It is now possible to say that there is undeniable evidence of a slow growth in incidence among gay men in the UK since 1999. HIV diagnoses among gay men have increased from just under 1400 in 1998 to over 2,100 (accounting for late reports) in 2004, a 36% increase. Some of this may be due to more gay men coming forward for tests.

However, real incidence among gay men, according to the UAPMP findings of the HPA, has certainly not decreased since the 1990s. Real incidence was worked out by taking the number of infections in GUM clinic attendees recorded by the UAPMP and then performing a second detuned assay, which fails to pick up on infections less than six months old (because these have lower antibody titres). It is important to emphasise that this is a relatively small sample and the results of the chart below do not reach statistical significance. However if they reflect the population of gay GUM clinic attendees at large, then during the last two years about 4% of gay men in London and about 2% outside London became infected with HIV. For comparison, this represents a considerably greater incidence than the estimated 1.2% a year among gay men in San Francisco discovered by similar means but less than half of the 8% incidence found in Baltimore.

Estimated HIV incidence in gay men attending GUM clinics, UK

(Source: HPA see http://www.hpa.org.uk/infections/topics_az/hiv_and_sti/publications/annual2004/fop_3_hiv.pdf)

At the moment it is probably too early to say that a second wave of HIV infection has hit gay men the low figures of the late 1990s could be the exception, and huge increase in sexually transmitted infections among gay men in recent years have only so far appeared to translate into a relatively modest increase in HIV infections. It is also worth pointing out that this relatively slow increase occurs against the background of a constantly increasing proportion of gay men in the population who have HIV.

Rates of prevalent diagnosed HIV infection acquired through sex between men have been calculated using the numbers of males aged from 15 to 49 years for each region/country of the UK, with the highest rates seen in the London Region (351 per 100,000), followed by the South East Region (65 per 100,000) and the North West Region (64 per 100,000). The true prevalence of diagnosed HIV infection in men who have sex with men is likely to be higher than these estimates, as they comprise only a small proportion of males aged from 15 to 49 years and the regional distribution is not uniform.

The annual London Gay Men's Sexual Health Survey (LGMSHS), which asks over 8,000 gay men in community venues about their sexual behaviour and asks them to provide an anonymous saliva sample to test for HIV prevalence, found an overall HIV prevalence in London of 10.9% when its findings were last published in a journal (in 2004, from a series of surveys ending in 2000).

The London Gay Men's Sexual Health Survey has now been expanded to include Brighton and Manchester. In March 2005 the lead researcher Julie Dodds told the 2005 8^th CHAPS gay mens prevention conference in Bristol that the most recent survey for which figures had been compiled (2002) found an HIV prevalence among gay men in Brighton of 14%, London of 12.5%, and in Manchester of about 8%. The lower prevalence in Manchester was largely due to the fact that this group was, on average, younger.

HIV prevalence is heavily skewed by age in the gay population, with the 2000 survey finding that only 1.6% of under-25s had HIV. This rose to a peak of 19.7% among men aged 35-40 and was lower in men above 40. This almost certainly reflects the peak incidence of HIV in gay men around 1990; peak prevalence in 2005 would probably be in the 40-45 age range. However it also attests to the fact that the more sex you have had, the more likely you are to have HIV, and older men are more likely to have it simply through the accumulation of sexual exposures.

It also found that HIV prevalence was skewed by ethnicity (with non-whites 50% more likely to have HIV) and education (men who left school at 16 had an 18% prevalence).

The UAPMP monitors HIV prevalence in patients attending 15 GUM clinics in England, Wales, and Northern Ireland, seven in London, and eight outside. The survey methodology has been described previously, and can be found in the Unlinked Anonymous HIV Prevalence Monitoring Programme Annual report and the PHLS website.

The UAPMP figures (see charts below) reveal that the increase in new diagnoses among gay men since 1999 has been echoed by an increase in prevalence among GUM clinic attendees but that this increase is exclusively restricted to London. A lot of this increase may be due to gay men previously diagnosed with HIV becoming more sexually active again and contracting STIs. However while it is too early to talk of an increase, there is certainly no evidence of a decrease in prevalence in men aged less than 25 years. The proportion of previously undiagnosed gay men under 25 has varied little between 1994 and 2003, staying at about 4% in London and about 1.6% outside London, indicating continuing transmission in this group of young homosexual and bisexual men. The proportion of men who were aware of their HIV infection prior to their clinic visit and who presented with an acute STI, has increased from 12% in 1994 to 35% in 2000.

