Gay men fall into different groups that are at radically different risk for acquiring HIV, Canadian research presented at the 11th International AIDS Impact Conference last week has found.
The findings of the SPOT Study, presented by Amélie McFadyen, a Canadian Research Chair in Health Education, are reminiscent of Australian research first presented at the previous AIDS Impact Conference (Mao), which showed that gay men adopted many different strategies of varying effectiveness to regulate their risk of HIV. But these findings quantify these different behaviours and their likely effectiveness in more detail.
The SPOT study is an ongoing research project offering free, anonymous, rapid HIV testing for men who have sex with men (MSM) in Montreal. Data on demographics and sexual and other behaviour were gathered from 1740 gay men and men who have sex with men (MSM), tested between July 2009 and January 2012 through structured interviews and self-administered questionnaires.
Participants were asked if they always, never or inconsistently used condoms; whether they ‘dipped’ in anal sex (i.e. withdrew before ejaculation); whether their partners’ status was HIV negative, positive or unknown; and whether, if their partner was HIV positive, whether they knew their viral load and if it was undetectable.
A large amount of data were collected in SPOT and some presented elsewhere at AIDS Impact, but the most intriguing analysis found that the men surveyed tended to fall into five ‘risk profiles’ of different sizes. The five risk profiles were created by a statistical approach known as latent class analysis. Some similarities were found amongst these profiles, but distinct strategies were clearly identifiable for each one.
This, the largest group, formed 52% of the men tested or about 900 individuals. This group was least likely to be diagnosed HIV positive (1.2%), but also least likely to have tested for HIV in the last year (47%). Nearly half of the group members (48%) were in a relationship, mainly with another HIV-negative man (84% of their partners were HIV negative at last test, 10% of unknown status and only 6% were HIV positive).
Whether in a relationship or not, they were less likely to have had one-off or occasional casual sex in the last three months than other groups (they had an average of 2.8 one-off or occasional partners in the last three months).
Over the last three months, 12% had had anal sex with a partner of unknown HIV status as a ‘top’ (i.e. they were insertive) and 8% as a ‘bottom’ (i.e. they were receptive – these are not mutually exclusive categories; many men did both). A third of those who had been bottoms had not used a condom (only a tenth of tops).
They had had very little sex with partners of known HIV-positive status (1.5% as top, 0.7% as bottom) of which half had been with partners they knew to have an undetectable viral load. Only 2 to 3% of this group reported sex in the previous three months that put them at risk of HIV.
This group was less monogamous than group one and less likely to know their partners’ HIV status, but maintained high levels of condom use. They formed nearly 18% of the SPOT participants, or about 300 individuals. Two per cent of this group tested HIV positive in SPOT and 57% had had an HIV test in the previous year. Thirty-nine per cent were in a relationship, 68% with a partner whose last test was HIV negative, 6% with an HIV-positive partner and 26% of unknown status (compared with 10% in group one).
They had had an average of 6.3 one-off or occasional partners in the last three months: 2.5 times as many as group one. Strikingly more men than in group one had had anal sex with a partner of unknown HIV status as a top (75%) or a bottom (78%) over the last three months, but they maintained high condom use (93% as a top, 100% as a bottom).
Again, few had had anal sex with partners known to have HIV (2.6% as top, only 0.6% as bottom) and they nearly always used condoms. This group had, on average, the highest educational level but were the least likely to have a primary relationship. They were also the second-most likely to have used post-exposure prophylaxis (PEP). Roughly 5% of this group reported sex in the previous three months that put them at risk of HIV.
Group three formed about 18% of the SPOT participants. They had, on average, the lowest educational status as a group. Like the two previous groups, they tended to avoid anal sex with men they knew to have HIV, but they were less monogamous than group one and had far lower condom use rates with unknown-status partners than group two. Forty-two per cent had a primary relationship, 56% with an HIV-negative partner and only 4% with a positive partner, so 40% did not know their partners’ status.
In this group, 2.6% tested positive in SPOT and 52% had had an HIV test in the previous year.
They had had an average of 6.3 one-off or occasional partners in the last three months. A similar proportion to group two had had anal sex with partners of unknown HIV status over the last three months as group two, though they were rather less likely to be bottom: 78% top, 50% bottom. However, their condom use rates were much lower: only 20% used a condom as top and 25% as bottom, and they were the most likely group to say they had had unprotected sex intentionally.
