Machine learning methods for identifying person-level mechanisms of alcohol use among sexual and gender minority intersections

点击翻译按钮获取中文摘要

The long-term objective of this Pathway to Independence Award is to support candidate Dr. McCabe in building an independent research program and to facilitate his transition into an independent faculty research position. To date, Dr. McCabe’s research has focused on 1.) refining quantitative methods applied in addictions research, and 2.) understanding individual differences in stress, developing self-regulation, and their associations with alcohol use (AU) among sexual minority and non-minority communities. Dr. McCabe seeks to expand his training in AU development, minority stress theory, and applied quantitative methods to a new emphasis on intersectionality and sexual and gender minority (SGM) AU risk, machine learning and multilevel methodologies, and ecological factors influencing AU disparities. This long-term objective will be achieved through a five-year training plan involving a carefully selected mentorship team as well as targeted coursework and hands-on training experiences. The goals of the proposed research are to 1) distinguish SGM subgroups and intersections at heightened risk for AU (e.g., bisexuals and trans persons, SGM young women of color), 2) assess the role of state policies in moderating AU risk, and 3) delineate moderators and mechanisms of heightened AU across SGM populations within and beyond the coronavirus pandemic. The mentored phase (K99) will involve cross-sectional analysis of the All of Us Research Program (AURP), a large (N=331,360) and diverse national dataset. Aim 1 will identify heterogeneity in alcohol and other substance use behaviors among sexual (1a; n=38,820 non-heterosexual) and gender minority (1b; n=2,660 transgender or nonbinary) communities. It will then test race/ethnicity and age as intersectional moderators of SGM inequities (1c) and state-level policies impacting SGM communities (1d; e.g., hate crime laws enumerating SGM identity) that further differentiate AU risk among SGM groups. During the independent phase, findings will be extended to address mediators and moderators of AU in the monthly AURP COVID-19 Participant Experience Survey (Aim 2; n=100,340) as well as the longitudinal, biennial AURP data that extends beyond the pandemic into 2027 (Aim 3). Aim 2 will test pandemic stressors as mediators of between-person AU among SGM intersections (2a) and examine intersectional (2b) and multilevel moderators (2c) of within-person AU. Aim 3 will test differences in post-pandemic recovery in AU among SGM intersections (3a) and determine pandemic mediators (3b) and moderators (3c) of this change. Findings will serve as the foundation for an NIAAA R01 submission during the R00 phase focused on geocoded neighborhood-level factors influencing developing alcohol risk across adolescence and young adulthood across SGM intersections. Mentors (Drs. Rhew, Lee, Helm) and consultants (Drs. Grimm, Bauer, Raifman) are committed to the candidate’s training, each providing unique expertise to the research and training plan. This award will support the candidate’s development as an independent cross-disciplinary prevention scientist in AU disparities and quantitative methods.