Abstract
Introduction:
Assessing the severity of kinetic tremor is important in clinical and research settings. Archimedes spirals are often used to assess tremor severity. Rating tremor from spirals has been based solely on visual information. However, one can often “hear” tremor.
Objectives:
We evaluated the type of information that could be gained from hearing tremors and assessed its validity using visualized tremor as a gold standard.
Methods:
52 essential tremor cases (94 spirals), 11 Parkinson’s disease patients and 11 controls (20 spirals each) were enrolled. During videotaping, cases drew Archimedes spirals in a standardized manner. Three experienced movement disorders neurologists first listened to videotaped spiral drawing (no visual) and assigned a tremor severity rating using Washington Heights Inwood Genetic Study of Essential Tremor (WHIGET) ratings (0, 0.5, 1, 1.5, 2, 3). Later, the neurologist watched the same videotape (no sound) and assigned a WHIGET tremor rating.
Results:
The Spearman’s correlation coefficient (visual vs. auditory rating) was 0.721 (p<0.001) and the weighted kappa statistic was 0.674, a level of agreement that is considered good.
Conclusions:
Auditory information provides in-time synchronous corroboration of what the rater is seeing. That is, it informs the visual data and provides construct validity. The results of this study have ramifications for observational and experimental therapeutic studies, which commonly rely on videotaped tremor examinations. Additional corroborative information is available if the sound is left on. Regardless of the decision to present sound to the rater or not, clear instructions and a standardized practice within the study are imperative.
Keywords: Essential Tremor, clinical, severity, rating
Introduction
Essential Tremor (ET) is a chronic and progressive neurological disease. The hallmark clinical feature of ET is kinetic tremor [1]. Assessing the severity of kinetic and other types of tremor is important in both clinical and research settings. In assessing the severity of kinetic tremor, one task that is generally performed is the drawing of Archimedes spirals [2]. To date, the rating of kinetic tremor from spirals as well as the rating scales designed to assign such ratings have been based solely on visual information [3]. During a recent study, the senior author began to observe that he could also “hear” tremor during the Archimedes spiral test and that there seemed to be an audible difference between a tremor that was severe and one that was mild. To our knowledge, there have been no studies that have 1) examined and quantified audible tremor or 2) assessed whether the information provided is of similar value to that provided visually. We designed the current study to formally evaluate the type of information that could be gleaned from hearing tremors and to assess the validity of such information, using visualized tremor as a gold standard.
Methods
Overview
ET cases in these analyses were enrolled in a large parent study - an ongoing, prospective, longitudinal study of cognitive impairment in ET (Clinical Pathological Study of Cognitive Impairment in Essential Tremor [COGNET]; National Institutes of Health Award #R01NS086736) from 2014 to present. Eligibility requirements for COGNET were (1) a diagnosis of ET; (2) a baseline age of at least 55 years; (3) no history of brain surgery as treatment for ET; and (4) an agreement to become a future brain donor. The study was approved by the Yale University, Columbia University, and University of Texas Southwestern Medical Center Institutional Review Boards. All cases provided written, informed consent.
A trained COGNET research assistant visited participants for in-home evaluations. Each visit included demographic and clinical questionnaires, and videotaped neurological evaluation, which included an assessment in each arm of postural tremor, tremor while pouring, drinking, using a spoon, and performing the finger-nose-finger maneuver, as well as the drawing of two Archimedes spirals with each hand, as previously described [4].
Based on their history and examination, all cases received clinical diagnoses of ET assigned by an experienced movement disorders neurologist using Washington Heights Inwood Genetic Study of Essential Tremor (WHIGET) criteria [2], which are both reliable [5] and valid [6]. All cases also fulfilled Consensus criteria for ET [7].
The current study was initiated within the parent study on all enrollees evaluated between August 2023 and March 2024, with the goal of assessing approximately 100 spirals over a nine-month period. During that time period, we ascertained 94 spirals from 52 individuals (i.e., most enrollees contributed both a right and a left hand spiral).
