In the first study to track chemsex over time, engagement with chemsex fell by two-thirds between 2015 and 2018, Janey Sewell and colleagues at University
College London report in the June issue of the International Journal of Drug Policy. Chemsex has been defined by researchers as the use of recreational drugs such as mephedrone, GHB/GBL and crystal meth to enhance sex.

Data come from a prospective cohort study, Attitudes to and
Understanding Risk of Acquisition of HIV over time (AURAH2), which recruited
HIV-negative or undiagnosed gay and bisexual men at three London and Brighton sexual
health clinics: 56 Dean Street, the Mortimer Market Centre and the Claude Nicol
Centre.

The data from the study may be unlikely to represent
national trends, as these three clinics have a reputation for providing
high quality services to gay men engaging in chemsex, with substance use and
sexual health being addressed holistically. The clinics are likely to both attract
more patients who engage in chemsex and also be better at supporting them than some
other services.

The participants were recruited between November 2014 and
April 2016, completed an initial questionnaire in the clinic, and were then
invited to complete quarterly online questionnaires for up to three years. Data
for this analysis come from 622 men who completed at least one online
questionnaire. As with all studies, ongoing engagement with the study declined
over time (458 completed the questionnaire 12 months after recruitment, 376 did
so after 24 months, and 72 did after 36 months).

Participants mostly identified as gay (95%), were university
educated (77%) and employed (89%). There was a wide range of ages, with a median
of 34 years. While 84% were white, 43% were born outside the UK.

In the first online questionnaire, 32% said that they had “used
drugs before or during sex (chemsex)” in the past three months. The most common
drug was mephedrone (25%), followed by GHB/GBL (20%) and crystal
methamphetamine (11%).

There was a steady decline in each quarterly survey: in the
last survey, three years after first recruitment, 11% reported any chemsex,
9.7% reported mephedrone use in chemsex, and 8.3% reported GHB/GBL in chemsex. Each of these falls was statistically significant.

However, there was no decline in use
of crystal meth over time.

A plausible explanation of this apparent fall could have been that
men engaging in chemsex would be more likely to drop out of the study, leaving ‘less
risky’ men to provide data. However, the researchers found that there was no association
between chemsex and not subsequently completing questionnaires. In addition, a sensitivity
analysis which only included those men who completed a survey in the last six
months of the study found a similar decline in chemsex over time.

Instead, the researchers offer three possible explanations
for the fall in chemsex:

  • The quality and effectiveness of the chemsex support
    provided in the three clinics.
  • Repeatedly completing questionnaires that encourage
    reflection on behaviour may have helped the study participants become more
    conscious of the consequences of their choices, which could have led to
    behaviour change.
  • The phenomenon described by statisticians as ‘regression to
    the mean’. A variable that is extreme on its first measurement will tend to be
    closer to the centre of the distribution for a later measurement. In other
    words, men may have been recruited to the study at a time of particularly high-risk
    behaviour, risk which would then tend to decline with time.

There were also some changes in sexual behaviour over time.
Men were a little more likely to have condomless sex with more than one partner
as the study progressed, but it was less likely that this sex involved a
partner of unknown HIV status or group sex. There were larger falls in
bacterial STIs (from 26% to 10%).

The researchers note that in
contrast to a previous report from St George’s Hospital in south London
, their
longitudinal study did not find an association between chemsex and a new
diagnosis of HIV. There were relatively few seroconversions (15) among study participants.