Motor patterns in Parkinson's disease: a data-driven approach

Abstract

To identify patterns of motor disturbances in Parkinson's disease (PD) and evaluate their relation with other PD domains. A cohort of 399 PD patients was randomly divided into two samples. Factors within the motor section of the SPES/SCOPA were identified by exploratory factor analysis on data from the first sample and next tested by confirmatory factor analysis in the second sample. Relations with other PD domains were evaluated by regression analyses. A four factor model was found to be valid. This included a tremor, a bradykinetic-rigid, and two axial factors. One axial factor ("rise", "gait", "postural instability") was associated with age and cognition, while the other axial factor ("freezing", "speech", "swallowing") was related to dopaminergic medication and complications of therapy. Both other factors showed no relevant associations with demographic and clinical characteristics. The identification of motor factors and their relation with other domains of the disease may help to elucidate the mechanisms responsible for these associations and provide an objective base for further research on subtypes in PD.