Navigating Dynamic Risks: Individual Risk Propensity Shapes Outcome Processing in the Columbia Card Task

Understanding why some individuals are more prone to problematic risk-taking requires examining how they process feedback during escalating risk scenarios. Traditional laboratory paradigms fail to capture two criticalaspects of real-world risky behavior: the dynamic nature of risk escalation and the orthogonal manipulation of risk parameters. We addressed these limitations using an ERP-adapted Columbia Card Task (CCT), enabling precise temporal analysis of neural responses as risks accumulate within decision episodes. Twenty-five participants completed the CCT while we recorded EEG, measuring early outcome detection (FRN/RewP, 240-340ms) and later evaluative processing (P3b, 300-500ms). We systematically manipulated gain amounts, loss amounts, and loss probability while examining how individual risk propensity—measured through actual task behavior—predicted neural feedback processing patterns. Critically, rather risk-averse and rather riskprone individuals employed fundamentally different temporal adaptation strategies that emerged dynamically as risks escalated. During early outcome detection, risk-prone participants showed progressive sensitization to negative outcomes across feedback sequences, while risk-averse participants showed habituation. These opposing patterns reversed for positive outcomes. During later evaluation, risk-prone individuals selectively maintained working memory updating for positive outcomes despite accumulating losses, while risk-averse participants enhanced negative outcome processing. These context-dependent effects were most pronounced under specific high-stakes conditions: smaller gains and larger losses amplified group differences in early processing, while probability information revealed that risk-prone individuals showed enhanced neural responses to improbable positive outcomes—providing a mechanism for persistent risk engagement despite unfavorableodds. Behavioral results revealed that risk-prone individuals achieved higher efficiency without performance costs, suggesting these neural patterns reflect adaptive individual differences rather than pathological dysfunction. Our findings fundamentally reframe risk propensity as a dynamic regulatory process triggered by escalating risk, with direct implications for identifying vulnerable individuals before maladaptive patterns crystallize.