Habitual sleep length and patterns of recovery sleep after 24 hour and 36 hour sleep deprivation.

Five long sleepers (LS) and 5 short sleepers (SS) were selected from 310 medical students. Nine regular sleepers (RS) were used as a control. The sleep was recorded during 3 reference nights, one recovery night after a 36 h sleep deprivation (R2), one morning sleep after a 24 h sleep deprivation (D1) and the night following D1(R1). According to previous data slow wave sleep (SWS) amounts were the same in the 3 groups while stage 2 and paradoxical sleep (PS) amounts increased with the sleep duration. The hourly distribution of intervening wakefulness and SWS were similar for all groups. When compared to RS or SS the hourly distribution in LS of PS was lower until the sixth hour. As a function of experimental conditions, sleep patterns of LS were the most affected. In R2 the sleep of LS more closely resembled that of RS or SS than in reference nights, while in R1 LS' sleep was the most disturbed. Morning sleep durations were very similar for all groups, but in LS intervening wakefulness was increased and PS was decreased when compared to RS and SS. Negative correlations (Spearman rank test) were found between the morning increase of body temperature after a sleep-deprived night and both TST and PS durations. In all recorded sleep periods, SWS amounts were positively correlated with prior wakefulness duration and the PS amount with TST.