Human Performance Deterioration Due to Prolonged Wakefulness Can Be Accurately Detected Using Time-Varying Spectral Analysis of Electrodermal Activity

Abstract

Objective:

The aim was to determine if indices of the autonomic nervous system (ANS), derived from the electrodermal activity (EDA) and electrocardiogram (ECG), could be used to detect deterioration in human cognitive performance on healthy participants during 24-hour sleep deprivation.

Background:

The ANS is highly sensitive to sleep deprivation.

Methods:

Twenty-five participants performed a desktop-computer-based version of the psychomotor vigilance task (PVT) every 2 hours. Simultaneously with reaction time (RT) and false starts from PVT, we measured EDA and ECG. We derived heart rate variability (HRV) measures from ECG recordings to assess dynamics of the ANS. Based on RT values, average reaction time (avRT), minor lapses (RT > 500 ms), and major lapses (RT > 1 s) were computed as indices of performance, along with the total number of false starts.

Results:

Performance measurement results were consistent with the literature. The skin conductance level, the power spectral index, and the high-frequency components of HRV were not significantly correlated to the indices of performance. The nonspecific skin conductance responses, the time-varying index of EDA (TVSymp), and normalized low-frequency components of HRV were significantly correlated to indices of performance (p < 0.05). TVSymp exhibited the highest correlation to avRT (–0.92), major lapses (–0.85), and minor lapses (–0.83).

Conclusion:

We conclude that indices that account for high-frequency dynamics in the EDA, specifically the time-varying approach, constitute a valuable tool for understanding the changes in the autonomic nervous system.

Application:

This can be used to detect the adverse effects of prolonged wakefulness on human performance.