Mitigating the Effects of Acute Vocal Exertion in Individuals With Vocal Fatigue

Abstract

Objectives/Hypothesis:

To investigate the effects of acute vocal exertion on individuals with vocal fatigue and to determine whether semi-occluded vocal tract exercises (SOVTEs) are more effective than vocal rest in mitigating acute effects.

Study Design:

Prospective, repeated-measures design.

Methods:

On consecutive days, 10 individuals (6 males, 4 females) with scores indicating vocal fatigue on the Vocal Fatigue Index completed two 10-minute vocal exertion tasks. Vocal rest or SOVTEs were interspersed in counterbalanced order between exertion tasks. Respiratory kinematic, acoustic, aerodynamic, and self-perceptual measures were collected at baseline, following vocal exertion, following SOVTE/vocal rest, and following the second exertion task.

Results:

Acute vocal exertion worsened phonation threshold pressure (P < .001) and vocal effort (P < .001) and reduced maximum fundamental frequency (P < .001). Speech was terminated at lower lung volumes following vocal exertion (decreased lung volume termination [LVT], P < .001). Exertion-induced changes in vocal effort and LVT were significantly reversed by both vocal rest and SOVTE. Detrimental changes in voice measures reoccurred following the second vocal exertion task. SOVTE and vocal rest protected against changes in respiratory kinematics when vocal exertion was resumed.

Conclusions:

Vocal exertion impacted laryngeal, respiratory, and self-perceptual measures in individuals with vocal fatigue. Both SOVTE and vocal rest partially mitigated changes in voice measures and prompted more efficient respiratory strategies that were maintained when vocal exertion resumed. These data increase our understanding of how individuals with vocal fatigue respond to vocal exertion tasks and offer preliminary guidance for optimal clinical recommendations.

Level of Evidence:

3

Editor’s Note:

This Manuscript was accepted for publication on May 07, 2021.

INTRODUCTION

Vocal fatigue is defined as self-perceptual or physiological deterioration in vocal function following prolonged voice use.^1-3 This voice complaint is common^4-6 and may be associated with vocal exertion.^7-10 In some individuals, vocal exertion (i.e. voice use that compromises optimal laryngeal function) results in chronic vocal fatigue, putting an individual at risk for voice disorders.¹¹ Although vocal exertion may induce changes in respiratory kinematics,^12,13 vocal effort,^14-16 aerodynamic,^17-19 and acoustic voice measures,^20-22 virtually all work has focused on vocally healthy individuals.²³ As such, this study sought to quantify and mitigate the effects of vocal exertion in individuals reporting frequent vocal fatigue.

In addition to mitigating the negative effects of vocal exertion in those with frequent vocal fatigue, it is also necessary to understand how efficient phonation might be facilitated when vocal exertion must be resumed. This is of clinical interest as vocal exertion may occur repeatedly in social and professional situations. Vocal rest is used to mitigate the effects of vocal exertion²⁴ as it may promote wound healing on a molecular level^25,26 and return aerodynamic and self-reported voice measures to near baseline levels.^10,17,19,27 Clinically, however, vocal rest may be impractical,^28-30 and the efficacy of short periods of vocal rest as frequently recommended in vocal hygiene protocols³¹ is relatively unexplored.³²

Emerging evidence suggests that semi-occluded vocal tract exercises (SOVTEs) may better mitigate the effects of vocal fatigue than vocal rest. Low-resistance SOVTEs allow for mobilization of laryngeal tissue, which ameliorates healing trajectory through attenuation of inflammatory response^33,34 and extracellular matrix synthesis.^26,35 Functionally, SOVTEs (e.g. straw phonation) can return measures such as phonation threshold pressure (PTP) and self-perceived vocal effort to baseline levels in healthy individuals.³² In addition, SOVTEs capitalize on source-tract interaction by promoting increased efficiency in airflow^36-38 and also reducing vocal effort.^39,40 Although there is evidence that both SOVTEs and vocal rest are beneficial,^19,26,27,32 it is unclear if these recovery strategies mitigate the effects of vocal exertion in individuals experiencing frequent vocal fatigue.

