The couple as context: Latino gay male couples and HIV

Abstract

HIV prevention efforts with gay men in relationships frequently omit primary partners. When they are considered, examinations of race/ethnicity are often overlooked despite higher infection rates among gay men of colour. Acknowledging both the need to contextualise the behaviours that may affect HIV risk for gay men of colour and the disproportionate impact of HIV on Latino gay men, the present study utilised semi-structured, qualitative interviews to explore relationship dynamics, sexual agreements and behaviours, and safer sex choices and HIV risk among nine Latino gay male couples. All participants were HIV-negative and in concordant negative relationships. Additionally, nearly all reported closed agreements. Analysis revealed participants engaging in four behaviours that may actively reduce their HIV risk: approaching sexual agreements from a practical standpoint, maintaining a high literacy around HIV, having exposure to social support groups for Latino gay men, and finding support in their relationship with another Latino gay man. Additional issues are raised where long-term HIV prevention is concerned. Intimate relationships are an important context for understanding both HIV risk and prevention among Latino gay men. Likewise, race/ethnicity provides an indispensable perspective on all research with gay couples. Future prevention efforts with gay men must strive to include both.

Introduction

Research with gay couples stretches back to the late 1970s and early 1980s (Bell and Weinberg 1978; McWhirter and Mattison 1984; Silverstein 1981). Findings from these studies identified different types of gay relationships and charted their development and maintenance over time. They showed that while some gay men are monogamous (hereafter referred to as a ‘closed agreement’), others permit sex with outside partners (hereafter referred to as an ‘open agreement’). These latter couples had understandings whereby sex outside the relationship was not considered cheating so long as each partner remained emotionally faithful to the other.

With the onset of the AIDS crisis in the 1980s, single gay men on the front lines of the epidemic became the focal point of prevention efforts and research involving gay couples was set aside. Importantly, however, just as single gay men found themselves at increased risk for HIV so too did many gay men who were in relationships. Their previous understandings about whether to allow sex with outside partners (hereafter referred to as ‘agreements’) now came with serious consequences.

By the late 1990s, HIV prevention efforts identified important distinctions between gay men in relationships and their single counterparts. Findings showed that gay men in relationships engage in higher rates of unprotected anal intercourse (UAI) with their primary partners than do single gay men with their casual partners (Chng and Géliga-Vargas 2000; Elford et al. 1999; Hays, Kegeles, and Coates 1997). Subsequent epidemiological studies confirmed primary partners as a significant source of many HIV infections (Davidovich et al. 2001; Kippax et al. 2003; Moreau-Gruet et al. 2001; Sullivan et al. 2009). Although once thought to be a haven of low HIV risk, the circumstances present in some relationships may actually increase it.

As the epidemic settles into its third decade important questions remain unanswered for gay couples. Recent studies involving gay men in relationships found that partners grow more permissive of UAI over time (Eaton et al. 2009; Prestage et al. 2008) and they point to gay men’s prioritisation of intimacy over safer sex (Frost, Stirratt, and Ouellette 2008) and de-emphasis of HIV prevention as a motivating factor for having an agreement (Hoff and Beougher 2010; Hoff et al. 2010) as possible explanations for this trend. These data, however, only scratch the surface, as basic information about agreements, how they operate, and their effect on HIV risk is sorely lacking.

Also lacking from research with gay men in relationships are analyses of race/ethnicity. While some studies present samples resembling the best-known estimates of the racial/ethnic makeup of the gay communities from which they recruited, others over-represent White gay men or omit reports of the racial/ethnic makeup of their samples, effectively whitewashing their participants (Young and Meyer 2005). This is troubling as HIV disproportionately impacts gay men of colour in the USA (Centers for Disease Control and Prevention (CDC) 2009; Harawa et al. 2004; Rhodes, Yee, and Hergenrather 2006). Given that gay men of colour do not differ significantly from White gay men in their reported levels of HIV risk despite their higher rates of HIV infection (Chng and Géliga-Vargas 2000; Ekstrand et al. 1999; Harawa et al. 2004; Millet et al. 2007), it is important to look beyond essentialising analyses that link one’s risk to one’s racial/ethnic identity and focus instead on contextual factors unique to gay men of colour that may affect their HIV risk.

