Ten percent of South African women report anal sex in the past three months

A cross-sectional survey has found that 14% of men and 10% of women in Cape Town, South Africa, report having heterosexual anal intercourse. Condoms are used at roughly equivalent frequency as for vaginal sex, it is reported in the online edition of Sexually Transmitted Infections.

The study’s authors believe that while anal intercourse needs to be addressed in behavioural interventions, it makes only a minor contribution to the South African epidemic.

However, the authors of an accompanying editorial argue strongly that this conclusion is premature, and point out that unaccounted-for anal intercourse could skew the findings of microbicide trials.

Anal intercourse between men and women has generally not received as much attention as anal intercourse between men. However, there is evidence (especially from the United States) that anal sex is practised by large numbers of sexually active adults, suggesting that it may play an important role in HIV transmission amongst heterosexuals.

Research on this topic in African countries has been limited, although a large household survey of South Africans aged 15 to 24 previously found that 3.6% of the sample reported anal sex. Young men reporting anal sex were more likely to be HIV-positive, to be older, to have had more sexual partners and to have had sex while using drugs or alcohol. Young women reporting anal sex were not, however, more likely to be HIV-positive.

For the new study, Seth Kalichman and colleagues recruited two convenience samples in Cape Town. In the first, 1360 adults were recruited from a busy sexually transmitted infection clinic. In the second sample, 3051 adults were recruited from public venues such as bus stops and shopping streets. Whereas the first sample was predominantly male and almost exclusively of black African ethnicity, the second sample was evenly split between men and women, and between people of black African and 'coloured' ethnicities. The median age of the participants was 30.

A self-completion questionnaire (in English, Xhosa or Afrikaans) asked about sexual behaviours in the past three months. The 6% of men who reported same-sex partners in that time period were excluded from the subsequent analysis, which focuses on heterosexual anal intercourse.

A total of 14% of men and 10% of women reported anal intercourse.

Over half the people who reported anal sex either had as much anal sex as vaginal sex, or more.

Condoms were used almost as often for anal intercourse as for vaginal intercourse. Men said that they used them on 75% of occasions they had anal sex (80% for vaginal sex), whereas women reported condom use on 50% of anal sex occasions (57% vaginal).

A number of demographic and behavioural factors had statistically significant associations with anal intercourse:

• Male sex
• Younger age
• Not married or living with a partner
• More sexual partners
• Ever using condoms, and recent use of condoms for vaginal sex
• Sexually transmitted infections
• Transactional sex
• Alcohol and drug use
• Ever tested for HIV
• Tested HIV positive

However the researchers do not report any multivariate analyses in order to clarify which factors were independently associated with anal sex.

Seth Kalichman and colleagues conclude that anal intercourse is practiced “at relatively low rates” by heterosexual men and women, and believe that anal intercourse probably accounts for a small number of infections in South Africa.

However, an accompanying editorial by Marie-Claude Boily and colleagues takes issue with these conclusions. They question whether the figures reported by Kalichman really are “relatively low rates”, and note that their own systematic review of HIV transmission risks found that receptive anal intercourse is around twenty times riskier for HIV acquisition than receptive vaginal intercourse.

Moreover, they point out that people reporting anal sex were more likely to be HIV-positive and to have more sexual partners in the past three months, both factors which would magnify the contribution of anal sex to the South African epidemic.

They also argue that continued infections due to anal intercourse could give the impression that an otherwise effective vaginal microbicide is less effective than it in fact is. They believe that more accurate estimates of anal intercourse are needed for trials of microbicides, in order to ensure that the studies have statistical power.