Abstract
Abstract Introduction The aim of this study is to determine what behavioral and psychological factors were predictive of high-risk drinking in students with and without a Substance Use Disorder (SUD) diagnosis. We also compared how these relationships differed between Fall 2019, before the COVID-19 pandemic, and Fall 2020, a period of increased financial stress, distance learning, and social isolation. Methods We used factor analysis to extract a wide range of behavioral and psychological predictor variables from the Fall 2019 and Fall 2020 American College Health Association’s National College Health Assessment III (ACHA-NCHA). Next, we used structural equation modeling to identify significant drivers of high-risk alcohol use (ASSIST alcohol subscale) in students with (n = 577) and without (n = 50,926) SUD diagnosis. Results When compared to 2019, both students with and without SUD diagnosis showed lower levels of high-risk drinking, more difficulties falling asleep, and longer sleep duration. In students without SUD, short sleep duration and loneliness were the primary predictors of past year high risk alcohol use. In students with SUD, long sleep latencies and loneliness (UCLA Loneliness Scale) were the primary predictors, and sleep factors predicted more variance in high-risk alcohol use than in the general college population. Students with SUD diagnosis who also had sleep latencies >30 minutes were significantly more likely to report binge drinking (>5 drinks in one sitting) than those with shorter sleep latencies. Conclusion Sleep factors and loneliness are large predictors of problematic alcohol use in both students with and without diagnoses of Substance Use Disorder. In the first year of the COVID-19 pandemic, students sleep patterns changed with students reporting longer sleep, but also greater symptoms of insomnia. Support (if any)