0097 REM-like Neural Activity Is Superior to NREM Parameters for Predicting Non-sleep Restricted Vigilance

Abstract

Abstract Introduction Currently unknown is whether REM-like neural activity reverses, incurs, or is neutral with respect to its effect on sleep pressure independent from NREM recovery. We investigated the relationship between a novel scalp-EEG measure of REM-like neural activity and next-day performance on the Psychomotor Vigilance Task (PVT) – a sensitive measure of sleep debt. Our EEG measure reflects the oscillatory-theta activity (OTA) that is the hallmark of REM sleep in animal studies. This method preserves amplitude information, isolates OTA from aperiodic theta-band power, and accounts for REM-like activity thought to occur outside of conventionally scored-REM epochs. We compared the PVT-association of OTA, traditionally measured EEG REM-theta, and traditional sleep parameters. To assess REM-like independent effects, we controlled for traditional parameters using a multivariate model. Methods Our method uses Irregular-Resampling Auto-Spectral Analysis to remove the aperiodic spectrum, followed by low-band power normalization to derive the relative oscillatory-theta activity (OTA). The average OTA was then used to predict next-day performance on the PVT. Traditional sleep parameters (sleep efficiency, total sleep time, %NREM, %N3, and delta-band power) were also examined. We combined data from two previously reported in-laboratory studies resulting in a sample size of 42 healthy young adult subjects. Analyses used non-sleep restricted overnight EEG recordings from a frontal channel. Results There was a substantial PVT-OTA association (OTA positively correlated with response-time and thus with reduced vigilance) during scored-REM epochs (r=0.44, p=0.003). This was stable irrespective of conventional sleep staging when using all sleep epochs (r=0.48, p=0.001), and OTA during scored-NREM (r=0.35, p=0.02). The effect was also stable after controlling for total sleep time, %NREM, and N3 delta-band power in a multivariate model (all-sleep PVT-OTA: r=0.5, p=0.004). Traditional sleep parameters were not significantly correlated with PVT performance. Conclusion OTA was a superior quantitative predictor of reduced next-day vigilance than traditional sleep parameters, and this persisted after controlling for NREM parameters. These findings are consistent with the hypothesis that periods of high REM-like activity are less restorative than other periods and may actually increase homeostatic sleep pressure. Support (If Any) This effort was supported by the Department of Defense Military Operational Medicine Research Program.