The Impact of Vocal Tremor on Deglutition: A Pilot Study

Abstract

Objective:

Vocal tremor (VT) poses treatment challenges due to uncertain pathophysiology. VT is typically classified into two phenotypes: isolated vocal tremor (iVT) and essential tremor-related voice tremor (ETvt). The impact of phenotypes on upper aerodigestive tract physiology during swallowing remains unclear. Qualitative and quantitative measures were employed to characterize tremor phenotypes and investigate the effects on swallowing physiology.

Methods:

Eleven ETvt participants (1 Male, 10 Female; $\overline{\mathrm{x}}$ age = 74) and 8 iVT participants (1 Male, 7 Female; $\overline{\mathrm{x}}$ age = 71) swallowed 20 mL boluses in cued and uncued conditions under standardized fluoroscopic visualization. Sustained/a/productions were captured to assess the rate and extent of fundamental frequency (F0) modulation. Penetration and Aspiration Scale (PAS) scores were obtained and swallowing biomechanics were captured using Swallowtail^™ software. Participants also completed the Swallowing Quality of Life (SWAL-QOL) questionnaire.

Results:

Hypopharyngeal transit was faster in both VT phenotypes compared with Swallowtail^™ normative reference data. Total pharyngeal transit times, however, were only faster in patients with iVT, relative to reference data. No significant differences were observed on the SWAL-QOL or PAS between tremor phenotypes. SWAL-QOL scores revealed that these patients rarely reported dysphagia symptoms.

Conclusions:

Subtle differences in swallowing patterns were observed across VT phenotypes, possibly related to adaptive mechanisms resulting in quicker pharyngeal bolus transit. Most patients did not report swallowing issues or dysphagia symptoms. This study is foundational for larger studies on this challenging population.

Level of Evidence:

4

INTRODUCTION

Voice tremor (VT) affects 1% of the global population,¹ is often treatment refractory, and results in substantive communication-related disability. VT poses pervasive treatment challenges due to a lack of clarity regarding distinct clinical phenotypes and putative variability in pathophysiology. Essential tremor (ET) is the most common disorder associated with VT in the literature, with laryngeal involvement occurring in 10%–62% of cases.^2-5 However, sparse data characterize the clinical presentation of essential tremor-related voice tremor (ETvt) in the context of deglutition. Further, recently updated classification criteria from the International Parkinson and Movement Disorder Society excluded isolated vocal tremor (iVT) as a clinical variant of ET. In contrast, the American Academy of Otolaryngology-Head and Neck Surgery employs the term laryngeal tremor, implying VT is isolated to the larynx. However, this term does not support clinical findings that tremor includes the intrinsic laryngeal muscles as well as the pharyngeal constrictors, tongue, soft palate, jaw, lips, and respiratory system musculature in this population.⁶ This nuanced discrepancy in classification adds complexity to the phenotypic landscape of ET, fueling debate regarding nomenclature and extent of tremor affecting upper aerodigestive tract structures and deglutition.

With regard to tremor and swallowing, Logemann and colleagues reported decreased esophageal transit in 10 patients with VT.⁷ The oral and pharyngeal phases of deglutition were described as “not functionally disruptive enough to cause patient complaints of dysphagia…”. Inclusion criteria were not specified, raising the possibility that these patients may have diverse tremor phenotypes. Regardless, little is known regarding the impact of tremor on swallowing physiology, particularly in the context of phenotypic variability.

The aim of this pilot study was to identify and characterize differences in swallowing timing between individuals with iVT, ETvt, and healthy controls. Given the precise and choreographed nature of deglutition, coupled with the impact of ETvt and iVT on upper airway structures, we hypothesize individuals with ETvt and iVT will present with variable swallowing-related measures during cued and uncued swallows. We sought to compare differences between tremor phenotypes and available normative reference data of neurotypical, geriatric adults. Specifically, we hypothesize differences between phenotypes and available reference data in oropharyngeal transit time (OPT), hypopharyngeal transit time (HPT), total pharyngeal transit time (TPT), and duration between complete airway closure relative to bolus arrival at the pharyngoesophageal opening (AC-PEO) in cued conditions. Within each phenotype, differences in timing parameters and severity of penetration/aspiration between cued and uncued swallows were characterized in addition to differences in self-reported Swallowing Quality of Life scores among individuals with different tremor phenotypes. Additionally, we postulate a relationship will emerge between the frequency/extent of tremor modulation and deglutative timing parameters. These data are foundational regarding the extent to which tremor influences oropharyngeal function during swallowing to guide a more full and well-designed study. Ultimately, we seek to enhance our understanding of clinical characteristics of VT in the context of deglutition, providing foundational insight to inform current clinical classification, and ultimately, to develop novel diagnostic and therapeutic strategies for this challenging patient population.

