Although Parkinson’s disease (PD) is diagnosed as a movement disorder, motor deficits are preceded by non-motor features including cognitive, cardiovascular, gastrointestinal, psychiatric and sleep disturbances. Characterizing patterns of such non-motor features in early, medication-naïve PD may allow for improved treatment.
This study will use cluster analyses based exclusively on non-motor features in medication-naïve PD patients. Cluster analysis groups similar objects in the same group; this study will define similarity by non-motor features. To see if these patterns of non-motor features in PD may help in diagnosis, we will use a novel analytic method called Multicategory Composite Least Squares Classifiers. Using data from participants of the Parkinson’s Progression Markers Initiative, we will develop strategies to recognize PD separate from motor features.
Relevance to Diagnosis/Treatment of Parkinson’s Disease:
Results of this project have potential to provide the first step in developing strategies of identifying PD prior to motor signs by using non-motor features and potential to improve treatment of PD at very early stages.
Novel recognition of non-motor subtypes in early PD will allow for development of targeted treatment strategies and improved prognosis. If patterns of non-motor features can distinguish PD from non-PD patients, this would open novel opportunities for earlier diagnosis and disease surveillance.