0335 Evaluating the Role of Circadian Rhythm Dysregulation in Alcohol Use Disorder (AUD) Relapse Probability

Introduction Identifying factors predictive of relapse in patients with alcohol use disorder (AUD) is essential to learning how to effectively treat this disorder. Our group previously found that following inpatient treatment, AUD patients with lower sleep regularity were more likely to relapse. The aim of this study is to discern whether circadian rhythm dysregulation is predictive of relapse. Methods A cohort of patients who received inpatient treatment for AUD (n=126) wore Philip Respironics actiwatches following discharge for four weeks. A subset of patients additionally wore the watches during prior discharge from inpatient treatment (n=24). Relapse status was assigned if a patient consumed any alcohol during the outpatient period of data collection. We calculated circadian rhythm nonparametric statistics including interdaily stability (IS) and intradaily variability (IV) for the inpatient and outpatient periods. Linear and logistic generalized mixed models were fit in R to estimate the effect of the inpatient-outpatient transition on circadian behavior, and of weekly circadian parameters on relapse probability. All analyses account for within-patient repeated measurements. Results Final sample size included 103 patients after actigraphy data filters were applied. Patients were 48.6±11.3 years of age and 32% female. Among patients with at least 4 days of inpatient data, IV was on average 0.068 units lower during the first week post-discharge (SE=0.018, F[1,29]=14.92, p< 0.001). By contrast, IS did not substantially change following discharge. Twenty-six patients (25.2%) experienced a relapse event. Relapse at any time during the study was associated with a mean 0.03 unit decrease in IS (SE=0.014, F[1,101]=4.29, p< 0.05), whereas relapse was not associated with a significant shift in IV. Conclusion Our preliminary analyses detected a change in circadian behavior (IV) at discharge, which may be an effect of scheduled activities that cease upon discharge. General alignment to a 24-h cycle, as captured by IS, was not disrupted by discharge. However, IS but not IV exhibited an association with relapse. Additional research on this topic will include clinical patient covariates and time to determine if IV and/or IS are predictive of later relapse and time until relapse. Support (if any) This research was supported by the National Institutes of Health Intramural Research Program.