Grasping the sporadic nature of electrodermal activity provides robust assessment of different cognitive load even in a perturbated status induced by sleep debt

Electrodermal activity (EDA) has long been a valuable tool in human monitoring, particularly for investigating emotional states in individuals. EDA serves as a reliable indicator of sympathetic autonomic responses in humans. Most EDA features described in the literature are issue from time-averaged analyses, based on as the tonic component of this biosignal. However, the inherently time-varying and sporadic nature of sympathetic responses suggests that focusing on the phasic component of EDA may offer a more suitable approach for capturing and analyzing the intermittent phenomena especially in medical and military contexts where tasks require substantial cognitive effort frequently performed under conditions of severe sleep deprivation. EDA signals were recorded from the plantar region in 32 participants (18 males, 14 females; age = 32 ± 7.8 years) during two MATB-II tasks designed to induce low and medium cognitive workloads. The experimental sessions were conducted over three consecutive days, during which sleep restriction protocols were implemented: 8 hours of sleep on the first day, total sleep deprivation on the second day, and 3 hours of sleep on the third day. Indices from EDA signal were calculated as averages over the different experimental tasks and over time to capture the dynamics of autonomic responses. To achieve this, frequency domain analyses were applied to EDA signal using methods reported in the literature (CVX, VFCDM, WPT). This study specifically focused on the TVSymp index, previously identified as the more sensible metric for characterizing individual emotional state. Perceived workload was evaluated using the NASA-TLX questionnaire, administered after each task. First, the results confirm the effect of progressive sleep debt on perceived mental demand (F = 4.06, p = 0.022), with a more pronounced difference observed when subjects performed tasks with a medium cognitive workload (F = 25.475, p < 0.01). The TVSymp index reliably captured the effect of cognitive load across the two experimental conditions (low vs. medium cognitive load in MATB-II tasks, F = 12.599, p < 0.001), irrespective of the level of sleep debt (F = 0.499, p = 0.609). In conclusion, the subtle and sporadic nature of sympathetic autonomic activation, reflecting individual responses to cognitive load variations, can be effectively assessed through frequency-varying analyses of EDA. Notably, the TVSymp index appears to be a robust marker for detecting subtle changes in cognitive demand, even when subjects undergo significant sleep deprivation. This research was funded by a CIFRE grant (2022/0869) awarded by the ANRT to THALES AVS FRANCE SAS, in support of the work of Boffet Alexis PhD student This abstract was presented at the American Physiology Summit 2025 and is only available in HTML format. There is no downloadable file or PDF version. The Physiology editorial board was not involved in the peer review process.