Common Neural Deficits in Social and Monetary Reinforcement Learning in Schizophrenia.

Background And Hypothesis: Individuals with schizophrenia (SZ) have well-documented behavioral and neural deficits to reinforcement learning (RL) from monetary feedback. Although they have a range of social functioning deficits, limited research has examined neural processes related to learning from social feedback. The present study examined how neural activation to social RL in SZ compares to activation to monetary RL.

Study Design: Thirty participants with SZ and 31 healthy controls completed a Probabilistic RL paradigm that included both social and monetary RL tasks in the scanner, each with positively and negatively valenced trials. Analyses included a region of interest-based approach to examine activation in areas associated with reward, learning, and social processes and assessed neural activation when making task choices (Choice) and upon feedback receipt (Outcome) for social versus monetary RL.

Study Results: Results indicated that across tasks, SZ had reduced signal in regions such as caudate and orbitofrontal cortex during Choice but not Outcome. In addition, patterns of neural activation were similar during social and monetary RL, for both valences, and brain activation during RL was largely unrelated to behavioral RL performance in either group.

Conclusions: Overall, the findings indicate that although SZ appear to have intact brain activation when receiving feedback about their choices, they may struggle to recruit the necessary circuitry in regions associated with generating expected values to successfully modulate their choices based on the received feedback. This aligns with literature on monetary RL in SZ and suggests that activation patterns may be similar during social RL.