Evaluation of tremor interference with control of voluntary reaching movements in patients with Parkinson's disease

Abstract

Background: A large population of patients with Parkinson's disease (PD) displays the symptom of resting tremor. However, the extent that resting tremor may affect the performance of movement control has not been evaluated specifically. This study aims at establishing methods to quantitatively evaluate motor performance in PD patients with tremor, and at analyzing the interfering effects of tremor on control of reaching movements.

Methods: Ten PD patients with tremor and Ten healthy control subjects were recruited to participate in this study. All patients and healthy control subjects performed point-to-point reaching movements with their tremor affected arm or preferred arm. We verified that a smoothing model of minimum-jerk trajectory (MJT) can be used to extract voluntary movement trajectory from tremor-corrupted movement trajectory in the reaching tasks by the patients. Performance indices of reaction time (RT) and movement time (MT) of reaching movements by the PD subjects with tremor were evaluated using MJT trajectories. Differences of RT and MT between the recorded trajectories and MJT in PD and control subjects were calculated to investigate the extent that tremor may affect their motor performance. Linear mixed-effects model was used to identify the contributions of tremor, bradykinesia and rigidity to the performance indices of RT and MT based on UPDRS scores. The power spectrum densities (PSD) of tremor were also evaluated using hand velocities to represent tremor intensity and to analyze their correlations with RT and MT.

Results: The MJT model demonstrated good fit to recorded trajectory with a more consistent estimation of motor performance for both PD and control subjects. The RT and MT of patients were found to be 43.4 and 79.5% longer respectively than those of healthy control subjects. Analysis of the linear mixed-effects model was not able to reveal that tremor, bradykinesia and rigidity each had a significant contribution to RT or MT in PD patients with tremor. However, the PSD of tremor was found to correlate significantly to RT, but not to MT, in both linear regression and linear mixed-effects model.

Conclusions: The minimum-jerk trajectory and power spectrum densities are effective quantitative tools for evaluating motor performance for PD patients with tremor. Resting tremor is one of the factors prolonging the initiation of voluntary reaching movement in these patients.

Keywords: Movement time; Parkinson’s disease; Reaching movements; Reaction time; Tremor.

Fig. 1

The experimental set-up and an example of data records in performing a reaching movement. In (a), the subject was seated in front of a table with his/her forearm wrapped with the apparatus on the table to record the kinematic and EMG data on the horizontal plane. In (b) and (c), the recorded trajectories (left column) and the sEMGs of six muscles (right column) were displayed

Fig. 2

Hand trajectories of reaching movements in a control and a PD subject respectively. In this figure, (a) and (b) display the recorded hand path trajectory and the MJT of one trial for a control and a PD subject. The arrow points to the direction of movement. (c) and (d) show the recorded hand path trajectory and the end-point error of all trials for a control and a PD subject

Fig. 3

Velocities with fitted MJT-velocity respectively of reaching movements in a control and a PD subject. (a-b) show the method of optimizing MJT. (c) and (d) illustrate the velocity profiles along with the instant of trigger for task executing, and the durations of RT and MT. In Fig. (c) the velocity of the MJT exhibits a good fit with the typical bell-shape to the recorded velocity in the normal subject. In (d), the velocity of the patient with PD shows significant oscillations from tremor, and the velocity of MJT could be fitted to the tremor-corrupted velocity with the bell-shape profile, which represents the trajectory of voluntary movement in the patient with PD

Fig. 4

The distribution of RT and MT with statistical results of control and PD subjects. (a) illustrates the distribution of RT and MT with the 95% confidence ellipse in control and PD subjects, and (b) shows the statistical comparison of RT and MT between control and PD subjects

Fig. 5

Statistical results of the changes of tremor PSD prior to and after voluntary movement. (a) exhibits that the number of tremor inhibited trials was significantly larger than non-inhibited trials. (b) displays the reduction of tremor component that the tremor component in the hand prior to voluntary movement is greater than that after voluntary movements. In (c) and (d), the energy component of tremor is reduced in most joints and sEMG of muscles in most patients, except for a few joints and sEMG

Fig. 6

The correlation between RT/MT and PSD of tremor prior to movement. The tremor intensity (PSD of tremor) was correlated to lgRT but not to lgMT