The first objective of this study was to investigate the effects of a 10-minute vocal exertion task on individuals reporting vocal fatigue. The Vocal Fatigue Index (VFI) was used to identify those experiencing frequent vocal fatigue.^41,42 Laryngeal and respiratory measures were selected based on previous sensitivity to exertion-induced voice changes^13,23 and mirrored those used in a companion paper examining healthy speakers.⁴³ The second objective was to compare the restorative effects of SOVTE and vocal rest. It was hypothesized that both regimens would mitigate the effects of exertion in the moment. The third objective was to determine whether SOVTE or vocal rest would have a protective effect when a second vocal exertion task was performed. It was hypothesized that SOVTEs would better protect against future vocal exertion.

MATERIALS AND METHODS

Participants/Design

The Purdue University Institutional Review Board approved all study procedures (IRB#1805020623). Ten young adults (Table I) were recruited for this study. All participants met the following criteria: 1) were over 18, 2) scored in the vocal fatigue range on the VFI (>24 on part 1, or >7 on part 2), 3) presented without gross laryngeal pathology on videostroboscopy, and 4) had normal spirometry scores. Exclusionary criteria included a history of 1) voice disorders, 2) asthma or respiratory disease, 3) head or neck cancer, and/or 4) formal vocal training.

Table I.

Population Demographics.

Subject Number	Sex	Age (yr)	VFI Part 1	VFI Part 2	SVC (% Predicted Value)	FVC (% Predicted Value)	Estimated BMI
1	F	19	12	8	94	95	22.7
2	F	18	24	3	96	99	27.4
3	F	18	24	2	80	80	20.2
4	F	19	24	3	82	84	16.4
5	M	18	14	9	82	83	19.5
6	M	27	25	4	88	84	21.5
7	M	18	27	6	93	93	19.7
8	M	20	13	8	98	95	28.5
9	M	25	24	2	86	80	25.8
10	M	21	24	3	87	86	20.9

VFI Part 1: Score >24 indicates vocal fatigue (scores indicating vocal fatigue in bold). VFI Part 2: Score >7 indicates vocal fatigue (scores indicating vocal fatigue in bold). SVC and FVC above 80% predicted values indicates normal score. BMI estimated from height and weight using Central for Disease Control BMI calculator found at https://www.cdc.gov/healthyweight/assessing/bmi/adult_bmi/english_bmi_calculator/bmi_calculator.html.

BMI = body mass index; FVC = forced vital capacity; SVC = slow vital capacity; VFI = vocal fatigue index.

This study employed a repeated measures research design. Participants completed two experimental sessions on consecutive days at similar times of day (±1 hour). On each day, participants completed two 10-minute vocal exertion tasks separated by SOVTE or vocal rest (Fig. 1). These strategies were performed on separate days (order counterbalanced). Laryngeal and respiratory measures were collected at baseline, after the first exertion task, after SOVTE/vocal rest, and after the second exertion task. Follow-up videostroboscopy was performed on both days to ensure no participant experienced laryngeal injury during the study.

Fig. 1.

Study design.

Vocal Exertion Task

Task development was guided by emerging evidence indicating that exertion-induced voice changes can occur quickly,^19,22,44 following sustained vowel productions.^18,43 The task consisted of 10 minutes of loud, sustained vowels held for maximum phonation time. Vowels were sustained for as long as possible, on one breath, at 50% pitch range (calculated in semitones using pitch glides to highest and lowest pitches sustained for 2 seconds). Target intensity was established by asking participants to produce vowels “twice as loud” as comfortable intensity. If below 80 dB sound pressure level (SPL) (as measured by sound level meter 24″ from the mouth), participants were asked to increase intensity until this threshold was reached. Target pitch, intensity, and vowel duration were established on practice productions before the task began. Verbal cues were given if productions fell below target values. Participants were cued to target pitch using a keyboard (Casio SA-76, Casio Computer Co. Ltd, Tokyo, Japan) prior to each production. Exertion task targets were identical across experimental days.

