men was more important than disclosing their homosexuality. For this reason, some strategically partnered with male sex workers (found in the streets, parks, online, or through gay-magazine ads), carefully negotiated discretion with male partners before meeting them, and preferred feminine men or transgender women. These men saw themselves as being ‘controlling’ and ‘dominant,’ often choosing male sexual partners (sometimes referred to as ‘bitches’) with whom they could establish unequal sexual partnerships, especially if partners were sex workers or very feminine. Choosing men who were feminine, but whom they could control and keep private, allowed discreet and straight men to manage their heterosexual social privilege while fulfilling sexual desire.

In addition, these men preferred Black and Latino sexual partners. A 54-year-old, straight-identified man explained, ‘Hispanic and Black chicks with dicks are more of a turn on than Caucasians for some reason:’

[In White transgender women,] their hormone level is not as strong. And that may have something to do with the melanin in the system. I don’t know, but when I see them, they never as busty as derriered as the Hispanic or the Black chicks with dicks…and well, the larger the penis is, the better.

His narrative suggests that men or ‘chicks with dicks’ who have accentuated feminine features (being ‘busty’ or having large ‘derrières’) and masculine features (a large penis) are more erotically desirable and that possession of these characteristics is seen as racialised. Thus, racialised sexual desire (i.e., finding Black and Latino men sexier) also shaped these men’s partnering strategies.

Juggling their masculine and heterosexual social appearance with their homosexual behaviour shaped this group’s engagement with health services and how these men perceived HIV risk, which in turn translated into their HIV prevention strategies. Seeking health services for routine check-ups was a sign of ‘weakness’ and femininity, and adherence to HIV testing was among other ‘emergency’ reasons for health-seeking (i.e., others included breaking a bone, being shot, extreme pain). Many of these men felt ‘in control’ and less vulnerable because they were tops in anal sex and because they were not gay. Some reported always using condoms with men or ‘chicks with dicks’ but not with their girlfriends. Nearly all men in this group recognised that they used condoms inconsistently, and because of this they perceived themselves at risk and got tested regularly. Reflecting on his own sexual identity, a discreet 46-year-old explained that men who feel they are not part of the gay risk ‘subgroup’ would not even consider PrEP as applicable to them:

‘Cause I’m not gay, that nigga’s gay – he’s suckin’ my dick,’ or ‘I don’t do that,’ but you don’t know what your partner’s doing, and I think that a lot of people are just in denial about their existence. And if you’re in denial about your existence, why would you wanna seek any help? ‘[HIV infection] is not gonna happen to me, or anybody I know.’

In fact, none of these men thought PrEP was a prevention tool that would work for them because of their association of HIV with gay men, taking medication as a sign of weakness and because of fear that if someone found their pills, they would be labelled ‘gay.’

Group 2 (N=7): Men in this group identified as gay, were bottoms in anal sex, and felt a sense of pride about their femininity

These men, most of whom adopted the identity of ‘the cunts,’ found masculine partners sexually desirable, and sought to hook-up with butch queens and trade men. They felt they belonged to a particular niche in the Black gay community, and often hung out with other men who were proud of their femininity. As a 22-year-old gay man explained, some of them called themselves ‘power bottoms’, expressing their feeling that they controlled their partners’ sexual pleasure and contradicting notions that bottoms had no power.

Identities such as ‘the cunts’ and ‘power bottoms’ reflect these men’s efforts to construct a powerful femininity, to resist stigmatisation, and to counteract the gender power inequalities they experience in relationships with trade or straight men. They generally felt a sense of belonging to a community of similar men, which afforded them social support. This was important to them because many of them described experiencing stigma and discrimination because of their feminine gender performance. When these men also hooked up with men who were hyper-masculine and discreet, they could not hang out with their sexual partners in public, meeting furtively in hotel rooms, parks and other discreet spaces. Although they resisted perceived gendered power imbalances, many of these men had experienced sexually violent situations (forced unprotected sex) with more masculine-appearing men, and they felt emotionally hurt by their partners’ inability to acknowledge them socially.

