Abstract :
[en] Background and aim. Aging affects the amount of cognitive resources available, and this can lead to decreased performance. Mental fatigue is another factor with a negative impact on cognition. However, it remains unknown how available cognitive resources interact with fatigue level across the lifespan. We investigated here how mental fatigue (induced by a cognitive task) affects cognitive performance according to age, and if eye-metrics (i.e., pupil response speed, quantity of long blinks) are a relevant marker of mental fatigue state over time.
Methods: 43 young, 25 middle-aged and 19 older people performed a 32-min mental fatigue induction task in condition of high or low cognitive load (HCL/LCL). Cognitive load was modulated by the stimuli time duration (STD), with shorter STD leading to higher load, and thus, higher mental fatigue across task. Images of the participant’s right eye were recorded by a camera positioned on a portable glass-like device. Before and after the task, participants rated on a visual analog scale their subjective level of fatigue. A mixed model analysis (p < .05) was conducted to determine the effects of group (young, middle-aged, older), condition (HCL vs. LCL) and time on task (4 blocks of 8 min) on mean accuracy. For a sub-sample of participants (N = 65), quantity of long blinks and pupil response speed were analyzed with mixed models.
Results: Participants reported more post-task fatigue, with less fatigue in the older. Mean accuracy performance was higher in the LCL condition compared to HCL, and during the first block compared to the following ones. Young people made more errors than the two other groups. No interaction was found in this model. For eye-metrics, more long blinks were recorded in the LCL condition, and a lower blink rate is observed in block 1 than blocks 3 and 4. The middle-aged and young groups increased their blink rates in the LCL compared to HCL condition, while the blink rate was similar between the HCL and LCL conditions in the older group. Pupil response speed (both constriction and dilatation) of the old group was slower in the HCL than LCL, as well as in comparison to young people, with the difference between the two groups being accentuated with increasing mental load.
Discussion. We succeed to induce a fatigue state with our task. Indeed, participants reported post-task fatigue and we observed lower accuracy in the HCL condition and for the last blocks. However, the absence of interaction on the accuracy measure suggests that the three groups vary the same way over blocks and conditions. Eye-metrics suggests that the LCL condition induced passive fatigue (by comparison to the more challenging HCL condition), as long blinks can be considered as indicator of drowsiness. Interestingly, pupil response velocity varies according to mental load (HCL vs LCL) in the older group only. These results are complex and indicate that the way fatigue affects cognitive performance vary across the life span, and that older people remains able to cope efficiently with a fatigue state.