Conflict of interest: The authors have no conflicts of interest to declare.
Assessment of sensory perception and processing using current perception threshold in Parkinson's disease
Article first published online: 17 NOV 2013
© 2013 Japanese Society of Neurology and Wiley Publishing Asia Pty Ltd
Neurology and Clinical Neuroscience
Volume 1, Issue 6, pages 209–213, November 2013
How to Cite
Ikeda, K., Deguchi, K., Kume, K., Kamada, M., Touge, T. and Masaki, T. (2013), Assessment of sensory perception and processing using current perception threshold in Parkinson's disease. Neurology & Clinical Neurosc, 1: 209–213. doi: 10.1111/ncn3.55
- Issue published online: 2 JAN 2014
- Article first published online: 17 NOV 2013
- Accepted manuscript online: 30 SEP 2013 09:25AM EST
- Manuscript Accepted: 18 SEP 2013
- JSPS KAKENHI. Grant Number: 24591297
- current perception threshold;
- epidermal nerve fiber;
- multiple system atrophy;
- Parkinson's disease;
- sensory processing
Although peripheral deafferentation is one of the causes of abnormal sensory processing in Parkinson's disease (PD), functional assessment of different-size sensory nerve populations remains insufficient.
We assessed the characteristics of sensory perception and processing in PD patients using current perception threshold (CPT).
A total of 28 patients with PD, 20 with multiple system atrophy (MSA) and 28 healthy controls underwent CPT examination using a Neurometer device, which can selectively stimulate A-β fibers at 2000 Hz, A-δ fibers at 250 Hz and C fibers at 5 Hz.
Significant differences in CPT for all stimulus frequencies were observed among the three groups. PD patients had significantly higher CPT for all stimulus frequencies than control individuals (P < 0.05). Also, PD patients had significantly higher CPT for 250 and 5 Hz than MSA patients (P < 0.05). At the optimum cut-off level, the CPT distinguished PD patients from MSA patients with a high sensitivity (82% at 250 Hz, 86% at 5 Hz) and specificity (75% at 250 Hz, 80% at 5 Hz). The CPT were not affected by clinical profile, cardiac sympathetic denervation or cardiovascular autonomic dysfunction.
Significantly increased CPT in PD indicate altered processing of sensory information involving A-β, A-δ and C fibers. The clear distinction of CPT between PD and MSA might reflect the involvement in small fiber dysfunction in PD, but not in MSA. These findings suggest that CPT testing might contribute to diagnosis of PD, as well as the understanding of sensory processing involving peripheral deafferentation.
Pain often described as a bizarre and unexplained painful sensation is one of the non-motor symptoms in Parkinson's disease (PD), and is experienced by up to 70–80% of patients.[1, 2] Although the exact mechanism of pain in PD remains controversial, impaired pain modulation of dopaminergic pathways involving the basal ganglia and the lateral and medial pain pathways,[2-7] dysfunction of monoamine systems such as serotonin and noradrenaline,[8, 9] and diminished activity of the dopaminergic descending inhibitory pathway have been estimated at the central (both cerebral and spinal) level.
At the peripheral level, significant losses of unmyelinated nerve fibers, free nerve endings and encapsulated nerve endings (Meissner corpuscles) have recently been shown in PD. Lewy body pathology, which might impact neuronal function and survival, was also observed in the peripheral nerves of skin in PD. These pathological findings suggest that nociceptive processing in PD is affected at multiple levels involving not only the higher interpretive centers, but also the peripheral nociceptors. Therefore, assessment of thinly myelinated A-δ nociceptors and unmyelinated C fiber nociceptors has enormous significance in the clarification of the pathophysiology of sensory dysfunction in PD.
For assessment of peripheral sensory nerves, analyzing the changes in the morphology and density of epidermal nerves is of immense diagnostic and prognostic value. However, strict counting rules and intensive training is required for accurate estimation of intraepidermal nerve fiber (IENF) density. In contrast, current perception threshold (CPT) testing with the Neurometer is suitable for use in screening for sensory perception and processing because of the following advantages: (i) A-β, A-δ and C fibers are measured quantitatively and selectively;, (ii) values of CPT show a high reproducibility independent of skin condition; and (iii) application of CPT testing is easy, non-invasive and less time-consuming than other methods.
The present study aimed to assess sensory perception and processing according to different-size sensory nerve populations in PD patients using CPT. We also used CPT to assess whether there is a difference in sensory perception and processing between PD and multiple system atrophy (MSA), because pain is fairly common in MSA as well as PD. In addition, we evaluated what factors might influence the results of CPT in PD patients.
