EARLY SEMESTER SLEEP VARIABILITY PREDICTS DEPRESSION AMONG COLLEGE STUDENTS

Abstract Introduction Sleep is a critical behavior predicting mental health and depressive symptomatology in young adults.The extant scientific literature generally focuses on self-reported sleep measures over relatively short time frames. Here, we examine whether actigraphy-measured sleep variables early in the academic semester predict depressive symptomatology at the end of the semester among first and second year college students. There is currently debate in the sleep literature about which sleep variables are the most robust predictors of depression among young adults. In this study, we evaluate total sleep time, midpoint sleep time, and sleep variability where variability is defined by the mean-squared successive difference (MSSD) of midpoint sleep as predictors of depression. Methods The sample consisted of 160 first and second year college students at a private American university. The students completed a beginning and end of semester assessment of depressive symptomatology using the Center for Epidemiologic Studies Depression Scale (CES-D), and wore a Fitbit throughout the semester to capture sleep features of interest: total sleep time (TST), midpoint sleep, and midpoint MSSD. Results When controlling for beginning of semester CES-D, early semester (weeks 3–6) midpoint sleep MSSD significantly (p < 0.05) predicted increased end of semester CES-D. These effects were specific to the sleep variability measure (MSSD). Total sleep time and sleep chronotype (i.e. midpoint sleep) were not significant predictors of end of semester depressive symptomatology. Conclusion Early semester sleep window variability among college freshmen, particularly during stressful midterm exams, is a robust risk factor for depression among college students. This work contributes to initial actigraphy studies suggesting that MSSD measures of sleep window variability foster increased mental health risks among young people. This work calls for further investigation to understand possible causal relationships between sleep variability and mental health. Support This work was supported by the Life@CMU project funded by the Carnegie Mellon University Provost’s Office.