Abnormal posture is often seen in patients with Parkinson disease (PD). Past studies have shown that a third of patients with PD have postural deformities.1,2 The clinical phenotype of abnormal posture is variable but can be categorized into two broad types: lateral flexion of the trunk (Pisa syndrome and scoliosis) and vertical flexion (camptocormia and anterocollis).1 Lateral trunk flexion, which occurs in both scoliosis and Pisa syndrome, is common in patients with PD. The prevalence of scoliosis is 8.5% to 60% in PD.1 A large multicenter study that focused exclusively on Pisa syndrome recently reported that this condition occurred in 8.8% of patients with PD and caused poor quality of life.2 Although many symptoms of PD can be satisfactorily controlled by medications, surgery, and physiotherapy, the management of postural deviations remains challenging, and thus postural abnormalities have attracted attention in recent years.1,3 Asymmetrical function of the basal ganglia is considered to be the main cause of lateral trunk flexion.3 Dysfunction of the basal ganglia facilitates the blink reflex by disinhibiting the spinal trigeminal nucleus through a pathway mediated by the superior colliculus.4
Brainstem excitability can be assessed by the blink reflex recovery curve (BRrc).5 Past studies have shown a disinhibited BRrc in some movement disorders, such as PD,6 dystonia,7,8 and blepharospasm.9,10 In addition, dystonia is often seen in PD and is believed to cause abnormal postures.1 Thus far there have been no electrophysiological studies examining the central nervous system function of patients with PD with abnormal posture. To determine with PD with abnormal posture is related to dystonia and abnormal function of the basal ganglia, we investigated whether a disinhibited BRrc was associated with abnormal posture in patients with PD, especially those with lateral trunk flexion.
We enrolled 21 subjects diagnosed with PD according to the United Kingdom Parkinson's Disease Society Brain Bank criteria.11 Patients with PD were classified into two groups: those with marked lateral trunk flexion (F-PD) and those with normal posture (N-PD). Patients were diagnosed with F-PD if they presented with lateral trunk flexion of at least 10° (Fig. 1). Causes of lateral flexion included Pisa syndrome and scoliosis. Pisa syndrome was almost completely alleviated by passive mobilization or spine positioning. Scoliosis was not relieved by passive movement or supine positioning, and patients had radiological evidence of a structural curve with axial vertebral rotation.1 The exclusion criterion was stage five on the Hoehn and Yahr scale (H–Y scale). None of our patients with PD had been diagnosed with other neurological diseases, such as dystonia and blepharospasm, before the onset of PD. In addition, none of the patients with lateral trunk flexion had other potential causes for this condition, such as congenital scoliosis or muscular disorders. In all cases, clinical variables were evaluated and electrophysiological examination was performed while patients were under the drug-ON condition. Ten age-matched healthy volunteers (controls; no medical history of neurological disease) were also enrolled to allow us to obtain normal BRrc values. Informed consent was obtained from all participants. The study protocol was approved by the Ethics Committee of the University of Miyazaki and was performed in accordance with the Declaration of Helsinki.
Clinical variables were determined from patients' medical records. In patients with PD, clinical evaluation included disease duration, H–Y scale, major motor symptoms (resting tremor, rigidity, finger taps, gait disturbance, postural stability, and body bradykinesia), motor fluctuation, and pharmacological treatment (levodopa and dopamine agonists). Major motor symptoms were assessed according to the Unified Parkinson's Disease Rating Scale (UPDRS) part III.
