|Home | About | Journals | Submit | Contact Us | Français|
As HIV research and prevention efforts increasingly target gay men in relationships, situational factors such as couple serostatus and agreements about sex become central to examinations of risk. Discordant gay couples are of particular interest because the risk of HIV infection is seemingly near-at-hand. Yet little is known about their sexual behaviors, agreements about sex, and safer sex efforts. The present study utilized longitudinal semi-structured, qualitative interviews to explore these issues among 12 discordant couples. Findings show that nearly every couple had agreements about reducing the likelihood of HIV transmission from one partner to the other. Negotiating these agreements involved establishing a level of acceptable risk, determining condom use, and employing other risk-reduction techniques, such as seropositioning and withdrawal. For half of the couples, these agreements did not involve using condoms; only two couples reported consistent condom use. Despite forgoing condoms, however, none reported seroconversion over the course of data collection. Additional issues are raised where long-term HIV prevention is concerned. Future prevention efforts with discordant couples should work with, rather than fight against, the couple’s decision to use condoms and endeavor to complement and accentuate their other safer sex efforts.
Despite three decades of prevention efforts infection rates among gay men remain stubbornly high (CDC, 2011). In San Francisco, one of the worst affected U.S. cities, prevalence has stabilized, albeit at hyperendemic levels (Scheer, et al., 2008). Researchers there, as elsewhere, are increasingly calling for contextualization of gay men’s sexual behaviors that acknowledge situational factors to explain why some gay men knowingly expose themselves to HIV risk. One such context is relationship status (Harawa, et al., 2004; C. C. Hoff, et al., 2006; Scheer, et al., 2008).
Examining HIV risk for gay men in relationships makes sense. Studies show that primary partners are a leading source of HIV infection (Davidovich, et al., 2001; Kippax, et al., 2003; Moreau-Gruet, Jeannin, Dubois-Arber, & Spencer, 2001; Sullivan, Salazar, Buchbinder, & Sanchez, 2009) and that gay men in relationships engage in high rates of unprotected anal intercourse (UAI) with their primary partners (Elford, Bolding, Maguire, & Sherr, 1999; Hays, Kegeles, & Coates, 1997; C. C. Hoff, et al., 2006; C. C. Hoff, et al., 1997; Kippax, et al., 2003; Lattimore, Thornton, Delpech, & Elford, 2010; Moreau-Gruet, et al., 2001; Prestage, et al., 2008; Prestage, et al., 2009; Remien, Carballo-Dieguez, & Wagner, 1995; Van der Bij, et al., 2007; Wagner, Remien, & Carballo-Diéguez, 1998). Several factors may have influence, including relationship length, inconsistent condom use over time, disinterest in using condoms, the desire for sexual and relational intimacy, and the de-emphasis of concerns about HIV (Davidovich, de Wit, & Stroebe, 2004; Eaton, West, Kenny, & Kalichman, 2009; Frost, Stirratt, & Ouellette, 2008; C. Hoff, Beougher, Chakravarty, Darbes, & Neilands, 2010; C. C. Hoff & Beougher, 2010; C. C. Hoff, et al., 1997; Moreau-Gruet, et al., 2001; Nieto-Andrade, 2010; Ostrow, et al., 2002; Palmer & Bor, 2001; Prestage, et al., 2008; Remien, et al., 1995; Theodore, Duran, Antoni, & Fernandez, 2004). Relationship dynamics, such as agreement type (e.g., open or closed) and couple serostatus (e.g., concordant negative, concordant positive, or discordant couples) offer additional context, as couples who allow sex with outside partners and/or couples where partners are different HIV statuses present different situational factors affecting HIV risk (Eaton, et al., 2009; C. Hoff, et al., 2010; C.C. Hoff, et al., 2009; C. C. Hoff, Coates, Barrett, Collette, & Ekstrand, 1996; C. C. Hoff, et al., 1997; Palmer & Bor, 2001; Pawlicki & Larson, 2011; Remien, et al., 1995; Sacco & Rickman, 1996; Shernoff, 2006; Wagner, et al., 1998; Wheldon & Pathak, 2009).
