Imbalance and lower limb tremor in chronic inflammatory demyelinating polyradiculoneuropathy

Imbalance is a prominent symptom of chronic inflammatory demyelinating polyradiculoneuropathy (CIDP). Although upper limb tremor in CIDP is described, lower limb tremor has not been assessed. The aim of this study was to determine whether lower limb tremor was present in CIDP and assess potential relationships with imbalance.


| INTRODUCTION
Chronic Inflammatory Demyelinating Polyradiculoneuropathy (CIDP) is an acquired auto-immune neuropathy characterized by progressive or recurrent symmetric proximal and distal weakness and sensory dysfunction developing over at least 8 weeks. 1 Imbalance, ataxia, and upper limb tremor are common clinical features of CIDP, contributing to disability. [2][3][4] Atypical CIDP subtypes, related to presence of nodal/ paranodal or myelin-associated glycoprotein antibodies also exhibit prominent tremor, ataxia, and imbalance. 5 The pathophysiological mechanisms underlying the development of imbalance in CIDP remain to be fully elucidated. A greater degree of distal lower limb muscle weakness has been suggested as a potential mechanism of imbalance. 6 Additionally, impairment of proprioceptive afferent inputs has also been proposed and is associated with a longer time to make anticipatory adjustments to postural perturbations. 7 Of possible relevance to postural imbalance in CIDP, lower limb tremor has been associated with postural instability in Essential Tremor 8 and primary orthostatic tremor. 9 The issue of whether lower limb tremor is a feature in CIDP, and whether it contributes to imbalance, remains to be established.
Balance can be objectively measured through analysis of posturography, using a force platform. Objective balance measures are collected through the piezoelectric transduction of forces exerted through the dedicated platform, enabling a calculation of ground reaction forces and center of pressure (CoP). The CoP is the center point where a participant is applying force through the lower limbs to maintain balance. The CoP moves as participants attempt to maintain position, resulting in a sway path, area, and velocity. 10,11 Although prior studies have reported abnormalities of sway parameters in peripheral neuropathy, 6,12,13 detailed analysis of the forces exerted through the platform to maintain stance has not been performed. Additionally, the relationship between balance and objective measures of lower limb tremor in CIDP has not been previously assessed.
Consequently, the aim of the present study was to determine whether lower limb tremor, as assessed by clinical and neurophysiological measures, was a feature of CIDP, and whether the presence of lower limb tremor contributed to postural imbalance.

| Patients
Consecutive CIDP patients were prospectively recruited from neuromuscular clinics between July 2019 and September 2022. CIDP patients were diagnosed by a neuromuscular expert and classified according to the European Academy of Neurology/Peripheral Nerve Society guideline on diagnosis of CIDP 1 and were only included if they fulfilled the criteria of typical CIDP. Patients with nodal/paranodal antibodies were excluded to minimize heterogeneity, particularly as these subtypes were recently classified as entities clinically distinct from typical CIDP. 1 Additional exclusion criteria were tremor-diagnosed prior to diagnosis of neuropathy; presence of a preexisting diagnosis of essential tremor, Parkinsonian tremor, or dystonic tremor; family history of these conditions in a first-degree relative; medication that could induce tremor, or the presence of medical conditions (such as hyperthyroidism) that could explain tremor. If CIDP patients were receiving IVIg, tremor analysis was performed just prior to their regularly scheduled dose, to correspond to a trough in treatment effect, and at the mid-point of the cycle, to correspond to a peak in treatment effect. Balance was only examined prior to IVIg (trough) to maximize sway and thus imbalance. 13 Prior to undertaking any assessment, written informed consent was obtained from all participants, and the study was ethically approved by the Western Sydney Local Health District Human Research Ethics Committee (2019/ETH08777/STE10638).

