This multimodal MRI study reports disease-specific glymphatic dysfunction and free-water patterns—cortical/midline in PD and cerebellar in MSA—with glymphatic–FW coupling mediating symptoms in PD but not MSA, and an integrated biomarker panel yielding AUC=0.994 for differential diagnosis.

Delivers a highly translatable imaging biomarker set for accurate PD vs MSA diagnosis and implicates glymphatic dysfunction as a potentially targetable mechanism in PD that could inform therapeutic development or repurposing efforts.

BACKGROUND: Parkinson's disease (PD) and multiple system atrophy (MSA) show overlapping clinical features, posing diagnostic challenges. This study investigates whether distinct patterns of glymphatic dysfunction and free water (FW) accumulation can differentiate their underlying mechanisms and serve as discriminatory biomarkers. OBJECTIVE: The objective of this study was to evaluate glymphatic function and FW pathology in PD and MSA, and to develop an integrated biomarker panel for differential diagnosis. METHODS: We conducted a cross-sectional and longitudinal neuroimaging study involving 231 participants: 74 healthy control subjects (HCs), 79 patients with PD, and 78 patients with MSA. Glymphatic function (diffusion tensor image analysis along the perivascular space [DTI-ALPS] index and choroid plexus volume [CPV]) and FW distribution were derived from magnetic resonance imaging. Diagnostic performance was evaluated using receiver operating characteristic curves. Mediation analyses explored relationships among glymphatic impairment, FW accumulation, and clinical symptoms. RESULTS: Both PD and MSA showed reduced DTI-ALPS and enlarged CPV versus HC. FW accumulation exhibited disease-specific patterns: cortical/midline in PD and cerebellar in MSA. Longitudinal analysis confirmed progressive FW accumulation in these regions. Spatial coupling between glymphatic dysfunction and FW was strong in PD but absent in MSA. FW mediated the relationship between glymphatic impairment and motor/autonomic symptoms in PD, but not MSA. The integrated model combining neuroimaging and clinical metrics showed excellent discriminatory power for PD and MSA (area under the curve = 0.994). CONCLUSIONS: PD and MSA exhibit distinct glymphatic-FW pathological profiles. The coupled mechanism in PD contrasts with the uncoupled pathology in MSA, reflecting divergent pathogenesis. Multimodal imaging biomarkers demonstrate high diagnostic accuracy, showing strong potential for differential diagnosis in clinical practice. © 2026 International Parkinson and Movement Disorder Society.