They had even less sex than groups one or two with partners they knew to have HIV, but their low condom usage rates with partners of unknown status meant that two-thirds of the group had had sex that put them at risk of HIV in the last three months, and numerically they formed the highest proportion of SPOT participants who were having sex that potentially put them at risk of HIV.
Group fourformed 9% of the SPOT participants or about 150 men. They also had high levels of unprotected sex with partners of unknown HIV status but had a lot more sex with men they did know had HIV, and in this group there was clear evidence of ‘viral sorting’ – they almost exclusively restricted unprotected sex with known HIV-positive partners to men with an undetectable viral load, so must have been having some discussion about HIV and viral load status. This was also the group most likely to define as gay rather than any other sexual orientation. They also had the highest alcohol use of any group.
However, they also had the highest proportion of regular partners of unknown status. Forty-two per cent had a primary relationship, 46% with an HIV-negative partner and 7% with a positive partner, so 46% did not know their partners’ status.
They had had an average of 8.7 one-off or occasional partners in the last three months: 2.5 times as many as group one. In total, 4.5% tested positive at SPOT – the highest diagnosis rate in the five groups – and 59% had had an HIV test in the previous year.
Sixty-nine per cent had had anal sex as top with partners of unknown status in the last three months, and 79% as bottom; only 26 and 19% respectively had used a condom in these encounters.
They hardly had any more sex with partners they knew to have HIV than the previous groups, if those partners had an unknown or detectable viral load – 0.6% as top and 3.8% as bottom, and all but one of these few encounters involved using a condom.
But they were strikingly more likely to have sex with men they knew had HIV and an undetectable viral load. Fourteen per cent (21 individuals) had had sex as top and 13.5% (20 individuals) as bottom with ‘undetectable’ men, and condoms had only been used 20% of the time in these encounters.
Despite this, their low condom usage rates with partners of unknown status meant that they shared with group three the highest rate of sex that put them at risk of HIV. Two-thirds fell into this category, but because only half as many belonged to this group as group three they formed a smaller proportion of the at-risk men in the whole SPOT cohort.
Group five,the smallest group, formed 4% of the SPOT group or about 70 individuals. This group was distinct in having by far the highest proportion of regular partners they knew to have HIV: 48% were in a relationship and of them, 50% had an HIV-positive partner, 43% an HIV-negative one and only 7% a partner of unknown status.
Not surprisingly, their HIV testing rate was very high – 74% had had a test in the last year – and their HIV diagnosis rate in SPOT was similar to group three at 2.8%. This group had the highest rates of use of most drugs except alcohol (group four) and cocaine (group three).
They had an average of 8.1 occasional or one-off partners in the last three months, and had somewhat lower rates of casual sex with partners of unknown status than any group except group one. Forty-six per cent had anal sex as top (condom use 45%) and 39% as bottom (condom use 61%) with partners of unknown status.
They had by far the highest rates of sex with partners they knew to have HIV or any group, but there was no evidence of ‘viral sorting’ – if anything, slightly the reverse, though the actual numbers are small for this group. Fifty-five per cent had had sex with an HIV-positive partner of unknown or detectable viral load as top in the last three months (62% with a condom) and 50% as bottom (63% with a condom); for partners with undetectable viral load the figures were 38% as top (67% with a condom) and 33% as bottom (78% with a condom). About 47% of this group had had sex that put them at risk of HIV in the last three months.
This group also had very high usage of PEP: 21% had ever used PEP compared with 17% of group four and 9 to 12% in the other groups.
A rough calculation shows that approximately 20 to 22% of SPOT participants had had sex in the last three months that put them at risk of HIV, and that groups three and four accounted for roughly 80% of these risky epsiodes.
Amélie McFadyen commented that her teams’ results showed that while gay men make strategic use of condoms, their risk-reduction strategies are diversified and complex and go beyond condom use. The groups showed evidence of various combinations of serosorting, viral sorting, negotiated safety, viral load, HIV testing and PEP.
“Public health messages, prevention programs, and counselling must be adjusted to account for this diversity,” she added.
McFadyen A et al.Beyond unprotected anal intercourse, what do HIV negative gay and bisexual men really do? : Latent class analysis draws a complex portrait of risk reduction strategies. Eleventh AIDS Impact Conference, Barcelona, abstract #461, 2013.
Mao LM et al. Rates of condom and non-condom-based anal intercourse practices among homosexually active men in Australia: deliberate HIV risk reduction? Sexually Transmitted Infections 87:489-493, 2011.