(Source: HPA see http://www.hpa.org.uk/infections/topics_az/hiv_and_sti/publications/annual2004/ua_gum_2003.pdf)

Many HIV-infected gay men have been engaging in unsafe sex that could have transmitted HIV to their HIV-negative partners. Against a background of a doubling in homosexually acquired gonorrhoea over two years 2000-2001, 26% of homosexual and bisexual men who were aware of their HIV infection prior to their GUM clinic attendance and 41% of those who were unaware of their HIV infection, were also infected with an acute sexually transmitted infection.

Using the proportion of undiagnosed HIV infections from this survey and data from the Survey of Prevalent Diagnosed HIV Infections, CD4 Monitoring and Natsal 2000, estimates of the total number of prevalent HIV infections in the UK have been made. At the end of 2002 over 20,000 men who have sex with men were estimated to be living with HIV infection in the UK, around one fifth of whom were undiagnosed.

A significant number of HIV-positive gay men still remain undiagnosed after a GUM clinic visit. In London in 2003, 79% of those who turned up at a GUM clinic and had HIV were already diagnosed. But of the 21% who were undiagnosed, more than half 54% - went away without their infection being diagnosed.

Outside London the HIV prevalence is far lower but the proportion of men turning up at GUM clinics who already know their status is lower too. Of the gay men with HIV who turned up at clinics outside London in 2003, 42% knew they had HIV. Of the 58% who did not, two-thirds were diagnosed at the visit but a third went away undiagnosed.

This is because HIV testing is still opt-in and is not included automatically among the STD tests offered at check-up. The uptake of voluntary confidential testing (VCT) for HIV in gay men, however, increased between 1997 and 2001 from 40% to 54% in London and from 60% to 64% elsewhere.

Cohort studies

The only UK study which has been able to track a cohort of gay and bisexual men is The National Gay Men's Sex Survey, carried out by Sigma Research.

This cannot be an objective survey of the gay male population of the UK, as like virtually any imaginable gay survey it depends on respondents who self-select as gay. Sigma recruit their respondents in three ways: 1) Via a booklet questionnaire picked up at community venues 2) via an online survey accessed via a number of gay websites and 3) up until 2004, via clipboard questionnaires at gay festivals.

Reports from this cohort may be able to offer a useful counterbalance to trends in voluntary HIV test reporting as a way of demonstrating trends in the incidence of HIV infections amongst gay and bisexual men, but neither can claim to capture an objective sample of gay men adjusted for confounding factors like age and sexual activity.

Both HIV prevalence and HIV testing history among the SIGMA cohort have remained remarkably similar in the eight years 1997-2004, with no apparent increase in either the number of testing or the number with HIV. (Source: Sigma Research)

However the composition of the Sigma cohort has changed over time. In recent years a higher proportion of respondents have come from the online part of the survey, relatively fewer from booklet questionnaires picked up at community venues, and the ‘festival’ section of the survey was dropped altogether for 2005 because it was labour-intensive. Online respondents have consistently tended to be younger and to have tested less often for HIV than respondents via booklets and festival questionnaires, and it is a least possible that the apparent lack of change in HIV prevalence Sigma found may be confounded by a change in the demographics of the respondents, due to the lower prevalence identified in younger men.

However as we see from the UAPMP GUM prevalence figures captured above, the increase appears to be an entirely London-based phenomenon, and as a national survey (unlike the London/Brighton/Manchester Gay Men's Sexual Health Survey), which does not concentrate solely on gay men socialising in urban ‘villages’, it may have captured a truer picture of HIV prevalence in the wider community. High HIV prevalence may be concentrating solely within certain urban communities of gay men with multiple partners.

Will we observe increases in HIV infection amongst gay men?

However the balance of available evidence suggests that, after a dramatic fall in new cases of HIV infection amongst gay and bisexual men during the mid1980s, new infections are happening more frequently.

