Sleep is regulated by complex biological systems and environmental influences, neither of which is fully clarified. This study demonstrates differential effects of partial sleep deprivation (SD) on sleep architecture and psychomotor vigilance task (PVT) performance using two different protocols (sequentially) that each restricted daily sleep to 3 hours in healthy adult men. The protocols differed only in the period of sleep restriction; in one, sleep was restricted to a 3-hour block from 12:00 AM to 3:00 AM, and in the other, sleep was restricted to a block from 3:00 AM to 6:00 AM. Subjects in the earlier sleep restriction period showed a significantly lower percentage of rapid-eye-movement (REM) sleep after 4 days (17.0 vs. 25.7 %) and a longer latency to the onset of REM sleep (L-REM) after 1 day (78.8 vs. 45.5 min) than they did in the later sleep restriction period. Reaction times on PVT performance were also better (i.e. shorter) in the earlier SR period on day 4 (249.8 vs. 272 ms). These data support the view that earlier-night sleep may be more beneficial for daytime vigilance than later-night sleep. The study also showed that cumulative declines in daytime vigilance resulted from loss of total sleep time, rather than from specific stages, and underscored the reversibility of SR effects with greater amounts of sleep., H. Wu ... [et al.]., and Obsahuje bibliografii a bibliografické odkazy