0100 Human Sleep Spindles Coupled To Hippocampal Sharp Wave Ripples Have Characteristic EEG Features. (12th April 2019)

• Record Type: Journal Article
• Authors: Coon, William G; Valderrama, Mario; Varela, Carmen; Amaya, Valentina; Henao, David; Stickgold, Robert; Wilson, Matthew; Manoach, Dara
• Is Part Of: Sleep. Volume 42(2019)Supplement 1
• Page Start: A41
• Page End: A41
• Publication Date: 2019-04-12
• Subjects: Sleep -- Physiological aspects -- Periodicals; Sleep disorders -- Periodicals; Sommeil -- Aspect physiologique -- Périodiques; Sommeil, Troubles du -- Périodiques; Sleep disorders; Sleep -- Physiological aspects; Sleep -- physiological aspects; Sleep Wake Disorders; Psychophysiology; Electronic journals; Periodicals; 616.8498
• DOI: 10.1093/sleep/zsz067.099
• Languages: English
• ISSNs: 0161-8105

Abstract

Abstract: Introduction: Sleep spindles mediate memory consolidation during sleep and are markedly reduced in schizophrenia. While spindle deficits correlate with impaired sleep-dependent memory, pharmacologically increasing spindle density in schizophrenia does not always improve memory. This may be because coupling with other NREM sleep oscillations like hippocampal sharp-wave ripples (SWRs) is required for spindles' optimal memory benefit and those spindles induced by these drugs are not coupled with SWRs. Identifying SWR-coupled spindles cannot currently be accomplished with scalp EEG alone since it cannot detect hippocampal activity. The EPILEPSIAE dataset of simultaneously recorded scalp EEG (to detect spindles) and intracranial EEG (iEEG; to detect SWRs) presents an opportunity to identify an EEG signature of SWR-coupled spindles. This would allow SWR-spindle coupling to be evaluated noninvasively using scalp EEG alone. Methods: To distinguish SWR-coupled from non-coupled spindles, we analyzed data from n=5 human subjects implanted with intracranial EEG (iEEG) electrodes in the hippocampus several days prior to invasive brain surgery. Polysomnography was acquired simultaneously. We identified sleep spindles in 17-21 EEG locations per subject and SWRs from 2-7 hippocampal channels per subject with automated algorithms. We then identified the subset of spindles coupled to SWRs and tested for significant differences in spindle features from non-coupled spindles. Results: SWR-coupled spindles were significantly longer (669ms ± 33ms [S.D.] vs. 603ms ± 16ms; p = 0.028; two-tailed paired t-test across subjects) and significantly faster (12.3Hz ± 0.9Hz vs. 11.7Hz ± 0.7Hz; p = 0.029; two-tailed paired t-test across subjects) than non-coupled spindles. Conclusion: Our findings suggest that a classifier could be trained to identify SWR-coupled spindles based on their EEG characteristics alone. This would allow a more refined characterization of spindle deficits in schizophrenia and testing of the effects of drugs on spindle-SWR coordination and memory. This, in turn, would substantially advance our ability to identify possible treatments for the cognitive deficits in schizophrenia, for which no remedies are available. Support (If Any): NIH-NHLBI 5T32HL007901-17; R01MH092638; K24MH099421