Recently diagnosed and younger HIV-positive gay men taking more sex risks in Boston study

A study from a community HIV clinic in Boston, Massachusetts, that treats gay men has found that sexual behaviour that increases the risk of HIV transmission was more common amongst younger gay men and was even more strongly associated with recent HIV diagnosis.

Younger and more recently diagnosed gay men were also more likely to have been diagnosed with an acute sexually transmitted infection in the last year than older gay men and those who had been diagnosed longer. Higher rates of risk behaviour were also, not unexpectedly, associated with methamphetamine, ketamine, poppers and poly-drug use.

This study’s findings conflict with a number of studies (e.g. Gorbach) that have found that people diagnosed with HIV, including young gay men, reduce their risk behaviour substantially in the months immediately following diagnosis, and that risk behaviour has a tendency to revert over time (e.g. McClelland – see aidsmap report).

The expectation that people will reduce their risk behaviour substantially after HIV diagnosis has been used by the United States Centers for Disease Control as one of the reasons to encourage frequent HIV testing in risk groups (CDC – see aidsmap report).

The authors of the present study comment: “These results underscore the need to develop more effective secondary preventions for HIV-infected [men who have sex with men], tailored to more recently diagnosed patients, particularly those who are younger and substance users.”

The study recruited 398 gay men attending Fenway Health, the largest HIV clinic for gay men in New England. Between July 2004 and August 2008 the men were tested for syphilis and for urethral gonorrhoea and chlamydia while medical records were reviewed for diagnoses of rectal and pharyngeal (throat) gonorrhoea and chlamydia.

This was a largely well, and well-educated, patient group. Three-quarters of the patients were white, half were graduates and 40% had an annual income over $40,000: the mean age was 41 (range 20-68), the mean length of time since HIV diagnosis was 8.6 years (range, two months to 22 years) and the mean CD4 count was 523. Two-thirds were taking antiretrovirals (ARVs) and 55% had a viral load under 75 copies/ml.

Thirty-six (9%) of the men had been diagnosed with an STI in the previous year, of whom 69% had syphilis, 33% gonorrhoea and 3% chlamydia. The methods used would tend to under-diagnose rectal and pharyngeal cases of gonorrhoea and chlamydia because many cases have few or no symptoms, so STI incidence may have been higher.

There was a high level of alcohol and drug use in the group: in the last three months 20% reported ‘binge drinking’ (defined as five or more drinks in one sitting), 23% had used methamphetamine and 57% other drugs.

Almost exactly half (50.5%) of the patients reported transmission risk behaviour in the last six months, which was defined as “unprotected insertive or receptive anal sex with…HIV negative or unknown status partners”.

As one would expect, reporting unprotected sex was associated strongly with STI diagnosis, as was drug use. Unprotected receptive sex nearly quadrupled the risk of STI diagnosis and insertive sex doubled it. Methamphetamine use more than tripled the likelihood of an STI, ketamine more than quadrupled it and poppers more than doubled it. Drug use also multiplied the likelihood of risky sex: poppers and methamphetamine fourfold and ketamine sevenfold.

Being younger and more recently diagnosed with HIV were associated both with STI diagnosis and sex risk behaviour. Being younger increased the chance of STI diagnosis by 60% and more recent diagnosis by 50%.

Youth increased the chance of reporting risky sex by 30% and more recent HIV diagnosis by 75%.

A problem with this study is that different risk factors were very highly correlated. For instance, having a detectable viral load increased the chance of reporting risky sex too, but in multivariate analysis this was almost entirely due to the fact that the more recently diagnosed were less likely to be on treatment.

They were also more likely to be young and to take drugs. In the case of drug use multivariate analysis reduced the proportion of transmission risk behaviour due to methamphetamine and other drugs so that ‘meth’ was associated with 2.5 times the chance of risky sex and other drugs doubled this risk.

Age and years since HIV diagnosis, though, proved impossible to untangle as they were so highly correlated. Both were significant predictors of risky behaviour but only years since HIV diagnosis were taken forward into the published multivariate analysis.

This found a strong link between recent diagnosis and risky sex: to be exact, the chances of reporting sex which risked HIV transmission fell by 40% for every 6.7 years diagnosed with HIV: roughly 6% a year. For age the link was somewhat weaker though not significantly so: they found a 24% drop in transmission-risk behaviour for every 8.4 years older: about 2.9% a year.

Because age and time since diagnosis were closely correlated, it is difficult to speculate on the causes of the risky sexual behaviour and subsequent decline; whether these were linked to youth and the effects of ageing, or specifically to do with coming to terms with having HIV.

As well as commenting on the need for better ‘positive prevention’ programmes in this population, the researchers also comment that a high prevalence of STIs in this population may have implications for biomedical approaches to HIV prevention, including ‘test and treat’ strategies, as it may weaken the impact of reducing viral load in the population.