Purpose: Existing theories to explain earlier age of menarche among some adolescents suggest that, in addition to heritable factors and nutritional status, stress, trauma, and family dysfunction can contribute to earlier onset of menarche. Given lesbian and bisexual girls are at greater risk of these stressors than their heterosexual peers, this study examined whether sexual minority girls report earlier menarche, and whether known exposure to violence, stress, and poorer family relationships mediate such a relationship.
Methods: Analysis of the BC Adolescent Health Survey of 2008, a province-wide, cluster-stratified random sample of students grades 7-12 in Western Canada. For this study, girls 15 and older who reported age of menarche were selected (n=6,812; mean age, 16.4 years, 4.8% identified as lesbian or bisexual). Measures included self-reported age at menarche, BMI, experience of sexual and/or physical abuse, an enacted stigma scale of harassment, assault, and discrimination, as well as family connectedness and family history of suicide. Analyses included general linear models and logistic regressions, depending on the type of variable, for mediation analyses among all variables, orientation, and age of menarche; Sobel tests of mediation were conducted with adjusted coefficients for the logistic regressions as suggested by Kenny (2006) derived from equations by McKinnon & Dwyer (1993). A multivariate general linear model identified key independent predictors; multiple mediator models using bootstrap estimates with age as a covariate identified the extent to which mediators explained the relationship between sexual orientation and age at menarche.
Results: Lesbian and bisexual girls reported significantly earlier age of menarche (11.8 vs. 12.4), significantly higher BMI (22.6 vs. 21.1), higher rates of all violence exposures and lower levels of family connectedness (all p<.01). All hypothesized mediation analyses were significant (p<.05, p<.01 or p<.001). In age-controlled multivariate analyses sexual orientation, BMI, sexual abuse history and family connectedness independently contributed to age of menarche (p<.05 or p<.001); in multiple mediator analyses, the total indirect effect as well as each individual indirect effect were all significantly different from 0, and the indirect effects accounted for 32% of the total effect of sexual orientation on age of menarche.
Conclusions: In this population-based study, lesbian and bisexual girls reported significantly earlier age of menarche on average. Exposure to violence, especially sexual and physical abuse, and poorer family relationships, in addition to BMI, helped explain this finding. Although cross-sectional studies are limited in their ability to establish causal influences, these results complement longitudinal research, which has a limited ability to examine sexual abuse as an influencing factor on menarche. Given the additional health and behavioural risks reported for early-maturing girls, stress-related earlier menarche may also help explain later health disparities among sexual minority girls compared to their heterosexual peers.