| Clinical assessments
All the participants underwent comprehensive clinical assessment including with clinical rating scales. Disability was calculated using the Inflammatory Rasch-built Overall Disability Scale (iRODS) for immune-mediated peripheral neuropathies. 14 Upper and lower limb power was assessed by the Medical Research Council (MRC) sum score 15 and the degree of sensory impairment by the Inflammatory Neuropathy Cause and Treatment (INCAT) sensory score. 16 An overall assessment of lower limb findings was calculated using the Neuropathy Impairment Score in the Lower Limbs (NIS-LL) scale. 17 Additionally, all the participants were assessed clinically for the presence of tremor and rated using The Essential Tremor Rating Assessment Scale (TETRAS). 18 This rating scale features both upper and lower limb assessments, so a separate subscore accounting for only the lower limb components was calculated for use in the present study and labelled the TETRAS-LL. Specifically, item 7 from the activities of daily living subscale, "Carrying food trays, plates, or similar items," and items 5 and 9 from the performance subscale, "Lower limb tremor" and "Standing tremor," respectively, were included. The maximum score on this subscale was 12.
Balance was measured using the Berg Balance Scale (BBS). 19 This scale assesses balance by performing functional tasks such as standing from sitting, controlled descent to a chair, turning, transferring, reaching, and bending. 19 The scale has been shown to have high reliability. 20 A score of 42, in combination with a history of imbalance, has a predicted probability of falling of 91%. 21 A BBS score ≤42 was therefore used to define CIDP patients with poor balance, while a BBS > 42 indicated good balance. Additional clinical markers including vibration sense, measured with a Rydel-Seiffer tuning fork, and gait speed, measured with a timed 10-meter walk test, were also collected as these have been shown to predict falls and imbalance. 22,23

| Nerve conduction studies
All the participants underwent conventional nerve conduction studies using established methods. 24 All studies were performed on a single machine using Synergy software (Nicolet EDX by Natus, Pleasanton, CA, USA). In all cases, the right lower limb was assessed. Motor nerve assessments included common peroneal, and tibial nerves, as well as tibial nerve F wave latencies. Orthodromic sensory nerve conduction of the sural nerve was also performed.

| Neurophysiological tremor studies
Tremor studies were performed using surface electromyography (EMG) with the patient initially seated in a large comfortable chair and then instructed to stand. Surface electromyography was recorded from the right leg and included the rectus femoris, biceps femoris, tibialis anterior, and medial gastrocnemius muscles. The right thoracolumbar paraspinal muscles were also recorded. All the data were recorded using the Porti-system from TMSi (Oldenzaal, The Netherlands). Recordings were for a duration of 30 s with a sampling frequency of 2048 Hz, analysis included a comparison between the first 15 s and the second 15 s to analyze for the effect of fatigue. Data acquisition was performed using a customized program written in MATLAB (Mathworks, Natick, Massachusetts, USA).
The conditions studied are outlined in Figure 1A. The tremor study was simultaneously video-recorded for blinded assessment of clinical tremor scores.

| Posturography analysis
Balance was measured using a Kistler force platform (9260AA6, Kistler Group, Winterthur, Switzerland). The participants were assessed in conditions 2, 3, 5, and 6 as per Figure 1A. Force in the anteroposterior, mediolateral, and vertical planes ( Figure 1B), as well as the center of pressure, was recorded for a duration of 15 s with a sampling frequency of 100 Hz.
The participants were instructed to stand as still as possible and to fix on a visual target 1 m away in the "eyes open" condition. A trial was discarded if the patient required external support to maintain their balance, and the condition was repeated. If two trials were unsuccessful, the condition was abandoned. The data were analyzed using Fast Fourier Transform (FFT) to examine for a dominant spectral peak in force generated through the platform. The center of pressure parameters were calculated according to previously published techniques. 11 Specifically, the path traveled by the center of pressure was measured by summating the minipaths between two consecutive sampling points using Pythagorean theorem, and the area of sway was calculated by summating the triangles formed by the mini path and the arithmetic mean point of the data ( Figure 1C). Additionally, the velocity of sway was computed by taking the distance over time of each mini path, and power spectrum density analysis was applied to examine for the peak velocity of sway. The anteroposterior direction has been shown to be the most reliable variable, 25 and so was the focus of the velocity analysis. Each parameter was also compared between eyes open and eyes closed conditions to assess the role of visual and somatosensory inputs. 26 All posturography assessments were performed blinded to the clinical findings and tremor analysis results.

| Data analysis
Statistical analysis was performed using R Statistical Software (version 3.6.3; R Foundation for Statistical Computing, Vienna, Austria).
Data were tested for normality using the Shapiro-Wilk test. The student t-test was used to compare means between groups in the case of parametric data, and the Wilcox test was used in the case of non-parametric data. Pearson's correlation coefficient was used to determine relationships between parametric variables and Spearman correlation if non-parametric. When comparing categorical variables, the Chi square or Fisher exact test was used. All the data are expressed as mean ± standard error of the mean (SEM) or median

| Patients
Twenty-five consecutive patients with CIDP were prospectively recruited. The mean age at assessment was 65.7 ± 2.7 years and BMI was 27.5 ± 0.7. The mean disease duration was 6.1 ± 0.7 years. The majority were receiving IVIg at the time of assessment (80%), while The center of pressure is sampled 100 times per second. Two adjacent sampling points are labeled p and p + 1. The line joining these points is the "mini path," all of which are summated to form the Sway Path. The AMP is the arithmetic mean point of all sampled points. The lines formed by connecting the AMP to sampling points p and p + 1 create triangles, the areas of which are then summed to form the Sway Area.
1 was treated with rituximab, 1 with mycophenolate mofetil and prednisone, and 3 were not receiving active therapy.

