Mirror movements in Parkinson's disease: effect of dopaminergic drugs

Mirror movements refer to involuntary movements that emulate voluntary activity in contralateral homologous body regions.¹ We recently described this phenomenon in the less affected limb in early asymmetric parkinsonism.² The effects of time and dopaminergic drugs have not been examined. We now report on the re‐evaluation of this clinical sign in a subset of patients with clinically definite Parkinson's disease who were undergoing dopamine replacement therapy, from our original cohort.

Methods

We recruited 13 treated patients for this longitudinal assessment from the initial cohort of 27 patients² with suspected mirror movements and recent‐onset asymmetric idiopathic Parkinson's disease, seen at the Movement Disorders Centre, Toronto Western Hospital, between September 2001 and January 2003. At follow–up, all patients had been receiving optimal doses of dopaminergic drugs for many months. Parkinsonian features were combined into axial and lateralised scores using related items of the motor subscale of the Unified Parkinson's Disease Rating Scale (UPDRS). The asymmetry index was calculated as the absolute difference between the sides divided by the sum of their scores (L−R/[L+R]). A larger asymmetry index indicates greater difference in disease burden between sides and therefore more asymmetry. The patients' total daily dose of dopaminergic drugs were converted to levodopa equivalents using a published formula.³ The evaluation was first performed at least 12 h after the last dopaminergic treatment (“off” condition), and immediately after obtaining maximal benefit after intake of the first morning dose of dopaminergic treatment (“on” condition) given as immediate‐release carbidopa–levodopa preparation in levodopa equivalent dosages.³ The four assessed tasks performed corresponded to UPDRS items 23 (finger tapping), 24 (hand movements), 25 (forearm pronation–supination) and 26 (foot tapping). These tasks were videotaped with a standardised protocol. Two investigators, blinded to the study, rated the associated mirror movements in amplitude, distribution and proportion of mirroring activity using our previously published clinical scale.² Simple and multiple linear regressions were used to test the effects of UPDRS total motor, axial, limb composite and lateralised scores on mirror movements. We evaluated changes in UPDRS and mirror scores with time and drugs using repeated‐measures analysis of variance, paired t test or sign‐ranked test as appropriate. Agreement between clinicians (inter‐rater reliability) was measured using the κ statistic.⁴ κ levels were interpreted as follows: <0.4, poor to fair agreement; 0.4–0.6, moderate agreement; 0.6–0.8, substantial agreement and >0.8, excellent (“almost perfect”) agreement.⁵ Data were analysed using StatView V.5.01 (SAS Institute, Cary, North Carolina, USA).

Results

The inter‐rater reliability for the raters blinded to the study was substantial for off and on assessments with chance corrected κ scores of 0.76 and 0.75, respectively. Table 1 lists the clinical and mirror movement features of the 13 patients (eight women) assessed. Tremors was the initial symptoms for most (10/13) of the patients. The side and nature of initial deficits, age and sex had no influence on mirror movement score, disease duration and UPDRS score. None of the patients had mirror movements on the more affected side. Table 1 shows the distribution of mirror movements. The mean time between the two assessments was 2.8 years (SD 0.8). At the second assessment, all patients were on dopaminergic treatment with a levodopa equivalent of 529 mg (SD 295). The off and on mirror movement scores were not significantly different from the first assessment. Among the independent predictors used for the assessments (table 1), the only significant difference was between UPDRS off and on scores (p<0.001). Therefore, we examined whether the changes in UPDRS scores correlated with changes in mirror movements. We found a highly significant inverse correlation between UPDRS change (UPDRS off−UPDRS on) and mirror movement change (mirror movements off−mirror movements on) (r = 0.83; p = 0.002). From off to on, patients with large improvement in UPDRS scores had more mirror movements whereas those with small improvement in UPDRS scores had less mirror movements (fig 1). Higher levodopa equivalents tended to be associated with more mirroring (F = 2.56; p = 0.138), but this trend was lost when adjusted for motor UPDRS scores. The asymmetry index predicted the severity of mirroring on the less affected or unaffected side in a linear fashion at initial assessment (r = 0.63; p = 0.02) but not during off (r = 0.07; p = 0.809) or on states (r = 0.07; p = 0.820) at the second assessment.

