Mechanisms and correlates of incentivized response inhibition in schizophrenia and bipolar disorder

When healthy individuals are incentivized on response inhibition tasks (e.g., Stroop), they recruit additional cognitive resources, enabling them to make faster, more accurate responses. Schizophrenia (SZ) and bipolar disorder (BP) are associated with poor response inhibition, but it is unknown whether SZ and BP show incentiverelated improvements to the same degree as healthy controls (HC). To investigate this question, reaction time data from an incentivized Stroop-style task were analyzed from 37 SZ, 26 B P, and 33 H C. We examined: 1) group differences in mean reaction time, 2) group differences in response caution and in rate of processing taskrelevant information derived from a computational approach (drift diffusion modeling), and 3) clinical and cognitive correlates of drift diffusion parameters in SZ and BP groups. When incentives were introduced, both HC and BP showed significantly faster response speed, but SZ did not show the same pattern of improvement as a function of incentives. Computational analyses indicated that groups did not significantly differ in response caution, but that both SZ and BP had a slower information processing rate compared to HC. In SZ, slow information processing rate was related to poor cognition; positive and negative symptoms were associated with impairments in information processing rate, but in opposite directions (i.e., increased information processing rate was associated with positive symptom severity; decreased information processing rate was associated with negative symptom severity). Our findings suggest impaired information processing rate may contribute to poor response inhibition in both SZ and BP, whereas response caution is intact in both disorders. However, SZ is distinguished from BP by a failure to enter an overall motivated state and decrease response speed when incentivized.