Diminished differentiation of rewards in individuals at clinical high-risk for psychosis

Abstract

Reward processing is impaired in people with schizophrenia, which may begin in the clinical high-risk (CHR) for psychosis period. The Monetary Incentive Delay (MID) task has been important in understanding the neural correlates of reward processing deficits in various psychiatric disorders. Previous research has found that CHR individuals have an imprecise mental representation of rewards, which leads to a diminished differentiation between rewards, though this has not been observed behaviorally. A total of 19 CHR individuals and 20 controls were given a novel variant of the MID task, designed to examine how modulating reward context may impact responses to reward cues, a process often referred to as "adaptive coding." Both groups appeared to update their behavior in response to the rewards available in this adaptive task. However, when compared to controls who showed a more graded decrease in response time to increasing reward contexts, CHR individuals appeared to have a sharp decrease in response time in the low reward context that is nearly stable across higher reward contexts. This is largely driven by the exponential component of the response time distribution, which is often interpreted to be more cognitively or effortfully influenced. Response times are related to negative symptoms, but not positive symptoms, disorganized symptoms, or estimated intelligence. Although an adaptive coding effect was not observed, these results provide novel insight into the reward processing mechanisms and volitional processes in the CHR population, as this was the first study to observe the diminished differentiation of rewards behaviorally.

Keywords: Anhedonia; Avolition; Clinical High-Risk for Psychosis; Reward Processing.

Fig. 1

The Adaptive Monetary Incentive Delay Task trial structure. Participants are presented with one of four Reward Context cues, which presents possible rewards for the trial. Immediately after, participants are explicitly shown the exact possible Reward Value they will be given if responding to the presented stimulus in time. At the end of the trial, participants are told if they won money on that trial or not

Fig. 2

a The interaction between reward context and group on RT. HCs display a graded decrease in RT as reward context increased, but CHR individuals show a sharp drop in RT in the presence of a reward. b The interaction between reward context and group on τ. HCs have a similar exponential component of their RT for all contexts, while CHRs have a larger τ for the no-reward context. Error bars are 99% confidence intervals