Reinforcement Learning and Decision Making in Anorexia Nervosa

Area of Science:

• Cognitive Neuroscience
• Clinical Psychology
• Computational Psychiatry

Background:

• Anorexia nervosa (AN) is characterized by rigid pursuit of weight loss despite negative consequences.
• Understanding the cognitive mechanisms underlying AN symptoms is crucial for developing effective treatments.

Purpose of the Study:

• To review recent literature on instrumental reinforcement learning in AN.
• To elucidate the mechanisms underlying AN symptoms, particularly decision-making processes.

Main Methods:

• Literature review of studies on instrumental reinforcement learning and decision-making in AN.
• Analysis of findings related to reward/punishment feedback, goal-directed learning, and rule-learning.

Main Results:

• Consistent evidence of impaired reward and punishment feedback learning in ill and weight-recovered AN states.
• Initial findings suggest reduced goal-directed learning, requiring replication.
• Mixed results regarding learning under changing rules and the role of fear/avoidance learning warrant further investigation.

Conclusions:

• Growing evidence supports altered reinforcement learning in AN.
• Computational modeling offers a path to nuanced characterization of cognitive processes.
• These insights can contribute to precision medicine for improved AN outcomes.