Gay men, risk and relationships

Relationships are good for you. People in a relationship tend to enjoy better health, suffer less depression and live longer than single people [1]. However, among gay men, an intimate relationship may also present a risk of HIV infection. Numerous studies have shown that gay men are less likely to use condoms with their regular partner than with a casual partner [1–3]. For many men, not using a condom with their regular partner is an expression of intimacy and trust [4–6]. Paradoxically, these two vital aspects of a loving relationship may act as a counterpoise with respect to HIV. Not all unprotected anal intercourse (UAI) presents an HIV transmission risk, however, particularly in a relationship. The risk can be eliminated if both partners test HIV negative, and agree only to have UAI with one another. This risk reduction strategy, known as negotiated safety, was first reported among gay men in Sydney [7]. Negotiated safety requires both men to be confident they are HIV negative (i.e. seroconcordant), and to comply with their agreement only to have UAI with each other. Seroconcordant UAI between two HIV-positive men also presents no risk of HIV transmission to an uninfected person, but raises the possibility of cross-infection for the positive men themselves. Any examination of sexual risk behaviour among gay men must therefore make the distinction between seroconcordant UAI, which presents little if any risk of HIV transmission, and UAI with a partner of unknown or discordant status, which presents a substantially higher risk. Surprisingly, not all researchers make this important distinction [8–10]. This crucial differentiation has been made in a study of Swiss gay men published in the current issue of AIDS [11]. In Switzerland, five out of six gay men (84%) with a steady partner reported the safe management of HIV risk within their relationship. Consistent use of condoms was the most common means of managing risk (reported by 43% of men), followed by negotiated safety (25%) and not practising anal sex (16%). The remaining men (16%) reported what the authors called ‘inadequate management of HIV risk’ with their steady partner. Most of these men (10%) chose to have UAI with a partner of unknown HIV status based on a personal assessment of past risks or because they were faithful. Overall, slightly more than one-third (36%) of Swiss gay men in a relationship reported UAI with their steady partner. However, the majority (25%) were HIV-negative men reporting UAI only with another HIV-negative man (negotiated safety), whereas a few (1%) were HIV-positive men also with seroconcordant UAI partners. It was the remaining 10% who reported high-risk UAI with their steady partner, i.e. with a person of unknown or (potentially) discordant status. Had the investigators used any UAI with a steady partner as a measure of risk they would have substantially overestimated the degree of sexual risk within a relationship. This highlights the importance of distinguishing seroconcordant UAI from UAI with a person of unknown or discordant status. How does the level of risk reported between steady partners in Switzerland compare with recent surveys among gay men in other countries? Despite some differences in survey design and methodology, two important findings emerge. First, in most countries [12–15] (P. Adam, personal communication) between one half and two-thirds of gay men had a steady partner at the time of the survey, highlighting the importance of the relationship as a context for high-risk behaviour. Second, the percentage of men in a relationship who reported high-risk sexual behaviour with their steady partner varied dramatically between countries. In Amsterdam, nearly 30% of gay men in a relationship reported high-risk sexual behaviour with their steady partner, i.e. UAI with a person of unknown or discordant HIV status [16]. In Germany, the corresponding figure was 23%[12]; in California and Oregon 21%[13]; in France 13% (P. Adam, personal communication); and in London and Sydney 12%[14,17]. Why are there such striking differences between countries in the level of sexual risk within a relationship? One, but not the only factor is the extent to which couples are aware of their own and their partner's HIV status. In Switzerland, just over half the couples said they knew each other's HIV status. In Amsterdam [16], the figure was substantially lower, whereas in Australia [15] it was considerably higher. This may reflect differences between countries in both the availability of HIV testing, as well as incentives for seeking an HIV test such as access to highly active antiretroviral therapy. Other factors include agreements between partners concerning unprotected sex [15], the disclosure of HIV status [18], and the extent to which health promotion programmes have focused on sex in relationships. The paper from Switzerland serves to remind us that mapping patterns of sexual risk behaviour is complex and requires a degree of sophistication. In behavioural research, investigators should distinguish men who are in a relationship from those who are not; seek information on UAI insertive and receptive, with both steady and casual partners; and ascertain the HIV status of the respondent as well as his sexual partners, in order to differentiate seroconcordant UAI from UAI with a person of unknown or discordant status. This distinction is particularly important when interpreting time trends in sexual risk behaviour. An increase over time in the percentage of men reporting UAI [19] does not, in itself, signify an increase in the level of sexual risk taking. It is the increase in the percentage of men reporting UAI with a partner of unknown or discordant HIV status that provides a marker of increasing high-risk behaviour [20]. This needs to be differentiated from any parallel increase in the proportion of men reporting seroconcordant UAI. Concerning couples, because sexual risk occurs between steady partners, any investigation of risk behaviour should preferably include rather than exclude [21] these men. A relationship may be a crucible of both security and risk. Behavioural research conducted over the past decade has consistently shown that a substantial minority of gay men report high-risk sexual behaviour with their steady partner [11–16,22–26] (P. Adam, personal communication). Interventions to reduce the risk of HIV transmission among gay men should therefore target couples as well as individuals [27], focusing on the quality of communication within the relationship and the degree of support partners provide one another [1]. A number of health promotion strategies have been developed, including mass media campaigns plus community and clinic-based interventions. Posters, cards and booklets focusing on risk within relationships were central to health promotion initiatives targeting couples in London (`Thinking it through') and Sydney (`Talk, test, test, trust'). A project in London currently allows couples, working in groups with a facilitator, to explore the potential for behaviour change and the maintenance of low risk within the relationship. A genitourinary medicine clinic provides sessions in which couples can be screened for sexually transmitted diseases/HIV together, and then discuss the test results with a counsellor as well as a physician. The effectiveness of these interventions requires further examination as does their transferability from one setting to another; or indeed, from one relationship to another. Not all relationships last for ever; in the Swiss study the median length was 3 years. Consequently, gay men need to be equipped with the skills to negotiate risk reduction not only with a current but also with a future steady partner. In this way, with the further development of couple-focused interventions, we may be able to anticipate a time when an intimate relationship protects against, rather than presents a risk for, HIV.