| Clinical findings
The CIDP cohort exhibited moderately severe neuropathy, as reflected by the clinical scores, with mean iRODS score 38.4 ± 1.1 (normal 48), mean MRC Sum Score 57.2 ± 0.7 (normal 60), mean INCAT Sensory score 7.9 ± 1.0 (normal 0), and NIS-LL score 22.7 ± 2.5 (normal 0). CIDP patients were further categorized into "good" (64%) and "poor" balance (36%) groups as defined by the Berg Balance Score (BBS). The mean BBS was significantly higher in the "good" balance (51.9 ± 1.0) compared with the "poor" balance (26.7 ± 3.4, p < .001, normal 56) group, as expected. The "poor" balance CIDP patients were older ("good" Age 60.4 ± 3.3, "poor" Age 75.0 ± 2.5, p = .002) although there was no significant difference in disease duration (Table 1). Those with "poor" balance exhibited greater disability as measured by the iRODS and NIS-LL scores (Table 1). Additionally, there was a strong correlation between BBS and age (R = À0.63, p < .001) as well as NIS-LL score (R = À0.67, p < .001). The 10 MWT and Rydel-Seiffer Score were significantly different between the two CIDP balance groups (Table 1)  There were no significant differences between the INCAT sensory or MRC sum scores and no difference in disease duration (Table 1). Separately, logistic regression analysis disclosed that age was a predictor T A B L E 1 Comparing CIDP patients with "good" and "poor" balance.

| Lower limb tremor in CIDP
Clinical assessment revealed that 32% of CIDP patients exhibited a postural lower limb tremor with legs outstretched, while three patients (12%) exhibited a tremor on standing. The TETRAS-LL score was 2.5 ± 0.3 overall; those classified in the "poor" balance group had a significantly higher TETRAS-LL score than those in the "good" balance group ("good" TETRAS-LL 2.1 ± 0.3, "poor" TETRAS-LL 3.6 ± 0.5, p = .037).
Surface EMG was most reliably recorded from the rectus femoris and tibialis anterior muscles. The mean spectral peak in each muscle during posture (legs outstretched) was comparable, being 12.5 ± 0.9 Hz  56, Figure 3A). Peripheral nerve conduction studies revealed that the tibial CMAP was significantly more likely to be absent in patients with tremor (Tremor 50% absent, No Tremor 12% absent, p = .037). Clinical measures of neuropathy severity did not significantly differ between patients with and without tremor (Table 2), except for the TETRAS-LL subscore which was significantly higher in patients with tremor ( Table 2).
Treatment with IVIg did not seem to exert a significant modulating effect on lower tremor (Table 2). Specifically, mean postural and orthostatic spectral peak frequency was not significantly different when assessed just prior to IVIg (trough) therapy and at the mid-point of the treatment cycle (peak). Similarly, clinical measures were not significantly changed following IVIg (Table 2).

| Balance findings
Force platform analysis disclosed a significantly higher sway path and area when assessing "eyes closed" in comparison with "eyes open" conditions (Table S1). Representative sway paths are shown in Figure S1A. Subgroup analysis revealed a significant increase in sway path and area in CIDP patients classified in the "poor" balance group, but only in the "eyes open" conditions (Table 1). This may be related to reduced sample size, as a significant proportion of CIDP patients classified in the "poor" balance group (55%, p = .002) were unable to complete posturography assessment in the "eyes closed" conditions.
Assessing ground reaction forces found comparable peak frequencies in the anteroposterior (0.7-1 Hz) and mediolateral (0.8-1.1 Hz) planes. There were also no significant differences between the peak sway frequency in the mediolateral or anteroposterior planes between the CIDP patients with "good" and "poor" balance ( Table 1).
There was therefore a dichotomy between those with lower frequency tremor (on surface EMG) associating with "poor" balance, and high frequency movements (on posturography) associating with "good" balance, depicted in Figure 3B.
The peak velocity of sway in the anteroposterior direction was also assessed. Two velocity peaks were evident, a lower (mean 1.1 ± 0.1 Hz) and higher (mean 3.0 ± 0.2 Hz) frequency peak. A representative plot is shown in Figure S1B. There was no significant difference in the velocity of sway between the CIDP patients with "good" and "poor" balance (Table 1).
Neurophysiology parameters were also compared between the "good" and "poor" balance groups. The common peroneal distal motor latency and leg conduction velocities were the only parameters that showed a significant difference between the two groups, sensory responses were not significantly different between the two groups (Table 1).