Sexual desire and gender inequalities led to power dynamics that sometimes made it difficult to negotiate condom use. Some of these men used condoms inconsistently when their potential partners were ‘trade’ (hypermasculine) because they found the masculinity of these partners to be highly desirable. Many of these men saw PrEP as a good prevention strategy. For example, a 22-year-old gay man, who describes himself as one of ‘the cunts,’ explained ‘Yes, I would use it [PrEP] if it’s to help keep me healthy and safe.’ The greater willingness to use PrEP that characterised men in this group seemed to reflect their sense of belonging to a community of similar Black gay men that attempted undercut masculine norms, including power imbalances when having sex with ‘trade’ men and the notions of ‘weakness’ that dissuaded other men from seeking health services.

Group 3 (N=8): Men in this group were gay and same-gender-loving men, were versatile in anal sex, and expressed both feminine and masculine gender performance

The diverse partnering strategies of men in this group, many of whom identified as ‘butch queens,’ varied depending on the relative masculinity and sexual identity of prospective sexual partners. Although the performance of an exaggerated version of masculinity made potential partners highly desirable for most men in our sample, for men in this group the value of masculine performance depended on whether their partners were open about their non-normative sexual identity. In fact, some men stigmatised straight-identified men, even though they were extremely masculine, referring to them as ‘DL’ (down low) and blaming them for sexual transmission to women.

Men in this group often saw themselves as belonging to the Black gay community. Some were also explicitly critical of the White gay community because they felt it ‘hypersexualised’ Black men and that they did not fit in unless they ‘queened out’ for white men. Several had a strong sense of community as ‘same-gender-loving’ men, as opposed to gay men, because this label was started in the Black community. Like several others in our sample, one 45-year-old man rejected the label ‘gay’ because racial solidarity among Black same-gender-loving men was a ‘political’ act:

I normally don’t like to refer to myself as gay. I think I just being a Black male; I like to identify with being a Black male. Gay always referred to being white, you know what I mean? When you’re a Black man I think for me it’s so important to identify with that first, right?

Their sense of racial community also affected their partnering strategies, as many of them only dated Black men. Although a few referred explicitly to political dimensions of racial solidarity, many men explained preferring other Black men as sexual partners in terms of sexual desire. Whereas several men affirmed stereotypes about a Black man having a large penis, for others – such as a 26-year-old gay man who, ‘love[s] Black men’ because he ‘like[s] the swag’ or ‘the way they carry themselves’ – sexual desire for Black men was fuelled by a shared sense of style and sensibility, social presentation, and cultural beliefs about sexuality.

Moreover, many of these men reported using condoms consistently and were apt to see a doctor regularly at community-based clinics and for routine HIV testing. However, some claimed they had condomless sex with men in the last year to pay bills or to have a place to stay for the night. Men sometimes chose their sexual positions strategically to reduce the risk of acquiring HIV. As a 31-year-old same-gender loving man explained, he is ‘not just a top’ and believes ‘risks are lower with HIV’ when he tops, so he is more careful when bottoming. He attributes tops becoming infected to ‘some of them eventually being penetrated at one point in time.’ In fact, vers tops perceived less HIV risk than men who were mostly bottoms, and sometimes reported not using condoms when they were bottoms because they only bottomed when they loved or trusted someone. Our findings about how men chose sexual position strategically depending on the kind of partnership (i.e., casual/one-time vs. steady with affective bonds) corroborate findings in the literature about strategic positioning in gay minority men and Black MSM (Marks et al., 2010; Van De Ven et al., 2002).

Thus, because of their sexual versatility (as bottoms/tops) and gender-awareness, these men thought critically about how strategic sexual roles and partnering strategies affected their willingness to take PrEP. Some of these men thought PrEP was a good idea for others (e.g., serodiscordant couples) but not for themselves because, as a 41-year-old gay man put it, taking a daily pill was ‘crazy,’ ‘a big commitment other than just putting a rubber on,’ although, ‘raw sex is nice.’ Many men did not think condoms would be used with PrEP, as a 29-year-old gay man explained: ‘people are just absolutely fascinated with just raw sex. So they would be like the first ones probably in line for PrEP. Like, oh my god, I can take this pill and have raw sex.’ In fact, very few men would use PrEP and condoms for the added protection and in case the condom broke. In addition to being ‘real’ about the possibility that men would not use PrEP and condoms, men in this group were also very reflexive about how masculinity and femininity could affect PrEP acceptance. Several likened taking Truvada daily to ‘the pill’ for contraception, and as one 22 year-old man suggested, messaging should, ‘call it “The Pill for Men”,’ to capitalise on gender ideology.

