Reinforcement Learning and Decision Making in Anorexia Nervosa.

Abstract

Purpose of review: We review recent literature on instrumental reinforcement learning involving decision-making in anorexia nervosa (AN) to understand mechanisms underlying symptoms of AN, such as rigid pursuit of weight loss despite negative consequences.

Recent findings: Relatively consistent findings indicate worse reward- and punishment-based feedback learning in the ill and weight-recovered states that is not observed in remitted samples. Initial studies suggest decreased goal-directed learning in AN, although this needs replication. Similarly, research is needed to clarify mixed findings related to learning under changing rules and the role of fear versus avoidance learning in AN. Growing evidence supports altered reinforcement learning in AN. Most studies examined the impact of outcome valence, changing rules, and habitual vs goal-directed control on learning. Computational modeling approaches can provide nuanced characterization of cognitive processes related to reinforcement learning and contribute to precision medicine efforts that may improve outcomes.