Spectral power and functional connectivity alterations in multiple system atrophy revealed by resting-state EEG

Multiple system atrophy (MSA) is an aggressive α-synucleinopathy characterized by motor and autonomic dysfunction. Elucidating neurophysiological patterns can provide crucial insights into the underlying neural mechanisms. Resting-state electroencephalography (EEG) and fMRI data were collected from 69 healthy controls and 123 MSA patients, comprising a discovery cohort ( n = 97) and an independent validation cohort ( n = 26). MSA patients showed reduced delta power in the cerebellar, frontoparietal, visual I, and limbic networks, with widespread increases in theta and high gamma power. Amplitude-based connectivity decreased across the alpha, beta, and gamma bands in subcortical, cerebellar, default mode, motor, and visual I networks, with increased theta-band synchrony involving motor, visual II, and frontoparietal networks. These EEG abnormalities were corroborated by fMRI connectivity, highlighting cross-modal consistency. Canonical correlation analysis revealed associations between EEG features and motor/non-motor symptom severity. Interpretable stacking models accurately classified MSA-C and MSA-P subtypes and predicted UMSARS progression in longitudinal follow-up. These findings identified aberrant neural activity and network dysfunction in MSA, highlighting EEG as a promising biomarker for disease monitoring.