METHODS

This prospective study was approved by the NYU Grossman School of Medicine Institutional Review Board (S21-00005) and conducted at a tertiary laryngology practice with a high volume of tremor patients. Participants who met inclusion criteria were provided information about the study and consented in compliance with federal institutional regulations governing clinical research policies. Participants included 11 patients with ETvt (1 Male, 10 Female; $\overline{\mathrm{x}}$ age = 74) and 8 patients with iVT (1 Male, 7 Female; $\overline{\mathrm{x}}$ age = 71), described in Table I. Inclusion criteria included 18–90 years of age with the ability to provide informed consent, a consensus diagnosis of iVT or ETvt between otolaryngology and neurology, no pre-existing dysphagia from other known causes, no confirmed or suspected neurological conditions other than VT, and not pregnant or breastfeeding. Participants could have received prior treatment for tremor, but must have been outside of the therapeutic window (i.e., returned to baseline following Botox injection). A consensus diagnosis of iVT or ETvt was established using the criteria described by Torrecillas and colleagues.⁸ This determination was based on comprehensive medical history gathered from primary care physicians, neurologists, and otolaryngologists, specifying the location of tremor to distinguish phenotypes. Phenotyping for iVT and ETvt was supported by self-reported tremor locations collected through the Quality of Life in Essential Tremor Questionnaire.

TABLE I.

Participant Demographics.

	ETvt (N = 11)	iVT (N = 8)	Normative data (N = 74)
Age
Mean ± SD	74.2 ± 6	70.63 ± 7.07	71.69 ± 5.22
Min-Max	59–86	52–82	65–88
Sex (N)
Male	1	1	38
Female	10	7	36

Modified Barium Swallow Studies (MBSS) were performed under continuous fluoroscopy (Siemens Cios ALPHA C-arm) captured at a rate of 30 frames per second employing TIMS software (TIMS Medical, Chelmsford, MA). Participants swallowed two repeated 20 mL 40% weight/volume VaribarTM (Bracco Imaging) suspensions, representing thin liquid bolus presentations (IDDSI level 0). Participants swallowed under two conditions: preceded by a cued 5-second bolus hold and an uncued swallow. The 20 mL bolus was selected to ensure participants could maintain volumetric control, but also to potentially introduce deglutition challenge. Cued conditions were elicited to identify challenges with coordination compared with naturally-elicited (uncued) swallows. Participants were instructed to swallow the liquid in a single swallow and self-fed all liquids using a medicine cup. Slight protocol deviations were made for one participant (i.e. decreased bolus size in cued presentation from 20 to 10 mL) to maximize patient safety and yield of clinical information.

Participants also completed the Swallowing Quality of Life (SWAL-QOL) questionnaire. The SWAL-QOL, first described and validated by McHorney et al,⁹ is widely used to quantify the impact of dysphagia on quality of life. The instrument includes 44 items across 10 domains (burden, eating duration, eating desire, food selection, communication, fear, mental health, social, sleep, and fatigue), with each item scored from 0 to 4 (worst-best). Total SWAL-QOL score is calculated by adding each domain score and dividing by 10 to provide scores ranging from 0 (worst) to 100 (best).

