Beyond Exploration: Heightened Recent Reward Reliance and Midfrontal Sensitivity to Outcome Differences in ADHD

Adaptive decision-making requires balancing the exploitation of known rewards with the exploration of alternative options, yet how attention-deficit/hyperactivity disorder (ADHD) shapes this trade-off remains unclear. Using a restless four-armed bandit task and electroencephalography (EEG), we investigated the neural correlates of exploratory decisions and how recent rewards shape choice behavior and feedback processing in 23 adults with self-reported clinical ADHD diagnoses and 27 controls. Across 450 trials with dynamically changing payoffs, choices were classified against the highest last-observed payoff, and behavioral modeling quantified the influence of reward history on subsequent decisions. Contrary to our predictions, exploration rates, task performance, and neural differentiation between exploration and exploitation did not differ significantly between groups. However, adults with ADHD were more likely to repeat an option after high reward feedback and switch after low rewards, driven primarily by the immediately preceding outcome. At neural level, ADHD participants also showed greater midfrontal sensitivity to unsigned changes between consecutive payoffs, expressed in a feedback-evoked response spanning 280 to 568 ms and carried by slow activity below 2 Hz. Together, these findings dissociate the raw frequency of exploration from the dynamic weighting of recent outcomes in ADHD. Rather than reflecting an indiscriminate drive for novelty, behavior was characterized by heightened recent reward reliance paired with amplified slow-wave midfrontal sensitivity to outcome shifts. Crucially, this profile emerged without compromising overall performance, suggesting that, in this volatile environment, ADHD was associated with an increased weighting of recent outcomes rather than with a performance deficit.