Social-cognitive determinants of condom use in a cohort of young gay and bisexual men

Abstract

The aim of this prospective study was to identify relevant determinants of young gay and bisexual men's (YGBM) condom use when having anal sex with casual partners. Respondents (185 YGBM in the midst of their coming-out; mean age 18.9 years) completed an online questionnaire on social-cognitive determinants of condoms use derived from the theory of planned behavior (Ajzen, 1991) at Wave 1. At six months follow-up (Wave 2) sexual behavior with casual partners was assessed. A total of 63 YGBM reported sex with a casual partner in the six months between Waves 1 and 2, of whom 49% (N=31) had anal sex. Of the YGBM who had anal sex, 42% (N=13) had unprotected anal sex. Condom use with casual partners was best predicted by the intention to always use condoms. Furthermore, attitude, descriptive and personal norms, and perceived control significantly predicted intention to always use condoms. Interventions, targeting YGBM, aiming to promote condom use with casual partners should focus on increasing attitudes and strengthening skills to negotiate and use condoms.

Keywords:

• adolescents
• condom use
• HIV/AIDS
• young gay and bisexual men
• theory of planned behavior

Acknowledgements

This study was financially supported by AIDS Fonds, the Netherlands.

Notes

1. Although some gay and bisexual youth in the Netherlands nowadays come out before the age of 16, it was deemed inappropriate to include them in this study as that would have required parental consent.

2. Initially, the relation between intention measured at Wave 1 and behavior measured at Wave 2 was analyzed for new partners because all social-cognitive determinants were assessed for anal sex with new partners. A subsequent explorative analysis of the relation between intention at Wave 1 and behavior with steady, casual, and regular partners at Wave 2 revealed a significant relation between intention and behavior only for casual partners. Because the amount of explained variance in behavior was 28% for casual partners and 11% for new partners, only the results for casual partners will be described in detail.

3. Because a substantial number of respondents indicated “Don't know” and “Not applicable” for the subjective norms items (see also Table 3), we also coded “Don't know” and “Not applicable” as missing. For the sub-group of respondents who do know what their social referents think they should do, subjective norms does significantly predict intention (β = 0.19; p=0.004).