0082 Naturalistic Sleep Restriction and Domain-Specific Cognitive Performance in Young Adults

Abstract Introduction Young adulthood is a period in which long-term sleep habits are established, yet many are at increased risk for insufficient sleep. Laboratory research shows that acute and chronic sleep restriction impair vigilance and attention, whereas higher-order cognitive processes remain intact. Less is known about how naturalistic variations in sleep relate to different cognitive domains. This study examined the relationship between sleep duration and architecture and cognitive performance in young adults. Methods Healthy young adults enrolled at a large public university (N=70, 20.4±1.9y, 42 female) completed an in-lab cognitive test battery and at-home sleep assessment. The cognitive battery included the psychomotor vigilance test (PVT; vigilance and attention), Digit Symbol Substitution Test (processing speed), N-Back Task (working memory), Motor Praxis Task (sensory motor speed), Visual Object Learning Task (visual learning and spatial working memory), Abstract Matching (abstraction), Line Orientation Task (LOT; spatial orientation), and Balloon Analog Risk Task (BART; risk decision making). Sleep was assessed via wrist-actigraphy, worn for 14 days 24/7, and two nights of at-home polysomnography. Partial Spearman’s correlations, controlling for time of test administration, examined associations between sleep and cognitive performance. Results In this sample, habitual nocturnal sleep time ranged from 2-9h/night (6.2±1.6). Longer rapid eye movement (REM) sleep duration and higher sleep continuity associated with better performance on the PVT (REM sleep and PVT Lapses: r=-0.23, p=0.05; sleep efficiency and PVT errors: r=-0.30, p=0.01; wake after sleep onset and PVT errors: r=0.30, p=0.02). Higher sleep continuity also associated with less risk-taking on the BART (sleep efficiency and BART pumps: r=-0.25, p=0.04; wake after sleep onset and BART pumps: r=0.25, p=.04). Longer total sleep time (r=-0.31, p=0.009), REM sleep duration (r=-0.29, p=0.020) and non-REM stage 2 sleep duration (r=-0.34, p=0.004) associated with faster reaction times on the LOT. Conclusion These findings contribute to growing evidence that vigilance and reaction time are the most sleep-vulnerable cognitive domains. Risk-taking behavior was also sensitive to sleep continuity. In this university-aged cohort, processing speed, working memory, abstraction, verbal learning, and sensory-motor speed appear to remain resilient to naturally occurring variations in sleep. Support (if any)