Discordant couples (i.e., one partner is HIV-negative, one is HIV-positive) offer unique opportunities for HIV risk and prevention. Discordant couples engage in high rates of UAI and do so for many of the same reasons as other couples (Bouhnik, et al., 2007; Denning & Campsmith, 2005; Lattimore, et al., 2010; Nieto-Andrade, 2010; Prestage, et al., 2008; Wagner, et al., 1998; Whittington, et al., 2002). They also habituate to risky sexual behaviors, but do so with potentially greater consequences (Cohen & Gay, 2010; Crawford, et al., 2003; Ostrow, et al., 2002; Palmer & Bor, 2001). Accordingly, they must adjust their sexual behaviors, agreements about sex, and safer sex efforts in the face of increased risk of HIV transmission from one partner to the other (Eaton, et al., 2009; Moreau-Gruet, et al., 2001; Nieto-Andrade, 2010; Theodore, et al., 2004). Discordant couples must also cope with the impact HIV/AIDS has on satisfaction, intimacy, and communication (Palmer & Bor, 2001; Remien, et al., 1995; Remien, Wagner, Dolezal, & Carballo-Dieguez, 2003). Despite these challenges, discordant partners remain committed to each other (C. Hoff, et al., 2010; Nieto-Andrade, 2010; Palmer & Bor, 2001; Wrubel, Stumbo, & Johnson, 2008). And with 10–25% of gay couples being discordant (C. Hoff, et al., 2010; Moreau-Gruet, Dubois-Arber, & Jeannin, 2006), it makes sense to tailor prevention efforts to their needs.
One leading prevention effort for couples is negotiated safety (Davidovich, et al., 2004; Elford, et al., 1999; Jin, et al., 2009; Kippax, et al., 2003). Although not foolproof, negotiated safety offers a degree protection – but only for concordant negative couples. Discordant couples are, effectively, left to find their own means of reducing risk.
This study explores sexual behaviors, agreements about sex, and safer sex efforts among discordant couples. Our aim is to understand which sexual behaviors discordant couples engage in and how they navigate risk and negotiate safety in order to highlight long-term HIV prevention concerns for and inform future interventions with this population.
This analysis utilizes longitudinal qualitative data, collected annually for three years, between April 2007 and June 2009, from 12 discordant couples. Participants belonged to a longitudinal qualitative study of gay couples (n = 40 couples) that included approximately equal numbers of each couple serostatus group. Participants were randomly selected from a larger, longitudinal quantitative study (n = 566) (Neilands, Chakravarty, Darbes, Beougher, & Hoff, 2010). The objective of both studies was to examine various relationship dynamics in gay couples and their relationship to HIV risk. Participants were contacted by phone to gauge their interest in participating in the qualitative arm (one interview annually for three years) that was supplemental to the quantitative arm (one computerized survey approximately every six months for three years). We invited a total of 63 couples to participate before reaching our goal of 40 couples. Eligibility for the quantitative arm required participants be at least 18 years old, be in their relationship at least three months, know their own and their partner’s HIV status, and identify as gay/bisexual. Couples with discrepant reports of partner serostatus were ineligible. Couples were eligible to participate if both partners met all eligibility criteria.
Each year, couples were given scheduled appointments for 60–90 minute semi-structured, qualitative interviews. Informed consent was obtained from each partner individually before the first interview. Participants were interviewed separately to encourage candid discussion of their relationship. Interviews examined the following topics: relationship history, a natural history of any agreements about sex, broken agreements, sexual behaviors, sexual and relational satisfaction, and HIV risk and safer sex efforts. Participants were paid $40.00 each as incentive.
Interviews were digitally recorded and transcribed verbatim. Transcripts were checked for accuracy (e.g., mistakes, misspellings, or omissions) and analyzed using codes developed from the study team’s previous research (C. Hoff, et al., 2010; C. C. Hoff & Beougher, 2010; Neilands, et al., 2010). An initial analysis of the transcripts, utilizing a Grounded Theory approach, also informed code development and their application (Denzin & Lincoln, 2003; Lindlof & Taylor, 2002). Using Transana version 2.3-MU (Woods & Fassnacht, 2007), transcripts from all three data collection points were pooled for analysis to provide a fuller picture of the couple, their sexual behaviors and agreements about sex, and changes over time.
Thirteen couples in the qualitative arm were discordant. For the present analysis, we utilized coded text from 12 of those couples. One couple, who broke up before their second interview, was excluded so that all data reflected a true longitudinal sample.
Of the 12 couples, seven included partners who were both white, one included partners who were both Latino, and four couples were interracial. One participant identified as bisexual; all others as gay. The mean age was 46 years (range: 31–66), age difference between partners was nine years (range: 1–22), and relationship length was 10 years (range: 10 months-33 years). All HIV-positive participants were on antiretroviral therapy (ART).
