Cross-sectional associations of gender identity and sexual orientation, with co-occurrence and clustering of health-related behaviours among British adolescents: Millennium cohort study.

Abstract

PURPOSE: We examined whether gender identity and sexual orientation were associated with seven health-related behaviours, and with co-occurrence and clustering of these behaviours among British adolescents. METHODS: Millennium Cohort Study (age 17 wave) provided data on the exposures, gender identity (male, female, genderqueer) and sexual orientation (heterosexual, bisexual, gay or lesbian, or other), and seven self-reported health-related behaviours (binge drinking, drug use, no consumption of breakfast, no consumption of fruits or vegetables, physical inactivity, poor sleep, and smoking or vaping). Poisson regressions examined associations between the exposures and single behaviours (reporting prevalence ratios (PRs)); and multinomial logistic regressions were used for behavioural cumulative co-occurrence score (reporting PRs). Cluster patterns were identified using Ward's agglomerative hierarchical cluster analysis while associations with cluster membership were performed using logistic regressions (reporting odds ratios (ORs)). RESULTS: Our sample included 6022 adolescents (55.4% female, 1.5% genderqueer, 11.6% non-heterosexual). Adolescents who identified as genderqueer had the highest prevalence of not eating breakfast (PR: 60.5% [95%CI 48.4-71.4]) and poor sleep (68.7% [95%CI 55.6-79.3]). Those who identified as bisexual had a higher PR of co-occurring behaviours (2.46 [95%CI 1.39-4.27]). Among the three clusters identified (1: Multiple risk behaviours; 2: Physical inactivity and binge drinking; 3: Poor diet and physical inactivity), adolescents who identified as genderqueer or other sexual orientation showed the highest prevalence in cluster 3. CONCLUSION: Gender and sexual minority British adolescents showed a higher prevalence of risky health-related behaviours, and higher risk of co-occurring behaviours. Physical inactivity and poor diet behaviours commonly clustered together for these groups.