A dose-response study of sleep loss and spontaneous sleep termination.

Recent concepts of sleep/wake regulation have emphasized circadian influences and largely disregarded homeostatic ones. The present experiment was designed to study sleep loss homeostasis while minimizing confounding circadian influences. Eight male subjects participated in the study. Night sleep was curtailed across four conditions to yield 0, 2, 4, or 8 hrs of sleep. The effects were studied on subsequent day sleep begun at 1100h and spontaneously terminated. Total sleep time (TST), Stage 2 (S2), and Stages 3+4 (SWS) showed very strong dose-dependent increases with increasing loss. REM sleep did not respond. After maximum sleep loss TST and S2 doubled whereas SWS increased fivefold. Sleep did not terminate until the prior loss of SWS had been recovered. The total SWS recovery approximately matched the loss. TST, S2, and REM failed to recover more than limited amounts of the loss. The results show that homeostatic influences on sleep may be much larger than usually acknowledged and that SWS closely, although not perfectly, reflects the “active component’ of sleep homeostasis.