Childhood gender non-conformity, sexual orientation and mental health problems among 18 to 89 year-old Danes.

Abstract

Children who do not comply with social norms and expectations concerning their biological sex are generally referred to as gender non-conforming1. Childhood gender non-conformity has been associated with various psychosocial stressors and mental health problems1, 2, but these associations remain to be investigated at a nationwide level. Prior research on this topic has focused primarily on non-probability-based samples of young people and sexual or gender identity minority groups, documenting strong links between childhood gender non-conformity and both non-heterosexuality and non-cisgender identity3-5. Same-sex oriented individuals considerably more often report childhood gender non-conformity than heterosexual peers, and it has been suggested that childhood gender non-conformity might, at least in part, explain the excess mental morbidity observed among sexual minorities6, 7. The aim of the present study was to investigate potential links between recalled childhood gender non-conformity and mental health problems in the general Danish population as well as in strata of heterosexual, homosexual and bisexual Danes. We used baseline questionnaire data from 27,548 individuals aged 18-89 years who participated in Project SEXUS, a prospective cohort study launched in 2017 with the aim of exploring the interplay between sexual and general health8, 9. The cohort was established using a probability-based sampling frame, and an individual weighting procedure ensured national representativeness with respect to key demographic factors (see supplementary information). Childhood gender non-conformity was defined using the following question: “How well, or how poorly, does the following statement fit you: As a child or young person, I had difficulties living up to other peoples’ perception of ‘a real girl’ (for respondents assigned female sex at birth) / ‘a real boy’ (for respondents assigned male sex at birth)”. Respondents who answered that the statement fitted them “well” or “very well” were considered as childhood gender non-conforming. We included measures of mental health problems ranging in severity from loneliness to suicidal thoughts/attempts. Respondents were asked how often they felt lonely, and we considered the responses “sometimes”, “often” and “always” as indicators of loneliness. To capture symptoms of depression or anxiety within the last 14 days, respondents were presented with the two-item Patient Health Questionnaire (PHQ-2) and the seven-item Generalized Anxiety Disorder (GAD-7) scale. Respondents with a PHQ-2 score ≥3 were considered to have depressive symptoms, while respondents with a GAD-7 score ≥5 were considered to have anxiety symptoms. Additionally, respondents were asked if they had ever received treatment by a doctor, a psychologist or a similar professional for a mental health problem, and they were asked if they had ever self-harmed without suicidal intent or if they had ever had suicidal thoughts with or without actual suicide attempts. We calculated prevalence estimates and performed chi-square tests to determine whether covariate distributions (age, educational attainment, difficulties paying bills within the last year, partner status, gender identity, sexual identity, same-sex sexual experience and attraction, and history of sexual assault before age 18 years) differed significantly between childhood gender conforming and non-conforming participants. Using binary logistic regression analyses, we calculated prevalence odds ratios (ORs) with 95% confidence intervals (CIs) for associations between recalled childhood gender non-conformity and the studied mental health outcomes using childhood gender conforming individuals as reference. Additionally, we performed analyses stratified on sexual identity categories (heterosexual, homosexual and bisexual) using gender conforming heterosexuals as reference. ORs were adjusted for age in 10-year categories, and fully adjusted ORs (aORs) were further adjusted for educational attainment, difficulties paying bills within the last year, partner status, gender identity, sexual identity, and history of sexual assault before age 18 years. We used demographically weighted data in all statistical analyses, which were carried out using the stats package in R (version 4.0.2). Of 27,548 study participants, 5,355 (demographically weighted proportion, 19.0%) reported having experienced difficulties living up to other people's perception of ‘a real girl’ or ‘a real boy’ during childhood or adolescence, with a larger proportion among women (21.2%) than men (16.9%) (p<0.001). More childhood gender non-conforming than conforming individuals reported ≤10 years of education, difficulties paying bills, and not having a spouse or a partner. Childhood gender non-conformity was reported markedly more often by non-cisgender respondents, individuals with a non-heterosexual identity, and those reporting any same-sex sexual behavior or attraction. Further, larger proportions of childhood gender non-conforming than conforming individuals reported having been the victim of a sexual assault before age 18 years (all p values <0.001) (see also supplementary information). Childhood gender non-conformity was associated with consistently greater odds of mental health problems in both age-adjusted and fully adjusted analyses, including loneliness (women: aOR=1.47, 95% CI: 1.34-1.62; men: aOR=1.45, 95% CI: 1.29-1.63); depressive symptoms (women: aOR=1.48, 95% CI: 1.32-1.65; men: aOR=1.51, 95% CI: 1.32-1.73); symptoms of anxiety (women: aOR=1.72, 95% CI: 1.56-1.90; men: aOR=1.63, 95% CI: 1.45-1.83); having ever received treatment for a mental health problem (women: aOR=1.54, 95% CI: 1.41-1.70; men: aOR=1.45, 95% CI: 1.29-1.62); self-harm (women: aOR=1.93, 95% CI: 1.66-2.25; men: aOR=1.83, 95% CI: 1.50-2.24); and suicidal thoughts/attempts (women: aOR=1.98, 95% CI: 1.78-2.19; men: aOR=1.54, 95% CI: 1.36-1.73). Several associations between childhood gender non-conformity and mental health problems remained statistically significant in analyses stratified by categories of sexual identity. Among heterosexual participants, mental health challenges were reported consistently more often by childhood gender non-conforming than conforming individuals, most notably so for self-harm (women: aOR=2.11, 95% CI: 1.79-2.48; men: aOR=1.81, 95% CI: 1.46-2.25). Using childhood gender conforming heterosexual peers as reference, childhood gender conforming homosexual participants generally did not exhibit statistically significantly increased odds of mental health problems. In contrast, childhood gender non-conforming homosexual participants had elevated odds of most mental health problems, with particularly high odds for suicidal thoughts/attempts (women: aOR=3.32, 95% CI: 1.89-5.82; men: aOR=2.62, 95% CI: 1.75-3.91) (see also supplementary information). For bisexual participants, odds of several mental health problems were increased among both childhood gender conforming and non-conforming individuals when compared to the reference group of childhood gender conforming heterosexuals. Particularly elevated odds were observed for self-harm (women: aOR=3.12, 95% CI: 2.06-4.71; men: aOR=5.27, 95% CI: 2.71-10.25) and suicidal thoughts/attempts (women: aOR=3.14, 95% CI: 2.12-4.65; men: aOR=2.33, 95% CI: 1.28-4.25) among gender non-conforming bisexuals (see also supplementary information). With approximately one-in-five study participants recalling difficulties living up to other people's gender-specific norms and expectations as a child or young person, our nationally representative findings demonstrate that childhood gender non-conformity is by no means a rare phenomenon restricted to specific subsets of the general population. While proportions of non-cisgender and non-heterosexual individuals reporting childhood gender non-conformity are clearly higher than corresponding proportions among cisgender and heterosexual peers, the vast majority of childhood gender non-conforming people in the general population are cisgender, heterosexual individuals. Importantly, we document that childhood gender non-conformity is linked to a considerably increased burden of mental health problems among both women and men, and among heterosexual, homosexual and bisexual individuals alike. These findings should raise awareness about the elevated burden of mental health problems among individuals with recalled childhood gender non-conformity, and stimulate initiatives to increase societal acceptance of gender diversity and eliminate bullying and violence against gender atypical children and adolescents.