Sport has become a popular tool for HIV prevention, based on claims that it can foster life skills that are necessary to translate knowledge, attitudes and behavioural intentions into actual behaviour. Empirical evidence of the effectiveness of sport-based HIV prevention programmes is, however, sorely lacking. We therefore conducted a cross-sectional survey assessing sexual behaviour and the determinants thereof among 454 youth of the Mathare Youth Sport Association (MYSA) in Kenya and a control group of 318 non-MYSA members. Multiple (ordinal) logistic regression models were applied to measure the association between MYSA membership and attitudes, subjective norms and self-efficacy related to condom use as well as sexual experience, age at sexual debut, condom use, history of concurrent relationships and number of partners in the last year. MYSA members were more likely to use condoms during the first sex act (odds ratio (OR)=2.10; 95% CI: 1.10-3.99). Consistent condom use with the current/last partner was 23.2% (36/155) among MYSA members vs. 17.2% (17/99) among the control group. Even after adjusting for media exposure - a factor associated with both MYSA membership and higher frequency of condom use - MYSA members were still found to use condoms more frequently with their current/last partner (adjusted OR=1.64; 95% CI: 1.01-2.68). Nevertheless, levels of condom use remain disturbingly low. More rigorous evaluations of sport programmes for HIV prevention are needed. When possible, programmes should be preceded by baseline assessments, trends in risk behaviour of the intervention group should be compared with those of a control group, and protocols for data collection and analysis should include measuring of and adjusting for potentially confounding factors.

No abstract