Although key informants thought that PrEP was best for ‘bottoms’ and people who already have ‘raw sex,’ they thought that more community-based education was necessary, like a social worker at an HIV prevention CBO explained: ‘They should have a professional…also want somebody that they can relate to, who are versed in PrEP and hear the good and the bad and the ugly, and then they’d be allowed to make a decision as to whether it’s for them. I think a lot of the times they don’t have much information or it’s rushed.’ Thus, for men who are open to consider PrEP as a prevention tool, community stakeholders thought clear messaging and community engagement would ‘demystify’ this new prevention tool and its proper use.

Group 4 (N=8): Men in this group had fluid sexual identity, versatile sexual role preference, and their gender performance largely depended on their social context or situation

This group includes men whose sexual identity is gay or openly bisexual in some contexts, such as bars, whereas it is straight or discreet in others. These men also visited different venues strategically when they wanted to be tops or bottoms in anal sex. For example, a 26-year-old man explained that he could choose to present themselves as ‘trade’ and seek to be the insertive partner through hook-up applications and online, whereas he was more of a ‘butch queen’ or ‘cunts’ at parties or clubs. A 47-year-old discreet man maintained a long-term relationship with an older man, but also had casual sex with men he met on hook-up websites. Thus, several men displayed strategic ways to fulfil their desire to be a top by adhering to masculine gender norms on smart phone apps, and more feminine (through their clothing, dancing, flirtatious character) in bars or parties.

The greater degree of sexual and gender fluidity in this group was associated with men’s socioeconomic disadvantage, including employment insecurity and being uninsured. Socioeconomic disadvantage led many of the men we interviewed to survive through sex work, which challenged their ability to negotiate condoms. A 24-year-old sex worker, for example, expressed his need to be the provider for his grandmother and ‘put food on the table.’ For sex workers, their sexual identity, gender performance, and sexual position depended on their clients. Male sex workers could charge more for bottoming if they presented themselves as masculine.

Socioeconomic disadvantage challenged condom negotiation and the importance of staying HIV-negative. Because these men had context-dependent gender presentation and sexuality, they participated in the gay scene and several were tested through HIV community-based organisations. An 18-year-old gay man, explained:

PrEP is more like a want. It’s not something you need. PrEP is more expensive than PEP (post exposure prophylaxis). I was told by…a youth counsellor at the Harlem Prevention Center. And he told me that PrEP can be up to thousands of dollars…We educate people to have protected sex, safe sex with condoms, safer sex, so if you are having safer sex and you just depend on PrEP…You don’t really need it. You want it… It’s like cable. You don’t need cable.

Many men in this group thought PrEP could be an effective tool for sex workers, but they were concerned about the cost, since many of them did not have health insurance. From key informant interviews with service providers at CBOs, it was clear that some providers believed PrEP was too expensive considering other available prevention methods and the cost of care for those with HIV and AIDS. Overall, some of the men who were at-risk (e.g., sex workers, socioeconomically disadvantaged men) encountered resistance about PrEP from providers because it was ‘too expensive’ and ‘unnecessary.’

Discussion

This study lays out numerous local or emic categories that exist within the externally created, etic category BMSM. By examining the meanings associated with these categories, and analytically grouping men into configurations constructed in relation not only to sexual identity, but also preferred sexual position and gender performance, we were able to unpack some key differences in social and cultural experiences that organise significant diversity within the otherwise undifferentiated category of BMSM – and that may ultimately affect the possibilities for advancing the use of PrEP and other methods in the HIV prevention toolkit. Like all prevention techniques and technologies, use of PrEP requires a complex set of ideas and practices to be put in place within the context of broader cultural meanings and social relationships, and its effective incorporation into the broader range of prevention options needed for diverse BMSM will only be possible if greater attention is given to this social reality. Advancing the use of PrEP for those who might benefit from it will only be possible if it is truly embedded within this social context and becomes meaningful within it. To accomplish this, our research suggests that a number of key considerations will be crucial: (1) recognising the importance of understanding local social and sexual subjectivities; (2) bringing a meaningful discussion of sexuality, sexual subjectivities, and sexual practice back into the response to HIV through community-based prevention efforts; and (3) reinforcing community-based prevention programs, even in times of constrained resources, and integrating biomedical prevention approaches into these services.