Among gay men of colour, Latino gay men (LGM) have been acutely affected by HIV (Díaz 1998; Díaz and Ayala 1999, 2001; Díaz, Ayala, and Bein 2004; Díaz, Ayala, and Marín 2000). The number diagnosed in the USA grows each year and they continue to account for a disproportionate percentage of new infections (CDC 2009). Some researchers have turned to examining the contexts within which HIV risk occurs to help explain the high rate of infection among LGM. These researchers studied structural factors, such as racism, homophobia, and poverty, to see how they shape the contours of the epidemic for LGM (Ayala and Díaz 2001; Díaz 1998; Díaz and Ayala 1999, 2001; Díaz, Ayala, and Bein 2004). Their findings locate HIV risk within specific situations whose circumstances make safer sex choices difficult. For example, a lack of self-control in sexual situations and the belief that HIV infection is inevitable inform LGM’s HIV risk more so than deficit models that suggest a lack of knowledge of or motivation to practice safer sex (Díaz and Ayala 1999, 2001). Recently, research has focused on immigrant/migrant status as a context for understanding HIV risk, with findings that show LGM who migrate to the USA being vulnerable to HIV infection due to differences in motivations for migrating, differences in sexual cultures, the lack of support upon arrival, and communication difficulties (Bianchi et al. 2007; Carrillo 2004; Carrillo et al. 2008; Philen 2006; Storey 2000).

In addition to the examples cited above, relationship status stands out as an important context for understanding HIV risk among LGM for two reasons: the number of men who report being in relationships and the effect of relationship status on HIV risk. Previous research shows between 40-60% of LGM have primary partners (Díaz et al. 1999; Jarama et al. 2005; Ramirez-Valles et al. 2008; Wilson et al. 2009). In addition to this, other research demonstrates that being in a relationship increases the rate or is predictive of UAI with primary partner (Chng and Géliga-Vargas 2000; Díaz et al. 1994).

While these studies point to specific contexts where HIV risk occurs, prevention efforts with LGM largely miss examining the couple itself as a context. Even where previous research addressed relationship status and agreement type, it relied on one partner for that data. Without both partners’ responses the picture is incomplete. As a result, it is difficult to tell how our knowledge of agreements and HIV risk may change once both partners are integrated into the analysis. This is of particular concern given that some studies differentiate low from high risk based on how participants identify their agreement type.

The present study explores Latino gay male couples in an effort to contextualise the behaviours that may affect their HIV risk. In doing so, it builds on the existing literature by examining relationship dynamics and agreements as a context for better understanding among safer sex choices and HIV risk among LGM, incorporating both partners in data collection, and making race/ethnicity an explicit focus of inquiry. Our aim is to examine specific contexts to address and respond to the disproportionate impact of HIV on LGM.

Method

Participants

Between December 2007 and March 2008, we recruited ten Latino gay male couples from the San Francisco Bay Area using active and passive recruitment strategies in community settings that were frequented by or served LGM. Two LGM were hired as field research staff to gain access to community settings and reach potential participants (Zea, Reisen, and Díaz 2003). Field research staff handed out study postcards at social venues, such as bars and cafes, at community health and HIV/AIDS service organisations, and at social support groups for LGM. When unable to hand out recruitment materials in person, field research staff left study postcards and hung tear-off fliers. We also recruited participants through postings on email listservs. All recruitment materials were printed in Spanish. Latino men of all races/ethnicities, HIV-negative and HIV-positive men, and gay and bisexual men were encouraged to participate.

Those interested called a toll-free hotline for more information. Callers were screened over the telephone for eligibility. Eligibility criteria required that participants be at least 18 years old, be in their relationships for at least three months, know their own as well as their partner’s HIV status, identify as gay or bisexual, and identify as Latino.¹ Additional eligibility criteria required that participants speak Spanish and be in relationships with other LGM. Speaking Spanish was an component of eligibility as we wanted to capture a sample of LGM who were on the periphery, rather than in the center, of San Francisco’s gay scene. Similar to studies of LGM that used language as a proxy for acculturation into the broader community (Díaz and Ayala 1999, 2001; Díaz et al. 1999; Jarama et al. 2005; Ramirez-Valles et al. 2010), we used language as a proxy for acculturation into the LGBT community. Being in a relationship with another LGM was a component of eligibility in an effort to keep the analytical focus exclusively on LGM.

