High-cadence dynamic cycling improves motor symptoms of Parkinson's disease (PD), such as tremor and bradykinesia. However, some participants experience greater benefits than others. To gain insight into how individual characteristics and cycling performance affects functional changes, data from two previous studies were used to build several preliminary predictive models. The purpose was to examine which variables contribute to greater improvement in symptoms after high-cadence dynamic cycling. We hypothesized that individuals with higher body mass index (BMI), increased age, more severe symptoms, and higher PD medication dosages were less likely to contribute effort during cycling. UPDRS-III was assessed before and after each session, and cadence and power were recorded every second. Entropy of cadence was calculated, and data were analyzed using analysis of variance and multiple linear regression. The multiple linear regression model of post UPDRS significantly (R2 = 0.81, p < 0.001) explained its variance, with pre UPDRS as the main predictor (p < 0.0001). The binomial logistic model of mean effort did not significantly (R2 = 0.36, p = 0.14) explain the variance. Post-hoc analysis found a significant (β = 0.28, p = 0.03) moderating effect of different levels of BMI on the association between mean effort and post UPDRS. These results suggest that BMI, effort, and baseline UPDRS levels can potentially predict individual responses to high-cadence dynamic cycling.
Keywords: BMI; entropy; mixed model analysis; movement disorder; rehabilitation.
Copyright © 2022 Gates and Ridgel.