0116 Slow oscillations promote long range effective communication: the key for memory consolidation in a broken down network

Abstract

Abstract Introduction The relation between slow oscillations (SOs, <1Hz) during non-rapid eye movement (NREM) sleep and systems-level memory consolidation is one of the most robust findings in cognitive neuroscience. However, NREM is a brain state seemingly unfavorable to systems consolidation because a hallmark characteristic of this state is a breakdown in connectivity and reduction in synaptic plasticity with increasing depth of sleep. Our study addresses this apparent paradox and how SOs orchestrate neural communication. Methods We employed generalized partial directed coherence to estimate directional causal information flow between EEG channels across the electrode manifold during SO and non-SO periods. We examined the magnitude of causal information flow over the phase of SOs and found two peaks of flow preceding and following the trough of the SO. We categorized source-sink pairs of flows into three groups based on distance between source and sink of information flow. All the peak flows in each group were averaged and we tested relation between averaged magnitude of the flow and overnight episodic memory improvement using correlation test. Results The results reveal that NREM generally (non-SO periods) and during the SO trough show dampened neural communication. Causal communication during non-rapid eye movement sleep peaks during specific phases of the SO ( before and after SO trough), but only across long distances. Correlation test results showed that episodic memory improvement was predicted by peaks of information flow with longest distances between sinks and sources, and not by any other phase of the SO or non-SO period. Conclusion This work introduces a non-invasive approach to examine information processing during sleep, a behavioral stage whose function, until now, has been understood only at a delay. The findings represent a conceptual leap in understanding how slow oscillations unlock memory consolidation in a broken down network which is by promoting long range effective communication. This research will promote further investigations of understanding how brain oscillations alone and in nested rhythms promote network communication, as well as to investigate how these properties vary and predict patterns of deficits in clinical populations and aging humans. Support (If Any)