0092 Subjective Insomnia, but not Objective Short Sleep Duration, is Associated with Impaired Sleep-Dependent Procedural Memory Consolidation

Abstract Introduction Sleep enhances procedural and declarative memory consolidation in healthy adults, but individuals with insomnia exhibit deficits in sleep-dependent memory processes. The mechanisms underlying these deficits are unclear. While subjective complaints define insomnia, some individuals also experience objectively short sleep (< 6 h/night). Here, we aimed to disentangle the effects of subjective insomnia and objective short sleep on overnight memory consolidation. Methods Thirty-eight participants were recruited into four groups: insomnia with short sleep (n=5), insomnia with normal sleep (n=13), short sleepers without insomnia (n=4), and good sleepers (n=16). Group classification was based on clinical interview and at-home actigraphy. Participants completed a four-day inpatient protocol that included training and immediate testing on the procedural Motor Sequence Task (MST) and declarative Word Pairs Task (WPT). Delayed retesting occurred after either a night of habitual-length sleep (monitored via polysomnography [PSG]) or an equivalent period of wakefulness. Linear mixed-effects models examined the interaction of condition (Sleep vs Wake), insomnia status, and habitual short sleep on percent change in performance from testing to retesting. Follow-up analyses evaluated potential moderation by PSG parameters. Results The three-way interaction among insomnia, short sleep, and sleep/wake condition did not predict changes in MST or WPT performance. However, a significant interaction between insomnia status and condition emerged for the MST (p=0.02), such that only participants without insomnia benefitted from sleep (Sleep–Wake=8.36% change; p=0.04) whereas participants with insomnia did not (Sleep–Wake=-5.20% change; p=0.22). On the WPT, no interaction was observed, but a significant main effect of condition indicated that all participants benefited from sleep relative to wake (Sleep–Wake=6.32% change, p=0.03). No PSG parameters, including total sleep time, significantly moderated these effects. Conclusion Here, only individuals without subjective sleep complaints benefited from a sleep-dependent boost in procedural memory, but all participants benefited in the declarative memory domain. Habitual sleep duration did not significantly influence outcomes, although there was a limited sample size for short sleepers. Future analyses will explore whether particular EEG features, such as cortical hyperarousal and spindles, mediate relationships between insomnia and performance. Support (if any) This work was supported by the NIH (R01NR018836).