Stress mediates the relationship between sexual orientation and behavioral risk disparities

Background: Growing evidence documents elevated behavioral risk among sexual-minorities, including gay, lesbian, and bisexual individuals; however, tests of biological or psychological indicators of stress as explanations for these disparities have not been conducted. Methods: Data were from the 2005-2010 National Health and Nutrition Examination Survey, and included 9662 participants; 9254 heterosexuals, 153 gays/lesbians and 255 bisexuals. Associations between sexual orientation and tobacco, alcohol, substance, and marijuana use, and body mass index, were tested using the chi-square test. Stress, operationalized as depressive symptoms and elevated C-reactive protein, was tested as mediating the association between sexual orientation and behavioral health risks. Multiple logistic regression was used to test for mediation effects, and the Sobel test was used to evaluate the statistical significance of the meditating effect. Results: Gays/lesbians and bisexuals were more likely to report current smoking (p < .001), a lifetime history of substance use (p <. 001), a lifetime history of marijuana use (p < .001), and a lifetime period of risky drinking (p = .0061). The largest disparities were observed among bisexuals. Depressive symptoms partially mediated the association between sexual orientation and current smoking (aOR 2.04, 95% CI 1.59, 2.63), lifetime history of substance use (aOR 3.30 95% CI 2.20, 4.96), and lifetime history of marijuana use (aOR 2.90, 95% CI 2.02, 4.16), among bisexuals only. C-reactive protein did not mediate the sexual orientation/behavior relationship. Conclusion: Higher prevalence of current smoking and lifetime history of substance use was observed among sexual minorities compared to heterosexuals. Among bisexuals, depressive symptoms accounted for only 0.9-3% of the reduction in the association between sexual orientation and marijuana use and tobacco use, respectively. More comprehensive assessments of stress are needed to inform explanations of the disparities in behavioral risk observed among sexual minorities.