Eleven Parkinson’s disease (PD) cases and 11 controls, enrolled in a study of the environmental epidemiology of tremor (2015 – 2020) [8], underwent the same evaluation as ET cases in COGNET, and were selected by the trained research assistant. The controls were selected randomly until we reached 20 spirals. The PD cases were selected to represent a range of action tremor severity during spiral drawing; here too, 20 spirals were obtained. PD diagnoses had each been assigned using London Brain Bank Criteria [9].
Piloting Procedures
Prior to formal initiation of the COGNET study, the research assistant piloted the procedure on 10 test subjects; these subjects were not included in the final analysis. We tested several different writing utensils (several types of pens, pencil), several different writing surfaces (several types of clipboards, flat hard surface) and several recording devices (video camera, iPhone). The procedure that provided the loudest and clearest sound for the recordings was a Sharpie s-gel black pen on flat surface (no clipboard) recorded with an iPhone 13.
Study Procedures
During each visit, demographic and clinical data were collected and a videotaped neurological examination was performed (see above). For drawing Archimedes spirals, participants were instructed to place the study pen on a designated point at the center of the blank 8.5 × 11 inch paper and create concentric circles slowly moving towards the outer edge of the paper, stopping at a designated endpoint. While drawing these spirals, participants were seated at a table, making sure they were steady and relaxed while completing the task. While participants completed the task, the paper was held in place by the trained research assistant administering the test to ensure the stability of the paper. Participants were instructed not to rest their wrist or arm on a flat surface while completing the task. The research assistant recorded the Archimedes Spiral task using an iPhone, which produced the clearest sound with the least background noise. The research assistant recorded the spiral with the device held 20 inches above the paper. The audio recorded from the videos were only used for assigning scores for audible tremor.
Spiral Analysis
There were three raters (E.D.L., V.S., and P.O.). For each rater, two separate grading sessions took place. First, a research assistant played the audio portion of the videos at full volume. Videos were randomized using a random digit table and presented to the neurologist in that order. The experienced movement disorders neurologists (E.D.L., V.S., and P.O.) could not see the visual part of the recording; this way, they only heard the sound of the recording. They wore headphones with full volume to properly hear every detail of the recorded audio. The recordings were replayed 2 – 3 times, if needed, to ensure confidence in the rating. While listening, the neurologist assigned WHIGET ratings - 0 (no tremor), 0.5 (very low amplitude tremor and almost never present), 1 (low amplitude tremor or intermittent tremor), 1.5 (moderate amplitude tremor and clearly oscillatory, but only sometimes of moderate amplitude), 2 (moderate amplitude and clearly and oscillatory and usually of moderate amplitude), 3 (large amplitude) and 4 (extremely large amplitude, unable to keep pen on paper) (see examples of each rating in Louis et al.[3]).
The second session, up to one month later, consisted of the rating the physical (i.e., visual) spiral by the same neurologist using the WHIGET scale. Spirals were randomized and de-identified such that they were presented in a different order than the auditory recordings.
Statistical Analyses
Spearman’s correlation coefficient was used to test the association between the numerical scores from the auditory and visual Archimedes spirals. In addition, the weighted Kappa coefficient was used to assess the agreement between the auditory and visual ratings, with values < 0.2 considered poor, 0.21 – 0.40 fair, 0.41 – 0.60 moderate, 0.61 – 0.80 good and 0.81 to 1.00 very good [10].
Results
The demographic and clinical characteristics of our cohort are shown (Table 1). A range of tremor severities (from WHIGET ratings of 0 through to 3) were present (Table 2) and Supplemental Tables (1 – 3).
Table 1:
Demographic and Clinical Information on Study Subjects
| Essential tremor | Parkinson’s disease | Controls | |
|---|---|---|---|
| n | 52 | 11 | 11 |
| Age (years) | 82.3 ± 4.4 (66 – 96) | 65.3 ± 10.6 (51 – 82) | 64.9 ± 12.2 (41 – 80) |
| Age of onset (years) | 39.0 ± 22.2 (4 – 76) | 57.0 ± 11.0 (42 – 73) | Not applicable |
| Tremor duration (years) | 43.1 ± 22.3 (6 – 87) | 7.9 ± 4.1 (1 – 14) | Not applicable |
| Total tremor score 1 | 23.0 ± 5.6 (11.5 – 36) | Not applicable | Not applicable |
Values are mean ± standard deviation (range)
The sum of 0 – 3 ratings in each arm of postural tremor and kinetic tremor (pouring, drinking, using a spoon, finger-nose-finger maneuver, and drawing of Archimedes spirals, range = 0 – 36.