Couples who gave discrepant reports of partner HIV status and couples where either partner identified as transgender were ineligible. Couples were eligible to participate only after both partners were screened and found to meet all eligibility criteria.

Procedure

Eligible couples were given scheduled appointments for 60-90 minute semi-structured, qualitative interviews that were conducted at the study offices in downtown San Francisco. The study team selected the qualitative interview as the format for this investigation because of its strengths examining the subjective experiences of LGM (Zea, Reisen, and Díaz 2003). On the day of the appointment, research staff members consented each partner individually. All study forms were printed in Spanish. Participants were interviewed separately to encourage candid discussion of their relationship free from the influence of their partner. Interviews queried racial/ethnic identity, a natural history of any agreements, whether those agreements had been broken, sexual experiences with primary and outside partners, HIV risk, and safer sex efforts.

Interviewers were trained to pose questions in an open-ended, conversational tone and probe for clarity when and where necessary. Training stressed the intent behind each question and encouraged interviewers not to stick too rigidly to the interview guide. This gave the interviews a natural flow and allowed participants to discuss what was salient and relevant. To help establish rapport with participants and mitigate any potential social desirability bias, interviewers were also self-identified as LGM (Berg et al. 2004). Each participant was paid US$40.00 upon completion of the interview.

Analysis

All interviews were digitally recorded, transcribed into Spanish, and translated into English producing two sets of transcripts: one in Spanish and one in English. The present analysis is based on the English transcripts. A research staff member fluent in both Spanish and English reviewed all transcripts for accuracy (e.g., mistakes, misspellings, omissions) as well as for faithfulness in translation. Transcripts were analyzed using codes that were developed based on an initial analysis of the interviews and the study team’s previous qualitative and quantitative research with gay couples (for more information, please see: Hoff and Beougher 2010; Hoff et al. 2009; Hoff et al. 2010). Codes were created to exhaustively cover all transcribed material. Research staff members coded all transcripts utilising a Grounded Theory approach to data analysis, which permits themes to emerge endogenously from the interviews (Denzin and Lincoln 2003; Glaser and Strauss 1967; Lindlof and Taylor 2002). Analysis was conducted using Transana (Woods and Fassnacht 2007) version 2.3-MU. Transcripts were paired by couple for analysis within and between couples.

For the present analysis, we utilised coded text from 18 of the 20 interviews (nine of the 10 couples). We removed one couple, a concordant positive couple, because, their serostatus was unique relative to the sample as a whole and therefore data from their interview would be difficult to interpret along data from the other interviews. Removing them permitted the present analysis to focus on the other participants, all of whom reported being HIV-negative. Therefore, all statistics listed below, excluding recruitment, are based on the remaining nine couples.

Recruitment and Demographics

Our goal for recruitment was 10 couples. Recruitment sources reported by participants included study postcards, tear-off fliers, and notification via email listserv. A majority of participants (seven couples) reported hearing about the study through recruitment efforts at social support groups for LGM. These groups include organisations with wide-ranging missions, from social networking to activism to HIV prevention and other health outcome foci.

Of the 18 individuals included in the present analysis, all identified as Latino and further identified as follows: 13 as Mexican, two as Peruvian, and one each as Guatemalan, Honduran, and Nicaraguan. All reported being HIV-negative and in concordant negative relationships. In terms of sexual identity, 15 identified as gay and three as bisexual. The age range of participants was 18-49 years, with five participants under age 30 and 13 who were older. The range of relationship length was six months to 13 years, with five couples who were together less than one year and 13 who were together three years or more.

Results

Upon analyzing the interviews, four behaviours emerged that may actively reduce the participants’ HIV risk, setting these Latino gay male couples apart from both previous research with gay couples and previous research with LGM. First, nearly every couple reported having a closed agreement. Beyond that, most couples approached those agreements with the understanding that breaks could happen, so they appended their agreements with clauses that gave them additional protection and peace of mind. Second, participants were knowledgeable of HIV and made choices that reduced their risk. Most reported a high awareness of HIV risk, using condoms with outside partners, and testing for HIV regularly – with some also testing with their partners at the start of their relationships. Third, a majority of participants had exposure to social support groups for LGM, where HIV prevention efforts are a likely component of group activity. Finally, all participants spoke highly of being in a relationship with another Latino and of the support being Latino men together in a relationship provided them.

