The proportion of gay men in the US who have had anal sex without a condom at least once in the past year has increased by 9 percentage points in the last six years, data from the country’s Centers for Disease Control and Prevention (CDC) reports.
The CDC report found that, by 2011, gay men formed more than half of all people diagnosed with HIV in all but two US states (South Dakota and Pennsylvania). Gay men especially predominated in the west-coast states (California, Oregon and Washington), where they formed more than 80% of the diagnosed population, while they formed less than 60% in the metropolitan Boston-to-Washington-DC axis, and in the southern states of Louisiana and Florida.
The proportion of men who had condomless sex in the last year rose from 48% in 2005 to 57% in 2011; this was highly statistically significant (p=<0.001).
In men who were HIV negative or did not know their status, there was a 10 percentage point rise from 47 to 57%, while in men who knew they had HIV, the rise in condomless sex was 7 percentage points, from 55 to 62%, and this was not statistically significant.
Relative to 2005, the proportion of men saying they had had condomless sex increased by 21 percentage points in HIV-negative or untested men, and by 12.5 percentage points in HIV-positive men.
The proportion reporting any condomless sex varied significantly by age in men with HIV, with 55% of 18- to 24-year-olds reporting it but 72% of 30- to 39-year-olds. There was no relationship to age in HIV-negative or untested men.
As has been reported elsewhere, significantly fewer black gay men reported condomless sex than white gay men, whether they had HIV or not.
In men with HIV, 56% of black men and 68% of white men reported condomless sex at least once in 2011; in HIV-negative and untested men it was 49% and 58% respectively, with even higher rates (62%) in Hispanic HIV-negative men.
The CDC also looked at the proportion of men who didn’t use a condom the last time they had sex and then looked at the proportion that was with partners of different or unknown HIV status; this was defined as risky sex.
Because respondents have provided blood samples for HIV testing since 2008, the CDC can now report whether this risky sex was reported by someone HIV negative, diagnosed HIV positive, or undiagnosed HIV positive. They found that the proportion of men who reported risky sex was essentially identical in HIV-negative and diagnosed HIV-positive men – 12 and 13% respectively. There was a slight fall between 2008 and 2011 in the proportion of HIV-positive men who knew their status who reported last-time risky sex, from 16 to 13% and, although this was not significant, it became significant in white HIV-positive men, where the fall was from 17 to 11% (p = 0.01).
In HIV-negative men, the proportion reporting risky sex did not change between 2008 and 2011.
In contrast, in men whose blood tested HIV positive but who did not know their status, the proportion reporting last-time risky sex was more than twice as high: in 2011, a third overall had risky sex last time they had sex and this had not changed from 2008. In white men, 37% reported having risky sex the last time they had sex.
This does not imply that not knowing one’s status leads to risky sex: rather, it shows that people who have HIV but don’t know it are, unsurprisingly, more likely to have taken sexual risks recently and are more likely to have been recently infected.
In this 'HIV-positive unknown' group, but not in others, there was an almost-significant association of risky sex with age, with 40% of 30- to 39-year-olds reporting risky sex (p=0.07).
The CDC also reported on HIV testing frequencies in survey participants, who were recruited in gay venues in major cities. It found that only 8% of gay men had never tested for HIV; that a third of HIV-negative or untested men had tested in the last three months, 50% in the last six months, and two-thirds in the last year; and that only 5% last tested more than five years ago.
HIV prevalence in men who had never tested was 12%, while the prevalence of undiagnosed HIV was even higher in men who last tested negative but had not taken a test for over five years – 17%.
Even in men who had last tested less than three months ago, however, 4% had undiagnosed HIV, and 7% of those who last tested between six months and two years ago had HIV.
There are a couple of limitations to the CDC’s study. They do not distinguish between receptive and insertive anal sex, and since a receptive HIV-negative partner is 18 times more likely to acquire HIV than if they are insertive, this may make a considerable difference to HIV risk – in other words, the CDC’s surveillance cannot gauge if ‘strategic positioning’ makes a difference.
Perhaps more significantly, they did not report on antiretroviral therapy or measure viral load among the diagnosed HIV-positive men. As many will have been on antiretroviral therapy and therefore effectively non-infectious, this means that the proportion of men having sex that really does risk HIV acquisition or transmission is overestimated in this report. The CDC does note that, for every 100 persons who are diagnosed with HIV, eight further infections are avoided simply due to those diagnosed starting antiretroviral therapy and also reducing their risk behaviour.
The CDC also notes that these figures are taken from cyclic surveillance of gay men in large US cities, and sexual risk and testing behaviours might differ considerably in rural areas. It also notes that the figures were collected before pre-exposure prophylaxis was licensed and that “for some [gay men] at high risk, taking pre-exposure or post-exposure prophylaxis can reduce risk”.
They recommend that gay men test for HIV at least once a year. However, given 7% prevalence in men testing less than every six months, and the fact that only 50% of gay men managed to test at least this frequently, this recommendation seems rather cautious.
Centers for Disease Control and Prevention HIV testing and risk behaviours among gay, bisexual and other men who have sex with men – United States. Morbidity and Mortality Weekly Report 62(47):958-962, 2013.
