Essential Tremor and Essential Tremor Plus Are Essentially Similar Electrophysiologically

Abstract

Background

The merits of classifying the heterogeneous group of essential tremors into essential tremor (ET) and essential tremor plus (ETP) are debated.

Objectives

We studied the electrophysiological and spiral characteristics of tremor in ET and ETP.

Methods

We reviewed standardized videos from a tremor database and clinically classified patients into ET, ETP, or dystonic tremor (DT). The following variables were derived from combined tri‐axial accelerometry–surface electromyography (EMG)—peak frequency, total power, peak power, full width half maximum, tremor stability index and EMG‐coherence. We analyzed hand‐drawn spirals to derive mean deviation, tremor variability, inter‐, and intra‐loop widths. We compared these variables among the groups.

Results

We recruited 72 participants (81.9% male) with mean age 47.7 ± 16.1 years and Fahn‐Tolosa‐Marin Tremor Rating Scale total score 31.1 ± 14.1. Patients with ET were younger (P = 0.014) and had less severe tremor (P = 0.020) compared to ETP and DT. In ETP group, 48.6% had subtle dystonia. Peak frequency was greater in ETP (7.3 ± 0.3 Hz) compared to DT (6.1 ± 0.4 Hz; P = 0.024). Peak power was greater in ETP and DT for postural tremor. Rest tremor was recordable on accelerometry in 26.7% of ET. Other variables were similar among the groups.

Conclusion

Electrophysiological evaluation revealed postural tremor of frequency 6 to 7 Hz in ET, ETP, and DT with subtle differences more severe tremor in ETP and DT, and higher frequency in ETP compared to DT. Our findings suggest a similar tremor oscillator in these conditions, supporting the view that these entities are part of a spectrum of tremor disorders, rather than distinct etiological entities.

Essential tremor (ET) is a common movement disorder characterized by isolated bilateral upper extremity postural tremor. ¹ The 2018 consensus statement on the classification of tremors defines ET as a syndrome of bilateral upper limb action tremor in the absence of other neurological signs, present for at least 3 years, with or without tremor at additional sites. ² Some authorities now consider ET as a syndrome comprising more than one condition. The 2018 consensus statement proposed that patients with subtle neurological signs in addition to postural tremor should be classified as having essential tremor plus (ETP). ² There is an on‐going debate about the merits of classifying the heterogeneous group of ET patients into these distinct classes. ³ Specifically, it is not known whether this subclassification captures the phenotypic heterogeneity in ET or whether ETP represents an etiologically distinct category from ET, overlapping with atypical or subtle forms of other closely related tremor syndromes like dystonic tremor. Subtle neurological signs are also difficult to elicit reliably, even for experienced raters and this may account for modest inter‐rater agreements when applying the newer classification scheme in practice. ⁴ , ⁵

Developing objective biomarkers for soft signs like dystonia and ataxia is one approach toward more accurate classification. ⁶ Electrophysiological tremor analysis may offer another approach, as relatively distinct electrophysiological signatures are known to exist in various tremor syndromes. For instance, the frequency and positional features of tremor on combined surface electromyography (EMG)–accelerometric tremor analysis are characteristically different in parkinsonian tremor, ET, and cerebellar tremor. ⁷ For orthostatic tremor and psychogenic tremor, electrophysiological tremor analysis forms part of the diagnostic standard. ⁸ , ⁹ Furthermore, subtle features like tremor variability (TV) were shown to be different between patients with ET and tremor associated with dystonia. ¹⁰ Specifically, dystonic tremor is thought to be jerky and irregular, with variability in amplitude and frequency. ¹¹ Apart from electrophysiological tremor analysis, hand‐drawn spirals from patients with tremor may show features distinct to tremor syndromes. ¹² For instance, patients with tremor associated with parkinsonism draw small spirals with reducing interloop distances, whereas patients with cerebellar ataxia draw large spirals with increasing between loop distances. ¹³ Presence of a clearly appreciable spiral axis is considered an early feature of ET, whereas this feature is less commonly observed in dystonic tremor. ¹⁴ , ¹⁵ In view of the above, we sought to investigate whether distinct electrophysiological and spiral characteristics of tremor can be detected in patients classified clinically as ET or ETP.

