745 Association of Longitudinal Sleep Profiles with Cognitive Functioning in Bipolar Disorder

Abstract

Abstract Introduction Emerging evidence suggests that some cognitive deficits in bipolar disorder may be attributed to sleep disturbance. However, current findings are limited by cross sectional analyses or longitudinal studies with sparse follow up on sleep quality. The current study examined longitudinal profiles of sleep as predictors of cognitive performance across a series of domains to better understand how sleep disturbance may impact in bipolar patients. Methods 241 bipolar disorder and 103 control participants had their sleep measured at baseline, 6- and 12-months using the Pittsburgh Sleep Quality Index (PSQI). Individuals with a “poor” categorical PSQI sleep rating (PSQI total score >5) at all time points were categorized as “Poor Sleepers” (n = 134). Those rated “good” at each time point were “Healthy Sleepers” (n = 107), and those with at least one discrepant sleep rating at any time point were labeled “Unstable Sleepers” (n = 103). Cognitive performance was assessed at 12-months, testing domains of memory, processing speed, and executive function. ANOVAs controlling for demographics, mood symptom severity, and diagnostic group examined effects of sleep profiles on cognitive function. Results Main effects for sleep group were found for the California Verbal Learning Test across all short and long delay recall conditions (F’s 4.45 – 5.8, p <.02 for all). The poor sleep group performed the worst, while no main effects for diagnosis were observed. Main effects of sleep group were found for time conditions of the Trail making task (F = 4.18, p = .016), but not for errors. Conclusion Sleep quality was more strongly related to verbal and visual memory than bipolar diagnosis, with weaker effects for executive function. Some cognitive domains in bipolar disorder may be more closely related to sleep disturbance than endogenous factors. Future studies should explore how improvements in sleep functioning relate to neurocognitive outcomes, and whether cognitive performance varies as a function of sleep instability independent of mood instability. Support (if any) Support: Heinz C. Prechter Bipolar Research Fund at the University of Michigan Depression Center, the Department of Veterans Affairs (IK2-CX001501; PI: Boland), the VISN 4 MIRECC, and National Institute of Mental Health (K23MH118580; PI: Goldschmied).