1060 Investigating Mechanisms in the Sleep-Alcohol Relationship: An Ecological Momentary Assessment Approach

Abstract Introduction Research has established sleep disturbance as a risk factor for alcohol misuse. Nevertheless, we do not understand the mechanisms for the sleep-alcohol relationship. We hypothesized that higher negative mood and lower positive mood were mechanisms for the relationship between sleep disturbance and alcohol misuse. Methods 64 participants (60.9 % female, mean age 22) completed daily diaries about their sleep and alcohol use, wore an Actiwatch, and responded to ecological momentary assessments of their mood for 14 days. Multi-level zero-inflated Poisson models tested whether negative and positive moods mediated the relationship between prior-night sleep disturbance and next-day alcohol use, including the likelihood of abstaining from alcohol and drinking quantity amongst individuals who do drink. Models assessed both within-subject (daily) and between-subject (average) effects because daily variations in sleep could predict alcohol use independently from average patterns. Race, age, sex (0=male; 1=female), and day type (0=weekday or 1=weekend) were controlled in the analyses. Results Shorter average total sleep time (IR=0.996, p < .05) and lower average sleep efficiency (IR=0.968, p < .05) predicted increased daily drinking quantity. Conversely, lower daily sleep efficiency predicted decreased daily drinking quantity (IR=1.020, p < .05). Lower sleep quality predicted greater negative mood on both the daily (B=-0.093, p < .05) and average level (B=-0.195 p < .05), but negative mood did not mediate the sleep-alcohol use relationship (daily 95% CI [-0.021, 0.009]; average 95% CI [-0.028, 0.184]) Daily positive mood mediated the relationship between daily sleep quality and daily alcohol abstinence (95% CI [-0.084, -0.009]). Lower daily sleep quality predicted lower daily positive mood (B = 0.106, p < .05), and lower daily positive mood predicted greater likelihood of abstaining from alcohol use (OR=0.95, p < .05). Conclusion Sleep disruption predicted alcohol use differently depending on whether it was acute or chronic. Negative mood did not mediate the sleep-alcohol relationship, and positive mood mediated the relationship in an unexpected way; lower daily sleep quality predicted lower daily positive mood, which predicted greater likelihood of abstaining from drinking. Overall, findings suggest that the sleep-alcohol use relationship is nuanced, varying with both positive mood and timescale. Support (if any) Idaho State University; PSI CHI.