Recovery Strategies

SOVTEs comprised of 10 minutes of straw phonation (protocol by Ingo Titze^38,45 in “Vocal Straw Technique,” National Center for Voice and Speech). Training and practice occurred before data collection. Straw dimensions were 19 cm length and 0.5 cm internal diameter. Vocal maneuvers in this protocol included sustained phonation, ascending and descending pitch glides, and phonating “Happy Birthday.” Video instructions guided participants to perform each maneuver in repeating 30-second increments. Participants were instructed to vocalize in comfortable pitch range and to breathe as often as necessary. Cues were provided to avoid vocal strain and promote forward resonance. A researcher was present to ensure accuracy. Participants were allowed 8 oz of water.

Vocal rest training also occurred prior to data collection. Participants were instructed to refrain from all vocalization during this 10-minute protocol. Participants were instructed to engage in mindful breathing (breathing focused on expansion of the abdomen at a relaxed pace), which was practiced before data collection. One participant coughed briefly, but no other vocalization occurred during this time. Participants were allowed 8 oz of water.

Outcome Measures

Outcome measures included 1) PTP at 80% pitch range (PTP80), 2) maximum and minimum fundamental frequencies of frequency range (F0), 3) cepstral peak prominence (CPP) on connected speech, 4) self-perceived vocal effort, 5) lung volume initiation (LVI), 6) lung volume termination (LVT), and 7) number of syllables per breath group. PTP, vocal effort, and CPP were included because these measures have consistently demonstrated sensitivity to exertion-induced voice changes.²³ F0 range was utilized as emerging evidence suggests this measure correlates with overall vocal health.⁴⁶ There is a paucity of respiratory kinematic data following vocal exertion, but LVI and LVT appear most useful in this context.^12,13 Syllables spoken per breath group was included to allow for interpretation of lung volume findings.

Voice Measures

Phonation threshold pressure.

PTP80 was collected using a circumferentially vented, calibrated pneumotachograph mask with a 2″ plastic tube (Glottal Enterprises, Syracuse, NY). Airflow and oral pressure were collected at a 4 kHz/s sampling rate via low- and high-bandwidth differential pressure transducers, calibrated prior to data collection (MCU-4 Calibration system, Glottal Enterprises). A measurement system (MSIF-2 system, Glottal Enterprises) connected pressure transducers to a digital multichannel hardware system (PowerLab 16/30 ADInstruments, Colorado Springs, CO). Participants wore nose plugs during data collection. A metronome cued participants to produce strings of the syllable /pi/ at a 1.5 syllable-per-second rate—as softly as possible without whispering. At each time point, five strings of /pi/ were collected at 80% pitch range cued with a keyboard (Casio SA-76). Target pitch was determined prior to data collection using soft pitch glides. Oral airflow was required to be under 10 mL/s. Data were analyzed using Labchart (LabChart 16/30 ADInstruments, Colorado Springs, CO) by selecting the middle three /p/ peaks from five syllable strings. Peak pressures were averaged to find PTP80 for each string.

Acoustic measures.

A head-mounted microphone (AKG C555 L, AKG, Vienna, Austria; 2″ mouth to mic distance) routed through an A/D converter (PowerLab 16/30 ADInstruments, Colorado Springs, CO) was used to make audio recordings at a 44 kHz sampling rate. Prior to data collection, a 90 dB calibration tone (Acoustic calibrator Model QC-10/QC-20, Quest Technologies, Oconomowoc, WI) was recorded to allow for calculation of SPL.

• Cepstral peak prominence (CPP) was calculated on the second 2 sentences of the rainbow passage using the Analysis of Dysphonia in Speech and Voice program (Model 5109, KayPENTAX, Montvale, NJ). Mean SPL and F0 were also analyzed using Praat to facilitate interpretation of CPP.

• F0 range was obtained by asking participants to glide to the maximum and minimum F0s they could sustain for 2 seconds on the vowel /i/ (excluding vocal fry). Praat (version 6.1.16) was used to measure F0 range.

Vocal effort.