A total of 28 patients with PD (mean [SD] age 66 years [11 years], range 43–83 years), 20 with MSA (63 years [8 years], range 49–77 years) and 28 age- and sex-matched healthy controls (64 years [8 years], range 49–82 years) participated in the present study. The mean duration of illness in the patients with PD (2.4 years [1.8 years],range 1–8 years) and MSA (2.0 years [0.9 years, 1–4) was equivalent. All patients with PD met the UK Parkinson's Disease Society Brain Bank criteria. They consisted of three with Hoehn and Yahr I, nine with II, 12 with three with IV and one with V. Based on their “tremor scores” and “non-tremor scores,” the patients were classified into three clinical subgroups: a tremor-dominant type (n = 2), akinetic-rigid type (n = 21) and mixed type (n = 5) subgroup. A total of 17 out of 28 PD patients had been treated with levodopa (n = 15), pramipexole (n = 9), ropinirole (n = 1) or selegiline (n = 3), alone or in combination, and the total daily levodopa equivalent dose was 466 mg (191 mg). All patients with MSA were diagnosed according to the consensus statement on the diagnosis of MSA. All patients were categorized as probable MSA. A total of 12 out of 20 patients had MSA of the cerebellar form (MSA-C), and the remaining patients had MSA of the parkinsonian form (MSA-P). None of the patients had a history of fluctuating cognition, visual hallucinations, diabetes, dysendocrinopathy, renal failure, chronic liver disease, alcohol abuse, vitamin B12 or folic acid deficiency, HIV or connective tissue disorders, or were taking medication known to affect the sensory and autonomic nervous system. Informed consent was obtained from all participants, and the protocol of the study was approved by the local ethics committee.
CPT testing with the Neurometer 3000 (Neurotron, Baltimore, MD, USA) was carried out in the supine position in a relatively quiet room. The stimulus site was near the external malleolus, which is the application site of sural nerve biopsy. The device emits sinusoidal alternating currents at 2000, 250 and 5 Hz at intensities from 0 to 10 mA. The current frequency at 2000 Hz is specific for measuring the function of A-β fibers; 250, for measuring Hz A-δ fibers; and 5, for Hz C fibers. This constant current output automatically compensates for alterations in skin resistance, and provides a standardized stimulus independent of skin thickness, degree of skin dryness or perspiration, or drying of the electrode paste. The electrical stimulus was initially increased until a specific sensation was reported by the participant. Short stimuli were then applied at progressively lower amplitudes until a minimal, but consistent, threshold was detected near the external malleolus bilaterally. The mean and the higher-side CPT values were evaluated in PD and MSA patients and healthy controls.
Evaluation of factors affecting the CPT
To clarify the factors affecting CPT in PD patients, we evaluated the relationship between CPT and other variables. For the clinical profile, the Hoehn Yahr stage, duration of illness, the influence of anti-parkinsonian drugs, sensory symptoms and more affected side of motor symptoms were used. In addition, the relationship between the “non-tremor score” in patients with akinetic-rigid type and CPT was assessed. The relationship between the “tremor score” in patients with tremor-dominant type and CPT was not assessed in the present study, because there were just two patients of this type. For assessment of cardiac sympathetic denervation, the heart-to-mediastinum (H/M) ratio was calculated in the early and delayed images of 123I-meta-iodobenzylguanidine (MIBG) myocardial scintigraphy. Cardiovascular autonomic function tests were carried out in order corresponding to bed rest on a tilt table for 20 min, a 60° head-up tilt (HUT) test for 10 min, the Valsalva maneuver for 15 s and active standing for 5 min, as previously described.
Data are expressed as means (SD). Overall differences in the means of a group were compared using the Kruskal–Wallis test followed by the Steel–Dwass test for multiple comparisons. The Mann–Whitney U test was carried out to analyze intra-group differences. The appropriate cut-off to differentiate case groups on the CPT values with the highest sum of sensitivity and specificity was established using receiver-operating characteristic (ROC) analysis. Associations between variables were studied by linear regression using Spearman's rank correlation coefficient (Z value). A probability level of P < 0.05 was considered a statistically significant difference.