Table 1 Characteristics of patients and mirror movements.

	At first assessment	At second assessment
Age (years)	63.4 (8.6, 43 to 78)	66.0 (8.4, 46 to 80)
Disease duration (years)	2.7 (1.4, 0.8 to 5.5)	5.5 (1.6, 3 to 9)
Levodopa equivalents (mg)*	46.4 (97.0, 0 to 300)	528.9 (294.7, 150 to 1000)
Motor UPDRS	21.3 (6.0, 9.5 to 28.5)	Off: 22.2 (8.9, 12.0 to 35.5)
		On: 17.2 (9.1, 7.0 to 29.5)
Lateralised UPDRS†	7.2 (3.2, 2 to 12)	Off: 6.0 (3.1, 3 to 11)
		On: 4.2 (1.8, 2 to 7)
Asymmetry index	0.50 (0.30, 0.12 to 1)	Off: 0.36 (0.21, 0 to 0.62)
		On: 0.38 (0.16, 0.14 to 0.67)
Axial UPDRS	6.7 (2.8, 3 to 12)	Off: 8.8 (5.2, 1 to 18)
		On: 7.7 (5.4, 0 to 16)
Mirror movement total score	13.4 (8.2, 0 to 25)	Off: 10.7 (8.2, 0 to 31)
		On: 11.4 (8.8, 0 to 23)
Distribution of mirror movements: (n/13)
Hands only	4	Off: 4
		On: 6
Foot only	1	Off: 1
		On: 6
Foot and hand	6	Off: 7
		On: 5
No mirror movements‡	2	Off: 1
		On: 1

Values in parentheses are SD, range.

UPDRS, Unified Parkinson's Disease Rating Scale.

*Only three patients were treated at first assessment, all were treated at the second assessment.

†Absolute side differences.

‡On the basis of video scoring these patients were found to have no mirror movements.

Figure 1 Regression plot of change in Unified Parkinson's Disease Rating Scale (UPDRS) and mirror movement (MM) scores. The horizontal line represents no change in MM. A negative change in MM indicates more mirroring and a positive change in MM indicates less mirroring.

Comments

We confirmed that mirror movements in the less affected limb are a feature of Parkinson's disease. They persist beyond 5 years of disease evolution, for at least 2–3 years after dopaminergic treatment is initiated, and their association with asymmetry is restricted to the earliest stages of clinical disease. Once treatment is started, the severity of mirror movements is influenced by dopaminergic drugs; mirroring was more prominent in patients whose motor response to dopaminergic treatment was greatest and less prominent in those whose response was less overt.

Although there was no overall change in mirror movement from the off to the on assessment, the inverse correlation between changes in UPDRS and mirror movements is intriguing. The reason for this correlation is not known. We speculate that in patients with large motor improvement, the improvement in bradykinesia and rigidity on the less affected arm may allow more mirror movements to occur. Although levodopa‐induced dyskinesia may increase with voluntary movement, none of our patients had dyskinesia and this is unlikely to be a contributing factor. Moreover, levodopa‐induced dyskinesia usually develops first on the most affected side,⁶ contralateral to the mirroring limb. In patients with small motor improvement, most of the treatment benefit may be on the more affected side. The reduction in mirror movements may be due to a reduced degree of asymmetry or a reduction in the effort required to perform the tasks used to assess mirror movements.

The potential use of mirror movements in the clinical evaluation of disease progression and the complete characterisation of the effects of dopaminergic dosage will require further studies with longer washout periods. Our mirror movement scale possesses substantial inter‐rater reliability for carrying out this objective.