Mirror Visual Feedback to Improve Bradykinesia in Parkinson's Disease

Abstract

Mirror visual feedback (MVF) therapy has been applied to improve upper limb function in stroke. When combined with motor training, MVF improves the performance of the trained and untrained hand by enhancing the excitability of both primary motor cortices (M1s). Bradykinesia is a typical feature of Parkinson's disease (PD), characterized by slowness in the execution of movement. This condition is often asymmetrical and possibly supported by a volitional hypoactivation of M1. MVF therapy could tentatively treat bradykinesia since the untrained hand, which benefits from the exercise, is generally more severely impaired in undertaking sequential movements. Aim of the study was to evaluate whether MVF therapy may improve bradykinesia of the more affected hand in PD patients. Twelve PD patients and twelve healthy controls performed for 10 minutes a finger sequence, receiving MVF of the more affected/nondominant hand. Before and after MVF training, participants performed a finger sequence at their spontaneous pace with both hands. M1 excitability was assessed in the trained and untrained hemispheres by means of transcranial magnetic stimulation. Movement speed increased after MVF training in either hand of both groups. MVF therapy enhanced cortical excitability of M1s in both groups. Our preliminary data support the use of MVF therapy to improve bradykinesia in PD patients.

Figures

Figure 1

Experimental paradigm.

Figure 2

Mirror visual feedback (MVF) training effect on behavioural data. Groups (Parkinson's disease (PD) patients, healthy subjects (HS)) and hand (trained, untrained) are indicated in the abscissa. Data recorded at baseline (before training) and after MVF training session are reported. Ordinate indicates the mean values of (a) number of finger movements performed in one minute during the assessments; (b) intertapping interval expressed in milliseconds; (c) touch duration expressed in milliseconds; and (d) % of correct sequences. Vertical bars indicate standard error of the mean (SEM). Asterisks indicate that in both groups the number of finger movements performed in one minute significantly increased and ITI significantly decreased after MVF training (∗p < 0.05; ∗∗p < 0.01).

Figure 3

Behavioural data, showing Δ score of the numbers of finger movements performed in one minute (number of finger movements/min after MVF training − number of finger movements/min before MVF training) in the trained and untrained hands of both Parkinson's disease (PD) patients (grey bars) and healthy subjects (HS) groups (white bars). Vertical bars indicate standard error of the mean (SEM).

Figure 4

Correlation analysis between individual changes in the number of finger tapping movements induced by mirror visual feedback practice and individual scores at the Parkinson's Fatigue Scale-16 (PFS-16) questionnaire in Parkinson's disease patients. There is a significant positive correlation between the improvement in the less affected/trained hand (a) and the clinical score (r = 0.64; p = 0.024), indicating that the higher the fatigability, the lower the performance improvement. In (b) the nonsignificant correlation between the improvement of the more affected/untrained hand and PFS-16 is depicted (r = 0.54, p = 0.07).

Figure 5

Input-Output (IO) curves measured in the first dorsal interosseus (FDI) muscle, of the trained (a and c) and the untrained (b and d) M1s before and after mirror visual feedback (MVF) training. Data of both groups, Parkinson's disease (PD) patients (a and b) and healthy subjects (HS) (c and d), who underwent MVF training, are shown. MEP amplitudes, in mV, are depicted from 90% to 130% S1mV (the stimulus intensity needed to evoke MEPs of approximately 0.8−1 mV peak-to-peak amplitude). Vertical bars indicate standard error of the mean (SEM).

Figure 6

Data from the control experiment (training without MVF). (a) shows the behavioural data, expressed as performance gain (Δ score of the numbers of finger movements performed in one minute) in the untrained hand of PD patients enrolled in the main experiment (PD mirror) and PD patients enrolled in the control experiment (PD control). In (b) the Input-Output (IO) curves of the untrained M1 of the PD control group, before and after training, are depicted. MEP amplitudes, in mV, are depicted from 90% to 130% S1mV (the stimulus intensity needed to evoke MEPs of approximately 0.8−1 mV peak-to-peak amplitude). Vertical bars indicate standard error of the mean (SEM). ∗p < 0.05.