Electronic Theses and Dissertations

Date

2024

Document Type

Dissertation

Degree Name

Doctor of Philosophy

Department

Psychology

Committee Chair

James Murphy

Committee Member

James P Whelan

Committee Member

Rory A Pfund

Committee Member

Philip Pavlik

Abstract

High-intensity drinking (HID; 8+/10+ drinks for women/men) is linked to higher BAC and negative alcohol-related consequences compared to heavy episodic drinking (HED; 4+/5+ drinks for women/men). This study used growth mixture modeling to identify HID trajectories, examined potential behavioral economic predictors (demand intensity, Omax, demand elasticity, and delayed reward discounting) of HID trajectory membership, and quantified associations between HID trajectories and alcohol-related outcomes (alcohol related consequences and life satisfaction). Participants were emerging adults from the community (N = 602; 42.7% male, 47.0% White, Black = 41.5%, Mage = 22.63) who reported recent heavy drinking at baseline. Assessments occurred every 4 months over 32 months (9 assessments). Single-group latent growth curve models were estimated for HID and HED. The linear model best described HID [χ²(32) = 35.31, p = .314; CFI = .997; RMSEA = .013], indicating a decrease in HID over time, though not statistically significant. HED fit best with a quadratic model [χ²(36) = 110.17, p <.001; CFI = .964; RMSEA = .059], indicating a deceleration in decrease over time. High correlations between HID and HED suggested multicollinearity, favoring the simpler HID-only model for clarity. A linear, 2-class HID model (AIC = 2323.24, BIC = 2358.36, LMR = 7.48, p <.05) was identified as the best most optimal fit to the data. The model was comprised of a no-HID class and a HID class with a non-significant, negative slope that indicated a decrease in HID over time (b = -1.32, SE = 1.37, p = .335). Higher baseline demand intensity, Omax and typical weekly drinking were significant predictors of HID trajectory membership. HID class membership predicted alcohol-related consequences at 32 months in univariate models, but not in multivariate models accounting for other drinking risk factors. HID did not significantly predict life satisfaction across relationship, family, and employment domains at 32 months in univariate models. In conclusion, HID trajectories in heavy drinking emerging adults can be effectively modeled using growth mixture modeling. Significant predictors of HID included behavioral economic variables and typical weekly drinking level, highlighting the importance of screening for these factors to identify higher-risk drinking profiles.

Comments

Data is provided by the student.

Library Comment

Dissertation or thesis originally submitted to ProQuest.

Notes

Embargoed until 07-30-2025

Download

Available for download on Wednesday, July 30, 2025

Share

COinS