Agreements

Eight couples reported having closed agreements (the remaining couple reported having no agreement). Protection against HIV motivated many couples with closed agreements. When asked how he prevents transmission, one participant, named Juan-Miguel² (aged 35), responded, “We simply do not have sex with other couples.” He continued, “It’s an agreement not to have sex with any other person.” Another participant, Marco (aged 49) was just as explicit, saying, “Yes, one of the reasons [we are monogamous] is HIV. He is negative and I am negative … and besides that, we have relations without condoms.”

Six couples appended clauses to their closed agreements which stated that ‘just in case’ sex outside the relationship occurred partners needed to use condoms. Some discussed further aspects of those clauses, such as the need for disclosure and no receptive anal intercourse with outside partners. Importantly, having this clause did not mean either partner condoned or desired sex with outside partners. As one participant, Nestor (aged 33), noted, “[W]e have monogamy. We’ve come to the agreement that if we’re going to do anything outside we’ll protect ourselves, but that’s not an invitation to do it.” Comments from another participant, Daniel (aged 18), suggested that while having a ‘just in case’ clause is one thing, invoking it is another. He said,

“I have no reason why I should sleep with another person and I don’t believe he has one either, because if there is a reason, it is because I am not satisfying him or he is not satisfying me.”

Some participants discussed their motivations for having this clause. Acknowledging the unknown, Ferran (aged 33) said,

“I know that I am weak … and maybe something happens … I will do my best to make sure it doesn’t happen, but many times I can’t predict what will happen in the future.”

Although not everyone discussed their motivations, those who did spoke to the same issue: being prepared just in case.

Breaks were part of many participants’ discussions of their agreements. One or both partners in five couples reported broken agreements. What constituted a break, however, varied widely. One participant, named Miguel (aged 21), said he kissed another man. Partners in one couple independently stated that they occasionally sought sex with outside partners after fighting with each other. Partners in another couple, Carlos and Nestor, reported being involved in the same break: a threesome with an outside partner that they later felt broke their closed agreement. Carlos (aged 21) said, “[S]imply put, it was me. It’s that easy. I had the opportunity, I wanted to do it, but I shouldn’t have done it.” Nestor (aged 33) discussed the fallout from the experience and the subsequent conversation, saying:

[T]he next day we talked, because it was important to me that he express what he was thinking. … [H]e told me that it was part of his fantasy [to have a threesome]. I felt bad because, at the end of the day, he wasn’t ready to do it.

When asked whether breaks had been disclosed a complex picture emerged, with some disclosing and others not. Both partners in the couple above said they discussed the break. The other four couples, however, reported multiple breaks with some breaks disclosed but most not.

HIV

Overall, the participants demonstrated a high awareness of HIV risk and regular utilisation of safer sex strategies, such as using condoms with outside partners and testing regularly. Seven participants tested for HIV along with their partners at the start of their relationships. Most said that they tested together when they first met, within the first several months of being in a relationship together, or before they had sex for the first time. Partners in one couple, Juan-Miguel and Daniel, who tested at the beginning of their relationship, explained that testing together was less about trust and more about starting their relationship off with a clean bill of health. Juan-Miguel (aged 35) explained, “So for my safety and also for his, so that he knows I am entering clean into this relationship with no diseases, and also to be sure he is clean. [That’s why] we decided to take these tests.” Daniel (aged 18) said,

At the beginning, when we got to know each other, it was not because we doubted each other. One day, [I] just said, ‘I am going to get tested.’ And he said, ‘Oh, I will go with you. It will also do me good and I will also get tested.’ And at that time we shared the results with each other.

Many who tested for HIV with their partners approached it matter-of-factly. One participant, Héctor (aged 33), said, “We got tested when we first met. I asked him to get tested. I said, ‘I’m gonna do the test. Let’s do the test together to be sure we’re both clean.’” His partner, Gabriel (aged 28), agreed, saying, “[H]e was going to get tested by himself anyway and I said that if he was going, why shouldn’t I go too?”

