Dynamics of the EEG slow-wave synchronization during sleep

Objective: To study the dynamics of spatial synchronization of the slow-wave activity recorded from different scalp electrodes during sleep in healthy normal controls. Methods: We characterized the different levels of EEG synchronization during sleep (in the 0.25-2.5 Hz band) of five healthy subjects by means of the synchronization likelihood (SL) algorithm and analyzed its long-range temporal correlations by means of the detrended fluctuation analysis (DFA). Results: We found higher-levels of interregional synchronization during 'cyclic alternating pattern' (CAP) sleep than during nonCAP with a small but significant difference between its A and B phases. SL during CAP showed fluctuations probably corresponding to the single EEG slow-wave elements. DFA showed the presence of two linear scaling regions in the double-logarithmic plot of the fluctuations of SL level as a function of time scale. This indicates the presence of a characteristic time scale in the underlying dynamics which was very stable among the different subjects (1.23-1.33 s). We also computed the DFA exponent of the two scaling regions: the first, with values approximate to 1.5, corresponded to fluctuations with period 0.09-0.75 s and the second, with values approximate to 1, corresponded to fluctuations with period 1.5-24.0 s. Only the first exponent showed different values during the different sleep stages. Conclusions: All these results indicate a different role for each sleep stage and CAP condition in the EEG synchronization processes of sleep which show a complex time structure correlated with its neurophysiological mechanisms. Significance: Very slow oscillations in spatial EEG synchronization might play a critical role in the long-range temporal EEG correlations during sleep which might be the chain of events responsible for the maintenance and correct complex development of sleep structure during the night. (c) 2005 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.