0096 Sleep Spindle Coherence And Density Predict Sleep-enhanced Learning In Schizophrenia. (27th April 2018)

• Record Type: Journal Article
• Authors: Coon, W G; Mylonas, D; Baran, B; Demanuele, C; Stickgold, R; Manoach, D
• Is Part Of: Sleep. Volume 41(2018)Supplement 1
• Page Start: A38
• Page End: A39
• Publication Date: 2018-04-27
• 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/zsy061.095
• Languages: English
• ISSNs: 0161-8105

Abstract

Abstract: Introduction: Schizophrenia (SZ) is marked by impairments in cognition, learning, and memory. Sleep spindles, a defining characteristic of Stage 2 NREM sleep, are thought to mediate memory consolidation during sleep and are markedly reduced in SZ. While spindle deficits correlate with impaired sleep-dependent memory in SZ, pharmacologically increasing spindle density does not improve memory performance. Because naturally occurring spindles involve the coordinated activity of multiple brain regions and memory should benefit from such an integrative process, we hypothesized that spindle activity that involves coherent cortical activity should relate to memory functions more closely than spindle activity in isolation. Methods: Twenty-three (23) SZ patients and 28 demographically-matched healthy controls (HC) spent two overnights in the MGH Clinical Research Center and were monitored with polysomnography. Subjects learned a motor sequence task (MST) before bedtime on the second night and were tested the following morning. We identified sleep spindles at electrode Cz with an automated detection algorithm and calculated (from Laplacian-transformed data) sigma band (12 - 15Hz) coherence during each spindle event (+/- 1-sec from spindle peak) as the average coherence derived from all possible electrode pairs (57 total). We then constructed linear models to predict overnight MST improvement as a function of spindle density (number per minute), spindle coherence, or both. Results: While spindle coherence alone did not predict improvement (F(1, 21)=1.11, p=0.304), a multivariate model including spindle density and spindle coherence did (F(3, 19)=3.60, p=0.033). More importantly, the multivariate model explained more than twice the variance in overnight improvement as the univariate model based on density alone (R 2 =0.362 vs. R 2 =0.178). No model was significant for HC subjects (p>0.05). Conclusion: Our findings suggest that it is not simply the density of sleep spindles but also their coordination across the cortex that is critical for effective memory consolidation during sleep. Additionally, these results are relevant to understanding previous findings that pharmacologically augmenting spindle activity in SZ does not improve memory; it may be necessary to both increase spindles and maintain or enhance their synchronous expression to restore memory function. Support (If Any): 5T32HL007901-17; R01MH092638; K24MH099421.