Probing midbrain dopamine function in pediatric obsessive-compulsive disorder via neuromelanin-sensitive magnetic resonance imaging

Abstract

Obsessive-compulsive disorder (OCD) is an impairing psychiatric condition, which often onsets in childhood. Growing research highlights dopaminergic alterations in adult OCD, yet pediatric studies are limited by methodological constraints. This is the first study to utilize neuromelanin-sensitive MRI as a proxy for dopaminergic function among children with OCD. N = 135 youth (6-14-year-olds) completed high-resolution neuromelanin-sensitive MRI across two sites; n = 64 had an OCD diagnosis. N = 47 children with OCD completed a second scan after cognitive-behavioral therapy. Voxel-wise analyses identified that neuromelanin-MRI signal was higher among children with OCD compared to those without (483 voxels, permutation-corrected p = 0.018). Effects were significant within both the substania nigra pars compacta (p = 0.004, Cohen's d = 0.51) and ventral tegmental area (p = 0.006, d = 0.50). Follow-up analyses indicated that more severe lifetime symptoms (t = -2.72, p = 0.009) and longer illness duration (t = -2.22, p = 0.03) related to lower neuromelanin-MRI signal. Despite significant symptom reduction with therapy (p < 0.001, d = 1.44), neither baseline nor change in neuromelanin-MRI signal associated with symptom improvement. Current results provide the first demonstration of the utility of neuromelanin-MRI in pediatric psychiatry, specifically highlighting in vivo evidence for midbrain dopamine alterations in treatment-seeking youth with OCD. Neuromelanin-MRI likely indexes accumulating alterations over time, herein, implicating dopamine hyperactivity in OCD. Given evidence of increased neuromelanin signal in pediatric OCD but negative association with symptom severity, additional work is needed to parse potential longitudinal or compensatory mechanisms. Future studies should explore the utility of neuromelanin-MRI biomarkers to identify early risk prior to onset, parse OCD subtypes or symptom heterogeneity, and explore prediction of pharmacotherapy response.

Fig. 1. Greater neuromelanin-MRI CNR among children with OCD.

Voxel-wise group difference results are summarized here (centered at MNI 7, −19, −16). The t-statistic for the association with an OCD diagnosis (present > absent) is displayed voxel-wise within the SN/VTA analysis mask. Only the positive effect was significant at the analysis-level. Voxels significant at p < 0.05 are in opaque colors; non-significant voxels are more transparent. The left side images are a zoomed inset; average contrast-to-noise (0–30%) within the SN/VTA analysis mask is shown on the right-side axial view.

Fig. 2. Group differences and symptom assocaitions with neuromelanin-MRI CNR.

Average neuromelanin-MRI CNR was estimated from the leave-one-out (LOO) analyses and used for post-hoc testing. Group differences based on OCD diagnosis are displayed in panel A. Panel B displays associations between neuromelanin-MRI CNR and worst-ever symptom severity within the OCD group based on the Children’s Yale-Brown Obsessive-Compulsive Scale (CY-BOCS). Panel C displays associations with illness duration (years since age when symptoms became distressing) among the OCD group. Associations with Obsessive-Compulsive Inventory-Children’s Version (OCI-CV) scores are presented in panel D with linear fits within group and a quadratic fit across the sample. Youth with OCD are displayed with blue triangle markers and those without are displayed by red circles.