Medial Frontal Theta Reduction Impairs Rule Switching via Prediction Error

Abstract

Cognitive flexibility, the ability to switch behavior in response to changing rules in an uncertain environment, is crucial for adaptive decision making. Prior research has hypothesized a key role of prediction error and theta oscillations in medial frontal cortex in this process. However, the causal link between such processes remains to be established. To address this, we combined neural stimulation, EEG, behavioral measurement, and computational modeling. Specifically, we applied high-definition transcranial direct current stimulation (HD-tDCS) to modulate theta oscillations as measured via EEG followed by a probabilistic reversal learning task in 48 adults (18 female and 30 male human subjects). We found that anodal stimulation reduced theta power and rule prediction error, and it increased the number of trials needed to reliably switch between rules. These findings support the role of rule prediction error signaling as a key mechanism linking neural oscillations to behavioral adaptation and highlight the importance of theta power and rule prediction error for cognitive flexibility.

Significance statement Cognitive flexibility—the ability to adjust behavior when rules change—is critical for adaptive behavior in uncertain environments. Although prediction error signaling and theta oscillations in medial frontal cortex have been proposed as key mechanisms, their causal relationship remains unclear. Here, we combine high-definition transcranial direct current stimulation (HD-tDCS), EEG, behavioral assessment, and computational modeling to test the causal contribution of those factors on cognitive flexibility. We show that anodal stimulation reduces frontal theta power and rule-level prediction errors, leading to more trials to commit to the new rule. These findings provide causal evidence that supports behavioral flexibility, advancing our understanding of the neural computations underlying adaptive decision making.