Most reported testing regularly, with the elapsed time between tests ranging from three to six months to annually. Nestor (aged 33) said that testing was part of his ‘just in case’ clause. If his partner ever discloses that sex with an outside partner has occurred he and his partner will resume using condoms until they get another HIV test. “[W]e are going to have to be protected … until we have the test again and know that we are safe.”

Participants also expressed a wide range of attitudes about HIV, including their beliefs about which agreement types they felt were risky and the inevitability of HIV infection. Echoing their motivations for having closed agreements, some expressed their belief that open agreements increased risk for HIV. When asked about what he considered was risky, Gabriel (aged 28) responded simply, “Having an open relationship.” Another participant, Manolo (aged 47), said, “For me the only safe, really safe way is a monogamous relationship.”

Three participants relayed their belief that HIV infection is inescapable. When asked why he worried about it when he is monogamous, avoids high HIV risk sexual behaviour, and tests regularly, one of them, Jaime (aged 30) said poignantly:

You never know. You can get infected because of anything … and that is why I am careful in all my relations. I have been careful, but sometimes, even when you want to avoid it, you won’t be able to.

Similar sentiments were expressed by Marco (aged 49), who said:

[Y]ou have to take care of yourself because no one will take care of you, you can be sure of that. If someone wants to stick it in you without a condom, run the other way because that son of a whore wants to ruin your life.

Social support groups

Social support groups for LGM featured prominently in most narratives, with participants reporting a range of experiences. Those who reported positive experiences said that they enjoyed attending the groups because they met new friends, felt more involved in their community, and were able to engage with others in thoughtful and important discussions about HIV and STDs. One participant, named Roberto (aged 21) said:

[My partner, José,] has attended for years, but when I arrived he brought me … and that’s when I realised that there were many people like me. … So then I started going too. I liked it because it is a very good environment and I meet new people, people like me.

Speaking to why his group was important to him, José (aged 36) said, “Well, it’s just that with gay Latinos you feel more, well, more at home.” Others reported mixed or negative experiences. These reports focused on group infighting or how some members used meetings to cruise for sex, which some participants felt disrespected their relationships. Of the cruising, Marco (aged 49) said:

[T]hey would throw you to the people that were called ‘the old guys’ and that was a pack of wolves, brother. They would bitch at each other; they would say bullshit and a half. ‘Here is my phone number.’ ‘Are you active?’ ‘Are you passive?’ ‘What is your dick like?’ So vulgar, brother, just filthy. That is the word I would use: filthy. And they would cruise and bitch. I left [and] didn’t return.

He added, “[N]one of them wanted to respect the relationship. A lot of them, when I told them that I had a relationship, they just went away. Then you realise that that friendship was not a friendship.” Another participant, Raphael (aged 40), also described the negativity, saying:

You don’t realise that it is recycled filth – filth meaning racism, sexism … and internalised homophobia. … Who needs enemies? We are doing to ourselves what supposedly they are doing to us somewhere out there. … I can just go to a gay bar or to a gay group and, sure enough, I will be stepped on and torn apart. … [T]hey don’t stop to think or to consider, ‘Oh my God, this is so toxic, so caustic.’

In response, various participants offered suggestions for improving the groups. Suggestions included coming together as a community, supporting and respecting one another, and making HIV prevention efforts more supportive and inclusive for couples.

Being a Latino gay couple

Present in every interview was the lived experience of being a Latino gay couple. Every participant commented positively on this experience and most felt it added something special to their relationships. One participant, Alejandro (aged 40) said, “For me, personally, it is better because you share the same culture and the same way of thinking.” His partner, Manolo (aged 47), shared similar sentiments,

“With him it’s even better because we speak Spanish all the time at home, we eat the same foods, we like the same kind of clothing, we attend religious festivals together and things like that.”

As these comments show, cultural overlap, such as a shared language, familiar cuisine, the same customs, and the same religious upbringing, was important. Also important was having similar values around family, relationships, and community. Raphael (aged 40) explained, “What I love about my relationship is that it reconfirms the love that I feel towards my community and my culture, towards my Latino culture.” Gabriel (aged 28) said:

Well, he and I being Latinos, we understand what a relationship is in some sense. Latino people are more connected to their family too. So I would say that if I’d ended up with someone who wasn’t Latino my beliefs would have been different, my life would have been different. It’s good to know that you’re with someone who understands your roots and holds your beliefs.