The Borg CR10 scale of exertion⁴⁷ was used to measure self-perceived vocal effort on the initial and final sustained vowels of the exertion task. The Borg scale was utilized because ratings from this scale correlate with auditory-perceptual and objective voice measures.⁴⁸

Respiratory Measures

Respiratory kinematics were collected using respiratory inductive plethysmography (Respitrace system, Ambulatory Monitoring, Inc., Ardsley, NY). Respiratory bands were placed around the rib cage (below the axilla) and abdomen (below rib 12).^49-51 Respitrace and audio signals were digitized synchronously using LabChart. A sampling rate of 1 kHz/s was used for respitrace system voltage.

The corrected sum of rib cage and abdominal displacements was used to estimate lung volumes.⁵² Correction factors were calculated during a period of rest breathing and “speech-like” breathing (breathing while silently reading the phrase “You buy Bobby a puppy now if he wants one” on each exhale)^49-51,53 while wearing nose clips. During these calibration procedures, rib cage and abdominal movements were recorded as well as lung volume as measured by a digital spirometer (FE141 ADInstruments, Colorado Springs, CO). The sum of rib cage and abdominal movements was compared to the volume output from the digital spirometer.^49,51,53-55 A least squares solution (Matlab pseudoinverse function) was used to determine calibration coefficients (k₁ and k₂) for rib cage and abdomen signals using the formula

$$\text{Spirometer}=k_1(\text{rib cage})+k_2(\text{abdomen}).$$ (1)

Lung volumes have been estimated within 5% accuracy from respiratory kinematic data using the least squares calibration method.^54,56 Lung volume was estimated during speech using these calibration coefficients in the formula

$$\text{Estimated lung}=k_1(\text{rib cage})+k_2(\text{abdomen}).$$ (2)

Maximum vital capacity was measured using the digital spirometer from three trials. The task required maximum inhalation followed by maximum exhalation.

Respiratory data collected during the rainbow passage were analyzed using a MATLAB algorithm (MathWorks, Inc., Natick, MA). Average end-expiratory level (EEL) was calculated from three troughs of rest breathing before each task.

• Syllables per breath group was the number of syllables produced on a single breath.

• LVI and LVT refer to lung volumes at which speech was initiated and terminated, respectively. LVI and LVT were measured from estimated lung volumes at the onset and offset of voicing for each utterance (identified using the acoustic signal) and were calculated as a percentage of vital capacity relative to EEL.

Statistical Analysis

Intra- and Inter-rater reliability values were calculated for 10% of data using two-way mixed, absolute single measure inter-class correlation coefficients (ICCs) computed in SPSS (Version 23, 2016). ICC values for intra- and inter-rater reliabilities were ≥0.90 for all voice measures and ≥0.88 for all respiratory measures, indicating good and excellent reliability, respectively.⁵⁷

Mixed linear models were performed using SAS (Version 9.4, 2013) for all measures. Fixed factors in the model were time (baseline, post exertion task 1, post recovery strategy, post exertion task 2), recovery strategy (SOVTE/vocal rest), and a time-by-recovery strategy interaction. Participant was included as a random factor to account for individual variation. Alpha level for significance was set at P < .01 to adjust for multiple comparisons, and Tukey adjustments were made for post hoc analyses. The repeated measure Cohen’s d calculator by Psychometric.de (https://www.psychometrica.de/effect_size.html) was used for significant post hoc tests.

RESULTS

Eight participants completed the vocal exertion task as designed. Two individuals reported vocal discomfort around the 5-minute marker and target pitch was reduced by three semitones to allow for completion of the full 10 minutes. For sustained vowels, average duration = 10.8 seconds (range 6.6–14.2), F0 = 269.3 Hz (range 145–390.4), and SPL = 90.4 dB (range 83.6–96.7). All participants completed SOVTE and vocal rest as described.

Voice Measures

Table II presents means and standard errors for all measures. Table III presents statistical summary of analyses. No significant effects were observed for CPP, minimum F0, LVI, or syllables per breath group. Only significant findings are discussed below.

Table II.

Means and Standard Errors for All Outcome Measures.