The importance of social and sexual subjectivities

First, we want to emphasise the importance of shifting the focus of our attention from broad, undifferentiated epidemiological categories, such as BMSM, and even from the more nuanced categories such as the diverse sexual identities that this broad category largely conceals, to a deeper understanding of the social and sexual subjectivities that shape the lives of these men. In fact, these configurations – which are intersectional in their inclusion not just of race and sexual identity but also through their inclusion of gender performance and sexual practices as axes of social diversity – tell us more about HIV vulnerability than broad (or even local) categories. The importance of these categories became more evident as plays on gender and sexuality were used strategically in seeking sexual partners, in deploying stigma, and in delimiting boundaries for belonging to the gay and/or Black communities.

Our results indicate that HIV, sexuality, and gender-related stigma may affect HIV prevention among men in each of these groupings differently. The most vulnerable group seems to be comprised of men with straight or discreet sexual identities, who report being tops in sex with men, and who have masculine ideologies and gender expectations. This group would be the least likely to take PrEP, especially because of the strongly held association of HIV with gay-identification among these men, the priority they give to being discreet about their sexual desires, and their disengagement with health care services because they deem health-seeking weak or feminine. Meanwhile, in seeking to secure their preferred sexual position as tops and feel dominant, straight-identified and discreet men interact sexually with sexually-versatile men, femme queens, transwomen, and partners who were often sex workers. The group of more feminine men and those who were sex workers were more open to the idea of taking PrEP and less mistrustful of health institutions.

Thus, understanding partnering strategies reveals important ways in which various prevention tools may function in a sexual scenario: one partner (e.g., a man who describes himself as straight and only assumes the insertive role) could choose to use condoms and an anally receptive partner may choose to be on PrEP. Men most affected by gender-related inequalities because of their feminine presentation may have less leverage to negotiate condom use and may thus benefit significantly from the option of taking PrEP. Studies that have begun to uncover ways in which PrEP affects sexual partnering through processes such as “biomed matching” (Newcomb, Mongrella, Weis, McMillen, & Mustanski, 2015) also need to consider the gendered dimension of sexual sorting processes.

Rather than stigmatising non-gay identified men as a ‘bridge population’ and as unfaithful ‘cheaters’ ‘on the down low’ – potentially as a response to unequal gendered power dynamics – our findings support research that calls for community-approaches to reduce stigma among this ‘hidden population.’ In fact, HIV prevention may reach hidden populations as well as address multiple forms of sexuality and gender-related stigma experienced by both openly gay men and discreet/straight men by understanding these sociocultural dimensions of sexual partnering, opening ways to engage potential sexual partners in different forms of community outreach.

To consider the ‘social reality’ of HIV prevention methods, it was therefore important to explore how gender performance and sexual practice together operate as organising principles of social life that generate social divisions and differentiations among BMSM. Gender is often mistakenly discussed as a way to distinguish biological difference rather than as a social differentiation, and thus public health discourse fails to recognise that gender is not reducible to male or female bodies but rather is continuously performed and enacted (Bridges & Pascoe, 2014; Connell, 2005). Few scholars have considered strongly ‘hybrid masculinities’: the distinct melding of masculine and feminine gender norms and expectations independent of sexuality or biological sex (Bridges & Pascoe, 2014; Bowleg, 2013; Connell, 2005). Ethnographic research has illuminated how normative expectations of masculinity lead some BMSM to monitor their gender performance (Balaji et al., 2012; Bowleg, 2013; Miller et al., 2005). We found that gender performance differences shaped key aspects of sexual practices and sexual cultures, but these differences were masked by the catchall category of BMSM. Categorical lenses such as BMSM do not allow us to see what is actually going on in men’s sexual experiences and gendered realities. This may have important implications for packaging HIV prevention to address the social factors that shape risk situations.

This study goes further to highlight the ways in which gendered performances may ‘play on’ or intentionally transgress normative gender expectations for what is ‘successful’ or ‘failed’ masculinity or femininity – and how this becomes critical (within the category of BMSM) when men use gender expression as a strategy for social differentiation and as a way to navigate power dynamics and social and sexual relationships. Populations such as the ones included in this study remain ‘hidden’ insofar as the research is unprepared to account for them because of limitations in close-ended surveys or other methods that are less engaged with communities. Ethnographic engagement was crucial and points to the conceptual and methodological contributions of this research. Our findings indicate that the erasure of difference and diversity, the silencing of sexuality, and waning effort to engage hard-to-reach communities through grassroots approaches is precisely the set of conditions that made it almost inevitable that infections would rise in the most marginalised and stigmatised communities – which are also precisely the communities that are least served by the unilateral biomedical administration of the epidemic.

