Functional EEG topography in sleep and waking: State-dependent and state-independent features
Introduction
Typical changes in the electroencephalogram (EEG) occur at the transition between waking and sleep, as well as at the transition between the major sleep states non-rapid eye movement sleep (NREMS) and rapid eye movement sleep (REMS). In fact, the scoring of human sleep is based to a large extent on differences in the amplitude and frequency of the EEG (Rechtschaffen and Kales, 1968). An increase in homeostatic sleep pressure induced by extended waking also gives rise to state-specific changes in the EEG. In NREMS, slow-wave activity (SWA; power in the 0.75- to 4.5-Hz band) is increased, whereas activity in the spindle frequency range (12–15 Hz) is reduced (Borbély et al., 1981, Dijk et al., 1990, Dijk et al., 1993, Finelli et al., 2001b, Knoblauch et al., 2002, Curcio et al., 2003; reviewed in De Gennaro and Ferrara, 2003, Borbély and Achermann, 2005). In waking, power in the theta range (5–8 Hz) is enhanced with progression of extended wakefulness (Cajochen et al., 1995, Cajochen et al., 1999, Aeschbach et al., 1997, Dumont et al., 1999, Finelli et al., 2000a, Strijkstra et al., 2003). The close association between SWA in NREMS and the duration of previous waking has led to the establishment of the two-process model of sleep regulation (Borbély, 1982, Daan et al., 1984, Borbély and Achermann, 2005).
Recording the EEG simultaneously from multiple sites made it possible to compute topographic power distributions and thereby to gain new insights into the dynamics of sleep. Specifically, the NREMS-REMS cycles were shown to be reflected by shifts in the power gradients along the antero-posterior axis (Werth et al., 1996). Cluster analysis of the NREMS EEG power maps revealed a segregation into different frequency bands corresponding closely to the traditional frequency bands (i.e., delta, theta, alpha, sigma, and beta) (Finelli et al., 2001b). This segregation appeared to be impervious to increased sleep pressure on the basis of both mean data (Finelli et al., 2001b) and individual maps (Finelli et al., 2001a). Therefore, different generators might underlie the frequency bands in the NREMS EEG. In addition, the increase in power in the low-frequency range in NREMS induced by prolonged waking was largest over frontal regions (Cajochen et al., 1999, Finelli et al., 2001b), whereas the corresponding decrease in power in the sigma band was most pronounced over the head vertex (Finelli et al., 2001b, Knoblauch et al., 2003). The power ratio recovery/baseline exhibited a topographic pattern similar to the power ratio between the first and second half of the baseline night (Finelli et al., 2001b). Thus, changes in sleep propensity are reflected in specific regional effects on EEG power. Nevertheless, these changes do not affect the topographical power distribution.
Functional neuroimaging studies with positron emission tomography (PET) complement these findings by showing that the prefrontal cortex is among the brain regions exhibiting the largest reduction of regional cerebral blood flow (rCBF) during NREMS (Maquet et al., 1997, Finelli et al., 2000b, Dang-Vu et al., 2005). Interestingly, a frontal deactivation could also be observed during REMS (Maquet et al., 1996, Finelli et al., 2000b), as well as in wakefulness after 24 h of sleep deprivation (Thomas et al., 2000). Thus, in addition to state-specific patterns of brain activation and deactivation, those and other imaging studies (reviewed in Maquet, 2000 and Nofzinger, 2005) highlight the existence of state-independent functional features of brain dynamics.
The main objective of this study was to explore and compare the functional topography of the EEG in all three vigilance states, NREMS, REMS and waking, and to investigate the response to sleep deprivation. Complementary to sleep data, waking data provide additional information to better interpret the causes of regional changes in the sleep EEG. The identification of state-dependent and state-independent features provides new insights into EEG dynamics.
Section snippets
Design of experiment
Eight right-handed, healthy male volunteers (mean age 23 years, range 21–25 years) participated in the study. After a baseline night (23:00 to 07:00 h), preceded by an adaptation night, subjects were kept awake for 40 h. Subsequent recovery sleep started at 23:00 h, and subjects were allowed to sleep 12 h. During the sleep deprivation period, subjects remained in the sleep laboratory and the surrounding area under constant supervision. The wake EEG was recorded at 3-h intervals starting at
Results
Sleep deprivation induced the typical changes in the sleep variables including a shortening of sleep latency, a reduction of waking after sleep onset and stage 1 and an increase of slow wave sleep, total sleep time and NREMS (Table 1, see also Table 1 in Finelli et al. 2000a). REMS and tonic REMS were not affected by sleep deprivation.
EEG power maps: state-independence and state-specificity
This is the first combined analysis of EEG topography of sleep (NREMS and REMS) and waking under conditions of normal and increased sleep pressure. The present study demonstrated that distinct frequency-dependent topographies of EEG power are present not only in NREMS (Buchsbaum et al., 1982, Zeitlhofer et al., 1993, Werth et al., 1997a, Finelli et al., 2001b) but also in waking and REMS. The question therefore arose to what extent the regional distribution of power is state-specific.
The
Acknowledgments
We thank Harry Baumann for his help with the experiment. The study was supported by the Swiss National Science Foundation grant 3100A0-100567.
References (57)
- et al.
Coherence analysis of the human sleep electroencephalogram
Neuroscience
(1998) - et al.
Dynamics of the human EEG during prolonged wakefulness: evidence for frequency-specific circadian and homeostatic influences
Neurosci. Lett.
(1997) - et al.
Sleep homeostasis and models of sleep regulation
- et al.
Sleep deprivation: effect on sleep stages and EEG power density in man
Electroencephalogr. Clin. Neurophysiol.
(1981) - et al.
Separation of circadian and wake duration-dependent modulation of EEG activation during wakefulness
Neuroscience
(2002) - et al.
Effect of total sleep deprivation on the landmarks of stage 2 sleep
Clin. Neurophysiol.
(2003) - et al.
Cerebral correlates of delta waves during non-REM sleep revisited
NeuroImage
(2005) - et al.
Sleep spindles: an overview
Sleep Med. Rev.
(2003) - et al.
Cortical EEG topography of REM onset: the posterior dominance of middle and high frequencies
Clin. Neurophysiol.
(2002) - et al.
An electroencephalographic fingerprint of human sleep
NeuroImage
(2005)