Nocturnal sleep latencies among alert, alert-deprived and sleepy subjects.

Twenty-seven healthy, asymptomatic subjects participated in a study to determine repeated nocturnal sleep latencies as a function of level of daytime sleepiness. Subjects were screened polysomnographically and were determined to be free of any sleep pathology. Nine subjects had a screening multiple sleep latency test (MSLT) of < or = 5 min and 18 had a MSLT of > or = 10 min. Subjects were assigned to 3 groups, sleepy (MSLT < or = 5 min), alert, and alert sleep-deprived groups (MSLT > or = 10 min). Repeated nocturnal latencies to sleep were determined at 2300, 0110, 0310 and 0510 h on 2 nights with at least 5 days between nights. The results showed a main effect of group on the repeated nocturnal sleep latencies. Post-hoc tests demonstrated longer repeated nocturnal latencies among alert subjects compared to the sleepy and alert sleep-deprived subjects. The latter two groups had comparable sleep latencies. A significant main effect of time of night was also shown. The 0510 h latency was significantly shorter than the latencies at 2300 and 0110 h. The 0310 h latency was significantly shorter than the 0110 h latency and comparable to the 2300 and 0510 h latency. Significant correlations were found between daytime MSLT, repeated nocturnal sleep latencies, and nocturnal sleep time. These results show systematic differences in repeated nocturnal sleep latencies among sleep and alert subjects and extend the already described characteristics of these subjects by comparing them to a group of alert sleep-deprived subjects.