Methods

We screened participants from a tremor database at the movement disorder clinic of the All India Institute of Medical Sciences, New Delhi. The database currently includes standardized video recordings, spiral drawings, and accelerometric/EMG recordings from 275 tremor patients recruited between November 2018 and July 2023. For this study, a single author (A.R.) screened the data for consecutive patients diagnosed with ET, ETP, or dystonic tremor (DT) and recruited in the database till April 2022. Standardized tremor video tapes of screened in participants were reviewed by three movement disorder specialists (K.B., A.L., and R.R.). Tremor was assessed in the following positions: at rest, posture 1 (hands outstretched and arms extended at elbows), posture 2 (arms abducted, elbows flexed), finger‐to‐nose testing and during writing, pouring tasks. Diagnosis of ET, ETP, or DT was ascertained according to the 2018 Movement Disorder Society consensus statement on tremor classification axis 1 definitions. ² We included patients with a consensus classification of ET, ETP, or DT by all three movement disorder expert raters. Participants with a diagnosis of DT were included as a comparator for analysis as previous studies have shown that DT versus ETP is a common ground for diagnostic uncertainty. ⁵ We were particularly interested in identifying any similarities that may exist between the ETP and DT cohorts. We excluded participants with a consensus alternative tremor syndrome diagnosis by the raters. Clinical severity of tremor was assessed using the Fahn‐Tolosa‐Marin Tremor Rating Scale (FTM‐TRS). ¹⁶ We used the Unified Dystonia Rating Scale (UDRS) to evaluate the severity of dystonia in patients with DT. ¹⁷

Subjects underwent tremor recording using combined accelerometry‐surface EMG with a sampling frequency of 2000 Hz and high pass filter 0.5 Hz. Medications for tremor, if any, were omitted on the morning of the study. Tri‐axial accelerometer (Biopac, Goleta, CA) was placed on the middle phalanx of the middle finger on the limb having more tremors visibly. In patients with bilaterally symmetrical tremor, the dominant upper extremity was recorded. Surface EMG electrodes were placed on the wrist flexor and extensor origins. Electromyography was recorded from the wrist flexors and extensors with sampling rate 2000 Hz and low pass filter 500 Hz, high pass filter 5 Hz. We recorded tremor in eight different positions‐rest (with and without mental arithmetic), posture 1 (arms stretched forward against gravity with and without mental arithmetic and with finger tapping), posture 2 (arms abducted with elbows flexed, with and without mental arithmetic), and after loading weight in posture 1. Time series data were stored for offline processing.

Data Analysis

We calculated the peak frequency (PF), total power (TP) (between 1 and 30 Hz) and peak power (PP) from the accelerometer time series data using power spectral density analysis individually for the three axes and for their resultant vector [√(x ² + y ² + z ²)]. The power spectrum was computed for fixed sliding time windows with a length of 2 seconds and 50% overlap using the Welch method. We used the first principal component derived from accelerometry data to calculate full width half maximum (FWHM) and tremor stability index (TSI). FWHM is the width of the distribution of the power spectral densities of the first principal component, corresponding to the point where the power was equal to half its peak value. ¹⁸ A smaller value for FWHM implies a narrower peak and, therefore, better spectral resolution. Tremor stability index signifies the stability of the tremor by quantifying the variability in its frequency over time. ¹⁹ We estimated TSI as described by Brittain et al (2015) ²⁰ the interquartile range of the distribution of the change in tremor frequency with respect to the instantaneous frequency of the first principal component. Electromyography data was pre‐processed using Butterworth high pass filter 1 Hz followed by full wave rectification before calculating the intermuscular coherence between flexor and extensor signals. ²¹ Scanned spiral images were processed using an in‐house algorithm to derive mean deviation (MD), TV, intra loop widths, and inter loop widths. Briefly, MD was calculated after radius‐angle transformation, as the total deviation from a best‐fit line for an ideal Archimedes spiral. The standard deviation (SD) of the total deviation from best fit represented TV. Loop width variables were calculated as the distance between adjacent loops in Cartesian co‐ordinates along four radii of each spiral.