Bringing sex back into HIV prevention

The concentration of the HIV epidemic in specific subpopulations, like BMSM, in the USA has happened over much the same period when community-based prevention programs have been scaled back and even curtailed as the result of funding cuts and a growing reliance on biomedical approaches such as test and treat and treatment as prevention. Whether intentionally or unintentionally, there has thus been a shift from more community-based approaches to more clinic-based approaches. This, in turn, has almost always been associated with a gradual process of marginalising, silencing and sanitising the direct and open discussion and exploration of sexuality as absolutely central to the development of meaningful prevention programs. Indeed, one of the reasons that biomedical approaches to prevention appear to have generated significant enthusiasm on the part of many public health policymakers and practitioners may be precisely because they are incorrectly perceived to make it possible to implement HIV prevention programs without having to engage with the lived realities of sexuality on the part of affected populations – realities that are almost always very distant from those of the policymakers and practitioners who supposedly serve them.

These tendencies run directly counter to many of the key lessons learned over the history of the HIV epidemic, both in the US and globally: that it is only by engaging with local realities, and in particular with local sexual meanings and sexual practices, that we can effectively address the epidemic (Aggleton & Parker, 2015). It was precisely because of this that community-based approaches to HIV emerged in the first place – and because of this that for more than three decades they have continued to be the most important settings for developing meaningful programs to address the epidemic among marginalised and excluded communities and populations. It was in these settings that sexuality could be openly addressed and discussed in meaningful ways, and that safer sex could be invented and reinvented ‘as community practice’ (Watney, 1990).

Our research underlines the fact that understanding local cultures of sexual desire still matters in what has come to be called the science of HIV prevention, even when we consider biomedical prevention methods, such as PrEP. In fact, sexual desire was linked to plays on domination and submission, among other sexually enacted power dynamics that reflected and affected men’s social relationships and their ability to negotiate HIV prevention strategies. Unless there is a new emphasis on, and a renewed approach to prevention that recognises the complexity of sexual communities and the need for community-based, community-owned, and community-led prevention programs, new approaches to HIV prevention will be destined to fail precisely because they will never be integrated into meaningful sexual and erotic practice. Such integration, on the contrary, will depend on identifying how best to apply new prevention tools within vulnerable populations, and to strengthen a kind of ‘treatment/prevention literacy’ at the local level that will make it possible for these populations to effectively employ them in practice (Aggleton & Parker, 2015).

Twenty-five years later, Simon Watney’s prescient understanding of safer sex as community practice continues to be as true today as it was then – precisely because it is only within the context of communities that otherwise marginalised and stigmatised sexualities can be openly discussed and addressed. In our sample, men who were more willing to consider using PrEP also showed a greater sense of belonging to a community of similar Black gay men, framed their social identities to undercut normative gender inequalities, and rejected gendered norms that they understood to dissuade other men from seeking health services (e.g., notions of health-seeking as ‘weakness’). It is only within a context that brings back a focus on sexual cultures of desire and sex education as community practice that it will be possible to overcome short-sightedness of biomedical approaches that alone do not strongly consider sexual partnering strategies, stigma, and socioeconomic inequalities.

Integrating biomedical prevention approaches into community-based programs

Finally, just as our findings emphasise the importance of bringing sex and sexuality back into meaningful HIV prevention programs – and highlight the fact that community-based services are best positioned to be able to do this – they also underscore the fact that HIV prevention in different venues must be differentially tailored precisely because of the trust (or lack of it) that different venues are capable of generating. For important sectors of the broader BMSM population, especially for straight-identified and other non-gay identified men who are socioeconomically disadvantaged, medical institutions and medical authority are clearly suspect. A long history of ethically-questionable practices (and sometimes outright discrimination and abuse) on the part of biomedical researchers and health practitioners have left their mark on the trust that many members of this population have in health services. Even with the best of intentions, there are limits to the extent that clinical services will be able to reach all segments of this population.