Table 2:
Comparison of visual and auditory WHIGET ratings (all raters and all three diagnoses)
| Auditory WHIGET Ratings | |||||||
| 0 | 0.5 | 1 | 1.5 | 2 | 3 | ||
| Visual WHIGET Ratings | 0 | 23 | 10 | 4 | 2 | 0 | 0 |
| 0.5 | 23 | 14 | 13 | 0 | 2 | 1 | |
| 1 | 22 | 14 | 26 | 12 | 8 | 0 | |
| 1.5 | 5 | 5 | 23 | 15 | 14 | 2 | |
| 2 | 4 | 2 | 8 | 17 | 38 | 13 | |
| 3 | 0 | 0 | 0 | 2 | 25 | 43 | |
Concordant ratings are bolded. One of the raters was unable to rate a small number (n = 12) of spirals.
- 159 (40.8%) of 390 ratings were concordant, and 139 (35.6%) of 390 were in adjacent cells; hence, 298 (76.4%) of 390 ratings were either concordant or in adjacent cells.
- Spearman’s r = 0.772, p < 0.001.
- Weighted kappa is 0.751 (95% CI = 0.705 – 0.797).
- Archimedes spirals are often used to assess tremor severity.
- Ratings are usually based on visual information, but one can often “hear” tremor.
- A neurologist assigned tremor severity ratings based on auditory data.
- Agreement between auditory and visual ratings was very high.
- Auditory assessment of tremor has been underappreciated.
We compared the visual and auditory WHIGET ratings for all raters and all diagnoses (Table 2), and for each rater separately (Supplemental Tables 1 – 3) as well as right and left hands separately. Our measures of agreement were: concordance, Spearman’s correlation coefficient, and Kappa statistic. For concordance, we assessed whether visual and auditory ratings were “concordant” (e.g., both visual and auditory rating were “2”) or whether they were in an adjacent cell in the Table (e.g., visual rating was “2” and auditory rating was either “1.5” or “3”).
Overall, the majority (83.6%) of visual and auditory ratings were either concordant or in an adjacent cell; similarly, the Spearman’s correlation coefficient was high (r = 0.721, p <0.001) and the weighted kappa statistic was 0.674 (“good” agreement) (Table 2). When we assessed agreement between visual and auditory ratings for each rater separately, the data were similar (Supplemental Tables 1 – 3). We also assessed agreement between visual and auditory ratings in the right and left hands separately, and the data were similar: For the right hand, 80 (40.0%) of 200 ratings were concordant, and 83 (41.5%) of 200 were in adjacent cells; hence, 163 (81.5%) of 200 ratings were either concordant or in adjacent cells. Spearman’s r (on right) = 0.723, p < 0.001; and weighted kappa (on right) = 0.712 (95% CI = 0.634 – 0.789). On the left, data were similar: 79 (41.6%) of 190 ratings were concordant, and 81 (42.3%) of 190 were in adjacent cells; hence, 160 (84.2%) of 190 ratings were either concordant or in adjacent cells. Spearman’s r (on left) = 0.811, p < 0.001; and weighted kappa (on left) = 0.783 (95% CI = 0.731 – 0.836).
We also assessed agreement in ET alone vs. all three diagnoses combined, and the percent concordance, Spearman’s r and kappa values were similar (data not shown).
Visual examples of spirals, along with their corresponding auditory information, are provided for a range of tremor ratings (video).
Discussion
To date, the rating of kinetic tremor from spirals as well as the rating scales designed to assign such ratings have been based solely on visual information. On a conceptual level, auditory information, though present, has never been the focus of systematic scientific attention, and whether it has any real value, totally unknown.