Table 1 shows the results of the CPT testing. A significant difference in the mean and higher-side CPT to 2000-, 250- or 5-Hz stimulation was observed among the three groups (P < 0.01 for mean or higher side at 2000 Hz; P < 0.0001 for mean side and P < 0.001 for higher side at 250 Hz; P < 0.0001 for mean or higher side at 5 Hz). PD patients had significantly higher CPT in the mean and higher-side for all stimulus frequencies than control individuals (P < 0.05). Also, PD patients had significantly higher CPT in the mean and higher-side for 250 and 5 Hz than MSA patients (P < 0.05).
|2000 Hz (A-β fibers)|
|Mean of both sides||371 (141)a||286 (94)||265 (79)a||< 0.01|
|Higher side||424 (168)a||330 (107)||295 (85)a||< 0.01|
|250 Hz (A-δ fibers)|
|Mean of both sides||86 (42)a b||40 (35)b||44 (30)a||< 0.0001|
|Higher side||107 (50)a b||49 (42)b||60 (44)a||< 0.001|
|5 Hz (C fibers)|
|Mean of both sides||62 (52)a b||19 (17)b||15 (9)a||< 0.0001|
|Higher side||75 (62)a b||23 (19)b||17 (10)a||< 0.0001|
As CPT to all stimuli were significantly higher in PD patients than in healthy controls, elevated CPT was interpreted to imply the sensory dysfunction in PD. Therefore, ROC analysis was carried out using the higher-side CPT. Sensitivity and specificity between PD and controls were maximums at CPT values of 356 at 2000 Hz, 67.5 at 250 Hz and 34 at 5 Hz. At these optimum cut-off levels, the CPT testing distinguished between PD and controls with a sensitivity of 64% and specificity of 82% at 2000 Hz, a sensitivity of 82% and specificity of 68% at 250 Hz, and a sensitivity of 82% and specificity of 93% at 5 Hz (Fig. 1). A total of ten out of the 12 PD patients (83%) with less than 1 year duration of illness exceeded the cut-off level at 250 and 5 Hz. The sensitivity and specificity between PD and MSA reached maximum values at CPT of 62 for 250 Hz and 31 for 5 Hz. At these optimum cut-off levels, the CPT testing distinguished between PD and MSA with a sensitivity of 82% and specificity of 75% at 250 Hz, and a sensitivity of 86% and specificity of 80% at 5 Hz (Fig. 1).
Relationship between the CPT and other factors
There was no significant relationship between the higher-side CPT in response to 2000-, 250- or 5-Hz stimulation and Hoehn & Yahr stage or duration of illness. Also, the higher-side CPT at 2000-, 250- or 5-Hz stimulation were not affected by a “nontremor score” in patients with akinetic-rigid type. The higher-side CPT in patients treated with any kind of anti-parkinsonian drug showed no significant differences compared with CPT in drug-naive patients (treated patients vs naive patients: 424  vs 425  at 2000 Hz, 98  vs 121  at 250 Hz, and 77  vs 72  at 5 Hz). There was no significant difference in CPT between the more affected side and less affected side (more affected side vs less affected side: 384  vs 364  at 2000 Hz; 84  vs 88  at 250 Hz; and 60  vs 64 (63) at 5 Hz). A total of 28 PD patients (43%) complained of sensory symptoms, consisting of pain (n = 9) and numbness (n = 3). Laterality of sensory symptoms was observed in five patients, four of whom had the sensory symptoms on the more affected side. There was no significant difference in higher-side CPT between patients with and without pain (patients with pain vs patients without pain: 424  vs 425  at 2000 Hz, 108  vs 106  at 250 Hz, and 61  vs 86  at 5 Hz).
The H/M ratio in the early and delayed images was 1.56 (0.29) and 1.37 (0.32), respectively. The reduction of systolic and diastolic BP during HUT and active standing were as follows: HUT –18/–5 (19/12); active standing –15/–3 (19/11). Just 21 patients with PD could adequately carry out the Valsalva maneuver, and the value was 1.39 (0.22). Also, the results of 123I-MIBG myocardial scintigraphy and cardiovascular autonomic function tests showed no association of CPT with any types of stimuli.
The principal findings of the present study were (i) CPT in response to 2000-, 250- and 5-Hz stimuli were significantly higher in PD patients than in healthy controls, (ii) CPT in response to 250- and 5-Hz stimuli were significantly higher in PD patients than MSA patients, and CPTs in response to these stimuli distinguished PD from MSA with high sensitivity and specificity;, and (iii) CPT in PD patients were independent of various clinical profiles, H/M ratio in 123I-MIBG myocardial scintigraphy and cardiovascular autonomic function.
As the sensory threshold in PD has been investigated by various experimental conditions,, the results remain controversial. In the OFF phase, heat or cold pain threshold,[3, 5-7, 21] and laser or electrical pain threshold[4, 5, 7, 22, 23] were lower in PD patients than in healthy controls. In the ON phase after anti-parkinsonian drug dosing or deep brain stimulation of the subthalamic nucleus (STN-DBS), aside from a couple of exceptions suggesting possible involvement of monoamine systems other than the dopaminergic system,[8, 9] the heat, cold[2, 3, 6, 24] or laser pain threshold and nociceptive flexion reflex threshold were increased relative to the OFF phase. The thermal threshold also showed improvement after STN-DBS. In addition, low-frequency repetitive transcranial magnetic stimulation over the motor cortex significantly increased CPT in response to 2000-, 250- and 5-Hz stimuli. Thus, the sensory thresholds in PD might be affected by the central mechanism involving the dopaminergic system.
