J Neuroendocrinol. 2026 Jul;38(7):e70235. doi: 10.1111/jne.70235.
ABSTRACT
Hyperthyroidism is associated with cognitive impairment and neuropsychiatric symptoms, yet its effects on white matter architecture and neuromolecular organization remain poorly understood. We examined how hyperthyroidism alters white matter wiring architecture and the neuromodulatory systems underlying these disturbances. Diffusion MRI-derived structural connectomes were analyzed in 30 patients with hyperthyroidism (HT) and 28 matched healthy controls (HC). Structural connections were classified into short-, medium-, and long-range pathways, and global and nodal network topology was quantified using graph theoretical analysis. Molecular associations were examined using nine PET-derived neurotransmitter receptor density maps through spatial correlation, multiple regression, and dominance analysis. Associations with thyroid hormone levels, clinical measures, and cognitive scores were evaluated using permutation-based statistical testing. Neurotransmitter-weighted network measures were computed by linking normative neurotransmitter density with graph theoretical metrics. Spatial autocorrelation was controlled using spin permutation testing, with results additionally corrected for multiple comparisons using false discovery rate (FDR) correction. Hyperthyroid patients exhibited significantly increased medium-range (p = .002) and long-range connectivity (p = .02), reduced modularity (p = .009), and increased characteristic path length (p = .032) as compared to controls, indicating disrupted segregation and efficiency. Nodal alterations were predominantly right-lateralized, involving cortical, subcortical, and cerebellar regions. Clinically, BMI predicted modularity (p = .01), while free thyroxine levels predicted characteristic path length (p = .04) and long-range connectivity strength (p = .03). The Nahor-Benson score was associated with increased medium-range connectivity (p = .02) and nodal degree (p = .03) in right cerebellar lobule IV-V. Neurotransmitter receptor distributions explained substantial variance in network topology (adjusted R2 = 0.45 in controls; 0.38 in hyperthyroidism), with dominant contributions from 5-HT1A (33%) and dopamine transporter (17%). Furthermore, 5-HTT-weighted within module degree z-score predicted thyroid-stimulating hormone (TSH) levels in hyperthyroidism (p = .03), linking hormonal dysregulation to molecular connectomic reorganization. Our findings suggest that excess thyroid hormone drives right-lateralized connectomic reorganization and increased metabolic demand, mediated by serotonergic and dopaminergic receptor architecture, linking endocrine dysfunction to large-scale brain network vulnerability. These insights inform mechanistic models and biomarkers for clinical translation.
PMID:42470173 | DOI:10.1111/jne.70235
