Objective: To evaluate the association between conventional clinical ratings of tremor and one week sensor-based monitoring  of tremor in patients with Parkinson’s disease (PwPD).

Background: Several methodological limitations render clinical ratings of tremor in PD less reliable. This study extends the initial validation of a sensor-based tremor amplitude measurement from a controlled to a home-based setting, incorporating a weekly percentage measure of tremor presence.

Method: Data were analyzed from 283 PwPD and 60 control participants of the Profiling Parkinson’s Disease (ProPark) study. Clinical tremor evaluation was performed using rest, postural and kinetic tremor items from the MDS-UPDRS III. Additionally, daily-life tremor presence (T%) and amplitude (TA) was determined using a wearable sensor recording angular velocity at the wrist for one week. Associations between clinical tremor evaluations and sensor outcomes were evaluated with a three-way independent ANOVA. Post-hoc pairwise Wilcoxon’s ranked sum tests with Bonferroni correction were used to compare sensor derived amplitude and presence of tremor between groups based on clinical evaluation and controls. The relationship between sensor-based tremor presence and amplitude was visually inspected.

Results: There was a significant association between most clinical evaluations of Postural, Kinetic and Rest tremors and the T% (F(4,332)=24.3, p<0.001; F(3,332)=10.5, p<0.001; F(3,332)=78.2, p<0.001, respectively) and TA (F(4,332)=9.7, p<0.001; F(3,332)=1.8, p=0.144; F(3,332)=3.7, p=0.011, respectively). Significant differences in TA measures were found between patients with postural tremor scores of 1 and 2, as well as among all groups and those scoring 3. Similarly, significant differences in T% were observed across all rest tremor groups, except between controls and those with a score of ‘0’, and between patients with scores of ‘1’ and ‘2’. Visual inspection revealed a U-shaped relationship between T% and TA, which progressively became more linear with heightened tremor presence.

Conclusion: Our study underscores the value of sensor-based measurements in quantifying tremor presence in daily life. However, accurate estimation of tremor amplitude hinges on the occurrence of adequate tremor episodes, as activities involving movement patterns within the 3.5 to 7.5 Hz frequency band may artificially inflate tremor amplitude estimates.

« Back to 2024 International Congress

MDS Abstracts - https://www.mdsabstracts.org/abstract/daily-life-sensor-based-measurement-of-tremor-presence-and-amplitude/