Differential Kinetics in Alteration and Recovery of Cognitive Processes from a Chronic Sleep Restriction in Young Healthy Men

Abstract

Chronic sleep restriction (CSR) induces neurobehavioral deficits in young and healthy people with a morning failure of sustained attention process. Testing both the kinetic of failure and recovery of different cognitive processes (i.e., attention, executive) under CSR and their potential links with subject's capacities (stay awake, baseline performance, age) and with some biological markers of stress and anabolism would be useful in order to understand the role of sleep debt on human behavior. Twelve healthy subjects spent 14 days in laboratory with 2 baseline days (B1 and B2, 8 h TIB) followed by 7 days of sleep restriction (SR1-SR7, 4 h TIB), 3 sleep recovery days (R1-R3, 8 h TIB) and two more ones 8 days later (R12-R13). Subjective sleepiness (KSS), maintenance of wakefulness latencies (MWT) were evaluated four times a day (10:00, 12:00 a.m. and 2:00, 4:00 p.m.) and cognitive tests were realized at morning (8:30 a.m.) and evening (6:30 p.m.) sessions during B2, SR1, SR4, SR7, R2, R3 and R13. Saliva (B2, SR7, R2, R13) and blood (B1, SR6, R1, R12) samples were collected in the morning. Cognitive processes were differently impaired and recovered with a more rapid kinetic for sustained attention process. Besides, a significant time of day effect was only evidenced for sustained attention failures that seemed to be related to subject's age and their morning capacity to stay awake. Executive processes were equally disturbed/recovered during the day and this failure/recovery process seemed to be mainly related to baseline subject's performance and to their capacity to stay awake. Morning concentrations of testosterone, cortisol and α-amylase were significantly decreased at SR6-SR7, but were either and respectively early (R1), tardily (after R2) and not at all (R13) recovered. All these results suggest a differential deleterious and restorative effect of CSR on cognition through biological changes of the stress pathway and subject's capacity (ClinicalTrials-NCT01989741).

Keywords: age; chronic sleep debt; executive processes; hormones; recovery; sustained attention; time of day.

Figure 1

Experimental design. We show here the experimental design with sleep time opportunities (gray bars with black lines), blood (red bars) and salivary sampling (black bars), KSS and maintenance of wakefulness sessions (MWT) and cognitive test sessions (C letter) during baseline days (B1 or B2), the 1st, the 4th and the 4th days of sleep restriction (SR1, SR4 and SR7) and during the sleep recovery days (R1 or R2, R3… R13).

Figure 2

Total and sleep stages (N1, N2, N3 and REM Sleep) durations during baseline (B1, B2), the sleep restriction (SRN1 to SRN7) and the following sleep recovery period (RN1–RN3, RN12 and RN13). (A) Total sleep time, (B) N1 sleep, (C,D) N2 and REM sleep, (E,F) N3 sleep and proportion of N3 sleep (% of total sleep time) are represented (gray bars) as mean of amounts in min (or % for F; ± SEM) for the 12 subjects. Significant differences compared to baseline values (B2): *p < 0.05; **p < 0.01; ***p < 0.001.

Figure 3

(A) Number of lapses responses (RT > 500 ms) and (B) response speed in the simple reaction time task (SRTT) task and (C) number of errors (“NoGo responses”) and (D) response speed of “Go” trials in the Go-NoGo task. All these variables are represented as the mean (± SEM) of the 12 subjects that performed at baseline (B2), during the sleep restriction period (SR1, SR4 and SR7), and the following period of sleep recovery (R2, R3 and R13). Significant differences compared to baseline values (B2): ^#p < 0.05; ^##p < 0.01; ^###p < 0.001.

Figure 4

(A) Number of lapses responses (RT > 500 ms) and (B) response speed in the SRTT task and (C) number of errors (“NoGo responses”) and (D) response speed of “Go” trials in the Go-NoGo task, that subjects (N = 12) performed in morning (8:30–9:30 am, gray bars) and evening (6:30–7:30 pm, black bars) at baseline (B2), during the sleep restriction period (SR1, SR4 and SR7) and the following period of sleep recovery (R2, R3 and R13). All these variables are represented as the mean (± SEM) of the 12 subjects that performed at baseline (B2), during the sleep restriction period (SR1, SR4 and SR7), and the following period of sleep recovery (R2, R3 and R13). Significant differences compared to baseline values (B2): ^#p < 0.05; ^##p < 0.01; ^###p < 0.001.

Figure 5

Main variables that are correlated with SRTT (A) and Go-NoGo variables (B). Each line represents a significant correlation (r coefficient and p values written closed to the lines) between two variables. The strength of the relationship is symbolized by the thickness of the arrow. Thin arrows represent weak but significant correlations (p < 0.05) and thicker arrows symbolize more significant ones (p < 0.01 and p < 0.001). Blue lines represent correlations between test (or MWT latencies) variables and age of subjects. Green lines represent correlations between test variables and MWT latencies. Red lines represent correlations between variables inside each cognitive test; SRTT (A) and Go-NoGo task (B). Black lines represent correlations between test variables and morning α-amylase concentrations in saliva samples.