Blink Reflex and Blink Reflex Recovery Curve
Electromyography responses were recorded using surface electrodes with Neuropack MEB-2200 (Nihon-Kohden, Japan) from the inferior orbicularis oculi muscles ipsilateral and contralateral to the electrical stimulation site. A ground electrode was placed on the nose. The electromyography responses were amplified and bandpass filtered (10 Hz–5 kHz). Electrical stimulation was applied to the supraorbital nerve at the supraorbital notch bilaterally with a monopolar stimulating electrode. All stimuli were 0.2 ms in duration, and the stimulus intensity was adjusted to obtain well-defined, reproducible ipsilateral and contralateral R2 responses in four consecutive trials. We used the R2 responses in which onset latency was over 30 ms.5 To avoid interference with the R3 response, we used an upper limit of 80 ms for the R2 endpoint.12 Subjects were studied at rest, in the supine position, with eyes gently closed. For BRrc recording, paired electrical stimuli were administered at interstimulus intervals (ISIs) of 200, 300, and 500 ms in accordance with a past study.9 To compensate for the variability of the responses, we applied each stimulus four times for each ISI at intervals of 15 to 20 seconds to minimize habituation, and the blink recordings were rectified and averaged. Onset latency and duration of R1 and R2 responses were determined by manual cursor marking of the beginning and end of responses. The area of a conditioned R2 response was calculated over the same duration as the unconditioned response. The BRrc was defined by plotting R2 area ratios (the R2 area of a conditioned response divided by the R2 area of an unconditioned response) for each ISI.
We compared the clinical characteristics of F-PD and N-PD patients, and also assessed differences in the blink reflex between the F-PD, N-PD, and control groups. The mean BRrc was defined as the mean recovery curve of the ipsilateral and contralateral stimulation sides. The difference in BRrc was defined as the recovery curve of the absolute difference between the ipsilateral and contralateral sides. Because asymmetrical basal ganglia activity is believed to cause lateral trunk flexion in patients with PD, we compared the absolute differences between F-PD and N-PD patients.
Details of the analyses of clinical features are shown in Table 1. A one-way analysis of variance (ANOVA) was used to assess differences between the F-PD, N-PD, and control groups in terms of blink reflex components (R1 latency, R1 amplitude, ipsilateral and contralateral R2 latencies, and ipsilateral and contralateral R2 areas). Repeated-measures ANOVA was performed with ISI (200, 300, 500 ms) as the repeated factor to assess the BRrc. Tukey correction was used for post hoc analysis. SPSS ver. 22 was used for the statistical analysis. P values less than 0.05 were considered statistically significant.
Eleven F-PD patients, 10 N-PD patients, and 10 controls showed no significant differences in age or sex. The F-PD patients included 10 with scoliosis and 1 with Pisa syndrome. No significant differences in clinical features were detected between F-PD and N-PD patients (Table 1).
In the study of right-side stimulation, there were no significant differences between the three groups regarding R1 latency, R1 amplitude, ipsilateral and contralateral R2 latencies, or R2 areas (Table 2). In the study of left-side stimulation, one-way ANOVA for the R1 amplitude showed significant differences between the three groups (P = 0.040). A post hoc test with Tukey correction showed a significant difference between the F-PD and N-PD groups (P = 0.044), but no significant difference between the F-PD and control groups (P = 0.126). There were no other significant differences in blink reflex parameters on left-side stimulation (Table 2).
Blink Reflex Recovery Curve
The contralateral and ipsilateral conditioned/unconditioned R2 ratios were almost identical; thus, we show only the results of the ipsilateral BRrc. Figure 2 shows typical examples of unconditioned (single pulse) and conditioned (paired pulses) blink reflex responses at ISIs of 200, 300, and 500 ms. Test stimulations were followed by conditioning stimulations after 200, 300, and 500 ms. At all three ISIs, conditioned responses were inhibited compared with unconditioned responses in N-PD and controls. However, conditioned responses were less inhibited at all ISIs in F-PD.
Figure 3A shows the mean BRrcs of F-PD and N-PD patients and controls. Repeated-measures ANOVA between the three groups showed significant effects of ISI (F = 29.30, P < 0.001) and group (F = 7.94, P = 0.002), but no significant interaction of group*ISI (F = 1.72, P = 0.158). A post hoc test with Tukey correction showed that the BRrc was significantly less inhibited in F-PD patients compared with both N-PD patients (P = 0.007) and controls (P = 0.004), whereas there was no significant difference in the BRrc between N-PD patients and controls (P = 0.979). Figure 3B compares the BRrcs on the convex and concave sides in F-PD patients. Repeated-measures ANOVA between the two groups showed a significant effect of ISI (F = 3.86, P = 0.0294), but no significant effects of group (F = 0.08, P = 0.784) or interaction of group*ISI (F = 0.24, P = 0.788).
