Sleep Quality and the Edge-On Edge-Off Cycle of Caffeine and Alcohol Use
Andrea M. Hussong, Amanda K. Haik, Michaeline Jensen, Jennifer M. Traver, Adrianna N. RichardsBackground
The Edge On-Edge Off (EO-EO) hypothesis posits that poor sleep quality leads to low daytime arousal, which individuals may combat with caffeine use later in the day (i.e., Edge-On). Greater caffeine use then leads to higher arousal in the evening, which individuals may take advantage of or combat by drinking alcohol (i.e., Edge-Off). Heavy drinking, in turn, disrupts sleep (potentially adding hangover effects to daytime fatigue) and continues the cycle, particularly in the absence of structural barriers to drinking.
Objective
The current study examines the behavioral predictions of an EO-EO cycle of alcohol use in which poor sleep quality predicts next-day caffeine use and, in turn, later day drinking. Drinking behaviors then perpetuate this cycle across days through the disruption of sleep.
Design
Data came from a 10-day timeline follow-back survey completed by 1496 college students (aged 18-28; 61% female-identifying; 15% African American, 65% White, 12% Asian, 5% multi-racial, and less than 3% American Indian or Pacific Islander).
Methods
We estimated Latent Curve Models with Structured Residuals to examine within-person links and serially mediated effects involving poor sleep quality, caffeine and alcohol use, and next night’s poor sleep quality, differentiating between cycles occurring during the week versus the college weekend. We also tested the same cycle with napping duration, rather than caffeine use, as the within-day mediator.
Results
Findings supported the behavioral predictions of the EO-EO cycle linking poor sleep quality and greater subsequent caffeine and alcohol use, forming a recurring cycle. Evidence for this cycle was limited to the college weekend. Napping duration did not mediate the association between poor sleep quality and greater drinking.
Conclusions
The EO-EO cycle was supported primarily during the college weekend. These findings offer a promising proof of concept regarding the EO-EO cycle and point to several future directions for research.