Session Information
Date: Thursday, June 23, 2016
Session Title: Other
Session Time: 12:00pm-1:30pm
Location: Exhibit Hall located in Hall B, Level 2
Objective: To develop statistical models of Parkinson’s disease (PD) progression based on a transversal assessment and to explore its predictive value in a three-year follow-up.
Background: Individual prediction of motor symptom progression in PD is currently not feasible. A better understanding of how PD symptoms progress would help patients make informed decisions and would support clinicians in treatment decisions and design clinical trials.
Methods: A cross-sectional evaluation of 300 consecutive PD patients was conducted using the Unified Parkinson’s disease Rating Scale (UPDRS) subscales II and III, the modified Hoehn & Yahr scale (H&Y), the Schwab and England Independence Scale (S&E), and the freezing of gait questionnaire (FOG-Q). Based on UPDRS-III, an axial index (i.e., items 18, 27, 28, 29, and 30) was calculated. UPDRS-III and H&Y was applied after 12h without antiParkinsonian medication (“OFF”) and one hour after the usual morning dose (“ON”). Multiple linear regression analyses (independent variables: disease duration, age>70, age at disease onset>55, tremor as the first symptom alone, and medication description) were used to develop statistical models of disease progression. Sixty-eight PD patients were reevaluated three years later. The regression coefficients from the transversal study were applied to analyze the reevaluation scores. The differences in the adjusted scores from the initial and the follow-up assessments were analyzed using one-sample t-test and descriptive statistics.
Results: For the cross-sectional assessment, the variance of test scores explained by the regression model ranged between 35% and 57%. The mean differences of the adjusted test scores were not significantly different from 0 (p>0.05). The frequency of patients scoring as expected at follow-up (i.e., adjusted score differences between -1 and +1) varied between the studied measures. The regression-based model predicted more accurately (i.e., 70%-72% of patients scored as expected at follow-up) the axial index and the S&E both OFF and ON conditions, and the FOG-Q. For the remaining measures, the frequency of patients scoring as expected ranged between 59% and 67%.
Conclusions: The study explored a regression-based approach to PD progression. The predictive value of the statistical models of PD progression was confirmed at the individual level in a prospective three-year follow-up.
To cite this abstract in AMA style:
A. Mendes, A. Gonçalves, N. Vila-Chã, M. Calejo, I. Moreira, J. Fernandes, J. Damásio, A. Bastos-Lima, S. Cavaco. Predictive validity of statistical models of Parkinson’s disease progression [abstract]. Mov Disord. 2016; 31 (suppl 2). https://www.mdsabstracts.org/abstract/predictive-validity-of-statistical-models-of-parkinsons-disease-progression/. Accessed September 7, 2024.« Back to 2016 International Congress
MDS Abstracts - https://www.mdsabstracts.org/abstract/predictive-validity-of-statistical-models-of-parkinsons-disease-progression/