Discussion
In this large study of 9-10 year olds, the neural correlates reward anticipation were similar to those previously observed in older adolescents and adults, extending the preliminary results of reward feedback in this cohort \cite{Casey2018}. There was, however, little evidence that individual differences in brain reward anticipation signals were associated with severity of psychotic-like experiences or number of depressive symptoms. There were no linear associations between brain reward anticipation and severity of psychotic-like experiences or number of depressive symptoms that survived correction for multiple comparison. There were a handful of associations that had p-values below 0.05 without multiple comparison correction, but none survived correction for multiple comparison. For example there were (uncorrected) negative associations between activations in left accumbens and right orbitofrontal cortex with psychotic like experiences; these associations was only present in the left, not right, accumbens, and the right not left orbitoftonal cortex, and are likely due to chance. There was a negative association between right orbitofrontal cortex activation and presence of anhedonia, and positive association between temporal lobe activations and presence of anhedonia, and between temporal lobe (temporal pole, fusiform gyrus, left middle temporal gyrus) activations number of depressive symptoms. None of these associations was significant after correction for multiple comparisons. Our results indicate that variations in brain reward anticipation signals at age 9-10 are not associated with the severity of psychotic-like experiences, or number of depressive symptoms.
There was a solitary finding that survived corrected for multiple comparison for the number of brain regions studied: namely children with presence of psychotic-like experiences had less activation in the right parsorbitalis (orbital aspect of inferior frontal gyrus) compared to those without such experiences. This isolated finding only in one hemisphere must be considered preliminary, as although we corrected for the number of brain regions studied, we repeated the analyses with several different psychiatric phenotypes and did not correct for these multiple phenotypic analyses. Furthermore, although a plausible post-hoc rationale could be constructed for the importance of the right parasobitalis in the genesis of psychotic like experiences, it does not correspond to the bulk of the prior findings in the literature relating reward anticipation to psychotic illness or psychotic psychopathology in older groups.