Project Summary
Exposure to racial discrimination has long been identified as a key social determinant of substance use that
stratifies by race/ethnicity. Starting in 2020, displays of racism and their impacts have intensified, and the recent
volatile climate during the COVID-19 pandemic provides a unique opportunity to assess mechanisms that link
discrimination and other social determinants of health with racial/ethnic disparities in drug use, but there has not
been the requisite science on this topic as it deserves. Existing research on the discrimination-substance use
association is plagued by both external validity (i.e., non-representative sampling with low generalizability to
overall U.S.) and internal validity (i.e., low temporal precision from annual inter-survey intervals to detect
prospective, potentially causal association). Additionally, there is little evidence to guide clinicians and
policymakers on interventions that can disrupt discrimination’s impact on substance use in the modern social,
economic, and public health context. This secondary analysis of the nationally-representative Understanding
America Study (UAS, n~9,000 U.S. adults) aims to elucidate the nature and mechanisms of the association of
discrimination with race/ethnicity inequalities in substance use. To chart the broad effects of the pandemic, this
study will harmonize longitudinally bi-weekly (March 2020-July 2021) and monthly (August/September 2021-
June 2022) survey waves with a total of 38 timepoints. Time-varying measures include 1) frequency of
discrimination experience, racism, and COVID-19 related discrimination, and 2) substance use (past-week use
of alcohol, cigarettes, e-cigarettes, cannabis, and other drugs, poly-substance use, and binge drinking) along
with key public health (e.g., COVID-19 illness and vaccination), economic (e.g., economic insecurity, housing
insecurity, and employment instability), and behavioral factors (e.g., resilience and coping strategies). We will
combine person-level data (38 survey waves) with state-level geocoded data on hate crime incidence, statutory
provisions, and other public health or economic context. Intensive longitudinal multi-level modeling will be
conducted to partition between- and within-subject variance and address between-person confounds, thus
overcoming the biases inherent in other annual-interval national substance use surveillance with low temporal
precision and recall errors. Through three inter-related aims and theory-driven hypothesis testing, we will 1)
examine whether within-person changes in discrimination are associated with substance use and if associations
are moderated by race and ethnicity (Aim 1), 2) determine person- and environment-level factors that exacerbate
or buffer against the association of discrimination with substance use (Aim 2), and 3) assess mechanisms that
explain race/ethnicity differences in the association of discrimination with substance use (Aim 3). Findings from
this study will provide timely evidence to guide clinicians and policymakers on interventions that can disrupt
discrimination’s impact on substance use in the modern social, economic, and public health context.