HAART, viral load and sexual risk behaviour

Since it became clear that highly active antiretroviral therapy (HAART) substantially reduces viral load as well as the morbidity and mortality associated with HIV/AIDS [1–3], there has been widespread discussion about the effect of HAART on sexual behaviour, particularly among gay men. It has been suggested that if people diagnosed with HIV believe a low viral load reduces infectivity they may be more likely to have unsafe sex [4]. In the wider community, there is concern that fear of AIDS as a life-threatening disease may dissipate leading to complacency around safer sexual practice [5]. In sum, advances in treatment may trigger risky sexual behaviour. There has indeed been a substantial increase in high risk sexual behaviour and sexually transmitted infections (STI) among gay men in many cities around the world since HAART first became available in 1996. This has been seen in Europe [6–12], Australia [13,14], Canada and the USA [15,16], among both HIV-positive as well as HIV-negative men. Some cities have also reported an increase in HIV incidence [17,18]. Has taking HAART or having a reduced viral load triggered risky sexual behaviour among HIV-positive gay men? In the current issue of AIDS, Van de Ven et al. report on the sexual behaviour of 119 Sydney gay men in HIV serodiscordant regular relationships, namely where one partner was HIV-positive and the other HIV-negative [19]. The majority of men (n = 82), interviewed between 2001 and 2003, said they did not engage in unprotected anal intercourse (UAI) with their regular partner. However, among the remaining 37 men who did report such UAI, this was more likely to occur where the HIV-positive partner was reported to have an undetectable viral load (odds ratio, 2.9; 95% confidence interval, 1.1–7.4; P < 0.05). The findings of Van de Ven et al. contribute further to our understanding of the role of HAART in relation to sexual behaviour. On the face of it, their paper suggests that having a reduced viral load may trigger risky sexual behaviour among HIV-positive gay men. While these findings resonate with those of other researchers [20–22], the authors point out that some studies found no association between viral load and sexual risk behaviour [23–25], including a community-based study in Sydney which they also conducted [26]. It was ‘against this background of mixed findings’ that Van de Ven et al. undertook the present investigation. Crepaz and colleagues have been able to throw light on these disparate findings in a recent systematic review of HAART and sexual risk behaviour [27]. They reviewed 25 papers published between 1996 and 2003 and conducted three meta-analyses to determine whether: (1) being treated with HAART; (2) having an undetectable viral load; or (3) holding specific beliefs about HAART and viral load were associated with an increased likelihood of engaging in unprotected sex. They found the overall association between having an undetectable viral load and sexual risk behaviour was non-significant [odds ratio (OR), 0.99; 95% confidence interval (CI), 0.82–1.21]. In other words HIV-positive people with an undetectable viral load were no more likely to engage in unprotected sex than those whose viral load was detectable. This absence of an association did not vary by risk group, country of study (USA or non-USA) and whether viral load was self reported or taken from medical records. They also found no significant overall association between taking HAART and unprotected sex (OR, 0.92; 95% CI, 0.65–1.31). People with HIV taking HAART were just as likely to engage in unprotected intercourse as those not taking HAART. Speculation that taking HAART or having an undetectable viral load might trigger risky behaviour among people with HIV is not supported by these meta-analyses. It would appear, therefore, that the recent increase in high-risk sexual behaviour among HIV-positive gay men can not be attributed to taking HAART or having an undetectable viral load per se. Nevertheless, the systematic review did find that the likelihood of unprotected sex was significantly higher among people who believed that HAART reduced HIV transmission or who were less concerned about HIV given the availability of HAART (OR, 1.82; 95% CI, 1.52–2.17). This association, which was only significant for gay men, was seen in US and non-US studies among both HIV-positive as well as HIV-negative and never-tested men. Since the primary studies in the meta-analytic review were all cross-sectional the authors recognized that they could not establish causality. Beliefs about HAART may have triggered high-risk behaviour or they may have been used as a post-hoc justification. Recent longitudinal studies have not resolved this issue since they have yielded conflicting results [28,29]. Interestingly, the many primary studies that have examined beliefs about HAART and sexual behaviour all found that only a minority of gay men believed HAART reduced HIV transmission or were less concerned about HIV in the light of HAART [5,7]. Consequently, even where a causal association exists between beliefs about HAART and sexual behaviour, the population-attributable risk associated with these beliefs will be modest [30]. Approximately 10% of the recent increase in high-risk behaviour might be explained by HAART-related beliefs; the remaining 90% can not be explained in this way. At a population level, therefore, it is unlikely that these beliefs alone can explain the increase in high-risk behaviour among HIV-positive as well as HIV-negative gay men since 1996 [7,31]. Where do we go from here? Clearly, the meta-analytic review examining the impact of taking HAART, viral load and beliefs about HAART on sexual behaviour needs to be updated regularly. The current review [27] was based on studies conducted up to 2001. Incorporating studies conducted since then, such as that by Van de Ven et al. [19], will allow us to examine whether the lack of an overall association between taking HAART, viral load and unprotected sex persists or not. The inclusion of longitudinal studies will be particularly valuable. As further primary research findings become available it will be possible for meta-analyses to make the distinction between HIV-positive men who only engage in UAI with a regular partner and those who report UAI with casual partners. The availability of HAART touches on larger social, behavioural and epidemiological issues concerning HIV/AIDS. Although HAART has had an exceptionally positive impact on HIV-related morbidity and mortality, it has also had unintended consequences. For individuals on treatment, this relates to adherence, side-effects and unforeseen longer-term health problems. The number of people with treatment-resistant strains of HIV is increasing and for people with such strains, fewer therapeutic options remain even when they present as treatment naive [32]. At a population level, the result of reduced mortality is more prevalent HIV infection. Although HAART seems to reduce individual infectivity, as more people are living with HIV and continue to have unprotected sex with both regular and casual partners, incident infection at a population level is likely to increase [33]. There are clear policy implications. Regular HIV testing of at-risk populations will lead to earlier identification of those who are HIV-positive so they may benefit from effective treatment. At the same time, interventions should target HIV-positive people to discourage further transmission of HIV and to reduce the likelihood of exposure to resistant virus [34]. Future research into gay men's behavioural responses to HAART could usefully be placed in the context of this broader policy agenda.