State-independent functional connectomic architecture captures transdiagnostic network alterations and clinical variation

Background: Psychiatric disorders show substantial overlap across diagnostic categories while differing markedly in clinical expression. Whether these shared abnormalities and individual differences are embedded in stable functional brain architecture rather than in state-specific connectivity remains unclear. Methods: We integrated functional connectomes across one resting-state and six task-based fMRI conditions in 377 participants from the YaleNeuroConnect cohort, including 223 healthy controls and 154 individuals with psychiatric disorders. Principal component analysis was applied to each participant's functional connectivity data across the seven states to derive a state-independent connectivity representation. Network-based statistics then identified components showing significant case-control differences. Cross-state consistency was assessed, and findings were validated in an independent UCLA cohort of 177 participants. Associations between component expression and clinical measures were further examined within the Yale psychiatric group. Results: Cross-paradigm connectivity revealed two reproducible patterns of state-independent network alteration. One pattern showed reduced connectivity among limbic, somatomotor and salience systems, whereas the other showed stronger cerebellar coupling with limbic and somatomotor systems. Both patterns showed consistent case-control differences across functional states and generalized to an independent UCLA cohort. Their clinical relevance, however, differed. Greater expression of the cerebello-cortical connectivity pattern was associated with lower positive affect, greater psychopathology and more everyday executive difficulties, whereas the reduced limbic-somatomotor-salience connectivity pattern showed no significant clinical associations. Conclusions: These findings delineate a stable reorganization of large-scale functional architecture across psychiatric disorders and suggest that broadly expressed psychiatric abnormalities and individual clinical variation may be differentially represented within the altered connectome.