Electrophysiological and spiral parameters were compared between the three groups ET, ETP, and DT. Log10 transformed values were used for comparisons of non‐normally distributed data. For each variable (PF, TP, PP, FWHM, TSI, MD, and TV) at four conditions (rest, P1, P2, and load) we performed one‐way analysis of variance (ANOVA) to detect any between‐group differences among ET, ETP, and DT. This was followed by multiple, pairwise comparisons using post hoc Bonferroni to identify which pairs were significantly different from each other. Differences were considered significant only if both between‐group and pairwise comparisons showed significant results. For the loop width variables, repeated measures ANOVA were used with FTM‐TRS for tremor severity as a co‐variate. Statistical analysis was carried out using Statistical Package of the Social Sciences (SPSS) software, version 20.0. (IBM, Armonk, NY) 20 (Chicago, IL).

Results

Between the period of March 2021 to May 2022, we screened videos of 80 participants with tremor. Seven participants were excluded (two tremor associated with parkinsonism, one indeterminate tremor, two neuropathic tremor and two patients had no apparent tremor in the video recordings). We included 72 participants with ET (n = 16), ETP (n = 35), or DT (n = 21). Clinical data and spiral parameters were available for all participants. Tremor analysis was performed in 58 (15 ET, 28 ETP, and 15 DT). Mean ± SD age of the study population was 47.7 ± 16.1 years, 59 (81.9%) were male and the mean disease duration was 9.5 ± 6.7 years. Mean total FTM‐TRS score was 31.1 ± 14.1. The mean UDRS score in the DT group was 7.5 ± 5.2. Baseline characteristics are shown in Table 1. Patients with ET were younger compared to the ETP and DT groups (P = 0.014). Greater proportion of ET and ETP were male, compared to DT (P = 0.002). In the ETP group 48.6% had subtle dystonia, whereas 22.9% had rest tremor. In the DT group, 33.3% had rest tremor on clinical examination. Tremor severity was lower in the ET group, compared to ETP and DT, although this was significant only on the FTM‐TRS part A (P = 0.020).

TABLE 1.

Baseline characteristics

Variable	ET (n = 16)	ETP (n = 35)	DT (n = 21)	P value
Age	39.8 (16.8)	47.0 (16.0)	55.0 (12.7)	0.014
Male sex, n (%)	15 (93.8)	32 (91.4)	12 (57.1)	0.002
Duration of symptoms, y	8.4 (8.9)	9.7 (6.1)	10.1 (5.8)	0.714
Subtle signs, dystonia	–	17 (48.6)	–	–
Subtle signs, rest tremor	–	8 (22.9)	–	–
Subtle signs, tandem ataxia	–	1 (2.9)	–	–
Subtle signs, others	–	4 (11.4)	–	–
FTM‐TRS total	25.8 (13.0)	30.5 (11.1)	36.3 (17.7)	0.071
FTM‐TRS part A	5.4 (4.1)	7.9 (4.5)	10.0 (5.5)	0.020
FTM‐TRS part B	11.8 (5.0)	13.9 (5.0)	16.6 (8.6)	0.073
FTM‐TRS part C	8.6 (5.8)	8.7 (4.6)	9.8 (5.7)	0.680

Note: Values represent mean (standard deviation).

Abbreviations: ET, essential tremor; ETP, essential tremor plus; DT, dystonic tremor; FTM‐TRS, Fahn Tolosa Marin Tremor Rating Scale.

