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

Sleep restriction, prevalent in modern society, impairs inhibitory control, but literature reveals inconsistencies in these effects. We hypothesized that these inconsistencies might stem from variations in how individuals perceive task demands, particularly expected or favoured responses. To test this, we investigated the impact of sleep restriction (4 h) on learning and inhibitory control in a predictable yet uncertain environment, focusing on built-up expectations. Using a two-session go/no-go task with varying probabilities of go stimuli (20–50–80%), we compared sleep-restricted (SR; N = 32) and well-rested (N = 34) participants. Both groups showed similar performance trends, with higher commission errors and faster reactions as go probability increased. Contrary to expectations, we found no group differences in inhibitory control. A computational model incorporating adaptive learning dynamics revealed that SR participants exhibited diminished perception of environmental volatility, suggesting challenges in detecting task environment changes and updating beliefs. These findings highlight sleep’s crucial role in maintaining cognitive function, particularly in dynamic environments, and offer new insights into the cognitive consequences of sleep loss.