Computational Mechanisms of Effort-Cost Decision-Making in Schizophrenia

Project Summary: Many people with schizophrenia experience reductions in motivation, which impair occupational functioning, reduce quality of life, and increase public health demands. Treatments for motivational impairments in SZ are largely ineffective, however, in part due to poor understanding of etiology. Recent work has provided strong evidence that abnormal effort-cost decision-making – calculations performed to weigh the “cost vs. benefits” of actions – may be a key contributor to motivational deficits in schizophrenia. Specifically, research has shown that people with schizophrenia are less willing than controls to expend effort to obtain rewards on experimental tasks, and that this deficit is related to motivational impairment. However, due to the use of imprecise experimental paradigms and analytic methods that are ill-suited to disentangle the contribution of component processes to effort-cost decision-making, it is unknown whether this reduction is driven by reduced sensitivity to the rewards or heightened sensitivity to the effort associated with actions. This knowledge has treatment implications, as interventions for targeting reward and effort sensitivity are different. We will use a combination of experimental tasks and associated computational modeling approaches, to quantify the relative contributions of effort and reward sensitivity to effort-cost decision-making in people with schizophrenia and healthy controls. We will also collect mobile-based assessments providing comprehensive phenotyping of motivational impairment experienced in daily life. We aim (a) to determine whether effort-cost decision-making deficits in schizophrenia reflects increased effort or reduced reward sensitivity, (b) to identify the neural substrates of effort-cost decision-making impairment, and (c) to establish whether effort measures correspond to measures of effort and reward in daily life. The fact that few researchers have been trained in both clinical phenomenology and computational modeling techniques limits the translational impact these approaches may have in understanding mental illness. With this in mind, the training plan is specifically-designed to provide hands-on instruction in 1) applying computational models to effort-cost decision-making to choice behavior, 2) integrating computational modeling with functional neuroimaging, and 3) relating computational modeling parameters to mobile-based assessments of daily motivational experience. Taken together, completion of the proposal will facilitate the applicant’s long-term goal of becoming an independent investigator examining the computational mechanisms of motivational impairment in various psychiatric conditions. Further, the data and skills acquired will position the applicant to competitively submit a transdiagnostic R01 proposal, designed to examine whether effort-cost decision-making impairments in psychiatric conditions characterized by avolition (e.g., major depressive disorder, schizophrenia) arise from similar or different computational mechanisms.