Date of Award

Winter 1-8-2021

Degree Type


Degree Name

Master of Public Health (MPH)


Public Health

First Advisor

Colin K. Smith

Second Advisor

Ike Okosun


INTRODUCTION: Income inequality has been previously shown to be related to adverse population health outcomes. A suggested etiology is that income inequality intensifies status anxiety, leading to unhealthy coping mechanisms such as substance use. Behavioral factors specific to Hispanic cultures may have the potential to build resiliency in adolescents against substance use related to status anxiety, but have not been considered as protective factors in large, nationwide studies on substance use among adolescents.

AIM: Determine the association between regional income inequality in the U.S. and substance use among 12th grade students, and determine whether this association is different for Hispanic students.

METHODS: Public survey data from 2012 – 2018 from the Monitoring the Future Survey, an annual, nationally representative survey on substance use and social and political views of secondary school and college students in the U.S., were used along with data from the U.S. Census Bureau on household income inequality for 4 regions of the U.S, represented by the Gini coefficient. Odds ratios, 95% confidence intervals, and p-values were calculated using binary logistic regression carried out in SAS 9.4 to determine the likelihood of substance use given the respective region’s Gini coefficients and controlling for confounders.

RESULTS: The Gini coefficient was negatively associated with substance use for the whole sample (OR <0.001), as well as for Hispanic (OR 0.002) and Black (OR <0.001) participants, and positively associated with substance use for White participants (OR 2.339), but was not significant (CI 0.032 – 170.112). When disaggregated by race only Future Plans and Father’s Education were consistently significant predictors for all three racial/ethnic groups and the whole sample. The model’s concordance statistic was 0.581, meaning it was able to correctly predict substance use among participants a little more than half the time.

DISCUSSION: Because the model’s predictive power was low, it is not sufficient for determining the true association between substance use and income inequality among the survey population, and differences among racial and ethnic groups could not be determined. Future research should look at specific cultural factors to determine whether they can build resiliency against substance related to status-anxiety at a population level.


File Upload Confirmation