Others spoke of having shared experiences, perspectives, and goals with their partners. Manuel (aged 31) explained:

I think it has all been good. Maybe if my partner were from another country it would be different because of the different cultures, different ideas. But, as we’re both Mexicans, we share many things. In his experiences that he has told me about, he identifies very much with things that have happened to me. We have a lot in common and everything has been positive, nothing negative.

Of future plans, one participant, Esteban (aged 34), remarked, “[W]e have plans to move back to Mexico. I think that if I had a partner from another country – how could I make that work? We would have to split up.” Raphael’s (aged 40) comments about his relationship were especially enthusiastic and summarised the sample’s attitudes as a whole:

It’s the kind of pride that gives me peace. We like the music, food, movies, our language. How we see things the same way. By sharing a plate of arroz con pollo, by giving each other little gifts, by commenting, ‘That looks good on you.’ Things that are everyday things but have a cultural brushstroke in them – I had never experienced that. And to be able to enjoy this community, this freedom to be gay, of the world of possibilities that opens up for you in this country, and especially in this city, as a gay immigrant, with a person who is also a gay immigrant at your side, it is fabulous, indescribable.

Discussion

Despite epidemiologic data that show high rates of HIV infection among Latino men in the USA (CDC 2009) and behavioural studies that indicate LGM are at increased risk for HIV (Díaz 1998; Díaz and Ayala 1999, 2001; Díaz, Ayala, and Bein 2004; Harawa et al. 2004; Rhodes, Yee, and Hergenrather 2006), our sample of nine Latino gay male couples engaged in four behaviours that may actively reduce their HIV risk.

While the typography of agreement types reported by participants largely fell in line with previous research, two important differences stood out. First, nearly every couple reported having a closed agreement. This places our sample in the upper end of the range of percentages of couples who have closed agreements as reported by previous research (Adam 2006; Hoff and Beougher 2010; Hoff et al. 2010; Prestage et al. 2006; Shernoff 2006). Second, many couples appended ‘just in case’ clauses, giving them back-up plans that could mitigate risk in the event sex with an outside partner occurs. Taken together, these differences suggest that these men approach their agreements from a practical standpoint as they balance their personal, romantic, and sexual needs with their concerns around HIV.

Literacy around HIV was high among participants, who reported few incidences of high HIV risk sexual behaviour, displayed a thoughtful understanding of HIV, described using condoms, and tested for HIV with some frequency. Importantly, these behaviours show participants taking an active role in both promoting their health and reducing their risk. They also demonstrate that deficit models of HIV risk, which overemphasise the individual and underestimate the effect of one’s environment, may not capture the full range of LGM’s experiences (Díaz 1998; Díaz and Ayala 1999, 2001).

That a majority of participants had exposure to social support groups for LGM further distinguishes the participants in our sample. Attendance at such groups may increase self-esteem and encourage self-efficacy when making safer sex choices (Ramirez-Valles et al. 2010), thus affecting the participants’ sexual behaviour by steering them away from potential HIV risk. Given that HIV infection rates remain a serious concern, further research into the effects of attendance at these groups, and whether behaviour change can be sustained over time, may help HIV prevention with LGM.

That every participant felt positively about being in a relationship with another Latino gay man is an important finding. It is clear from the data that this is of great value to the participants and may benefit their sense of wellbeing and relationship satisfaction. It may also lay the foundation for a relationship that provides support for reducing HIV risk (Hoff et al. 2007; Julien, Chartrand, and Bégin 1996) and resiliency against hopelessness and fatalism (Díaz and Ayala 2001). It also raises questions about whether differences exist for LGM in interracial relationships, as these men face a different set of relational circumstances as a result of having non-Latino partners (Bianchi et al. in press). Future inquiries that compare and contrast these two groups, and tease apart the challenges to and the benefits of different racial/ethnic couplings, may assist future HIV prevention efforts.

Approaching agreements from a practical standpoint, maintaining a high literacy around HIV, having exposure to social support groups for LGM, and having a supportive partner may have a beneficial effect on these LGM’s efforts to remain HIV-negative. However, there are three underlying issues of concern where long-term HIV prevention is a goal.