Reports of rectal gonorrhoea in men increased fivefold in 1990 after a consistent decrease between 1982 and 1989. Rectal gonorrhoea is believed to be a reliable marker for unprotected anal intercourse and was chosen as the single best marker of behaviours indicative of risk of sexual transmission of HIV in the Health of The Nation targets in 1992. Reports of gonorrhoea among men who have sex with men reported by GUM clinics rose by 165% (1495 to 3964) in the UK between 1995 and 2004. However the 2004 figures was slightly lower than the 2003 figure (by 3%).  

Recent outbreaks of syphilis in Brighton, Manchester and London have been concentrated in gay men. Syphilis cases in men in general increased 17-fold from 109 cases in 1995 to 1976 cases in 2004, and increased 28-fold in gay men, from 36 cases in 1995 to 1065 cases in 2004. Between the years 1998 and 2004, syphilis cases in gay men went up exponentially, doubling every year. Last years increase of 36% in men and 29% in gay men is therefore something of a slowing down of the rate of increase.

Gonorrhoea is also used as an epidemiological marker for unprotected sex. However this does not mean it is a marker for HIV-serodiscordant unprotected sex. If a significant proportion of gay men are choosing preferentially to have unprotected sex, when they have it, with men of their own HIV status, then the gonorrhoea statistics may over-predict rises in HIV. So while gonorrhoea incidence in gay men went up 165%, HIV incidence has only gone up by about 40%.

But gone up it has: and evidence from both the national Gay Mens Sex Survey and the London Gay Men's Sexual Health Survey suggest that the amount, both of unprotected anal intercourse and serodiscordant anal intercourse has gone up. Here, for instance are the findings from the Sexual Health Survey up to 2000:

(Source: BMJ Journals. See Dodds in references below)

There are probably several reasons why HIV infection may be increasing again amongst gay men:

• Some gay men may be choosing to have unsafe sex in circumstances which they mistakenly believe to be `low risk'. For example, Project SIGMA has found that gay men are more likely to have unprotected anal sex with regular partners (e.g. in relationships, reflecting the power of emotional attachment) than with casual partners. Recent research suggests that some men may be basing their decisions about whether or not to practise safer sex on more or less informed guesses about their own and/or their partners' HIV antibody status.
• Other research suggests that gay men are basing decisions about whether to have unprotected sex on the basis of their or their partners HIV viral load, mistakenly assuming that an undetectable viral load in plasma also means undetectable HIV in semen. In fact researchers have found only a weak or even no correlation between plasma and seminal viral loads, and from 12% to 20% of gay men have more HIV in their semen than in their blood.
• Today, unsafe sex is more risky than it ever was, because, as the overall prevalence among gay men increases, there is a greater likelihood that one's partner (or oneself) will be HIV-positive.
• Accidents do happen. Many men report occasional `slip-ups', whether these were due to problems such as condoms failing or being unavailable in the `heat of the moment', or to the effects of drink and/or drugs.
• Studies have shown that a key factor behind the widespread uptake of safer sex among gay men in the 1980s was the perception among gay men that safer sex was the expected sexual norm among their friends and partners. The sense of `newness' and urgency of AIDS and the widespread sense of a community mobilising to protect itself were to some extent lost during the late 1980s.
• Safer sex campaigns for gay men were all but ignored in the late 1980s and early 1990s. Even organisations such as The Terrence Higgins Trust, which are popularly seen as `gay organisations', provided little or no new prevention initiatives for gay men during these years. The need to inform gay men of all ages who are just beginning their sexual careers, and to provide ongoing support and advice for those who have already adopted safer sex, was almost entirely ignored.
• HIV prevention campaigns directed specifically at gay men with HIV have been very rare, despite the fact that HIV positive people are in a disproportionately powerful position to reduce HIV transmission, by being the people who both have reliable knowledge of their HIV status and can choose to disclose it.

Targeted mass media campaigns have been a strength of the CHAPS initiative, which has been delivering HIV prevention and health promotion to gay men in England and Wales since the mid' 1990s.

References

Dodds JP. Increasing risk behaviour and high levels of undiagnosed HIV infection in a community sample of homosexual men. Sex Transm Infect  80: 236-240, 2004.

Grassly NC. Host immunity and synchronized epidemics of syphilis across the United States. Nature 433: 417-421, 2005.