Using Novel Risk Factors to Predict College Student Binge-Drinking: A Behavioral Economics and Satir Growth Model Perspective

Background: College student binge drinking is a complex, multifaceted behavior. Little has changed over the past 15 years, binge-drinking rates remain high with approximately 40% of undergraduate students reporting an episode within the past 30 days. The intractability of college student binge drinking rates suggests that critical risk factors remain unaddressed by prevailing prevention and intervention efforts. Interventions targeting the novel risk factors of alternative rewards, commitment and consistency, and delay discounting demonstrate promising treatment effects. However, all identified intervention studies used small samples, limiting the generalizability of findings. To provide stronger evidence for interventions that target these novel risk factors, this study analyzed the relationships between student binge drinking and alternative rewards, commitment and consistency, and delay discounting using the National Epidemiologic Survey on Alcohol and Related Conditions-III (NESARC-III), a national dataset containing information regarding college student drinking behavior.

Methods: Data and sample: NESARC-III is a cross-sectional dataset containing national data of 36,309 individuals, aged 18 years and older living in the United States. This study sampled 509 participants, aged 18- to -25-years-old, that reported past-year college enrollment.

Measurement and analysis: An index and scales created with NESARC-III survey questions measured the individual-level predictors of alternative rewards, commitment and consistency/congruence, and delay discounting. Sociodemographic and interpersonal covariates were included. Outcomes of interest included binge drinking, heavy binge drinking, frequency of drinking personal largest consumption of alcohol, and lifetime alcohol use disorder diagnosis. Binary logistic regression was used to predict the effects of individuals’ alternative rewards, commitment and consistency/congruence, and delay discounting on the outcome variables of interest.

Results: Model 1 findings suggested that every one-level decrease in delay discounting increased the odds of individuals’ binge drinking by 84% (β=-1.86, OR=.16, p=.01, CI=.04, .67). Model 2 findings suggested that every one-level decrease in commitment and consistency/congruence (β=-4.35, OR=77.78, p=.05, CI=.698, .867) and delay discounting (β=-4.06, OR=.02, p=.00, CI=.00, .24) increased the odds of individuals’ heavy binge drinking. Model 3 findings suggested that every one-level decrease in alternative rewards (β=-.73, OR=2.08, p=.040, CI=1.03, 4.19), commitment and consistency/congruence (β=-1.10, OR=3.01, p=.032, CI=1.10, 8.26), and delay discounting (β=-1.95, OR=.14, p=.006, CI=.03, .57) increased the odds of individuals drinking their personal largest consumption of alcohol. Model 4 findings suggested that every one-level decrease in delay discounting increased the odds of individuals’ lifetime alcohol use disorder by 97% (β=-3.47, OR=.03, p=.00, CI=.01, .12). Significant covariate findings will also be discussed.

Implications: Using a large-scale secondary dataset (NESARC-III), this study provided stronger evidence for interventions targeting the novel risk factors of alternative rewards, commitment and consistency, and delayed discounting. These novel risk factors warrant further evaluation in future studies. Both clinicians who help college students with alcohol-related issues and researchers developing binge-drinking interventions would do well to explore the impact of these novel predictors.