REM sleep - by default?

Elements of three old, overlapping theories of REM sleep (REM) function, the Ontogenetic, Homeostatic and Phylogenetic hypotheses, together still provide a plausible framework - that REM (i) is directed towards early cortical development, (ii) "tones up" the sleeping cortex, (iii) can substitute for wakefulness, (iv) has a calming effect. This framework is developed in the light of recent findings. It is argued that the "primitiveness'' of REM and its similarity to wakefulness liken it to a default state of "non-wakefulness" or a waking antagonist, anteceding "true" (non-REM) sleep. The "toning up" is reflected by inhibition of motor, sensory and (importantly) emotional systems, together pointing to integrated "flight or fight" activity, that preoccupies/distracts the organism when non-REM is absent and wakefulness unnecessary. Dreaming facilitates this distraction. In rodents, REM can provide stress coping and calming, but REM deprivation procedures incorporating immobility may further enhance stress and confound outcomes. REM "pressure" (e.g. REM rebounds) may be a default from a loss of inhibition of REM by non-REM. REM can be reduced and/or replaced by wakefulness, without adverse effects. REM has little advantage over wakefulness in providing positive cerebral recovery or memory consolidation. (C) 2000 Elsevier Science Ltd. All rights reserved.