46. regional brain iron dysregulation associated with genetic risk for schizophrenia: A UK biobank quantitative susceptibility mapping study in 64,226 individuals

Background Iron is the most abundant trace metal in the brain and is essential for neurotransmitter synthesis, mitochondrial function, oxidative phosphorylation, and ATP production. Iron homeostasis is crucial, and its dysregulation may affect neurobiological pathways in psychotic disorders, including dopaminergic neurotransmission, white matter integrity, mitochondrial function, oxidative stress, and neuroinflammation. Neuroimaging studies in schizophrenia highlight alterations in iron levels in subcortical regions, with recent evidence linking iron deficits to elevated dopamine synthesis capacity. However, confounders in patient studies include medication effects and illness duration. Examining brain iron in unaffected individuals with genetic risk for schizophrenia allows identification of mechanisms preceding illness onset, in an antipsychotic-naive cohort. We conducted a whole-brain voxel-wise analysis of associations between schizophrenia polygenic risk scores (PRS) and brain iron, measured by quantitative susceptibility mapping (QSM), in the UK Biobank. Methods UK Biobank Population Archive of Neuroimaging Data Organised for Rapid Analysis (PANDORA) QSM data from 64,226 individuals were analysed. Participants were healthy European ancestry individuals. Participants with a psychotic or affective disorder, neurodegenerative or neurological condition were excluded. PRS were calculated using the PRS-CS method with summary statistics from the Psychiatric Genomics Wave 3 GWAS. PANDORA enables efficient, large-scale analysis through supervoxel-based computation, with projection to full voxel-wise resolution (182 × 218 × 182 voxels, 1 mm³ isotropic). General linear modelling tested associations between polygenic risk scores and QSM values, controlling for smoking, alcohol use, BMI, over 200 in-built imaging confounds, and 10 genetic principal components. Two-tailed detection used the F-statistic, with the direction determined by the T-statistic. Multiple comparison correction employed false discovery rate (FDR)(q < 0.05). Significant voxels were clustered (k ≥ 20, 26-connection). Results We identified 12,086 significant voxels in which schizophrenia genetic risk was associated with brain iron (FDR q < 0.05): 6399 voxels (53%) positive and 5687 (47%) negative. Cluster analysis identified 27 positive and 31 negative clusters (k ≥ 20). Positive associations were mainly in the subcortical basal ganglia, especially the bilateral pallidum, followed by the putamen and white matter. Negative associations were more widely distributed, with white matter dominance, limbic involvement (bilateral amygdala), thalamus, and cortical regions, including the subcallosal and insular cortices. Discussion Schizophrenia genetic risk is linked to bidirectional regional iron dysregulation in unaffected individuals. Divergence between positive (basal ganglia-dominant) and negative (white matter/limbic-dominant) associations suggests distinct mechanisms. Basal ganglia localisation aligns with involvement of dopaminergic pathways. In minimally myelinated subcortical regions, increased susceptibility likely reflects iron in dopamine-rich areas. White matter involvement may reflect effects on oligodendrocyte iron metabolism and effects of myelin. These patterns differ from patient deficits, suggesting temporal trajectories from genetic risk to illness. Cross-sectional design limits causality. Multimodal and longitudinal studies are needed.