Analysis of the qualitative data revealed that all but one couple had agreements about reducing the likelihood of HIV transmission from one partner to the other (hereafter referred to as “safety agreements”). Negotiating these agreements involved establishing a level of acceptable risk, determining condom use, and employing other risk-reduction techniques.
Nearly every participant described the discussions they had with their partner about what felt safe and what felt risky. Drawing a line between the two formed the boundary of the couple’s safety agreements. One HIV-positive participant explained,
We talk about the things that are risky. … [F]rom the beginning basically, those things were brought up: What was risky that we know and [what was] risky that we know but that we would be willing to take the risk.
Similarly, another HIV-negative participant commented, “I know that fucking him without a condom is unsafe. But we’ve decided … that would be a tolerable risk area for us.” Participants stressed the importance of agreeing on where they drew the line. One HIV-negative participant said, “If it was just his decision, without any input from me at all, we wouldn’t be together. It’s mutual between the two of us.”
In many cases, however, the HIV-negative partner appeared to be the one leading the discussion. One HIV-positive participant offered the following sentiments:
When we first started, we used condoms. Then, he didn't [want to]. … Although I prefer to use them, I've adapted to what he wants. And … it feels safe within the dynamics of our relationship because we're open and honest and we have these agreements. … But our doctor doesn't think we should be doing this.
Couples used condoms consistently, over time and across sexual behaviors, or they used them inconsistently or not at all. Two couples reported using condoms whenever they had anal sex. (Four couples reported not having sex at the time they were interviewed.) Comments from the HIV-negative partner in both couples reflected a hard resolve to remain HIV-negative. One said, “We will not engage in unprotected anal sex. … I’m not interested in becoming HIV-positive and that’s not a difficult thing to negotiate for us.” The other HIV-negative partner was similarly clear: “We take precautions and I’m careful. Period. We are never going to put me in a position where I can get HIV.” He continued, “HIV changes a lot of things... it's here, now it's next to me, and it's scary. So, it cannot be another way anymore … it's about being safe.” His HIV-positive partner offered similar sentiments, “[W]hen we have anal sex we use condoms. … Since we know that there's still risk, even though it's very low, because my HIV status is undetectable.”
Six couples reported using condoms inconsistently or not at all. The choice to forgo condoms was grounded in the couple’s level of acceptable risk, sexual preferences, and personal feelings. One HIV-positive participant described:
We used condoms in the beginning and eventually decided [not to use them]. He was comfortable with it because we talked about it and, based on the literature we read, the chances of conversion for the top, unprotected, are a lot less than for the bottom… it was an informed decision by the two of us.
An HIV-negative participant said:
I’ve yet to incorporate them in a hot way. They just break everything for me. Not only do I not get my erection, I lose desire completely. So I stopped using them because it destroyed what good sex we could have. And I haven’t worked on it. I don’t want to work on sex. I want it to be natural and condoms are not natural to me.
Most couples employed a variety of other techniques to reduce HIV risk. These include seropositioning (i.e., the HIV-negative partner is insertive), monitoring the HIV-positive partner’s viral load, withdrawal before ejaculation, and avoiding sex when sick or injured (e.g., cut in mouth). Most used these techniques in combination. One HIV-negative participant, who does not use condoms and bottoms during anal sex, said:
He’s got a lot of T-cells, his viral load is undetectable, he doesn’t have precome, he has difficulty orgasming. So we feel there’s some time to play that allows me to orgasm and we take care of him afterwards. And often, it’s just pleasurable enough for him, at least he says, to go as far as we do. I mean, he doesn’t need to orgasm [in me] to enjoy it.
Some couples spoke of minimizing the frequency, duration, and intensity of anal sex as a way of reducing HIV risk. As one HIV-negative participant described it, “He’s always concerned because sometimes he wants to play rough. But when he plays rough he’s scared that something can happen to me.” His HIV-positive partner similarly responded, “I’m a top and he’s a bottom. So we have to be careful. … I’m not that aggressive with him as I think I would have been if that equation didn’t exist.”
Despite these risk-reduction efforts, many of the participants expressed discomfort with and anxiety over their sexual behaviors. One HIV-negative observed:
Well, there’s a feeling I have that it’s like playing with a loaded gun. And that’s not good. There was a lot of guilt on both our parts earlier on, but we don’t have anal sex very often… So, we’re trying to minimize our chances, but we’re taking a lot of chance, so that part of it doesn’t feel good and I wish we didn’t have to think that way.
Nearly every couple had safety agreements and negotiating them involved establishing a level of acceptable risk, determining condom use, and employing other risk-reduction techniques. Safety agreements were remarkably stable over the course of data collection. Half of couples reported using condoms inconsistently or not at all. Their motivations reveal a preference for other risk-reduction techniques. No seroconversions were reported.
