HIV risk behaviour in gay men: on the rise?

In most developed countries it was gay men who were affected earliest and most severely by HIV. In the early 1980s, incidence rates of HIV of 20% a year were described, leading to a prevalence of up to 50% in gay men in some US cities.¹ Death and disability became part of life in gay communities. As soon as AIDS was recognised as being a sexually transmissible infection, education and prevention campaigns were rapidly implemented. In many countries, these were mobilised by gay organisations in the face of government inaction. These campaigns led to large scale declines in the practice of unprotected anal intercourse, which had previously been the norm, and declines in the incidence of sexually transmissible infections, including HIV.¹,² This behaviour change was so successful in reducing the incidence of HIV that gay communities became the example of what is possible in HIV prevention in the community.

Since the mid-1990s, new combination antiretroviral therapies have led to major declines in rates of death and serious morbidity.³ Twenty years after the start of the HIV epidemic, a new generation is growing up with HIV. Within gay communities, there is some evidence that optimism over HIV treatments is associated with a rise in risk behaviours.⁴

The uptake of condom use by gay men has been remarkable, but whether this is now the cultural norm is uncertain. In this issue of the BMJ, Dodds et al report slightly increasing levels of unprotected anal intercourse among gay men in the United Kingdom (p 1510).⁵ Although most gay men still use condoms for anal sex most of the time, similar trends are reported from the United States and Australia.^6–8 Rates of gonorrhoea are also increasing among gay men in these countries.⁶,⁹,¹⁰ Despite these trends there is no clear evidence of an increase in the incidence of HIV. This may be related to the difficulty in measuring the incidence of HIV infection. Few countries monitor the incidence of HIV infection at a population level. However, data from Australia's surveillance system for newly acquired HIV infection do not show a recent increase in HIV incidence in homosexual men.⁹ On the other hand, there has been a report of increasing HIV incidence in a cohort of gay men attending one clinic in the United States.¹¹ In the United Kingdom, the number of HIV diagnoses reached its highest annual level in 1999, but this was largely due to an increase in heterosexually acquired infection, and there was little evidence of a change in rates of diagnoses of HIV infection in homosexual men.¹²

There are at least two reasons why an increase in unprotected anal sex among gay men may not increase the incidence of HIV. Firstly, as the HIV viral load in semen is reduced to below detectable levels in most men being given combination antiretroviral therapy, treatment with these agents may decrease infectivity.¹³ Secondly, the increases in unprotected anal intercourse may not put the participants at risk for new HIV infection.

Much of the unprotected anal intercourse reported by gay men is between seroconcordant partners, where there is no possibility of a new HIV infection.¹⁴ In Australia, however, increases in unprotected anal intercourse have occurred both in HIV positive and in HIV negative men, and with both casual and regular sexual partners.⁸,¹⁵ Thus it is likely that some of the increase in unprotected anal intercourse does involve sexual encounters where there is a risk of new HIV infection.

There are several challenges in preventing a resurgence of HIV in gay men. Firstly, we need better systems for the timely reporting and measurement of trends in risk behaviours and in the incidence of HIV infection. Monitoring systems for risk behaviours for HIV need to take into account the context of the sexual encounters: for example, whether the partner was casual or regular, and whether or not the HIV status of the partner was known. Secondly, there is a challenge to educationalists to design and implement behaviour change programmes that work in the new context of HIV infection, and a challenge to researchers to evaluate these programmes to ensure that the most effective interventions are broadly implemented. Thirdly, there is a challenge to gay communities around the world to recognise and respond to this threat.

In a situation where the immediate, overwhelming threat of death from AIDS is no longer present, promoting condom use is likely to be much more difficult than in the 1980s. However, if antiretroviral therapy becomes less effective because of viral resistance, then the rate of infection may well increase and current levels of unsafe sexual behaviour may lead to an increased incidence of HIV infection.

Papers p 1510