Purpose: The function of working memory (WM) is impaired by total sleep deprivation (TSD) and cognitive load. However, it is unclear whether the load modulates the effect of TSD on WM. We conducted a pilot study to investigate the effects of 36 h of TSD on WM under different load levels. Materials and methods: Twenty-two male students aged 18–25 years were enrolled, who underwent two types of sleep conditions (baseline and 36 h TSD), where they performed two N-back WM tasks (one-back task and two-back task) with simultaneous electroencephalography recordings. Results: Repeated measures analysis of variance (ANOVA) indicated that, with the increasing load, the reaction time increased and the accuracy decreased. After TSD, the correct number per unit time decreased. The significant interaction effect of the P3 amplitudes between the load level and the sleep condition showed that the reduction in the amplitude of P3 in the two-back task due to TSD was more obvious than that in the one-back task. Conclusions: Our results provided evidence for the moderation of load on the impairment of TSD on WM. The degree of TSD-induced impairment for a higher load was greater than that for a lower load. The current study provides new insights into the mechanisms by which sleep deprivation affects cognitive function.

Sleep deprivation impairs cognitive function and is accompanied by a simultaneous compensatory effect, one of the brain's capacities to maintain function in emergency situations. However, the time course of the compensatory effect is unclear. In this study, 22 male participants completed a pronunciation working memory task that included congruent and incongruent stimuli trials with EEG recordings before and after total sleep deprivation (TSD). Behavioral performance analysis showed that after TSD, the participants’ reaction time (RT) was shortened, but accuracy was reduced significantly. Analysis of event-related potential (ERP) results showed that the amplitude of N2 (an early visual ERP) was larger (i.e., more negative) after TSD than at baseline. A significant interaction between congruency and sleep condition was seen. Compared to that before TSD, the increase in amplitude of P3 (a stimulus-induced positive deflection component) under an incongruent stimulus was larger than that under a congruent stimulus after TSD. Moreover, a significant negative correlation was found between P3 amplitude and RT. Our results suggest that TSD impairs cognitive function. Meanwhile, the brain activates a compensatory mechanism after TSD, which is comprehensive during the conflict-detection and information-updating stages. This study provides a fresh viewpoint for understanding how TSD affects cognitive function.