 |
|
|
|
|
|
|
||||
|
|
|
Oral presentation to the Gay and Lesbian Medical Association conference in Vancouver in August 2000.
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| With REGULAR Partners | insertive | receptive | |
| Without ejaculation: | 30% | 34% | |
| With ejaculation: | 24% | 27% | |
|
At all with a REGULAR partner: 49% |
|||
| With CASUAL Partners | insertive | receptive | |
| Without ejaculation: | 16% | 14% | |
| With ejaculation: | 10% | 6% | |
|
At all with a CASUAL partner: 26% |
|||
|
|
|||
One of our earliest published findings was the high level of risk behaviour we found among young gay and bisexual men in Vancouver. We have consistently seen high levels of risk behaviour within this cohort over the past five years.
As shown in this slide, among men with regular partners, almost half reported at least one episode of unprotected anal sex in the previous year, either insertive or receptive. Among men with casual partners, over a quarter reported at least one episode of unprotected anal sex in the previous year.
Note that for those with casual partners, significantly more of them engaged in unprotected anal sex without ejaculation. Since this practice doesn't protect against the transmission of HIV and other STDs, prevention messages are needed which target specific sexual activities among young gay and bisexual men.
| Adjusted odds ratio | |
| Education < high school | 2.40 |
| Low social support | 1.65 |
| Used poppers in past year | 2.40 |
| Nonconsensual sex over age 12 | 1.85 |
In our first published paper by Steffanie Strathdee and colleagues, which appeared in the Journal of AIDS and Human Retrovirology in 1998, we examined the social determinants of sexual risk-taking.
In comparing risk-takers to non-risk-takers, we found that risk-takers were significantly more likely to have less formal education, less social support, and higher depression scores than non-risk-takers. They were also more likely to use recreational drugs, particularly poppers. Risk-takers were also significantly more likely to have experienced non-consensual sex, particularly as a youth or an adult.
Steffanie Strathdee et al, CMAJ 2000
| INSERTIVE | RECEPTIVE | |||||
| At follow-up | At follow-up | |||||
| No | Yes | No | Yes | |||
| At baseline | No | 113 | 43 | No | 99 | 47 |
| Yes | 20 | 47 | Yes | 24 | 53 | |
|
|
||||||
| Odds ratio: 2.2 |
|
|||||
We also looked at trends in risk behaviour over time. We particularly wanted to know how many of the participants who had always used condoms for anal sex in the year prior to baseline then went on to have unprotected anal sex in the year after entering the study, as this would indicate that their risk for getting HIV was increasing rather than decreasing.
This analysis is from our third published paper, again by Strathdee and colleagues, which appeared in the Canadian Medical Association Journal in January of this year.
We compared participants who increased their risk behaviour by going from safe to unsafe from one year to the next, with those who went in the opposite direction (that is, those who were at risk before their baseline visit who consistently used condoms for anal sex in the year after baseline). We found that among those who changed their behaviour from one year to the next, twice as many increased their risk as decreased them. This trend held true for both insertive and receptive anal sex with regular partners which are the two cases shown on this slide and also for insertive anal sex with casual partners (but not for receptive anal sex with casual partners).
This disturbing trend towards high-risk sex underscores the urgent need for targeted prevention efforts among young gay and bisexual men.
|
With REGULAR partners:
|
With CASUAL partners:
|
The question on everyone's mind, of course, is why are young gay and bisexual men continuing to engage in high risk sexual behaviours despite a decade of HIV prevention efforts? Some have suggested feelings of invincibility among young people in general, a sense of fatalism or inevitability among young gay men in particular, or a lack of direct experience with the reality of AIDS among this age group.
While there are no easy answers, our findings do provide the most commonly cited reasons for having unprotected anal sex, which differ somewhat depending on whether the risk behaviour is with regular or casual partners.
The most common reasons cited for not using condoms for anal sex with regular partners were involvement in a long-term and/or monogamous relationship, a desire for increased intimacy, and the assumption that both partners were HIV-negative. Among men with casual partners, the assumption that both partners were HIV-negative still figured prominently, along with getting "carried away," being intoxicated and not caring at the time as the most common reasons given for not using condoms for anal sex. For both regular and casual partners, other common reasons related to increased comfort and pleasure.
