Effects of circadian misalignment on cognition in chronic shift workers

Abstract

Shift work is associated with increased human operational errors, presumably due to the circadian timing system that inhibits optimal cognitive function during the night. Circadian misalignment, which is the misalignment between the circadian pacemaker and behavioral/environmental cycles, impairs cognitive performance in non-shift workers. However, it remains uncertain whether the adverse cognitive consequences of circadian misalignment are also observed in chronic shift workers. Thus, we investigated the effects of circadian misalignment on cognitive performance in chronic shift workers. Using a randomized, cross-over design that simulated day shift work (circadian alignment) and night shift work (circadian misalignment), we show that circadian misalignment increases cognitive vulnerability on sustained attention, information processing and visual-motor performance, particularly after more than 10 hours of scheduled wakefulness. Furthermore, their increased levels of subjective sleepiness and their decreased sleep efficiency were significantly associated with impaired sustained attention and visual-motor performance. Our data suggest that circadian misalignment dramatically deteriorates cognitive performance in chronic shift workers under circadian misalignment. This increased cognitive vulnerability may have important safety consequences, given the increasing number of nighttime jobs that crucially rely on the availability of cognitive resources.

Figure 1

Within-subject, randomized, cross-over study design. Circadian alignment (upper panel) and misalignment (lower panel) protocols. For the former, scheduled sleep times were maintained between 11PM to 7AM across all days, while for the latter these timings were inverted by 12 h after Day 1. During Day 2 for the aligned condition, the Psychomotor Vigilance Task (PVT) and Probed Recall Memory (PRM) with Presentation phase (PP) and Recall phase (RP) were conducted at 2PM and 6PM, the Addition Task (ADD) at 12PM and 4PM, and the Unstable Tracking Task (TKT), Digit Symbol Substitution Task (DSST), Performance evaluation and effort scales (PEERS) and Karolinska Sleepiness Scale (KSS) at 12PM, 2PM, 4PM and 6PM. For the misaligned protocol, timing of cognitive testing during Day 2 was inverted by 12 h. Light levels were 90 lux to simulate typical room light intensity, 450 lux for 30-minute periods to simulate the morning commute preceding the simulated day shift and following the simulated night shift, 4 lux to permit assessment of dim-light melatonin levels, and 0 lux during scheduled sleep.

Figure 2

Cognitive performance in chronic shift workers under circadian alignment/misalignment. (A) Sustained attention (PVT 10% slowest reaction times and PVT lapses) worsened under circadian misalignment, particularly after 11 h of scheduled wakefulness, as compared to circadian alignment. (B) Information processing (number of correct DSST responses/min) performance improved only under circadian alignment, particularly when DSST assessments occurred 11 h after scheduled wakefulness, as compared to circadian misalignment. (C) Visual-motor performance (number of TKT losses) became progressively worse under circadian misalignment, particularly after 7 h of scheduled wakefulness, as compared to circadian alignment. White and black circles correspond to individual data under circadian alignment and misalignment conditions, respectively. Data correspond to mean ± standard error of the mean, *p < 0.05 (see results for statistics).

Figure 3

Subjective ratings of sleepiness and performance in chronic shift workers under circadian alignment/misalignment. (A) Subjective sleepiness (KSS) indicated higher levels of sleepiness during circadian misalignment, particularly after 9 h of scheduled wakefulness, as compared to circadian alignment. (B) Subjective ratings of performance (PEERS) did not differ between circadian alignment and misalignment conditions. White and black circles correspond to individual data under circadian alignment and misalignment conditions, respectively. Data correspond to mean ± standard error of the mean, *p < 0.05 (see results for statistics).

Figure 4

Association of misalignment-induced change in subjective sleepiness with cognitive performance in chronic shift workers. (A,B) Correlations between effect of misalignment for subjective sleepiness levels (x-axis: KSS levels) and for sustained attention (y-axis: (A) PVT slowest reaction times, (B) PVT lapses), (C) information processing (y-axis: DSST ratio of number of correct responses/minute), and (D) visual-motor performance (y-axis: TKT number of losses). X-axis and Y-axis correspond to the change from circadian alignment to misalignment conditions (see results for statistics).

Figure 5

Association of sleep efficiency with cognitive performance in chronic shift workers. (A) Sleep efficiency was lower during circadian misalignment, as compared to when the same individuals were under circadian alignment. Correlations between sleep efficiency levels (x-axis: difference of sleep efficiency between circadian misalignment and alignment conditions) and sustained attention (y-axis: (B) PVT slowest reaction times, (C) PVT lapses), (D) information processing (y-axis: DSST ratio of number of correct responses/minute), and (E) visual-motor performance (y-axis: TKT number of losses). X-axis and Y-axis correspond to the difference between circadian misalignment and alignment conditions. For panel A, white and black circles correspond to individual data under circadian alignment and misalignment conditions, respectively. Data correspond to mean ± standard error of the mean, *p < 0.05 (see results for statistics).