Reduced Reverse Replay in Anxious Individuals Impairs Reward Prediction

Abstract Anxiety traits can manifest as a tendency to overlook positive information, impacting memory and decision-making. Using EEG-based neural decoding, we examined future reward prediction in participants with varying levels of trait anxiety. In a resting period, following value learning, we found an increase in reverse replay for a task sequence that led to reward. This reward-related replay was significantly reduced within individuals exhibiting higher trait anxiety. Furthermore, this reduction correlated with a distorted neural representation of the reward-predictive sequence, an effect mirrored behaviorally in an attenuated preference for stimuli within the same sequence. These findings indicate that higher trait anxiety may impede formation of reward-predictive representations by reducing offline replay during rest. More broadly, our demonstration of sequential replay using EEG enables examination of offline learning and memory consolidation in wider populations, particularly those with psychopathology.