A licensed Speech-Language Pathologist (SLP) with 5 years of experience in evaluating dysphagia rated each swallow using an 8-point ordinal penetration–aspiration scale (PAS)¹⁰ to quantify the depth of bolus invasion into the airway and any physiological responses to airway invasion. Swallowtail^™ (Belldev Medical, Arlington Heights, IL) was employed to analyze oropharyngeal transit time (OPT), hypopharyngeal transit time (HPT), TPT, and duration between complete airway closure relative to bolus arrival at the pharyngoesophageal opening (AC-PEO). Methods for timing analyses were developed by Leonard and Kendall^11,12 and defined in Table II. Timing measures from cued conditions were compared with normative reference data provided by Swallowtail^™ based on the research conducted by Leonard and Kendall.^11,13 Normative reference data included 74 (36 Male, 38 Female; $\overline{\mathrm{x}}$ age = 71.69) participants, all of whom swallowed 20 mL of thin liquid following a 5-second hold.

TABLE II.

Definitions of Swallowtail^™ Timing Parameters.

Timing measure	Operational definition
OPT	Duration of bolus transit through the oropharynx, from the bolus head’s first movement past the posterior nasal spine that leads to a swallow to the first exit of the bolus from the valleculae.
HPT	Duration of bolus transit through the hypopharynx, from the time the bolus head exits the valleculae to the time of bolus tail clearance of the pharyngoesophageal opening.
TPT	Sum of oropharyngeal and hypopharyngeal transit times.
AC-PEO	Duration between complete airway closure relative to bolus arrival at the pharyngoesophageal opening.

A licensed SLP recorded participants sustaining the vowel/a/at a comfortable pitch and loudness, following protocols by Patel et al.¹⁴ and Lester et al.¹⁵ Recordings were acquired using The Speech Lab system (CSL Model 4500B, PENTAX Medical), with the microphone placed 4–10 cm from the lips at a 45° angle and a sampling rate of 44.1 kHz. Recordings were edited to obtain the middle 2 s of the/a/production and analyzed for a rate of F0 modulation (in Hz) and extent of F0 modulation (in %) using algorithms published by Markus Brückl¹⁶ using Praat.¹⁷

Statistical analyses were conducted using GraphPad Prism (v.10.0; GraphPad Software, San Diego, California, USA). Nonparametric methods were used when the assumptions of homogeneity of variance and/or Gaussian distribution were not met. PAS and SWAL-QOL scores were compared between tremor phenotypes using Mann–Whitney tests. Timing parameters within phenotypes in cued conditions were compared with Swallowtail^™ normative reference data using Brown–Forsyth and Welch ANOVA tests with Dunnett’s multiple comparison tests. Timing parameters within phenotypes during uncued swallows were compared using unpaired t-tests with the option of Welch’s correction or Mann–Whitney tests. Paired t-tests or Wilcoxon tests were utilized to compare timing measures and PAS scores within each phenotype between individual cued and uncued swallows. Additionally, Pearson correlations were calculated to examine the relationship(s) between acoustic parameters of tremor (i.e., rate and extent of F0 modulation) and Swallowtail^™ timing parameters within each phenotype.

To establish inter-rater reliability of PAS scores and Swallowtail^™ timing parameters, 10% of the MBSS files were randomly selected and redundantly analyzed by a second licensed SLP with over 5 years of experience treating and evaluating dysphagia. PAS scores were dichotomized in a binary fashion (1–2 and 3–8); the scores of 1 and 2 represent functional safe swallowing. The scores of 3 or higher reflect clinically significant penetration and/or aspiration. Cohen’s kappa coefficient for PAS scores revealed perfect agreement on dichotomized scores (κ = 1) between raters. Intra-class correlation coefficients for Swallowtail^™ timing measures revealed strong agreement (r = 0.73–0.97) between raters.

RESULTS

SWAL-QOL

The distribution of measures for all parameters is presented in Table III. SWAL-QOL scores for ETvt participants ranged from 66 to 97 ($\overline{\mathrm{x}}=88.6$; SD = 9). For iVT participants, scores ranged from 79.9 to 100 ($\overline{\mathrm{x}}=92.5$; SD = 6.6). No significant difference was observed between phenotypes (U = 28; p = 0.2).

TABLE III.

Measure Distribution for Study Parameter Across All Participants.

