U.S. Teen Mental Health at the Intersections of Race, Gender, and Sexual Orientation
Repository for a paper in the special issue on social epidemiology methods at the American Journal of Epidemiology
Title: Intersectional inequities in suicide ideation by race, sexual orientation, and gender among U.S. high school students pre and post 2020: An application of random effects intersectional MAIHDA
Abstract: Quantifying intersectional health inequities and examining changes over time are foundational to social epidemiology. I-MAIHDA (intersectional multilevel analysis of individual heterogeneity and discriminatory accuracy) is a recent innovation that simplifies quantitative intersectional analyses while providing methodological improvements over conventional approaches. We illustrate the use of logistic I-MAIHDA with random effects to estimate intersectional inequities in suicidal ideation among U.S. high school students before and after 2020 by race, sexual orientation, and gender, using 2017-2021 data from the Youth Risk Behavior Surveillance System. The U.S. faces a youth mental health crisis made worse by the many disruptions of 2020, including the beginning of the COVID-19 pandemic. Before 2020, we found substantial inequities in ideation, ranging from 9.8-12.7% among heterosexual boys to over 50% among bisexual Multi-race/Other and White girls. We also found notable changes pre/post 2020. Strata at the lowest (heterosexual boys) and highest risk (bisexual girls) showed little change, while middle risk-ranked strata (Black Other/Questioning and lesbian girls, White Other/Questioning boys and girls, and Multi-race/Other gay boys) reported large increases in ideation. Our findings suggest worsening teen mental health in the 2017-2021 period, particularly among racial and sexual orientation minorities. We illustrate the value of I-MAIHDA for understanding changes in intersectional health inequities.
Code for the analyses
CleanAndMergeData.do: Stata code for cleaning and merging dataMELogisticModels.do: Stata code for running mixed effects logistic I-MAIHDA modelsModels_AllOutcomes.R: R code for running mixed effects logistic I-MAIHDA models (used in Manuscript/Supplemental)Visualizations.R: R code for visualizationsHausmannTestReDo.R: Custom R code for doing Hausmann test for models running using glmmTMB R packageCatepillerPlots_glmmTMB.R: Custom R code for generating caterpillar plots for models running using glmmTMB R package
Analysis ready data files are too big to share on GitHub but can be requested.
On-going work in progress developing interactive YRBSS outcome results
Intersectional Youth Mental Health Interactive [ link ] (preview image):
