The electroencephalographic substratum of the awakening

Abstract

The aim of the present study was to characterize the regional electroencephalographic substratum of the awakening process by means of a Hz-by-Hz EEG spectral power analysis. For this purpose, we recorded a group of 25 female subjects who slept for at least two consecutive nights in the laboratory. The postsleep waking EEG was compared to the one recorded during the presleep wakefulness from four midline derivations (Fz–A1, Cz–A1, Pz–A1, Oz–A1). Results indicated that the first 10 min after awakening are characterized by an increase of EEG power in the low-frequency range (1–9 Hz) compared to the corresponding presleep waking period, and by a significant decrease of EEG power in the beta range (18–24 Hz). As regards topographic differences, the increase of EEG power upon awakening in the delta–theta range showed a parieto-occipital prevalence. Moreover, the occipital derivation showed a larger decrease of power in the beta range as compared to the other derivations.

In conclusion, the EEG substratum of the sleep offset period is characterized by a pattern of increased EEG power in the delta–theta and low-alpha bands, and of decreased power in the beta range. This pattern could be considered as the spectral EEG signature of the sleep inertia phenomenon. The state of postsleep EEG hypoarousal does not subside in the first 10-min period after awakening considered in the present analysis. Finally, according to our results, the more posterior scalp locations show stronger EEG signs of sleep inertia, and could be the last ones to properly wake up.

Introduction

Sleep inertia denotes a period of transitory hypovigilance, confusion and impaired cognitive and behavioral performance that immediately follows awakening. While the variables influencing sleep inertia have largely been explored (reviewed in [12], [37]), its physiological underpinnings have received scarce attention. Nevertheless, the physiological data available clearly show that the sleep-to-wake transition is not a rapid shift from one state of consciousness to another, but a complex process that takes some time to be completed. As an example, it has incidentally been reported that cerebral blood flow (CBF) [28] and CBF velocity [22], [26] immediately following awakening are reduced in comparison to daytime wakefulness. In addition, Hajak et al. [22] showed that, upon morning awakening, subjects required up to half an hour to reach CBF velocity values corresponding to the waking state of the previous evening. The delayed increases in CBF velocity after awakening strongly suggest an uncoupling of cerebral electrical activity and cerebral perfusion. More recently, H₂¹⁵O PET was used to assess changes in regional CBF upon awakening from stage 2 sleep [3]. It was found that global CBF rates did not change during the first 20 min of wakefulness, indicating that the dissipation of sleep inertia is not a function of generally increasing levels of brain activation. Rather, the reactivation of local brain areas was critical. In fact, CBF was most rapidly reestablished in centrencephalic regions (e.g., brainstem and thalamus), suggesting that the reactivation of these regions underlies the reestablishment of conscious awareness. Then, across the ensuing 15 min of wakefulness, further increases in CBF were evident primarily in anterior cortical regions (prefrontal association cortices).

Electrophysiological studies also support the notion that brain activation levels upon awakening largely differ from those characterizing wakefulness. Visual evoked potentials (VEP) recorded upon awakening have long been recognized as having decreased amplitude and increased latency of 100–300 ms components relative to waking [5]. Moreover, following one-fourth of the arousals from slow-wave sleep (SWS), VEP contained an apparent carry-over of typical SWS components. No similar changes in VEP were observed after awakenings from REM sleep [5]. A decrease of P300 amplitude immediately upon awakening as compared to the presleep measure has also been observed [39]; this effect persisted for up to 15 min following awakening. A similar amplitude drop of the P300 wave has recently been found after sudden awakening following as short as 3 min of sleep [4]. In addition, an amplitude reduction of the N1–P2 complex of the auditory evoked potential (AEP) during the sleep–wake transition as compared to presleep wakefulness levels has been reported [13]. Moreover, during recovery sleep the N1–P2 amplitude decreased at parieto-occipital sites, while it increased at the frontal midline derivation as compared to baseline, as a possible effect of a compensatory effort of the frontal areas for the increased homeostatic drive for sleep during the recovery night [16].

The only study that examined EEG power spectra during behaviorally identified (button pressing to stop a tone) spontaneous arousals from sleep, showed a non-predicted, gradual and continued drop of power in the delta, theta and sigma bands, extending well into the first few minutes of wakefulness [29].

Altogether, these studies confirm that a state of cortical hypoarousal characterizes the waking period in the first few minutes after awakening from sleep. Moreover, it seems that the transition between sleep and wakefulness may show large regional differences, since the reactivation of some subcortical and cortical areas is faster than with other brain areas.

The first aim of the present study is to assess, for the first time, the EEG correlates of the awakening period after a whole night of sleep by means of a Hz-by-Hz EEG spectral power analysis. Moreover, the presence of regionally graded differences along the antero–posterior axis in the EEG substratum of the sleep offset period have also been evaluated.