Tremor Analysis

PF was greater in ETP (7.3 ± 0.3 Hz) compared to DT (6.1 ± 0.4 Hz; P = 0.024) when combined across all the four conditions tested (F[1, 2] = 3.954, P = 0.025). The PF was similar between ET (6.6 ± 0.4) and ETP (P = 0.408). The frequency difference between ETP and DT was not limited to a specific within‐subject condition and was a consistent overall group effect (Table 1, Fig. 1). Compared to the ET group, postural tremor at PF (PP) was greater in amplitude in ETP and DT at posture 1 (ET vs. ETP, P = 0.009; ET vs. DT, P = 0.024) and posture 2 (ET vs. ETP, P = 0.041; ET vs. DT, P = 0.009) (Table 1), consistent with the greater tremor severity in ETP and DT on clinical assessment using FTM‐TRS part A (Table 1). On accelerometry, using PP as a measure of tremor amplitude, 26.7% patients with ET, 21.4% with ETP and 46.7% patients with DT had rest tremor amplitude at least 50% of the postural tremor amplitude at posture 1 (P = 0.215). Total power, FWHM, and TSI were similar in all the groups and across conditions of rest, P1, P2, and load (Table 2). Electromyography coherence between flexors and extensors was similar among ET, ETP, and DT (Table 2).

FIG. 1.

Estimated marginal means of TP, PP, PF, FWHM, TSI, and EMG coherence in ET, ETP, and DT. Condition 1—rest, 2—P1, 3—P2, 4—load. Error bars represent ±2 standard error. DT, dystonic tremor; EMG, electromyography; ET, essential tremor; ETP, essential tremor plus; FWHM, full width half maximum; PF, peak frequency; PP, peak power; TP, total power; TSI, tremor stability index.

TABLE 2.

Tri‐axial accelerometry–surface EMG tremor analysis in ET, ETP, and DT

Variable	ET (n = 15)	ETP (n = 28)	DT (n = 15)	P value**	P value*
PF, rest (Hz)	6.7 (2.1)	7.7 (2.3)	6.2 (2.1)	0.075	0.025
PF, P1 (Hz)	6.9 (1.8)	7.3 (1.6)	6.0 (1.2)	0.031
PF, P2 (Hz)	5.7 (1.8)	6.8 (2.0)	5.6 (1.3)	0.083
PF, load (Hz)	7.2 (2.0)	7.3 (1.5)	6.4 (1.2)	0.219
Log10TP, rest	−6.4 (1.5)	−6.0 (1.2)	−5.8 (1.6)	0.667	0.076
Log10TP, P1	−6.1 (1.1)	−5.1 (1.3)	−5.1 (1.5)	0.045
Log10TP, P2	−5.7 (1.0)	−4.8 (1.4)	−4.5 (1.6)	0.032
Log10TP, load	−5.7 (1.3)	−4.7 (1.3)	−4.9 (1.4)	0.084
Log10PP, rest	−6.7 (1.4)	−6.2 (1.5)	−5.9 (1.8)	0.360	0.032
Log10PP, P1	−6.4 (1.3)	−5.2 (1.4)	−5.2 (1.6)	0.015
Log10PP, P2	−5.9 (1.2)	−4.8 (1.5)	−4.4 (1.8)	0.011
Log10PP, load	−6.0 (1.5)	−4.6 (1.3)	−5.0 (1.6)	0.020
FWHM, rest	2.8 (2.5)	2.3 (1.0)	2.1 (1.4)	0.234	0.052
FWHM, P1	2.9 (2.0)	2.6 (1.8)	1.9 (0.6)	0.168
FWHM, P2	2.0 (1.4)	2.1 (1.2)	2.3 (1.5)	0.371
FWHM, load	3.5 (2.2)	2.0 (1.5)	2.5 (1.8)	0.924
TSI, rest	13.7 (9.1)	11.3 (8.3)	10.8 (6.4)	0.060	0.076
TSI, P1	32.1 (47.4)	13.1 (17.6)	10.5 (10.5)	0.368
TSI, P2	13.0 (13.4)	9.5 (11.1)	5.4 (5.3)	0.125
TSI, load	26.0 (48.8)	13.7 (35.0)	9.6 (8.8)	0.104
EMG coherence, rest	0.31 (0.23)	0.29 (0.22)	0.30 (0.28)	0.587	0.702
EMG coherence, P1	0.23 (0.13)	0.18 (0.07)	0.22 (0.14)	0.381
EMG coherence, P2	0.17 (0.05)	0.18 (0.06)	0.22 (0.15)	0.235

^* P value for overall effects. P value denotes result of one‐way ANOVA.

