The effects of exercise upon sleep: a critical review.

The topic of exercise and sleep is often seen to address the question of sleep function. If sleep were a special state for body restitution, then daytime exercise would be expected to cause an increase in tissue repair during subsequent sleep. However, comprehensive measures of body restitution have yet to be taken during this sleep. Those changes which do occur are associated with the sleep EEG, where the main effect in man is an increase in SWS, primarily with trained (fit) subjects. As the sleep EEG is not a measure of body restitution, but only an indication of brain functioning, such SWS changes can only be described in terms of the effects of exercise upon the brain. Major methodological problems pervade exercise and sleep studies, particularly, differences in sleep stage scoring criteria, and "noise" due to within-subject variability of the sleep EEG, often coupled with insufficient recording nights. Between-study comparisons indicate that trained subjects may have higher baseline SWS values than have untrained subjects, although the few studies of both subject types indicate little difference. Following acutely increased exercise (usually only one day), untrained subjects may show transient SWS elevations limited to the first part of sleep, but no whole-night changes. Trained subjects are more likely to show the latter response. Total energy expenditure of the exercise seems to have a lesser SWS effect than does rate of expenditure. The greater endurance of trained subjects seems to produce a greater likelihood of heat and otherr stresses; this may be an important key to the SWS increases, particularly as body heating without exercise produces similar SWS changes. Little is known of subjective sleep effects of exercise. By EEG criteria, exercise does not seem to improve sleep, but as good sleepers are typically used, there is little room for such improvement. Shorter-term studied of bed-rest and reduced exercise in healthy subjects result in little effect on sleep, although there are changes under chronic conditions. No study has looked at the sleep effects of becoming trained. Findings from animal studies indicate an increase in non-REM sleep following exercise. The deeper form of this sleep is either unaffected or remains unassessed. However, sleep deprivation and othe stressors may have been confounding factors here. Finally, several proposals for human research are made, and further aspects of the topic of body restitution and human sleep function are discussed.