Estimating substance use disparities across intersectional social positions using machine learning: An application of group-lasso interaction network.

OBJECTIVE: An aim of quantitative intersectional research is to model the joint impact of multiple social positions on health risk behaviors. Although moderated multiple regression is frequently used to pursue intersectional research hypotheses, such parametric approaches may produce unreliable effect estimates due to data sparsity and high dimensionality. Machine learning provides viable alternatives, offering greater flexibility in evaluating many candidate interactions amid sparse data conditions, yet remains rarely employed. This study introduces group-lasso interaction network (glinternet), a novel machine learning approach involving hierarchical regularization, to assess intersectional differences in substance use prevalence. METHOD: Utilizing variable selection and parameter stabilization functionality for main and interaction effects, glinternet was employed to examine two-way interactions between three primary social positions (gender, sexual orientation, and race) predicting heavy episodic drinking, cannabis use, and cigarette use prevalence. Analyses were conducted using the All of Us Research Program (N = 283,403), a national sample with high representation from populations historically underrepresented in biomedical research. Results were replicated using holdout cross-validation and compared against logistic regression estimates. RESULTS: Glinternet prevalence estimates were more stable across discovery and replication samples relative to logistic regression, particularly among sparsely represented groups. Prevalence estimates for cigarette and cannabis use were elevated among sexual minority and White cisgender women compared to heterosexual and non-White women, respectively. CONCLUSIONS: Glinternet may improve upon traditional moderated multiple regression methods for pursuing intersectional hypotheses by improving model parsimony and parameter stability, providing novel means for quantifying health disparities among intersectional social positions. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

The Inequitable Burden of the COVID-19 Pandemic Among Marginalized Older Workers in the United States: An Intersectional Approach.

OBJECTIVES: The COVID-19 pandemic has profoundly affected the lives of people globally, widening long-standing inequities. We examined the COVID-19 pandemic's impact on employment conditions by race/ethnicity, gender, and educational attainment and the association between such conditions and well-being in older adults in the United States. METHODS: Using data from the Health and Retirement Study respondents interviewed between May 2020 and May 2021 when they were ≥55 years of age, we examined intersectional patterns in COVID-19-related changes in employment conditions among 4,107 participants working for pay at the start of the pandemic. We also examined the compounding nature of changes in employment conditions and their association with financial hardship, food insecurity, and poor self-rated health. RESULTS: Relative to non-Hispanic White men with greater than high school education (>HS), Black and Latinx men and women were more likely to experience job loss irrespective of education; among those who did not experience job loss, men with ≤HS reporting Black, Latinx, or "other" race were >90% less likely to transition to remote work. Participants who experienced job loss with decreased income or continued in-person employment with decreased income/shift changes had greater prevalence of financial hardship, food insecurity, and poor/fair self-rated health than others. DISCUSSION: The impact of COVID-19 on employment conditions is inequitably patterned and is associated with financial hardship, food insecurity, and adverse health in older adults. Policies to improve employment quality and expand social insurance programs among this group are needed to reduce growing inequities in well-being later in life.