Different learning aberrations relate to delusion-like beliefs with different contents

Area of Science:

• Cognitive Neuroscience
• Psychiatry
• Computational Psychiatry

Background:

• The prediction error model of delusions is successful but struggles to explain diverse delusion content.
• Persecutory delusions and paranoia involve unfounded beliefs about others' harmful intentions.
• Other delusions include external control over thoughts/actions or specific personal meaning of world events.

Purpose of the Study:

• To compare learning in paranoid and non-paranoid delusion-like beliefs using distinct cognitive tasks.
• To investigate how learning deficits relate to specific delusion types and predict psychosis.
• To explore dissociable deficits in learning and belief updating.

Main Methods:

• Comparison of learning in probabilistic reversal learning and Kamin blocking tasks.
• Assessment of clinical high-risk status and paranoia levels.
• Fitting task-specific computational models to behavioral data.

Main Results:

• Clinical high-risk status alone did not alter probabilistic reversal learning task results.
• Paranoia correlated with excessive switching in reversal learning and inappropriate learning in blocking tasks.
• Non-paranoid delusion-like beliefs showed specific impairments in cue combination learning.

Conclusions:

• Paranoia is linked to low learning rates across tasks.
• Non-paranoid delusions relate to specific cue-updating impairments.
• Dissociable learning deficits in paranoia and other delusions may predict psychosis.