How we fall asleep: regional and temporal differences in electroencephalographic synchronization at sleep onset

We hypothesized that the brain shows specific and predictable patterns of spatial and temporal differences during sleep onset (SO) reflecting a temporal uncoupling of electrical activity between different cortical regions and a dissociated wakelike and sleeplike electrocortical activity in different cortical areas.

We analyzed full-scalp electroencephalographic (EEG) recordings of 40 healthy subjects to investigate spatial and temporal changes of EEG activity across the wake-sleep transition. We quantified EEG sleep recordings by a fast Fourier transform (FFT) algorithm and by a better oscillation (BOSC) detection method to the EEG signals, which measured oscillatory activity within a signal containing a nonrhythmic portion.

The most representative spatial change at SO is the frontalization of slow-wave activity (SWA), while the θ activity, which mostly shares a similar temporal and spatial pattern with SWA, exhibits a temporo-occipital diffusion. The time course of these oscillations confirms that the changes of the dominant waves coexist with topographic changes. The waking occipital prevalence of α oscillations is progressively replaced by an occipital prevalence of θ oscillations. On the other hand, more anterior areas show a wide synchronization pattern mainly expressed by slow waves just below 4 Hz and by spindle oscillations.

The whole pattern of results confirms that the centrofrontal areas showed an earlier synchronization (i.e., they fall asleep first). This finding implies a coexistence of wakelike and sleeplike electrical activity during sleep in different cortical areas. It also implies that the process of progressive brain disconnection from the external world as we fall asleep does not necessarily affect primary and higher-order cortices at the same time.