Parkinson’s disease (PD) and its progression to dementia (PDD) are among the most devastating neurodegenerative disorders; no disease-modifying therapy currently exists. Timely intervention and efficient clinical-trial stratification hinge on early prediction, yet available diagnostics lack sensitivity in the pre-symptomatic phase, and established biomarkers fail to reliably forecast conversion to dementia. By integrating large-scale multi-omics data, this project will meet this urgent need by forecasting PD onset and dementia risk years before clinical symptoms emerge. Using the unparalleled multimodal resources of UK Biobank-structural and diffusion MRI, plasma proteomics, and metabolomics-we will discover precise, biologically grounded signatures for early detection and prognosis. Leveraging the cohort’s scale, we will fuse these omics features with our in-house imaging and metabolomics datasets already published (PubMed: 39394257, 39528560) and build Cox and machine-learning models that accurately identify individuals at highest risk for PD and PDD. The project will thereby transform PD management by (1) enabling therapeutic intervention during the neuroplasticity window, (2) guiding targeted prevention for high-risk individuals, and (3) providing personalized stratification for emerging disease-modifying trials. We will deliver clinically actionable tools that mitigate the mounting personal and societal burden of PD and dementia.
