Pupil dilation during negative prediction errors is related to brain choline concentration and depressive symptoms in adolescents

Depressive symptoms are associated with altered pupillary responses during learning and reward prediction as well as with changes in neurometabolite levels, including brain concentrations of choline, glutamate and gamma-aminobutyric acid (GABA). However, the full link between depressive symptoms, reward-learning-related pupillary responses and neurometabolites is yet to be established as these constructs have not been assessed in the same individuals. The present pilot study, investigated these relations in a sample of 24 adolescents aged 13 years. Participants completed the Revised Child Anxiety and Depression Scale (RCADS) and underwent a reward learning task while measuring pupil dilation and a single voxel dorsal anterior cingulate cortex (dACC) MEGA-PRESS magnetic resonance spectroscopy scan assessing choline, glutamate and GABA concentrations. Pupil dilation was related to prediction errors (PE) during learning, which was captured by a prediction error-weighted pupil dilation response index (PE-PDR) for each individual. Higher PE-PDR scores, indicating larger pupil dilations to negative prediction errors, were related to lower depressive symptoms and lower dACC choline concentrations. Dorsal ACC choline was positively associated with depressive symptoms, whereas glutamate and GABA were not related to PE-PDR or depressive symptoms. The findings support notions of cholinergic involvement in depressive symptoms and cholinergic influence on reward-related pupillary response, suggesting that pupillary responses to negative prediction errors may hold promise as a biomarker of depressive states.