Figure 4 shows the average of the absolute difference between the right and left BRrcs of each F-PD and N-PD patient. Repeated-measures ANOVA between F-PD and N-PD patients showed a significant effect of group (F = 6.64, P = 0.018) but no significant effects of ISI (F = 2.30, P = 0.114) or interaction of group*ISI (F = 0.82, P = 0.922). Thus, the BRrc in F-PD patients was more disinhibited than in N-PD patients and controls, and this disinhibition in F-PD was asymmetrical.
This study indicated that in F-PD patients, the BRrc (1) was more disinhibited than in N-PD patients and controls, (2) did not correlate with the convex side of the trunk during lateral flexion, and (3) was asymmetrically disinhibited.
None of the parameters of the conventional blink reflex, except for the R1 amplitude with left-side stimulation, showed any significant difference between the F-PD, N-PD, and control groups. The R1 amplitude was significantly reduced with left-side stimulation in F-PD patients, but the reduction was not associated with clinical symptoms or the direction of lateral trunk flexion. The R1 amplitude with right-side stimulation was also reduced in F-PD patients, but not significantly so. Although a past study found that the R1 component showed no pathological change in PD,13 R1 amplitude reduction with bilateral side stimulation might reflect mild brainstem dysfunction in patients with PD with lateral trunk flexion.
A recent study reported disinhibition of the BRrc in PD, especially tremor-dominant PD6; thus, the BRrc in PD patients had been believed to be specifically associated with the severity of tremor. In this study, however, the BRrc in N-PD patients was not disinhibited. The UPDRS part III tremor score (items 20 and 21) of the PD patients in the past study was 13.4 ± 5.2; by contrast, in our F-PD and N-PD patients, the resting tremor scores were 1.7 ± 1.9 and 1.5 ± 1.2, respectively. Because the tremors in our patients with PD were well controlled by anti-Parkinson medications, the BRrc in N-PD patients was not disinhibited and the relationship between the abnormal BRrc and lateral trunk flexion in F-PD might have been analyzed more easily.
In this study, the mean BRrc was disinhibited in F-PD but not N-PD patients. Past studies relevant to abnormal posture have shown that the BRrc is disinhibited in cervical and generalized dystonia.7,8 In addition, BRrc inhibition is significantly disturbed in patients with primary torsion dystonia and is progressively normalized after deep brain stimulation of the globus pallidus internus, which improves dystonia.14 Disinhibition of the BRrc is related to dystonia as well as lateral trunk flexion in PD. The mechanism underlying lateral trunk flexion in PD may partly involve dystonia.
This study detected no correlation between the side of BRrc disinhibition and the side of trunk convexity. There have been no past reports about the laterality of the BRrc. Although the convex side was reported to be associated with more striatonigral denervation on the contralateral side,3,15 a recent study reported no relation between the convex side of the trunk and the affected side of the brain.2 Although it has been hypothesized that the peripheral nervous system, including the musculoskeletal system, plays a pathological role in dystonia, it is believed to be more likely that the primary causes are central nervous system abnormalities, including asymmetry of basal ganglia outflow and defects in the central integration of sensory information.3 Compensation for the laterality of dysfunction is also believed to contribute to pathogenesis; thus, the pathophysiology of the direction of trunk deviation is complicated.
This study indicated that the BRrc was asymmetrically disinhibited in F-PD but not in N-PD patients. In PD, excitability of the blink reflex is disinhibited through the following mechanism.4,16 Dopamine depletion facilitates the activity of the substantia nigra pars reticularis, which inhibits the superior colliculus through GABAergic output. The loss of superior colliculus excitability decreases the nucleus raphe magnus activity, which removes inhibition of the spinal trigeminal nucleus. As a consequence of this pathway, the blink reflex becomes hyperexcitable. In a rat model of hemiparkinsonism, greater depletion of striatal dopamine caused more severe postural deviation.15 Furthermore, another animal study suggested that asymmetry of dopaminergic activity between the two sides of the basal ganglia led to postural deviation.3 Thus, the asymmetrically disinhibited BRrc in F-PD patients may reflect the fact that asymmetrical dopamine depletion in the basal ganglia disinhibited brainstem function, and that this depletion is related to lateral trunk flexion.