	ETvt (N = 11)	iVT (N = 8)	Normative data (N = 74)
PAS cued
Median	1	1
Min-Max	1–8	1–2
PAS uncued
Median	1	2
Min-Max	1–7	1–3
SWAL-QOL
Mean ± SD	0.89 ± 0.09	0.93 ± 0.07
Min-Max	0.66–0.97	0.8–1.0
OPT cued (seconds)
Mean ± SD	0.5 ± 0.3	0.3 ± 0.17	0.48 ± 0.34
OPT uncued (seconds)
Mean ± SD	0.42 ± 0.03	0.24 ± 0.08
HPT cued (seconds)
Mean ± SD	0.72 ± 0.14	0.65 ± 0.14	0.87 ± 0.38
HPT uncued (seconds)
Mean ± SD	0.71 ± 0.08	0.71 ± 0.1
TPT cued (seconds)
Mean ± SD	1.21 ± 0.37	0.95 ± 0.18	1.41 ± 0.68
TPT uncued (seconds)
Mean ± SD	1.13 ± 0.32	0.95 ± 0.1
AC-PEO cued (seconds)
Mean ± SD	0.25 ± 0.18	0.13 ± 0.14	0.16 ± 0.29
AC-PEO uncued (seconds)
Mean ± SD	0.2 ± 0.17	0.17 ± 0.16

Penetration–Aspiration Scale

PAS scores for cued swallowing conditions in ETvt participants ranged from 1 to 8 (Med = 1) and 1 to 2 in iVT participants (Med = 1). For uncued conditions, PAS scores ranged from 1 to 7 for ETvt participants (Med = 1) and ranged from 1 to 3 in iVT participants (Med = 2). No significant differences were observed for cued (U = 34; p = 0.46) or uncued (U = 30; p = 0.26) conditions between phenotypes. Additionally, no significant differences were observed in PAS scores within ETvt (W = −4; p = 0.63) or iVT (W = 10; p = 0.31) phenotypes when comparing cued and uncued conditions.

Timing Parameters

Cued TPT was significantly faster in iVT ($\overline{\mathrm{x}}=0.95$; SD = 0.18) relative to reference data ($\overline{\mathrm{x}}=1.41$; SD = 0.68) (F[DFn, DFd] = 7.54(2, 35.68); p < 0.001; Figure 1A). Similarly, cued HPT was significantly faster in both ETvt ($\overline{\mathrm{x}}=0.72$; SD = 0.14) and iVT ($\overline{\mathrm{x}}=0.65$; SD = 0.14) phenotypes compared with reference data ($\overline{\mathrm{x}}=0.87$; SD = 0.38) (F[DFn, DFd] = 8.438 (2.000, 47.65); p = 0.03; p = 0.005; Figure 1B). No significant differences were observed for OPT (F[DFn, DFd] = 1.92 (2, 22.3); p = 0.17) or LVC-PEO (F[DFn, DFd] = 1.43 [2, 30.2]; p = 0.26) between phenotypes and reference data. No significant differences were observed in OPT (t[DF] = 1.9 [12.04]; p = 0.079), HPT (t[DF] = 0.16 [17]; p = 0.88), TPT (t[DF] = 1.7 [12.9]; p = 0.12), or LVC-PEO (U = 42; p = 0.89) between phenotypes during uncued swallows. Similarly, no significant differences were observed for timing parameters between cued and uncued conditions in either phenotype (Supplemental Table 1).

Fig. 1.

Bar graph representing mean and SD of differences in (A) total pharyngeal transit time and (B) hypopharyngeal transit time (in seconds) between ETvt (White), iVT (Black), and normative reference data (Gray). * = p < 0.05. ** = p < 0.01. *** = p < 0.001.

Acoustic Correlates

In ETvt, a significant negative correlation was observed between the rate of F0 modulation and HPT during uncued conditions (r = −0.65, p = 0.03), indicating a moderate relationship. In iVT, a significant positive correlation was observed between the extent of F0 modulation and OPT in cued conditions (r = 0.94, p = 0.002), indicating a strong relationship. All correlation coefficients are provided in Supplemental Table 2.

DISCUSSION

Vocal tremor affects approximately 1% of the global population,¹ presenting a considerable treatment challenge due to its often-refractory nature resulting in communication-related disability. The lack of clarity regarding distinct clinical phenotypes and underlying pathophysiological variability further complicates treatment approaches. As noted previously, only a single paper nearly 40 years ago documented mild swallowing deficits, primarily in the esophageal phase, in patients with ET.⁷ However, that study did not characterize the clinical phenotype of patients to determine the potential correspondence between phenotype, tremor severity, and swallowing biomechanics.