	SOVTE With Straw, Means (SE)				Vocal Rest, Means (SE)
Measure	Baseline	Post Exertion 1	Post SOVTE	Post Exertion 2	Baseline	Post Exertion 1	Post Vocal Rest	Post Exertion 2
PTP 80 (cmH20)	5.08 (0.50)	6.33 (0.59)	5.61 (0.50)	6.36 (0.60)	5.20 (0.35)	6.56 (0.67)	6.00 (0.63)	6.66 (0.69)
CPP (dB)	5.69 (0.30)	6.11 (0.38)	5.91 (0.33)	6.04 (0.33)	5.85 (0.37)	6.15 (0.39)	5.99 (0.33)	6.39 (0.32)
Maximum F0 (Hz)	715.9 (58.8)	622.5 (52.8)	671.4 (60.9)	646.8 (66.9)	688.1 (61.1)	591.4 (50.5)	605.9 (49.6)	601.2 (56.4)
Minimum F0 (Hz)	109.2 (12.9)	131.9 (17.5)	117.0 (16.2)	121.1 (18.6)	121.1 (14.9)	133.9 (18.4)	135.6 (18.3)	120.2 (16.8)
Vocal effort (Borg CR10)	2.5 (0.70)	6.90 (0.76)	3.14 (0.73)	8.13 (3.09)	2.41 (0.55)	7.09 (0.56)	2.72 (0.62)	6.95 (0.66)
LVI (%VC)	14.5 (2.05)	12.9 (2.08)	18.4 (2.26)	17.0 (1.8)	15.9 (2.3)	15.2 (2.7)	17.5 (1.9)	19.3 (2.8)
LVT (%VC)	−1.41 (2.2)	−3.59 (2.6)	3.28 (1.9)	1.19 (2.9)	−.73 (2.3)	−2.14 (2.7)	3.62 (1.3)	2.19 (3.4)
Syllables	16.1 (1.27)	15.3 (1.19)	15.5 (1.13)	15.5 (1.17)	17.1 (1.7)	15.0 (0.77)	15.6 (1.10)	17.5 (1.65)

SOVTE = semi-occluded vocal tract exercise; PTP = phonation threshold pressure; CPP = cepstral peak prominence; LVI = lung volume initiation; LVT = lung volume termination; %VC = percent vital capacity.

Table III.

Statistical Summary.

	Interaction Between Time and Recovery Strategy		Main Effect of Time (Exertion)		Main Effect of Recovery Strategy
Measure	F	P	F	P	F	P
PTP 80 (cmH20)	0.183,27	.910	9.983,27	<.001 *	0.551,9	.478
CPP (dB)	0.253,27	.857	2.023.27	.134	2.01,9	.191
Maximum F0 (Hz)	0.183,27	.905	2.403,27	<.001 *	7.341,9	.096
Minimum F0 (Hz)	1.213,27	.324	1.903,27	.153	3.091,9	.112
Vocal effort (Borg CR10)	0.343,27	.795	44.53,27	<.001 *	0.341,9	.575
LVI (%VC)	1.033,27	.383	1.083,27	.302	0.901,9	.444
LVT (%VC)	0.043,27	.989	5.373,27	.005 *	0.221,9	.652
Syllables	1.023,27	.394	1.003,27	.407	1.851,9	.182

^*Indicates statistical significance (significant values in bold)

PTP = phonation threshold pressure; CPP = cepstral peak prominence; LVI = lung volume initiation; LVT = lung volume termination; %VC = percent vital capacity.

PTP80.

A significant main effect of time (F_3,27 = 9.98, P < .001) was observed for PTP80 (Fig. 2). This measure significantly increased after exertion (t_3,27 = −4.59, P < .001, Cohen’s d = 1.63). This increase in PTP80 was partially reversed following SOVTE and vocal rest; data decreased to values between baseline (t_3,27 = −2.33, P = .116) and post exertion (t_3,27 = 2.27, P = .131). PTP80 again increased significantly above baseline following the second vocal exertion task (t_3,27 = −4.73, P < .001, Cohen’s d = 1.06).

Fig. 2.