The use of auditory information is not new in neurology. For example, during electromyographic (EMG) studies, the “diver bomber sound” can provide clues about the underlying diagnosis. In general, as humans, we are highly receptive to aural information. With this in mind, given the high illiteracy rates during his times, Shakespeare did not intend that his plays would simply be read in books; rather, they were composed so that they could be conveyed in the form of spoken words (i.e., to be heard in a theater). The aural experience enhances their comprehension.
The central aims of this study were first to raise awareness of and draw attention to the auditory information in spiral drawing. We demonstrated that auditory data can actually provide sufficient information for a neurologist to assign ordinal tremor severity ratings that are highly comparable to those assigned using the gold standard, i.e., visual information.
We are not proposing that auditory rating of spirals should replace visual ratings in research settings or clinical practice. We are making the point that auditory information provides in-time synchronous corroboration of what the rater is seeing. That is, it informs the visual data and provides construct validity. In assigning tremor scores, raters are generally trying to decide which of two adjacent ratings best applies, and the decision can be a difficult one filled with uncertainty and equivocation; the auditory information might be of value in swaying the rater towards one of the two ratings.
The results of this study have ramifications for observational and experimental therapeutic studies, which commonly rely on videotaped tremor examinations. What is apparent is that additional corroborative information is available if the sound is left on, offering an “advantage” to raters. Regardless of the decision to present sound to the rater or withhold it by turning the sound off, a clear set of instructions and a standardized procedure within the study design are imperative.
It should be acknowledged that the visualization of tremors can provide additional information, other than that of tremor severity. This includes information about the shape of the tremor waves, for example. Such information cannot be gathered from auditory assessments alone. That is to say, our study assessed the value of each type of rating in terms of assessing the severity of tremor alone. Furthermore, our study assesses one means of assessing tremor severity, the spiral; it is important to note that other methods are commonly used and are of value. Some of these can even generate sound (e.g. pouring water).
This study had a number of limitations. First, there were no WHIGET tremor ratings of 4. These are assigned when the study subjects is not able to keep the pen on the paper, generally due to a severe tremor in the z plane (in and out of the paper). This tremor severity is rare and did not occur in our cohort. Nevertheless, the sound produced is distinctive, and it is likely, although not proven, that auditory and visual information would have provided similar information. Second, we limited our assessment of kinetic tremor to the task of drawing an Archimedes spiral; it is highly likely that information of similar value could be obtained during other tasks such as the pouring of water between two glasses. Third, we used a reliable [5] and valid [6] tremor rating scale (WHIGET) [2], which we have used in numerous studies over the past 30 years, and one with a substantial to near perfect level of agreement to the Tremor Research Group Essential Tremor Rating Scale (TETRAS) [11; 12]. Although we could have used both scales, the added value of a second scale would likely have been negligible. Finally, future studies may wish to quantitatively analyze the recorded sound (i.e., objective sound analysis) as a potential way to expand the impact of audible information.
The study also had a number of strengths. First, to our knowledge, this was the first time this scientific question has been asked and our data allow us to address it. Second, tremor ratings were provided by three expert movement disorders neurologists. Third, we assessed patients with two tremor disorders, ET and PD, as well as controls, thereby not limiting ourselves to a single condition. Fourth, the WHIGET tremor rating scale was used; the scale is both reliable [5] and valid [6] and scores correlate highly with those of the TETRAS [11; 12]. Finally, the use of participants from the primary COGNET study allowed for a wide range of tremor severities that spanned from no tremor to severe tremor.
Supplementary Material
Visual examples of spirals, along with their corresponding auditory information, are provided for a range of tremor ratings. Use of headphones, with full volume, will allow the listener to properly hear every detail of the recorded audio. At the start of each visual example, there is some downtime before the subject starts drawing a spiral. Spirals begin at the following times: 0 at 00:05, 0.5 at 00:12, 1 at 00:27, 1.5 at 00:39, 2 at 00:51, and 3 at 1:17.00.
Funding
This work was supported by National Institutes of Health Award #R01NS086736. NIH played no role in the design, collection, analysis, or interpretation of the data, or in the writing of this manuscript. The authors declare that there are no conflicts of interest relevant to this work.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Ethical Compliance Statement
This study was approved by the University of Texas Southwestern Medical Center Institutional Review Board, approval numbers STU-2020-0565 and STU-2020-0564. Written informed consent was obtained from all participants.
