Abstract
Background
College life is characterized by marked increases in alcohol consumption. Extraversion and neuroticism are associated with alcohol use problems in college and throughout adulthood, each with alcohol use patterns consistent with an externalizing and internalizing pathway respectively. Students higher in extraversion drink more frequently and consume more alcohol, while neuroticism is paradoxically not consistently associated with elevated alcohol use.
Objective
This study examined whether students higher in neuroticism may drink the day before stressors, namely tests and assignment deadlines.
Method
Multilevel generalized linear models were performed using data from a longitudinal study of first-time, first-year undergraduates assessing alcohol use across four years of college, with daily diary bursts each semester.
Results
Students higher in extraversion had heavier alcohol use and greater alcohol use problems in their fourth year of college. Neuroticism was not associated with drinking behaviors or with drinking before a test or assignment, but was associated with greater fourth year alcohol problems. Students lower in extraversion who reduced heavy drinking the day before academic events had fewer alcohol use problems at the fourth year of college relative to students higher in extraversion.
Conclusions
Students higher in extraversion appear to exhibit a continuity of established alcohol use patterns from adolescence, predisposing them to a more hazardous trajectory of college alcohol use. Characteristics of low extraversion may afford some protection from alcohol-positive college culture. High neuroticism appears associated with a hazardous trajectory of college alcohol use, but continued research into situational factors of alcohol use in high neuroticism is warranted.
Declaration of interest
The authors declare that they have no conflict of interest. The authors alone are responsible for the content and writing of the article.
Funding
Data used in this article were gathered with the support of funding from the National Institute on Alcohol Abuse and Alcoholism (R01AA016016) to J. L. Maggs.
Notes
1 The mean fixed slope across participants was added to reduce collinearity between students, and each slope was multiplied by 100 and centered to assist coefficient and confidence interval interpretation.
2 We considered multilevel structural equation modeling as an alternative approach that could retain all data, but found that the associations and suite of interactions we planned to test were analytically intractable in this framework.
3 Changes in average drinking due to academic events were converted into drinking probability using raw AE predicted scores with the function Y = exp[AE/D slope].