Many young gay and bisexual men who engage in unprotected anal sex appear to be practising a form of "negotiated safety": they are far more likely to dispense with condoms with regular partners than with casual partners; and although the reasons for having unprotected anal sex differ between regular and casual relationships, in both cases one of the most common reasons cited was the assumption that both partners were HIV-negative.
As eligible Vanguard participants can't have previously tested positive, our current baseline HIV prevalence rate of 2.65% can't be extrapolated to the wider community.
More significant is our HIV infection rate. We initially found HIV infection rates of over 3% per year, which is double the average infection rate seen in the American HIVNET cohorts for this population.
The infection rate in our cohort when calculated as a fraction of 100 person-years of observation time has since declined in our cohort, as can be expected in this type of longitudinal study. Nonetheless, we continue to see new infections among our participants. We currently have 19 seroconverters, for a current overall infection rate of 1.16% per year.
| Study year | HIV rate |
| All years: | 0.9% |
| 1st | 1.0% |
| 2nd | 0.9% |
| 3rd | 0.2% |
| 4th | 1.0% |
| 5th | 2.0% |
Some of our participants are also at risk for HIV through injection drug use, and the recent outbreak of HIV among injection drug users in Vancouver no doubt has had an impact on the infection rate within this study.
When we remove from the analysis participants who reported injection drug use, our overall infection rate drops to just under 1% per year.
When we break down those infections into an annualized incidence rate, however, we see a significant and troubling increase in infection over the past three years, from 0.2% in the third year of the study to ten times that rate (or 2%) in the past year.
This is consistent with reports from other cities, including San Francisco and Toronto, of increasing HIV incidence among young gay and bisexual men.
| HIV- (n=1325) |
HIV+ (n=26) |
|
| Education < high school | 12% | 27% |
| Unstable housing | 6% | 19% |
| Ever been paid for sex | 25% | 50% |
| Recent crack/cocaine use | 30% | 54% |
| Unprotected anal receptive sex | 40% | 60% |
|
All p-values < 0.05 |
||
We then wanted to find out which risk factors are predictive of seroconversion among young gay and bisexual men.
In order to increase our statistical power, research associate Amy Weber combined our data with that of the Omega Cohort, which is a very similar study of gay and bisexual men in Montreal, Quebec. Together we were able to analyse the questionnaires and test results of nearly 1400 participants.
We found that participants in the two studies who were HIV-negative at baseline and who subsequently seroconverted were significantly more likely:
After controlling for other factors, however, unprotected anal sex did not emerge as an independent predictor of seroconversion.
The only factor that did emerge as an independent predictor of seroconversion was having ever been paid for sex.
| Predictors of HIV prevalence at baseline: | |
| For all men (n = 1373) | |
| Education < high school | 2.33 |
| Recent crack/cocaine use | 2.23 |
| Ever been paid for sex | 2.08 |
| Predictors of HIV seroconversion: | |
| For all men (n = 1351) | |
| Ever been paid for sex | 3.06 |
| For men reporting anal receptive sex (n = 1127) | |
| Unprotected anal receptive sex | 2.33 |
| Unstable housing | 4.88 |
We then calculated odds ratios for both HIV prevalence at baseline and HIV seroconversion during the course of the two studies.
We found that participants who tested HIV-positive at baseline were over twice as likely to have dropped out of high school, to have used crack or cocaine, and to have ever been paid for sex.
Participants who seroconverted during the course of the two studies were three times as likely to have ever been paid for sex.
Among those who had engaged in anal receptive sex, participants who had had unprotected anal sex were, not surprisingly, over twice as likely to seroconvert.
