Skip to content

Abstract Search

Primary Submission Category: Methodological approaches to studying public health

Innovating Quantitative Intersectional Methods to Identify Mental Health Disparities from Adolescence to Adulthood

Authors:  Talia Kieu, Deshira Wallace,

Presenting Author: Talia Kieu*

Those with multiple marginalized identities are more likely to experience severe chronic depression than those with a single or no marginalized identities. The Intersectionality Framework suggests that individuals’ overlapping social identities, referred to as social location, yield varying levels of risk and protective factors. Mixed findings on depression disparities may be due to a focus on the influence of multiple identities in isolation (additive effects), rather than social locations (intersectional effects). This study aims to conduct 3 cross-sectional analyses to estimate the % of variability in depression attributable to the intersectional (between-group) effects vs. additive (within-group) effects of sex, racialization, sexual orientation, and socioeconomic status across the life course. I analyzed Waves 1 (adolescence), 4 (young adulthood), and 5 (adulthood) of the National Longitudinal Study of Adolescent to Adult Health (n=12,160). Using the novel Multilevel Analysis of Individual Heterogeneity and Discriminatory Accuracy (MAIHDA) approach, in which individuals are clustered in social location groups (akin to geographic locations), I constructed a cluster variable comprised of 24 unique combinations of identities listed above. For each wave, I estimated the % of variance attributable to between-group differences in a baseline model, and a model adjusted for additive effects. The remaining variance after accounting for the additive effects of these identities represents their intersectional effects. In preliminary analyses, 6.21%, 7.81% and 8.33% of the total variance in depression is attributable to the intersectional effects of social identities in adolescence, young adulthood, and adulthood, respectively. By using multilevel modeling, this study overcomes the limitations of traditional additive approaches. Findings will have implications for community-engaged research by tailoring mental health interventions to address intersectional vulnerabilities.