In contrast, the tactile threshold transmitted by A-β fibers was increased in both the ON and OFF phases,[11, 23, 26] and correlated with loss of myelinated fibers. The impairment in mechanical pain transmitted by A-δ fibers correlated with loss of epidermal nerve fibers. These findings suggest that the sensory problems in PD can be attributed to a peripheral deafferentation, as well as affected central modulation. Thus, the abnormalities of CPT in response to 2000-, 250- and 5-Hz stimuli in the present study might reflect the peripheral deafferentation involving A-β, A-δ and C fibers in PD.
Although it has been reported that pain is more common in MSA than in PD, the difference in pain processing between the two disorders remains unknown. In the present study, CPT at 250 Hz (for A-δ fiber stimulation) and 5 Hz (for C fiber stimulation) discriminated both disorders with high sensitivity and specificity. This finding suggests the possibility that peripheral deafferentation is present in PD, but not in MSA. The hypothesis is consistent with a previous report that dermal small-fiber density and the number of Schwann cells are significantly reduced in PD compared with MSA.
The CPT to 2000-, 250- and 5-Hz stimuli in PD patients were not affected by the severity or duration of illness. The elaborate analysis showed that the CPT in response to 250- and 5-Hz stimuli in the present study exceeded the cut-off level in 83% of PD patients with less than 1 year's duration of illness. This finding is consistent with previous reports showing the presence of sensory symptoms including pain and loss of autonomic innervation in the skin at an early stage of the disease. In combination with cardiac sympathetic denervation, orthostatic hypotension or baroreflex failure,[29, 30] CPT testing might provide a novel biomarker for detecting early-stage PD.
The CPT were not affected by a “non-tremor score” in patients with akinetic-rigid type PD. Akinetic-rigid type PD is characterized by degeneration of the dopaminergic projections to the dorsal putamen. Assuming that a “non-tremor score” reflects the degree of pathological changes of the dorsal putamen, the dorsal putamen within the basal ganglia might not play an important role in the CPT. In the present study, tremor-dominant type PD characterized by degeneration of the caudate nucleus and the lateral putamen was not assessed. To estimate the impact of the CPT on the basal ganglia, the relationship between CPT and the “tremor score” should be assessed in patients with tremor-dominant type.
As was seen in the present study, sensory symptoms including pain are clinically predominant on the more affected side in PD patients. This is in line with the observation in hemiparkinsonian patients that a facilitation of nociceptive spinal responses and abnormalities in laser-evoked potentials and pain perception were prevalent on the affected side. The nociceptive withdrawal reflex threshold in patients with bilateral parkinsonian signs, however, did not show a predominance of the more affected side. In a similar fashion, no bilateral difference of the CPT findings was observed in the present study, in which a large majority of patients had bilateral parkinsonian signs. As the IENF density was significantly lower on the more affected side, the CPT findings could reflect both a combination of the peripheral deafferentation involving A-β, A-δ and C fibers and the altered processing of sensory information transmitted along each different-size sensory nerve population at the central level.
There were several possible limitations to the present study. First, the IENF density was not evaluated in this study. To clarify the significance of CPT testing in PD, studies of the relationship between the CPT and the IENF density should be carried out. Second, the present study was a cross-sectional study involving a small number of patients. As PD diagnosis might change at follow-up, a longitudinal study should be carried out in a larger group of patients with PD. Third, unlike the estimation of IENF density, the results of CPT should be interpreted as an indicator of comprehensive sensory information processing with respect to each sensory nerve population.
In summary, PD patients showed a significant increase in CPTs in response to 2000-, 250- and 5-Hz stimuli, suggesting altered processing of sensory information involving A-β, A-δ and C fibers. Of those, CPT for 250- and 5-Hz stimuli discriminated between PD and MSA, probably because of the difference in cutaneous nerve fiber density. As CPT abnormality was observed even in the early stage of PD, CPT testing might contribute to the diagnosis of early-stage PD. To clarify whether CPT testing is useful for the diagnosis of PD, as well as the understanding of sensory processing involving peripheral deafferentation, a prospective study for a much larger number of patients with PD and other parkinsonian disorders is warranted.
This work was supported by JSPS KAKENHI Grant Number 24591297.
- 25Changes in sensory functions after low-frequency repetitive transcranial magnetic stimulation over the motor cortex. Tokai J. Exp. Clin. Med. 2009; 34: 122–9., , .