First, those with closed agreements who believe that open agreements are riskier may be lulling themselves into a false sense of security (Chng and Géliga-Vargas 2000). Studies examining the role of primary partners in HIV transmission continually find that they are a leading source of new HIV infections (Davidovich et al. 2001; Kippax et al. 2003; Moreau-Gruet et al. 2001; Sullivan et al. 2009). While not all of the infections reported in these studies are the result of broken closed agreements, some are. And given that half of the couples in our sample reported broken agreements may increase their HIV risk. That nearly every partner who reported breaks only sometimes or never disclosed is disconcerting. From an HIV risk perspective, when partners do not disclose they may inadvertently place their partners at increased risk. From a relationship satisfaction perspective, undisclosed breaks may generate feelings of guilt and create distance and tension between partners (Hoff and Beougher 2010). Those in relationships may find it helpful to regularly check in with their partners about their agreements, especially when doubts arise. LGM in particular may be wrestling with sexual silence and have trouble approaching their partners if they feel the need to talk (Carrillo 2002; Díaz 1998). While some couples may agree and find it preferable not to talk about sex with outside partners, choosing not to disclose when rules are broken is a different form of silence and may jeopardise the safety and satisfaction agreements have to offer. As a result, LGM may need additional support around disclosure and mitigating the potential emotional fallout.

Second, reports by three participants that HIV infection is inevitable could undermine their efforts to practice safer sex (Díaz and Ayala 1999, 2001). Although HIV infection is far from inevitable, some LGM nonetheless embody a sense that their fate on this issue has already been decided. That sense, coupled with experiences of homophobia and racism, may increase the possibility of that fatalism becoming a self-fulfilling prophecy. While low in number in our sample, fatalism among LGM has been well-documented elsewhere (Ayala and Díaz 2001; Campo et al. 2005; Díaz 1998; Díaz and Ayala 1999, 2001). Those who feel HIV infection is inevitable may need additional support to change their attitudes around HIV and keep them HIV-negative.

Third, based on comments from those involved in social support groups for LGM, not everyone received the support they needed. Group dynamics that were antagonistic to couples alienated some participants, which led them to distance themselves from the groups or leave them altogether. Critically, those who distanced themselves from the groups also distanced themselves from the groups’ HIV prevention efforts. That exposure to these groups may actively reduce HIV risk raises concerns around the long-term HIV prevention needs for those who have left them or who have no access to them. To increase their effectiveness, social support groups should broaden their focus to include those in relationships. Integrating issues relevant to LGM in intimate relationships may benefit individuals like those in our sample as well as single LGM.

There are several limitations to this study. First, a majority of participants were recruited from social support groups for LGM. Exposure to these groups may have increased the participants’ willingness to take part in the study, readiness to talk candidly about their sexual behaviour, and likeliness to have internalised safer sex practices. While this may have a limiting effect on the generalisability of the findings for LGM as a whole, it illuminates the relationship as an important context for better understanding sexual behaviour HIV risk for LGM. Second, all participants were residents of the San Francisco Bay Area. As such, they may differ from LGM living in suburban or rural areas in the USA, where gay communities are smaller, hidden, or nonexistent and access to HIV prevention messages is more sporadic. Third, all participants in the present analysis were HIV-negative and in concordant negative relationships. Concordant positive and discordant couples may demonstrate different relational and sexual dynamics and it would be inaccurate to assume the motivations behind their agreements and safer sex efforts are the same. Fourth, the HIV status of participants was self-reported; no actual testing occurred. We forewent testing because we are most interested in how one’s perception of his own as well as his partner’s HIV status guides his sexual behaviour.

The present study explores relationship dynamics, agreements, and HIV risk among Latino gay male couples. Building on previous research, it incorporates both partners in data collection, makes race/ethnicity an explicit focus of inquiry, and uses intimate relationships as a context for understanding HIV risk for LGM. We uncovered four behaviours that may help couples in our sample remain HIV-negative and discussed three important concerns where long-term HIV prevention with LGM is a goal. Future research with gay couples should continue to include race/ethnicity as an important factor of analysis for understanding HIV risk as well as safer sex choices. Additional qualitative studies involving Asian and Black gay men, as well as interracial couples, are warranted and should be pursued.