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.
Declaration of interests
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
References
- [1].Louis ED, The primary type of tremor in essential tremor is kinetic rather than postural: cross-sectional observation of tremor phenomenology in 369 cases. Eur J Neurol 20 (2013) 725–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [2].Louis ED, Ottman R, Ford B, Pullman S, Martinez M, Fahn S, and Hauser WA, The Washington Heights-Inwood Genetic Study of Essential Tremor: methodologic issues in essential-tremor research. Neuroepidemiology 16 (1997) 124–33. [DOI] [PubMed] [Google Scholar]
- [3].Louis ED, Zhao Q, Meng H, and Ding D, Screening for Action Tremor in Epidemiological Field Surveys: Assessing the Reliability of a Semi-Quantitative, Visual, Template-Based Scale for Rating Hand-Drawn Spirals. Tremor Other Hyperkinet Mov (N Y) 2 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
- [4].Berry DS, Cosentino S, and Louis ED, A prospective cohort study of familial versus sporadic essential tremor cases: Do clinical features evolve differently across time? J Neurol Sci 454 (2023) 120854. [DOI] [PubMed] [Google Scholar]
- [5].Louis ED, Ford B, and Bismuth B, Reliability between two observers using a protocol for diagnosing essential tremor. Mov Disord 13 (1998) 287–93. [DOI] [PubMed] [Google Scholar]
- [6].Louis ED, Wendt KJ, Albert SM, Pullman SL, Yu Q, and Andrews H, Validity of a performance-based test of function in essential tremor. Arch Neurol 56 (1999) 841–6. [DOI] [PubMed] [Google Scholar]
- [7].Bhatia KP, Bain P, Bajaj N, Elble RJ, Hallett M, Louis ED, Raethjen J, Stamelou M, Testa CM, Deuschl G, P. Tremor Task Force of the International, and S. Movement Disorder, Consensus Statement on the classification of tremors. from the task force on tremor of the International Parkinson and Movement Disorder Society. Mov Disord 33 (2018) 75–87. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [8].Louis ED, Meyers JH, Badejo OM, Cristal AD, Hickman R, and Factor-Litvak P, Comparative Burden of Subclinical Tremor in a Cohort of Normal Individuals Stratified by Familial Risk for Essential Tremor. Neuroepidemiology 50 (2018) 41–46. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [9].Hughes AJ, Ben-Shlomo Y, Daniel SE, and Lees AJ, What features improve the accuracy of clinical diagnosis in Parkinson's disease: a clinicopathologic study. Neurology 42 (1992) 1142–6. [DOI] [PubMed] [Google Scholar]
- [10].Cohen J, Weighted kappa: nominal scale agreement with provision for scaled disagreement or partial credit. Psychol Bull 70 (1968) 213–20. [DOI] [PubMed] [Google Scholar]
- [11].Elble R, Comella C, Fahn S, Hallett M, Jankovic J, Juncos JL, Lewitt P, Lyons K, Ondo W, Pahwa R, Sethi K, Stover N, Tarsy D, Testa C, Tintner R, Watts R, and Zesiewicz T, Reliability of a new scale for essential tremor. Mov Disord 27 (2012) 1567–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [12].Hernandez B, Berry DS, Grill N, Hall TM, Burkes A, Ghanem A, Sharma VD, and Louis ED, WHIGET and TETRAS Ratings of Action Tremor in Patients with Essential Tremor: Substantial Association and Agreement. Tremor Other Hyperkinet Mov (N Y) 14 (2024) 14. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Visual examples of spirals, along with their corresponding auditory information, are provided for a range of tremor ratings. Use of headphones, with full volume, will allow the listener to properly hear every detail of the recorded audio. At the start of each visual example, there is some downtime before the subject starts drawing a spiral. Spirals begin at the following times: 0 at 00:05, 0.5 at 00:12, 1 at 00:27, 1.5 at 00:39, 2 at 00:51, and 3 at 1:17.00.