Bisexuality and condom use among men: The importance of multiple sexual orientation measures

Background: Are bisexual men at elevated risk of unprotected sex than gay/homosexual and heterosexual men? Although findings from convenience samples suggest that they are, researchers have overwhelmingly used behavioral measures of bisexuality. Recent studies, however, propose that subjective measures of sexual orientation also are significant for this outcome. This study examined sexual identity, attraction, and behavior in relation to condom use among bisexual men in the United States. Methods: Data came from the National Survey of Family Growth, a nationally-representative probabilistic sample. Condom use at last sexual contact was specific to bisexually-active men's male and female partners. In a nested design, logistic regression models measured bisexual identity, attraction, and behavior in relation to condom use. Multivariate analyses controlled for identity, behavior, attraction, sociodemographic factors, and risk covariates. Results: Based upon the men's most recent male partners, bisexual (versus heterosexual) attraction was associated with an elevated odds of condom use. Identity, attraction, and behavior did not distinguish bisexual men from gay/homosexual men. Bisexuality measures were not significant upon controlling for sociodemographic and risk factors. Based upon their most recent female partners, having bisexual (versus heterosexual) attraction remained positively associated with condom use; however, behavioral bisexuals (versus heterosexuals or homosexuals) were also more likely to use condoms. These differences persisted in fully adjusted models. Conclusions: In studies of health, it is important to employ multiple measures of bisexuality rather than single measures alone. Even so, this study highlights the significance of examining specific behaviors when determining risks among bisexual men.