In order to
better understand the Ugandan HIV epidemic, researchers have been conducting
studies with six ‘most at-risk populations’, one of which is men who have sex
the hostile social environment, men who have sex with men can be difficult to
reach. The researchers therefore used respondent-driven sampling in order to
recruit research participants. This is a modified form of snowball sampling
an initial set of respondents recruit people they know, who then recruit other
people they know, and so on. A mathematical model is then used to weight the
sample to compensate for non-random recruitment patterns, so the results should
be less prone to bias.
began in May 2008 and only a few weeks later LGBT activists were arrested at an
HIV conference in Kampala.
Recruitment slumped at this point, but did recover by July and August. When a
wave of arrests and alleged police abuse occurred in September recruitment
slumped once again, never to fully recover before the survey closed in April
these incidents, the researchers were able to recruit 303 men who had had anal
sex with a man in the previous three months. Those participating were
predominantly young (50% under 25), with a median of eleven years of schooling, and
almost all were Ugandan.
confidentiality, no names or contact details were collected from participants
(they were identified with a number and a fingerprint). Information was
collected using a self-completion computer survey (ACASI), in either English or
majority (78%) had had sex with a woman at some time; 29% had fathered children; and 16%
were currently living with a female partner.
often a mismatch between the sexual orientation terms that men most identified
with and their reported attraction to men and women:
- Whereas 56% identified with
‘gay’ or ‘homosexual’, 70% said they were attracted mostly or only to men.
- 37% identified as ‘bisexual’,
but 12% were attracted to both men and women.
- 7% identified as ‘straight’ or
‘heterosexual’, while 19% were attracted mostly or only to women.
or transactional sex was common: 42% had ever sold sex to a man, and 25% to a
overall HIV prevalence amongst adult males in Kampala is 4.5%, it was 13.7% in this sample.
had inaccurate perceptions of their own HIV status – only one in ten of those
with HIV were aware of the fact. Moreover, one in ten of the whole sample
thought they had HIV when in fact they were HIV-negative.
lacked basic information about HIV transmission risks. When asked whether
insertive or receptive anal sex was riskier, 11% answered that neither activity
posed a risk. Only 11% correctly answered that receptive sex is riskier.
never used condoms. Condom use was higher with male partners than female
partners; higher with steady partners than casual partners; and lowest of all
with commercial sex partners. Most men used lubricants although these were very
researchers wished to identify the demographic or behavioural characteristics that were most strongly associated with HIV infection. In multivariate
analysis, factors such as condom use or numbers of partners were not
significantly associated with having HIV. In fact, only two factors were: age
and homophobic abuse.
Men aged 25
or over were four times more likely to have HIV (odds ratio 4.3, 95% confidence
interval 1.5 to 12.8). Amongst men over 25, HIV prevalence was 22.4%.
Men who had
ever experienced violence or abuse because of their sexuality were five times
more likely to have HIV (odds ratio 4.8, 95% confidence interval 1.8 to 13.1).
Of the whole sample, 37% had been physically abused at some point, 37% had been
blackmailed and 26% had been forced to have sex.