| DISCUSSION
Postural lower limb and slow orthostatic tremor was evident in approximately one third of CIDP patients and the presence of tremor associated with poor balance. These findings are novel in the CIDP cohort, for whom only upper limb tremor has previously been assessed. Further, posturography assessment disclosed a dominant high-frequency spectral peak in the vertical plane in 44% of CIDP patients, and it appeared to be an adaptive mechanism for maintaining balance.

| Pathophysiological mechanisms underlying tremor in CIDP
Upper limb tremor has been reported in 60%-80% of patients with inflammatory neuropathy, including CIDP, 2,3,27 and is termed neuropathic tremor. While upper limb neuropathic tremor was previously reported in the current CIDP cohort, 2 lower limb tremor has not been previously assessed in CIDP. The present study reports the novel finding of lower limb tremor evident in 32% of CIDP patients, being either postural or slow orthostatic (<12 Hz), with a small proportion (12%) exhibiting both types of tremor.
The mechanism underlying generation of neuropathic tremor is debated, although a complex interplay between peripheral nerve dysfunction and a central processor has been suggested. 27 Mistimed peripheral inputs were originally shown to drive tremor generation by confusing a central processor into producing corrective movements. 27 A recent study demonstrating an existence of a tremor frequency gradient along the length of the upper limb supported a partial role for mistimed peripheral inputs, although the absence of correlation between sensorimotor nerve dysfunction and tremor frequency suggested additional mechanisms. 2 The present study argues against a significant role of peripheral mistimed inputs in generation of lower limb tremor in CIDP given the absence of a tremor frequency gradient and lack of association between tremor and neurophysiological measures of peripheral nerve dysfunction.
A central generator has been reported as a potential mechanism for neuropathic tremor. 28,29 Lower frequency tremor (4-6 Hz) has been previously reported, 27,30 suggesting that the central generator may be located in either the cerebellum or midbrain. 9 Of relevance, cerebellar dysfunction has been implicated in inflammatory neuropathies, 28,31 as have thalamic abnormalities, 29 including reports of improvement following thalamic deep brain stimulation. 32

| Balance and tremor in CIDP
Balance is maintained by a combination of factors including ankle strength, visual processing, the vestibular system, somatosensory inputs, and the cerebellar system. 35 Lower limb tremor may also contribute to imbalance via central mechanisms. Specifically, lower limb peak tremor frequencies recorded in the current CIDP cohort were similar to tremor frequency recordings in Essential Tremor (ET), and dysfunction of cerebellothalamic networks has been implicated in postural imbalance mediated by lower limb tremor in ET. 8,9 Consequently, it is possible that cerebellothalamic networks may be implicated in postural imbalance evident in CIDP, although functional neuroimaging studies may be required to confirm this suggestion.
The velocity of sway in the present cohort also suggests the potential role of the cerebellum in poor balance. A prior analysis of patients with cerebellar atrophy reported two velocity peaks in anteroposterior sway, 25 at $1 Hz and between 2 and 3 Hz, contrasting with vestibular disease ($1 Hz peak). A delay in long loop cerebellar postural reflexes and increased reliance on the other mechanisms of balance were implicated as likely mechanisms. 25 In the present study,

| Adaptive mechanisms in imbalance
A proportion of CIDP patients exhibited a high-frequency peak ($16 Hz) on posturography when assessing forces in the vertical plane. Similar peaks are evident in primary orthostatic tremor and healthy controls when posture is perturbed. 40 A potential mechanism for this is an increased cortical drive to lower limb muscles, coupled with increased cortical activity in the beta band (15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30), as reported in experimentally induced imbalance in healthy controls. 41 Corticomuscular coherence studies also reported a shift towards cortical control in the 6-15 Hz and 16-40 Hz bands in healthy controls as the complexity of standing tasks increased. 42 Taken together, these findings suggest that increased cortical drive in a high-frequency band may be an adaptive response to postural imbalance. In the present study, high-frequency peaks were mostly evident in CIDP patients categorized as exhibiting "good" balance, supporting the notion of an adaptive response. Specifically, postural imbalance in CIDP may lead to increased cortical drive to lower limb muscles resulting in joint stiffening, increased stability, and improvement in balance. The cortical drive is possibly impaired in CIDP patients classified in the "poor" balance group, and elucidating the pathophysiological mechanisms that impair the cortical drive could be of therapeutic importance.
The present study has some limitations. The CIDP patients were