But an even stronger association was found between seroconversion and housing stability: among participants who engaged in receptive anal sex, those with unstable housing were nearly five times as likely to seroconvert as those with stable housing.
| Stable (n=57) |
Unstable (n=53) |
|
| Education < high school | 31% | 54% |
| Ever in jail | 39% | 65% |
| Sex with both men and women | 35% | 58% |
| Sex with both men and women | 35% | 58% |
| Recent crack use | 28% | 47% |
| Injection drug use | 18% | 45% |
|
All p-values < 0.05 |
||
Since sex trade involvement and lack of stable housing appear
to be key risk factors for HIV among young gay and bisexual men,
we had a closer look at the interaction between the two by comparing
sex trade workers on the basis of their housing stability.
Compared to sex trade workers with stable housing, those with unstable housing were significantly more likely to have dropped out of high school, to have ever been incarcerated, to have sex with both men and women, and to use crack and injection drugs.
As we showed in the previous slide, unstable housing is strongly predictive of seroconversion among young gay and bisexual men in general, and here we see that among sex trade workers in particular, lack of stable housing is associated with other risk factors for HIV and, in general, poorer social determinants of health.
| RoC* (n=624) |
Aboriginal (n=57) |
|
| Been paid for sex (ever) | 21% | 65% |
| Been paid for sex (in past year) | 11% | 53% |
| Non-consensual sex (ever) | 33% | 47% |
| NCS by family member | 9% | 23% |
| NCS under age 12 | 16% | 26% |
|
* Rest of Cohort. |
||
One in ten Vanguard participants is Aboriginal, and Aboriginal men who have sex with men may be particularly at risk for HIV infection. These data are from our second published paper by Katherine Heath and colleagues, published in the International Journal of STD and AIDS in 1999.
We found that Aboriginal men who have sex with men face social and economic disadvantages in almost every measure of health, including poverty, mental health issues, drug and alcohol use, and history of sexual abuse.
As shown on the slide, Aboriginal participants were significantly more likely to have been paid for sex, and to have been sexually abused. Almost half the Aboriginal participants reported having been sexually abused, compared to one-third of the rest of the cohort. Aboriginal men were also more likely to have been sexually abused by a family member and at a younger age than non-Aboriginals.
Interesting, however, we did not find elevated sexual risk taking behaviour among Aboriginal participants. Although Aboriginals were more likely to test positive for HIV at baseline, they were no more likely to seroconvert during the study. Given the known social determinants of HIV risk, however, it's clear that Aboriginal men who have sex with men require culturally-specific HIV prevention efforts, including access to sexual abuse counselling.
| Prevalence | Incidence | |
| Gay men | 1.31* | 0.76* |
| Bisexual men | 1.43* | 0.96* |
|
|
||
| Rest of Cohort | 1.13 | 0.93 |
| Sex trade workers | 7.48 | 5.01 |
|
|
||
| Stably housed STWs | 7.14* | 3.48 |
| Unstably housed STWs | 7.48* | 8.96 |
|
* Differences not statistically significant. |
||
We did, however, find elevated HIV rates among other sub-populations within the cohort.
Bisexual men had marginally higher HIV prevalence and incidence than did gay men, but this difference wasn't statistically significant.
Sex trade workers, however, had dramatically higher HIV prevalence and incidence than the rest of the cohort: seven and half percent compared to just over one percent prevalence, and five percent per year incidence compared to under one percent for the rest of the cohort.
When we compared sex trade workers with stable and unstable housing, we again noticed a dramatic increase in HIV incidence. Although the difference in prevalence between these two groups isn't statistically significant, as you can see the incidence jumps from three and a half percent to nearly nine percent per year for sex workers without stable housing.
This infection rate is similar to that seen among all gay and bisexual men at the height of the epidemic in the mid-1980s.
 |
 |
 |
 |
But it's not all bad news. In our latest published paper by Kevin Craib and colleagues, which appeared in the journal AIDS in February of this year, we compared Vanguard participants with participants in the Vancouver Lymphadenopathy-AIDS Study (the VLAS), which is the earlier cohort of 1000 gay men that started in 1982.
We found that young gay and bisexual men in the mid-90s were much more likely to use condoms than were their counterparts in the mid-80s.
