Institute of Pharmacology, University of Zurich, Zurich, Switzerland
hree processes are involved in sleep regulation: (1) A homeostatic process determined by the history of sleep and waking; (2) a circadian process, controlled by an endogenous pacemaker which is basically independent of sleep and waking; and (3) an ultradian process within sleep which is represented by the alternation of nonREM sleep and REM sleep. In the two-process model, the interaction of a homeostatic and circadian process specifies the timing and propensity of sleep. This model has gradually gained wide acceptance. EEG slow-wave activity, the physiological marker of nonREM sleep homeostasis, proved to be highly dependent on prior sleep and waking in both humans and animals. Moreover, in recent studies using the forced desynchrony protocol, the homeostatic and circadian facets of sleep could be separated and analyzed. Finally, extensive, elegant electrophysiological studies demonstrated that slow-wave activity has a counterpart in the fluctuations of the membrane potentials of thalamocortical neurons, a consequence of the advanced state of hyperpolarization that is typical for deep nonREM sleep. An intermediary degree of hyperpolarization (i.e. superficial nonREM sleep) is accompanied by fluctuations in the spindle frequency range. This opened the possibility of incorporating different segments of the EEG in a functional analysis of sleep.
Recent experiments with recordings from multiple derivations revealed that the spectral power of the sleep EEG differs along the antero-posterior axis. These differences were shown to be frequency-specific as well as state-specific, and to exhibit consistent variations across and within nonREM sleep episodes. A key question raised by the findings is whether cortical regions may be differently involved in the sleep process.
Another recent development is the attempt to characterize the processes underlying REM sleep regulation. It was demonstrated that a three-night selective REM sleep deprivation results in an increase in "REM sleep pressure" during the deprivation period but only in a short-lasting REM sleep rebound during recovery. However, longer lasting effects in the recovery period were revealed by a spectral analysis of the REM sleep EEG. The incorporation of REM sleep regulation and brain topography in an elaborate model constitutes one of the future challenges in sleep research.
