Insights into psychosis risk: Unveiling impaired reinforcement learning through a behavioral and functional near-infrared spectroscopy-based optical neuroimaging study

This is the authors' abstract. We don't add key points for this paper.

BACKGROUND: Reward processing is crucial for learning, motivation and decision-making, and can be disturbed in the development of psychosis. Previous research has linked cognitive impairment and abnormalities in brain function to the clinical high risk state of psychosis (CHR-P). However, the extent to which processes of cognitive flexibility in response to reward feedback are impaired in CHR-P compared to healthy control (HC) individuals is largely unknown. METHODS: To address this knowledge gap, we conducted a probabilistic reward task in 59 people with CHR-P (age: 18.89 ± 0.66 years, 54% female) from the Bern Early Recognition and Intervention Center, and 24 HC (age: 19.37 ± 1.20 years, 50% female). In addition, we recorded the prefrontal neurovascular response of the subjects using functional near-infrared spectroscopy (fNIRS) neuroimaging during task performance. RESULTS: Behavioral results of the probabilistic reward task showed that CHR-P subjects had a significantly lower propensity for reward feedback compared to HC subjects, especially in the later course of the task (p = 0.018, Cohen's d = 0.58). When comparing the fNIRS measurements, we found a significantly lower task-induced increase in total hemoglobin concentration ([tHb]) in CHR-P subjects compared to HC subjects (p = 0.049). CONCLUSIONS: Our findings indicate impaired reward feedback processing in CHR-P subjects, suggesting reduced sensitivity to reward. This is reflected in decreased task-induced response in cerebrovascular [tHb], indicating lower task-relevant prefrontal activities compared to HC subjects. These findings suggest that impaired processing of reward feedback may contribute to reward learning deficits and inflexible cognition in CHR-P subjects.