There were some limitations to our study. The number of cases was small; thus, we examined the F-PD group, which included patients with scoliosis or Pisa syndrome. None of the F-PD patients underwent dopamine transporter imaging, although asymmetry of basal ganglia output might be related to lateral trunk flexion.
Despite these limitations, our study indicated that the BRrc in F-PD patients was significantly disinhibited, and the laterality of this inhibition was significantly greater than in the other groups. We supposed that the asymmetrically disinhibited BRrc in F-PD patients was associated with lateral trunk flexion. Examination of the BRrc may permit early detection of asymmetrical basal ganglia dysfunction that can eventually cause lateral trunk flexion. Early detection of postural abnormalities could contribute to appropriate treatment choices in patients with PD.
1. Doherty KM, van de Warrenburg BP, Peralta MC, et al. Postural deformities in Parkinson's disease. Lancet Neurol 2011;10:538–549.
2. Tinazzi M, Fasano A, Geroin C, et al. Pisa syndrome
in Parkinson disease
: an observational multicenter Italian study. Neurology 2015;85:1769–1779.
3. Castrioto A, Piscicelli C, Pérennou D, Krack P, Debû B. The pathogenesis of Pisa syndrome
in Parkinson's disease. Mov Disord 2014;29:1100–1107.
4. Basso MA, Evinger C. An explanation for reflex blink hyperexcitability in Parkinson's disease. II. Nucleus raphe magnus. J Neurosci 1996;16:7318–7330.
5. Kimura J. The blink reflex. In: Kimura J, ed. Electrodiagnosis in Diseases of Nerve and Muscle: Plinciples and Practice. 2nd ed. Philadelphia: FA Davis, 1989; 307–331.
6. Nisticò R, Salsone M, Vescio B, et al. Blink reflex recovery cycle distinguishes essential tremor with resting tremor from de novo Parkinson's disease: an exploratory study. Parkinsonism Relat Disord 2014;20:153–156.
7. Nakashima K, Rothwell JC, Thompson PD, et al. The blink reflex in patients with idiopathic torsion dystonia. Arch Neurol 1990;47:413–416.
8. Pauletti G, Berardelli A, Cruccu G, Agostino R, Manfredi M. Blink reflex and the masseter inhibitory reflex in patients with dystonia. Mov Disord 1993;8:495–500.
9. Schwingenschuh P, Katschnig P, Edwards MJ, et al. The blink reflex recovery cycle differs between essential and presumed psychogenic blepharospasm. Neurology 2011;76:610–614.
10. Conte A, Defazio G, Ferrazzano G, et al. Is increased blinking a form of blepharospasm? Neurology 2013;80:2236–2241.
11. Gelb DJ, Oliver E, Gilman S. Diagnostic criteria for Parkinson disease
. Arch Neurol 1999;56:33–39.
12. Ellrich J, Hopf HC. The R3 component of the blink reflex: normative data and application in spinal lesions. Electroencephalogr Clin Neurophysiol 1996;101:349–354.
13. Valls-Solé J, Muñoz JE, Valldeoriola F. Abnormalities of prepulse inhibition do not depend on blink reflex excitability: a study in Parkinson's disease and Huntington's disease. Clin Neurophysiol 2004;115:1527–1536.
14. Tisch S, Limousin P, Rothwell JC, et al. Changes in blink reflex excitability after globus pallidus internus stimulation for dystonia. Mov Disord 2006;21:1650–1655.
15. Herrera-Marschitz M, Utsumi H, Ungerstedt U. Scoliosis
in rats with experimentally-induced hemiparkinsonism: dependence upon striatal dopamine denervation. J Neurol Neurosurg Psychiatry 1990;53:39–43.
16. Basso MA, Powers AS, Evinger C. An explanation for reflex blink hyperexcitability in Parkinson's disease. I. Superior colliculus. J Neurosci 1996;16:7308–7317.