Objective: To develop optimal models of patients’ perception of non-motor and motor function on objectively measured Parkinson’s disease (PD) severity.

Background: There is increased interest from clinical and regulatory bodies in the voice of the patient and its relationship to clinician administered measures of PD disease severity.  Understanding the specific self-reported non-motor and motor functional impairments that contribute to PD severity is important to treatment efforts.   Optimal model development addresses this question in an agnostic method.

Method: The validated International MDS-UPDRS Translation database was used, providing complete examinations on 6684 persons from 19 languages(n=6,684). Patient reported functional impairment in non-motor (Part 1) and motor (Part 2) domains were used to model overall PD severity as measured of the clinical examination from the MDS-UPDRS (Part 3). Automatic Linear Modeling using an All Possible Subset – Best Model algorithm was used based on minimization of the Akaike Information Criterion. The resultant optimal model was replicated using the Overfit Prevention Criterion (OPC). Significance of item contribution was set at p < 0.05.

Results: The optimal model for non-motor functional impaired was significant (F= 211.436 [10,6662], p < 0.0005) with 10 items identified as significantly important (in order of importance): Urinary Function, Cognitive Impairment, Constipation, Hallucinations and Psychosis, Fatigue, Apathy, Depression, Daytime Sleepiness, Pain, and Light Headedness. The optimal model for motor functional impairment was also significant (F= 487.468 [13,6659], p < 0.0005) with 13 items identified as significantly important (in order of importance): Tremor, Walking and Balance, Get Out of Bed, Hygiene, Handwriting, Turn in Bed, Speech, Eating, Saliva and Drooling, Freezing of Gait, Dressing, Chewing and Swallowing, and Hobbies. Of the two models, the patient's perception of motor function was superior to their perception of non-motor function according to AIC values (34,356.445 vs 36,973.152). OPC indicated successful replication for both models.

Conclusion: Patient’s perception of non-motor and motor function can successfully model PD severity as rated by examiners. These results demonstrate the linking of self-reported non-motor and motor function to objectively determined PD motor severity.

« Back to 2019 International Congress

MDS Abstracts - https://www.mdsabstracts.org/abstract/modeling-the-effect-of-patients-perception-of-non-motor-and-motor-function-on-parkinsons-disease-severity/