Quantitative susceptibility mapping of brain iron in adult ADHD

Attention-deficit/hyperactivity disorder (ADHD) is a neurodevelopmental condition that frequently persists into adulthood and is linked to alterations in fronto-striatal circuitry and dopaminergic signaling. Brain iron is essential for dopamine synthesis and neural metabolism and can be indexed in vivo using quantitative susceptibility mapping (QSM), which reflects regional magnetic susceptibility. Pediatric studies have mostly reported reduced tissue iron susceptibility in ADHD, but data in adults are limited. This study investigated regional susceptibility in adults with ADHD, its relationship to current and retrospective ADHD symptoms, and depression.

Twenty-five adults with ADHD and 24 healthy controls underwent 3T MRI, including high-resolution QSM. Mean susceptibility was extracted from 89 cortical and subcortical regions of interest. For each ROI, group effects were estimated using linear regression with heteroskedasticity-consistent (HC3) inference. To account for group differences in depressive symptoms, additional models included Beck Depression Inventory (BDI-II) scores. Dimensional associations with ADHD symptoms were tested using covariate-adjusted ROI-symptom correlations for current symptoms (CAARS) and retrospective childhood symptoms (WURS-k). Benjamini-Hochberg false discovery rate (FDR) correction was applied within each analysis across ROIs.

No ADHD-control group difference survived FDR correction in any model. At the uncorrected level, adults with ADHD showed lower susceptibility in ventral temporal and posterior midline regions, most consistently in the fusiform gyrus and posterior cingulate cortex (small-to-moderate effect sizes across models). When additionally adjusting for depressive symptoms, nominal group differences remained confined to ventral temporal/temporo-limbic and orbitofrontal regions (fusiform, entorhinal cortex, medial orbitofrontal cortex), but again did not survive FDR. Covariate-adjusted ROI-symptom associations did not meet FDR significance; nominally, higher ADHD symptom burden (particularly impulsivity) was associated with lower susceptibility in posterior midline regions (posterior cingulate, precuneus) and ventral temporal cortex (fusiform).

In this adult sample, QSM provided no robust evidence for widespread or regionally specific alterations in brain iron susceptibility in ADHD after multiple-comparison correction. Nevertheless, the reproducible pattern of nominal effects-centered on ventral temporal and default-mode network hub regions-suggests that inter-individual variation in cortical susceptibility may relate to clinical heterogeneity and neurodevelopmental timing rather than constituting a strong diagnostic signature. Larger, medication-stratified and developmentally informed studies are needed to clarify whether subtle iron-related susceptibility patterns track symptom dimensions, treatment exposure, and longitudinal trajectories.