In the current study, we sought to characterize swallowing function across tremor phenotypes and investigate the potential interplay between tremor and deglutition. Of note, no patients reported overt dysphagia, and generally, SWAL-QOL scores ($\overline{\mathrm{x}}=0.90$ across all participants) and PAS scores (MED = 1 across all participants) were generally good. However, the oldest ETvt participant (86 y/o) had the highest PAS score (7), lowest SWAL-QOL score (66), and longest time between bolus entering the pharyngoesophageal opening and complete airway closure during cued (0.5 s) and uncued (0.46 s) swallows. Collectively, these findings suggest the possibility of an additive effect of tremor and advancing age. The prevalence of dysphagia in older adults ranges widely from 11% to 80% depending on clinical setting, assessment instruments employed, and dysphagia criteria.^18,19 Variations in swallowing biomechanics in this study underscore the need to discern between naturally occurring age-related changes in swallowing (i.e., Presbyphagia) and overt dysphagia associated with specific tremor phenotypes. Presbyphagia is characterized by subtle alterations in the swallowing physiology including slower oropharyngeal transit^20,21 and vestibule closure.^21,22 Although previous studies consistently reported age-related changes in swallowing, when accounting for the absence of other health conditions and medications, swallowing tends to remain preserved and nonpathological even into late age.^23,24 The participant in the current study may be an illustrative case among ETvt patients, hinting at potential heightened susceptibility to swallowing-related dysfunction. This observation highlights the importance of further exploring age-related factors in the manifestation, impact, and clinical management of VT on swallowing function. However, it is important to exercise caution generalizing from a single case, particularly considering the potential absence of available data on undisclosed comorbidities in this participant’s medical history.

Distinctive patterns emerged in the relationship between tremor characteristics and swallowing biomechanics across phenotypes. In ETvt, a faster tremor rate (indicating more frequent laryngeal oscillations during sustained phonation) was associated with faster hypopharyngeal transit in uncued swallows, potentially indicating a more efficient swallow. Conversely, individuals with iVT with greater extent of tremor modulation (signifying increased oscillatory movement of laryngeal structures during sustained phonation) tended to have slower OPT in cued swallows, suggesting subtle differences in oral coordination during swallowing. Interestingly, patients with iVT exhibited faster TPT in cued swallows compared with normative reference data, suggesting a potentially compensatory mechanism to improve swallowing efficiency. Both iVT and ETvt phenotypes had faster cued HPT compared with normative reference data. These variations may stem from differences in underlying neural mechanisms, motor coordination, and/or adaptive responses specific to each tremor phenotype.

Both cued and uncued swallows were included to identity potential differences between reflexive and volitional swallows within each tremor phenotype. Significant differences in swallowing timing parameters or airway infiltration were not observed between cued and uncued conditions within each phenotype. These findings putatively suggest reflexive and volitional swallows may share commonalities in neural control and coordination across phenotypes. From a clinical perspective, participants did not report swallowing problems. Subtle differences in swallowing biomechanics do not appear to functionally impact deglutition. These findings may still be relevant as indicators of early stage changes, but generalizability is limited due to the relatively small sample size. Additionally, displacement measures were not considered. Displacement measures related to UES narrowing and reduced pharyngeal constriction could provide insight into the bolus velocity findings in the current study. Future studies will incorporate displacement measures to enhance the understanding of swallowing dynamics. In addition, control data were obtained from a geographically distinct cohort, and the literature did not support a priori power calculations. However, this work is foundational for further investigation on a larger scale.

CONCLUSION

VT phenotypes exhibited subtle differences in swallowing biomechanical patterns, suggesting potential adaptive mechanisms that contribute to faster pharyngeal bolus transit. Additionally, tremor rate and extent of modulation may distinctly influence temporal aspects of deglutition across VT phenotypes. Variations in swallowing biomechanics among VT phenotypes underscore the potential distinct influence iVT and ETvt have on the upper aerodigestive tract. Despite these differences in swallowing patterns, the majority of patients did not report dysphagia symptoms. This study is foundational for larger studies on this challenging population.