This slide is a bit busy, but if you focus just on the left-hand bars in each of these four graphs, you can see that Vanguard participants were significantly less likely than VLAS participants to report "never" using condoms for anal sex.
In all cases, the majority of VLAS participants in the mid-80s reported "never" using condoms for anal sex, whereas the majority of Vanguard participants reported using condoms for anal sex at least some of the time.
While we may take it for granted nowadays that gay and bisexual men use condoms for anal sex, it should be remembered that this is the result of a significant change in normative sexual behaviour within a community that prior to the AIDS epidemic had no use for condoms.
 |
Even though the half of the Vanguard participants did not always use condoms when they had anal sex, we can still see a significant reduction in HIV incidence between young gay and bisexual men in the 1990s compared to those in the '80s.
Taking the date of questionnaire completion as the zero point for each participant, and following each participant for 44 months of observation, a marked difference emerges in the HIV incidence patterns for the two cohorts.
At 36 months, the cumulative rate of HIV incidence in the Vanguard Project was 3.6%, compared to 7.9% in the VLAS.
Although Vanguard participants reported more sexual partners and a higher frequency of anal intercourse than VLAS participant, they were nonetheless significantly less likely to seroconvert in the mid-1990s than their VLAS counterparts a decade prior. This can presumably be attributed to their increased use of condoms for anal sex.
 |
 |
Next we merged all the data from the two cohorts and stratified HIV incidence based on reported condom use with casual sexual partners.
Significantly higher HIV incidence can be seen among participants in both cohorts who reported "never" using condoms for anal intercourse with casual partners, compared to those who "ever" used condom for anal sex.
This is true for those who reported either insertive or receptive anal sex with casual partners.
Although young gay and bisexual men are not always using condoms, even inconsistent condom use among participants in either of these cohorts appears to have had a protective effect against HIV seroconversion.
 |
Concern has been expressed that recent advanced in antiretroviral therapies may cause gay and bisexual men to be less concerned about HIV.
We recently had the opportunity to to examine this question in collaboration with researchers conducting similar studies of young gay and bisexual men in London, England, and Sydney & Melbourne, Australia.
Last year, all three studies incorporated the same questions into their questionnaires regarding gay and bisexual men's attitudes towards safer sex and HIV in light of these improved HIV treatments.
The results in this somewhat obscure slide were presented as a discussion poster at the recent International AIDS Conference in Durban, South Africa, and are explained in more detail in our Five Year Report and on our web site. Suffice it to say that in all four cities, only a minority of young gay and bisexual men expressed optimism in light of the new HIV treatments; most participants remained cautious about the risk of unprotected sex.
Striking variations in the relationship between this so-called "HIV optimism" and both HIV status and sexual behaviours indicate that gay and bisexual men's responses to the new drug treatments vary widely between countries. This absence of clear universal trends emphasises the importance of local monitoring of the knowledge, attitudes and behaviours of young gay and bisexual men.
We are still seeing high levels of sexual risk behaviours among young gay and bisexual men, and in our cohort we have seen a trend towards increasing risk behaviour.
And we are still seeing disturbingly high HIV infection rates among young gay and bisexual men, and recent data both from our cohort and other studies suggests that HIV incidence is escalating among this population.
Marginalised sub-populations within this community are at particular risk for HIV infection and/or have specific risk profiles, including Aboriginals, bisexual men and sex trade workers.
Finally, we can't overstate the importance of the social determinants of health that place young gay and bisexual men at increased risk for HIV infection, particularly low education, low social support, low self-esteem, mental health issues (particularly depression), sexual abuse and unstable housing.
Our findings suggest that prevention efforts are urgently needed that specifically target young gay and bisexual men.
Prevention efforts must take into account social support, socio-economic factors and social connectedness, which are key to HIV prevention. Renewed prevention efforts urgently needed. Prevention efforts must: specifically target young gay and bisexual men; be culturally specific; target populations at elevated risk (e.g. sex trade workers); take into account sexual abuse a nd abuse counselling; and take into account the social and sexual realities of young gay and bisexual men.
INTRODUCTIONS