Predictions under sleep restriction: The effects of partial sleep deprivation on learning and inhibitory control

Sleep restriction, prevalent in modern society, impairs cognitive functions such as learning and inhibitory control. However, literature reveals inconsistencies in these effects, suggesting underlying mechanisms are not fully understood. We hypothesized that these inconsistencies might stem from variations in how individuals perceive task demands, particularly expected or favored responses. To test this idea, we investigated the impact of sleep restriction (4h) on learning and inhibitory control in a predictable yet uncertain environment, focusing on the role of built-up expectations. Using a Go/NoGo task with varying probabilities of Go-stimuli (20%-50%-80%), we compared the performance of Sleep-Restricted (SR; N=32) and Well-Rested (WR; N=34) participants across two sessions. Participants were unaware of the changing probabilities. Both groups showed similar overall performance trends, with higher rates of commission errors and faster reaction times as the probability of Go increased. However, the WR group demonstrated a greater ability to adapt to changing probabilities. Remarkably, we found no group differences in inhibitory control. To further investigate these findings, we employed a computational model incorporating adaptive learning dynamics. Our model revealed that SR participants exhibited a diminished perception of environmental volatility, suggesting they may face challenges in detecting relevant changes in the task environment and consequently updating their beliefs. In conclusion, our study provides evidence that sleep restriction alters task perception and impairs belief updating. These findings, alongside previous research, highlight the crucial role of sleep in maintaining optimal cognitive function, particularly in dynamic environments.