Back

Six Latent Variables Underlie MDS-UPDRS Scores and Reveal a Dissociation Between Patient- and Clinician-Assessed Parkinson's Disease Symptoms

Kumar, B. S.; Humphries, M. D.

2026-08-11 neurology
10.64898/2026.08.10.26360084 medRxiv
Show abstract

The Movement Disorder Society's Unified Parkinson's Disease Rating Scale (MDS-UPDRS) is the global standard for characterising Parkinson's Disease (PD) in clinical contexts. However, the specific symptom phenotypes it captures remain poorly understood, potentially limiting its value for diagnosis, prognosis, and stratifying patients. To address this, we developed a spectral estimation approach to find the unique latent variables captured by the 60 scores of MDS-UPDRS parts I, II, and III from 852 sporadic PD patients. Our analysis revealed six latent variables that robustly captured variation between patients and generalised across cohorts. The primary variable encoded symptom laterality, while others encoded distinct clinical features including tremor severity, and revealed an unexpected dissociation between patient self-reported symptoms and clinician-assessed symptoms, highlighting potential gaps in how PD is currently evaluated and understood. Our findings open the door to precise MDS-UPDRS phenotyping of patients for treatments and clinical trials.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.