An innovative methodological approach to analyzing alcohol use transitions in the context of prescription drug misuse and other health behaviors

Project Info

Type Of Project: Research
Target Population: Young, Adults , Adults
Funding Agency: Health and Human Services, Department of-National Institutes of Health
Project Locations: Flint, MI
Principal Investigators: Jason Goldstick
Lead Institution: University of Michigan Medical School

For corrections or updates to your project please send required information to or click here.

This proposal describes a plan to study what drives transitions in alcohol misuse in youth, and how those transitions coalesce with other health behavior transitions. Of particular concern is the recent increase in the prescribing, and misuse, of prescription opioids and sedatives, which has facilitated the emergence of a new crisis related to alcohol misuse: its contribution to overdose risk when used in combination with prescription opioids and/or sedatives. Prior research has identified individual-, peer-, parental-, and community-level correlates of alcohol misuse, but little empirical research focuses on what drives transitions in alcohol use behavior. Filling this gap has great potential to inform intervention and prevention design. Specifically, factors that facilitate transitions into more problematic drinking patterns are important targets for prevention, while those associated with sustained problematic drinking over time are targets for intervention. To address this knowledge gap, we will analyze data collected during the NIDA-funded (R01) Flint Youth Injury study, a longitudinal study of 600 drug-using youth recruited from an Emergency Department in Flint, Michigan. Using an innovative analytic technique, we will model behaviors over the follow-up period (roughly 6-month follow-ups for two years, for a total of five measurements) as continuous-time Markov Chains with covariate-modulated transition probabilities. In addition to the direct modeling of how covariates affect transition rates, a key advantage of this modeling framework is the elegant handling of missing time points and variation in the exact follow-up schedule.

Contact Info:

Contact Email: