Staying vigilant during recurrent sleep restriction: dose-response effects of time-in-bed and benefits of daytime napping

Abstract

Study objectives: We characterized vigilance deterioration with increasing time-on-task (ToT) during recurrent sleep restriction of different extents on simulated weekdays and recovery sleep on weekends, and tested the effectiveness of afternoon napping in ameliorating ToT-related deficits.

Methods: In the Need for Sleep studies, 194 adolescents (age = 15-19 years) underwent two baseline nights of 9-h time-in-bed (TIB), followed by two cycles of weekday manipulation nights and weekend recovery nights (9-h TIB). They were allocated 9 h, 8 h, 6.5 h, or 5 h of TIB for nocturnal sleep on weekdays. Three additional groups with 5 h or 6.5 h TIB were given an afternoon nap opportunity (5 h + 1 h, 5 h + 1.5 h, and 6.5 h + 1.5 h). ToT effects were quantified by performance change from the first 2 min to the last 2 min in a 10-min Psychomotor Vigilance Task administered daily.

Results: The 9 h and the 8 h groups showed comparable ToT effects that remained at baseline levels throughout the protocol. ToT-related deficits were greater among the 5 h and the 6.5 h groups, increased prominently in the second week of sleep restriction despite partial recuperation during the intervening recovery period and diverged between these two groups from the fifth sleep-restricted night. Daytime napping attenuated ToT effects when nocturnal sleep restriction was severe (i.e. 5-h TIB/night), and held steady at baseline levels for a milder dose of nocturnal sleep restriction when total TIB across 24 h was within the age-specific recommended sleep duration (i.e. 6.5 h + 1.5 h).

Conclusions: Reducing TIB beyond the recommended duration significantly increases ToT-associated vigilance impairment, particularly during recurrent periods of sleep restriction. Daytime napping is effective in ameliorating such decrement.

Clinical trial registration: NCT02838095, NCT03333512, and NCT04044885.

Keywords: napping; recovery sleep; sleep restriction; sustained attention; time on task; vigilance.

© Sleep Research Society 2022. Published by Oxford University Press on behalf of the Sleep Research Society.

Figures

Figure 1.

Experimental protocols. The study protocol lasted for 14 (9 h group) to 15 days (remaining groups). (A) The 9 h group maintained the same 9-h TIB (23:00–08:00; black bars) throughout the protocol (*day R22 is not applicable to the 9 h group because of its shorter protocol duration). All other groups started with two adaptation and baseline nights (B1 and B2) of 9-h TIB per night (black bars). This was followed by the first cycle of sleep manipulation for five nights (M11 to M15) and recovery sleep for two nights (R11 and R12; TIB = 9 h; black bars). The second cycle consisted of just three nights of sleep manipulation (M21 to M23) and two nights of recovery sleep (R21 and R22; black bars). During both sleep manipulation periods, participants had nocturnal TIBs of 5 h (5 h group; red bars), 6.5 h (6.5 h group; orange bars), or 8 h (8 h group; blue bars). Additionally, nap groups were given a nap opportunity starting at 14:00 following each sleep manipulation night. (B) Two nap groups had 5 h TIBs on the manipulation nights and were allowed to nap for either 1 h (5 h + 1 h group; shaded dark red bars) or 1.5 h (5 h + 1.5 h; dark red bars). (C) Another nap group had a 6.5-h TIB at night and 1.5-h nap opportunity (6.5 h + 1.5 h; brown bars) each day during the manipulation periods. A cognitive test battery (green bars) was administered daily at 10:00, 15:00–16:15, and 20:00, except during the first and last days of the protocols.

Figure 2.

Time-on-task effects on the number of lapses in the Psychomotor Vigilance Task (PVT) as a function of nocturnal TIB and daytime napping. Left panel: The means and standard errors of the number of PVT lapses in each 2-min bin, averaged across the three tests each day, are plotted from the second baseline day (B2) to the day after the first recovery night in the second cycle (R21). Right panel: The changes from the first to the fifth bin (time-on-task effect) in the number of PVT lapses, averaged across the three tests each day, are shown. The least square means and standard errors estimated with general linear mixed models are illustrated. Shaded gray areas indicate the sleep manipulation periods (M11 to M15 and M21 to M23). (A) Data are shown for the 9 h group (black), 8 h group (blue), 6.5 h group (orange), and 5 h group (red). (B) Data are plotted for the 5 h group (red), the 5 h + 1 h group (dark red open circles with dotted line), and the 5 h + 1.5 h (dark red filled circles with solid line). ^^^p < 0.001, ^^p < 0.01, and ^p < 0.05 for significant contrasts between the 5 h group and the 5 h + 1 h group. ***p < 0.001, **p < 0.01, and *p < 0.05 for significant contrasts between the 5 h group and the 5 h + 1.5 h group. (C) Data are plotted for the 6.5 h group (orange) and the 6.5 h + 1.5 h (brown). ***p < .001, **p < .01, and *p < .05 for significant contrasts between the 6.5 h group and the 6.5 h + 1.5 h group.