For new approaches to HIV prevention, treatment and care to be accepted as legitimate and incorporated into community practice, information about such approaches needs to disseminated and appropriated at the community level – and in ways that communities can take ownership rather than rather than treating it as nothing more than ‘following the doctor’s order’. This sort of situational and contextual tailoring of prevention messages, and the building of confidence and trust needed to provide the social context for the adoption of new technologies, requires a reaffirmed commitment to community-based and community-led initiatives – precisely at a time when funding cutbacks and narrowing of available services appear to be pushing in exactly the opposite direction (Capacity for Health Project, 2013; Garcia et al., forthcoming; Hampton, 2011; Weinberg, 2013). Taking seriously this recognition of diversity and of the relevance of community-based expertise about this diversity suggests that PrEP (or any other new technological innovation) will not be a universally efficacious magic bullet. Rather, we need to understand it, and how best to use it, as a new tool in a growing toolkit of possible prevention approaches. It will be the right tool for some people, depending on their specific circumstances, particularly when it is presented as a form of combination prevention that is sensitive to cultural nuances and differences.

Considering that the meaningful engagement of local communities has been crucial to rolling out prevention and treatment campaigns globally, biomedical prevention should seek to address social, cultural and economic vulnerabilities that drive HIV at the same time that technologies such as PrEP are offered as a prevention option. A growing body of literature has emerged on the importance of understanding sexual and gender diversity among MSM of color and MSM in the Global South (Cáceres, Aggleton, & Galea, 2008; Johnson, Jackson, & Herdt, 2000; Nyeck & Epprecht, 2013). This study highlights the importance of recognising this diversity when making public health recommendations about the use of biomedical prevention approaches globally, which often rely on narrow assumptions about sexuality and gender in MSM. For instance, in 2014 the World Health Organization announced that PrEP is highly appropriate to combat HIV among MSM (WHO, 2014). The WHO’s recommendation appears to be a direct statement based on epidemiological data, but is very much in line with the focus of this special issue: the kinds of ‘troubles with categories’ mentioned above stirred confusion in communities throughout the world about how this overarching statement applied to the diversity of MSM, sexual practices, and partnering strategies (Heitz, 2014).

From the perspective of many clinical practitioners and biomedical institutions, the kind of local sexual categories and cultures that our research helps to illuminate are unfortunately seen more as impediments to effective prevention programs and strategies. This is hardly a new development; it is something that has existed since the very beginning of the epidemic, evident in the long delays that characterised the implementation of organised public health responses to HIV and AIDS, and in the inability of official programs to engage with the sexual cultures and meanings of non-normative communities and populations most at risk. It is precisely because of this that community-based venues and approaches have repeatedly shown themselves to be the most effective context for the delivery of HIV prevention information and technologies to marginalised and stigmatised populations; new approaches such as PrEP or other biomedical prevention techniques and technologies in the expanding HIV prevention toolkit are no exception (Arnold et al., 2012; Kerrigan et al., 2015; Parker, 1996).

While biomedical approaches are undeniably useful tools to prevent HIV and mitigate its effects on the most vulnerable populations, the historical importance of community-led HIV prevention indicates that the effectiveness of biomedical prevention will hinge on better understanding diversity within local cultures of sexuality and gender and addressing sociocultural vulnerabilities by leveraging local expertise. When looking to package HIV prevention, public health interventions focused mostly on biomedical approaches would benefit from looking back at the historical record to address elements of stigma associated with sexuality, gender, and HIV that also affect the acceptance of biomedical technologies such as PrEP.

In light of this, our empirical and theoretical argument is supported by the undeniable diversity in both local sexual subjectivities and sexual partnering patterns, which demonstrates the important forms of heterogeneity obscured by a category such as BMSM. Our findings suggest that identity and partnering patterns are inextricable from broader social context and shape risk behaviours and men’s willingness to accept different prevention methods. Our analysis begins to explore ways that BMSM navigate through these complex social universes, as members of the various intersecting communities that cross and intersect in a social space that remain largely underexplored and undifferentiated in public health practice. It is precisely this sort of insight that serves as a point of departure for what could be truly culturally meaningful ways of responding to the epidemic, including culturally appropriate ways of making biomedical prevention available, and renewed community ownership of the premises and processes through which an expanded toolkit of prevention options might offer, to those most in need of them.