0089 Influence of REM Sleep on Working Memory Precision and Capacity

Abstract Introduction The rigidity or fluidity of working memory (WM) capacity has been debated, with some studies suggesting that capacity is strictly limited by the number of to-be-remembered objects. While others indicate that prioritizing certain items in a memory array (and therefore creating more precise representations) can influence the number of items we are able to store. This creates a tradeoff between memory precision and capacity. Sleep, specifically NREM, has been shown to enhance WM using complex sequences or updating tasks (e.g. OSPAN, n-back). However, no studies have examined whether sleep plays a role in the trade-off between capacity and precision. This study bridges this gap by examining how a nap versus wake affects WM capacity and precision measured from a continuous color recall task. Methods 64 healthy adults (18-35 yrs) completed a WM continuous recall task before and after a period of rest. Participants were randomized into three rest groups: a 60-minute NREM nap with polysomnography (PSG), a 90-minute NREM+REM nap with PSG, and an unmonitored active wake period. Using linear models (LMs), we predicted post-sleep WM capacity and precision, with rest group and sleep architecture (total sleep time (TST) and total time spent in NREM and REM) as predictors. Results The NREM+REM group had moderately higher post-nap improvement in precision than the NREM group. There were no significant differences in precision changes between the AW group and either nap group, and no differences in capacity between sessions across the three groups. Time in REM, but not NREM, correlated with precision improvement. Finally, we found that increased REM sleep moderately predicted a tradeoff between improvement in precision over capacity (i.e. capacity change - precision change). Conclusion Our results show that while WM capacity does not improve with a nap, precision is influenced by sleep. We specifically found that REM sleep supports WM precision improvement and the trend toward precision improvement over capacity improvement. Our results provide preliminary evidence supporting a relation between sleep and WM capacity and precision. Support (if any) University of California - Irvine Internal Grant