There are limitations to this study. Though couples were randomly selected for the qualitative arm, they were part of a larger quantitative study whose sample was recruited via a convenience strategy. Similarly, all couples were residents of the San Francisco Bay Area. Therefore, generalizations should be made cautiously. Finally, the HIV status of participants was self-reported – no actual testing occurred. We forwent testing because we are interested in how one’s perception of serostatus guides sexual behavior.
This study significantly expands what is known about the HIV risk and safer sex efforts of discordant couples. While previous research verified the existence of agreements about sex and charted their typography, it focused on whether couples allow sex with outside partners. Data from this study demonstrates that discordant couples also make agreements about safety, which, with epidemiological support, could expand negotiated safety to include discordant couples by adding rules such as periodic STD testing for both partners and regular viral load testing and ART adherence for the HIV-positive partner. Lacking other strategies to model, this study shows discordant couples striking out on their own, sometimes against their own doctors’ recommendations and oftentimes against conventional HIV prevention messages, negotiating agreements that balance their sexual and relational needs with their concerns about HIV.
Important issues remain where long-term HIV prevention is concerned. The decision to forgo condoms demands scrupulous adherence to ART, close monitoring of viral load, and regular testing for STDs and prompt treatment if they occur (Hallett, Smit, Garnett, & de Wolf, 2011; Vernazza, Hirschel, Bernasconi, & Flepp, 2008; Whittington, et al., 2002). Adroit, forthright communication between partners is also important to maintain both partners’ health (Darbes, Chakravarty, Beougher, Neilands, & Hoff, in press; Remien, et al., 2003). Furthermore, it is imperative that discordant couples have accurate and complete information about the relative risk and safety of the other risk-reduction techniques they employ (Cohen & Gay, 2010; Jin, et al., 2009; Persson, 2011; Vernazza, et al., 2008). For example, couples spoke of limiting the frequency, duration, and intensity of anal sex to avoid HIV transmission, however, no data is available on these methods and their efficacy is dubious at best. Couples must separate fact from fiction to make knowledgeable decisions about their level of acceptable risk.
Safety agreements offer unique opportunities to reduce risk among discordant couples. Interventions could target the negotiation process to help them make the safest agreements possible. Ideally, this would occur at the beginning of the relationship as safety agreements, and the decision to forgo condoms, gel quickly (Davidovich, et al., 2004; C. C. Hoff, et al., 2006; Remien, et al., 1995). That most couples adhered to their safety agreements over time suggests partners are committed to each other’s health and wellbeing. Interventions could latch onto this to sustain safer sex efforts over time. That condom use was conspicuously absent from most safety agreements exposes a critical disconnect between many of those who dispense HIV prevention messages and many of those for whom those messages are targeted. Typically, sexual behaviors are presented as increasing one’s HIV risk. The participants, however, framed them as decreasing HIV risk. There is truth to both perspectives.
On the one hand, some gay men overestimate how safe and underestimate how risky their sexual behaviors are, especially when they make assessments retrospectively. Additionally, some lose sight of their cumulative risk over time by focusing on individual risky episodes. On the other hand, some counselors and health practitioners, who offer absolutes as prevention strategy and eschew the relative protection afforded by certain behaviors, preclude a more honest discussion of risk. Unfortunately, this dynamic frames condom use as a choice between having a sexually and interpersonally satisfying relationship and one’s long-term health. Certainly, condoms work for some discordant couples. However, for the majority of couples who do not use them, prevention messages promoting condoms may be ignored. Prevention efforts should work with, rather than fight against, the couple’s decision to use condoms (Mao, et al., 2011; Persson, 2011). Rather than ask discordant couples to use condoms every time they have sex, future interventions could implement a situational approach by asking them to use condoms when the HIV-positive partner is insertive or when either partner has anal sex with outside partners. Research suggests that condom use may increase if couples use them in a targeted fashion (Hallett, et al., 2011). Future interventions should endeavor to complement and accentuate discordant couples’ other safer sex efforts.
The authors extend their thanks to the participants, to Research Assistants Brianne Amato, Luis Ciprez, Rand Dadasovich, Michael Díaz, Anthony Freeman, Walter Gómez, Binh Nguyen, and Brad A. Vanderbilt, and to Postdoctoral Fellows Carmen Gómez Mandic and Anu Manchikanti Gómez. This research was supported by grants RO1 MH 75598 and MH 65141 from the National Institute of Mental Health.