BACKGROUND: Schizophrenia is associated with abnormalities in neurodevelopmental processes. Furthermore, dysfunctional neural circuits involved in reward processing may be linked to the development of symptoms in schizophrenia and are predictive of long-term functional outcome. It is however unknown whether neural signatures of reward anticipation are detectable in children with high psychotic-like experiences. METHODS: Using data from the ABCD study 4.1, we defined a healthy control (N = 50) and a high psychotic-like experience (N = 50) group with a Prodromal Psychosis Syndrome (PPS) score >
3 and distress score >
6 at baseline (9-10 years) and 2nd year follow-up (11-12 years). While undergoing functional MR-imaging, all children completed the Monetary Incentive Delay task. Using the preprocessed ABCD-data, we explored whether behaviour and brain activations for reward and loss anticipation in areas underlying reward processing differed between groups and time-points. Furthermore, we investigated whether those brain activations that showed differences between the groups were predictive of later PPS scores. Additionally, we also employed computational modelling to assess response vigour. RESULTS: While response times did not differ, the computational model revealed that response vigour for salient cues was significantly lower in the high PLEs compared to controls at baseline. We also found that children with high PLEs demonstrated lower activation during reward anticipation in the anterior insula at the baseline time-point
the nucleus accumbens, the putamen, the dorsolateral (dlPFC) and the ventral medial prefrontal cortex at the 2nd year follow-up, and in the caudate at both timepoints, compared to controls. Regression analysis revealed that deactivations in the left anterior insula and left dlPFC, was predictive of later PPS scores. CONCLUSION: This study reveals that neural alterations during reward anticipation are detectable in children with high PLEs. These dysfunctions in neural activation patterns may serve as potential predictive biomarkers for psychosis.