^** P value for univariate analysis among groups at each individual condition.

Abbreviations: EMG, electromyography; ET, essential tremor; ETP, essential tremor plus; DT, dystonic tremor; PF, peak frequency; TP, total power; PP, peak power; FWHM, full width half maximum; TSI, tremor stability index.

Spiral Analysis

Mean deviation (ET: 37.3 ± 17.3, ETP: 41.9 ± 21.3, DT: 47.4 ± 22.7; P = 0.343) and TV (ET: 70.9 ± 25.0, ETP: 70.4 ± 9.7, DT: 74.8 ± 9.2; P = 0.524) were similar among the groups. There was no difference in the intraloop widths (Fig. 2) (F[2, 68] = 0.465, P = 0.630) and interloop widths (F[2, 68] = 0.465, P = 0.630) among the groups.

FIG. 2.

Interloop and intraloop widths in ET, ETP, and DT. DT, dystonic tremor; ET, essential tremor; ETP, essential tremor plus.

Discussion

We conducted a cross‐sectional study to explore the electrophysiological characteristics of tremor in a cohort of ET, ETP, and DT. In this study, tremor in both ET and ETP showed similar electrophysiological characteristics. Both groups of patients had a predominant postural tremor with mean frequency of ~6 to 7 Hz, which is within the typical frequency range of ET (5–10 Hz). Patients with ETP had higher frequency tremor compared to DT. ETP and DT groups had more severe postural tremor, both on clinical rating and power spectrum of the tremor on accelerometric analysis, which is a surrogate measure for tremor amplitude. Other tremor features such as TV (FWHM, TSI) and flexor‐extensor EMG coherence were similar among all groups. Spiral features—MD, TV, and loop widths were indistinguishable among the groups.

In this cohort, disease duration and postural tremor severity were greater in ETP compared to ET. This may support the argument that “plus” features appear along with the progression of ET, and depict a “state‐condition” of more advanced or widespread ET pathology than a “trait‐condition.” ³ , ²² Longitudinal studies or pathological studies of ET and ETP matched for disease duration are required to confirm or refute this observed association, which has also been reported in prior retrospective studies. ²³ , ²⁴ The higher tremor frequency in ETP compared to DT may partly be explained by the younger age of the ETP group; in both ET and physiological tremor frequency is known to reduce with increasing age. ²⁵

Although visible rest tremor necessitated a classification as ETP, on electrophysiological analysis, there were no significant differences in TP and PP at rest between ET and ETP. At rest, tremor could be recorded on accelerometer in equal proportions in all three groups of patients (more frequent in DT, although statistically insignificant); the amplitude of rest tremor was nearly an order of magnitude less than that of postural tremor. Importantly, 26.7% patients clinically classified as ET also had significant rest tremor recordable on accelerometry. The frequency of rest tremor was similar to the postural tremor in both groups. This suggests that minimal/subclinical rest tremor maybe present in patients clinically classified as ET, although the larger samples and healthy controls may need to be studied to confirm this observation. Of note, tremor irregularity has been previously shown to distinguish hand tremor in cervical dystonia from ET. ²⁶ In this study however, TV assessed using two different methods (FWHM and TSI) was similar in both groups.

Interestingly, TV and other parameters were similar to the other groups (ET, ETP) in the DT cohort as well. This is divergent to published data that shows a variable frequency and burst instability in DT. ¹⁰ The DT group was recruited predominantly for tremor, which was the prominent symptom; whereas the dystonia was more often picked up on examination and the UDRS scores suggested mild–moderate severity. The majority (nearly 50%) of our ETP group also comprised of patients with subtle dystonia. Taken together with these clinical similarities, our data suggests that the underlying electrophysiology and subsequently the substrates for tremor are similar in ET, ETP, and DT, at least when considering patients with prominent tremor in DT and patients with subtle dystonia related ETP. These results support a common mechanism and origin of tremor in all these conditions, which does not seem to be modified by the presence of additional clinical signs.

