Modeling Decision-Making in Schizophrenia: Associations Between Computationally Derived Risk Propensity and Self-Reported Risk Perception
Emma N Herms,Joshua W Brown,Krista M Wisner,William P Hetrick,David H Zald,John R Purcell
DOI: https://doi.org/10.1093/schbul/sbae144
2024-09-06
Abstract:Background and hypothesis: Schizophrenia is associated with a decreased pursuit of risky rewards during uncertain-risk decision-making. However, putative mechanisms subserving this disadvantageous risky reward pursuit, such as contributions of cognition and relevant traits, remain poorly understood. Study design: Participants (30 schizophrenia/schizoaffective disorder [SZ]; 30 comparison participants [CP]) completed the Balloon Analogue Risk Task (BART). Computational modeling captured subprocesses of uncertain-risk decision-making: Risk Propensity, Prior Belief of Success, Learning Rate, and Behavioral Consistency. IQ, self-reported risk-specific processes (ie, Perceived Risks and Expected Benefit of Risks), and non-risk-specific traits (ie, defeatist beliefs; hedonic tone) were examined for relationships with Risk Propensity to determine what contributed to differences in risky reward pursuit. Study results: On the BART, the SZ group exhibited lower Risk Propensity, higher Prior Beliefs of Success, and comparable Learning Rates. Furthermore, Risk Propensity was positively associated with IQ across groups. Linear models predicting Risk Propensity revealed 2 interactions: 1 between group and Perceived Risk, and 1 between IQ and Perceived Risk. Specifically, in both the SZ group and individuals with below median IQ, lower Perceived Risks was related to lower Risk Propensity. Thus, lower perception of financial risks was associated with a less advantageous pursuit of uncertain-risk rewards. Conclusions: Findings suggest consistently decreased risk-taking on the BART in SZ may reflect risk imperception, the failure to accurately perceive and leverage relevant information to guide the advantageous pursuit of risky rewards. Additionally, our results highlight the importance of cognition in uncertain-risk decision-making.