Disparities in Adverse Childhood Experiences By Sexual Identity: An Intersectional Analysis of the Add Health Study

Background and Purpose: Adverse childhood experiences (ACEs) are known to be unequally distributed in the United States by race/ethnicity, economic status, and gender. Emerging evidence suggests that individuals who identify as lesbian, gay, or bisexual report more ACEs than do their heterosexual peers, though it is unclear whether and why there are differences in exposure to specific adversities. Moreover, little is known about the extent to which ACEs vary across interlocking intersections of race/ethnicity, income, gender, and sexual identity.

Methods: Data were derived from the National Longitudinal Study of Adolescent to Adult Health (Add Health), a multi-wave panel investigation. Primary ACE variables were measured at wave 1 spanning grades 7-12, and supplemental ACE data were collected at waves 3 (age 18-26) and 4 (age 24-32). Eleven self-reported ACEs were coded dichotomously, including five forms of child maltreatment and six other household and community adversities: alcohol/drug abuse, mental illness, incarceration, divorce/separation, parent/sibling death, and violent crime victimization. Wave 4 survey responses were used to categorize four sexual identities: straight, mostly straight, gay, bisexual. Wave 1 data were used to code self-reported gender (man; woman), race/ethnicity (Black, Hispanic, White), and childhood poverty status (income-to-needs ratio below national poverty threshold or receipt of means-tested benefits).

Bivariate analyses were performed to describe variation in each ACE and a cumulative ACE score by sexual identity. ANCOVAs were run to test whether ACE scores varied by sexual identity, gender, race/ethnicity, and poverty; analyses were repeated after adding interaction terms between sexual identity and gender, race/ethnicity, and poverty. Last, a multilevel analysis of individual heterogeneity and discriminatory accuracy (MAIHDA) was conducted to explore variation in ACE scores across strata composed of multiple intersecting social categories.

Results: In the full sample (N=12,519), unadjusted mean ACE scores were higher among individuals who identified as bisexual (M=2.12), mostly straight (M=2.00), and gay (M=1.98) than individuals who identified as straight (M=1.56). All ACEs except physical neglect, violent crime, and parent/sibling death were more prevalent among sexual minority participants than straight participants. Main-effect ANCOVA results verified the ACEs-by-sexuality differences and showed that ACEs were also more prevalent among non-White than White individuals and poor than non-poor individuals (p< .05). ACEs did not differ by gender overall, though both gender and poverty significantly moderated the association between sexual identity and ACE scores (p< .01).

MAIHDA results confirmed the main-effect ANCOVA results. There were no significant between-strata differences in ACE scores. However, an estimate of proportional change in variance indicated that 21.8% of sample variation in ACE scores was attributable to between-strata effects, signifying considerable heterogeneity associated with the intersection of social categories above and beyond the independent effects of each category.

Conclusions and Implications: Individuals who identify with a sexual minority status appear to endure high levels of childhood adversity, though the explanatory mechanisms are uncertain and likely heterogeneous. To further advance research on disparities, we call for more regular use of quantitative intersectional approaches such as MAIHDA that can unpack complex associations between sexual identity and other social categories.