Clinical studies using the 2018 criteria to reclassify patients with ET have consistently found that over half of the patients now require to be reclassified as ETP. ²⁷ , ²⁸ , ²⁹ Such a distinction solely on clinical grounds is difficult to perform reliably, with overlapping classifications not only between ET and ETP, but also with DT. ⁵ Agreement on the presence of subtle neurological signs may be poor, even for expert raters. ⁴ The challenges posed by the new classification to case definition for clinical practice, epidemiology, clinical trials, and genetics have been extensively highlighted. ²² , ³⁰ Our results lend support to these concerns by showing that the electrophysiological characteristics of tremor are essentially similar in these two distinctly defined entities. One possibility is that the techniques we have used cannot adequately appreciate the clinical differences that we observed and further covert features may need to be explored, for instance, using machine learning techniques in larger tremor datasets. The stronger possibility is that ET, ETP, and DT are part of the same spectrum of tremor disorders, with less or more pronounced clinical differences, but essentially the same tremor features as detected by electrophysiology. Although encouraging the detection and systematic recording of subtle additional signs is an important outcome of the new classification, especially given the spectrum of motor and non‐motor manifestations now known to be associated with ET, it needs to be reconsidered whether the distinction into two unrelated categories on their basis is meaningful.

There are several limitations to our study. The sample size is small, and larger studies with longitudinal follow up are required to understand whether the subtle signs detected evolve into distinct electrophysiological entities over time. Because of limited sample size, data were not analyzed separately for different subcategories of ETP, such as those with subtle dystonia, ataxia, or rest tremor. Another limitation is the young age of the participants with ET; it would be of value in the future to extend these studies to an older group of patients. We performed single upper extremity recordings and some additional parameters such as coherence between limbs were not assessed. The methodologies used for tremor analysis at different centers vary and data from other cohorts using different recording and analysis methodologies may reveal additional differences not detected in this study.

Conclusion

Electrophysiological evaluation revealed a predominant postural tremor of frequency 6 to 7 Hz in ET, ETP, and DT. Subtle signs resulting in a classification of ETP seem to be more prominent in tremor of longer duration. Notwithstanding the presence of additional clinical features, electrophysiological, and spiral tremor characteristics are essentially similar among these three tremor syndromes.

Author Roles

(1) Research Project: A. Conception, B. Organization, C. Execution; (2) Statistical Analysis: A. Design, B. Execution, C. Review and Critique; (3) Manuscript Preparation: A. Writing of the First Draft, B. Review and Critique.

R.R.: 1A, 1B, 1C, 2A, 2B, 2C, 3A, 3B.

A.R.: 1B, 1C, 2A, 2B, 2C, 3A, 3B.

A.V.: 1C, 2A, 2B, 2C, 3A, 3B.

A.L.: 1A, 1B, 1C, 2C, 3A, 3B.

N.T.: 1C, 2C, 3A, 3B.

A.D.: 1C, 3A, 3B.

D.B.: 1C, 3A, 3B.

D.M.R.: 2C, 3A, 3B.

A.K.S.: 2C, 3A, 3B.

K.B.: 1A, 1B, 1C, 2C, 3A, 3B.

Disclosures

Ethical Compliance Statement: This study was approved by Institutional Ethics Committee No. IEC‐85/05.02.2021, RP‐13/2021 and registered in CTRI. Written informed consent was obtained from all subjects before recruitment. We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this work is consistent with those guidelines.

Funding Source and Conflict of Interest: The study was funded by AIIMS‐University College London Collaborative Research Project. The authors declare that there are no conflicts of interest relevant to this work.

Financial Disclosures for the Previous 12 Months: The authors declare that there are no additional disclosures to report.