Means and standard deviations for phonation threshold pressure at 80% pitch range at all time points.

F0 range.

A significant main effect of time (F_3,27 = 2.40, P < .001) was observed for maximum F0, which decreased following exertion (t_3,27 = 4.52, P < .001, Cohen’s d = 1.02). This decrease in maximum F0 was partially reversed following SOVTE and vocal rest as data increased to values greater than those observed post exertion (t_3,27 = −1.95, P = .231), but less than those observed at baseline (t_3,27 = 2.57, P = .074). Maximum F0 was again significantly truncated compared to baseline values following the second vocal exertion task (t_3,27 = 3.35, P = .012, Cohen’s d = .96).

Vocal effort.

A significant main effect of time (F_3,27 = 44.53, P < .001) was observed for vocal effort (Fig. 3). Ratings increased following exertion (t_3,49.04 = −6.41, P < .001, Cohen’s d = 1.98), significantly returned to baseline following both SOVTE/vocal rest (t_3,49.04 = 5.66, P < .001, Cohen’s d = 1.77), and significantly increased again following the second exertion task (t_3,49.04 = −6.35, P < .001, Cohen’s d = 1.97).

Fig. 3.

Means and standard deviations for vocal effort as measured by the Borg CR10 scale of exertion at all time points.

Respiratory Measures

Lung volume termination.

A significant main effect of time (F_3,27 = 5.37, P = .005) was observed for LVT (Fig. 4). LVT decreased following exertion (t_3,27 = 1.04, P = .027, Cohen’s d = .35) and increased above baseline following both SOVTE/vocal rest (t_3,27 = −3.67, P = .005, Cohen’s d = .79). No significant changes were observed following the second exertion task. LVI followed similar trends at all time points, but no significant main effect of time was observed.

Fig. 4.

Means and standard deviations for lung volume initiation and termination at all time points.

DISCUSSION

This study demonstrated changes in PTP, maximum F0, vocal effort, and LVT following vocal exertion. Both SOVTE and vocal rest strategies returned vocal effort to baseline and increased LVT above baseline levels. PTP and maximum F0 returned toward—but did not reach—baseline. Both strategies had a protective effect on LVT in the second exertion task.

These data suggest that exertion-induced changes can occur quickly in individuals reporting frequent vocal fatigue. The increase in PTP induced by vocal exertion suggests increased laryngeal viscoelastic properties, possibly secondary to mild exertion-induced laryngeal edema.^11,58 This may also explain the decrease in maximum F0, as increased laryngeal mass could restrict the thinning and elongation of the vocal folds required to elevate pitch.⁵⁹ This finding corroborates previous work associating vocal fatigue with difficulty navigating upper F0 range.^2,60 In fact, inability to produce high-pitched phonation may separate healthy and disordered voices.^60-62 No significant changes in CPP were observed, potentially because inflammatory responses were mild such that habitual speech was less impacted than more demanding voice abilities such as producing soft voice or upper F0 range. Further study may consider whether exertion-induced edema played a role in these findings. In addition, further investigation should determine whether these effects were specific to the vocal exertion task utilized in this study.

Individuals reporting frequent vocal fatigue also terminated speech at lower lung volumes following vocal exertion. Our previous study demonstrated that following the same task, vocally healthy individuals initiated and terminated speech at higher lung volumes,⁴³ potentially facilitating production of higher subglottal pressure without increasing reliance on muscular forces.^49,50 In contrast, individuals with vocal fatigue made no such adjustment despite increases in driving pressure for phonation (PTP). The lack of adaptation may lead to laryngeal strain as individuals strive to increase subglottic pressure without the benefit of increased lung volume.¹² It should be mentioned that all lung volume values observed in this study fell within normal range; additional work with larger sample sizes may confirm the significance and implications of these differences in respiratory strategy.

No significant differences in the effects of SOVTEs and vocal rest were observed. This finding differs from a previous report.³² The fact that both strategies worked similarly may reflect the short-duration exertion task.^17,19 It is also noteworthy that although both strategies produced improvements in PTP and maximum F0, these measures did not return to baseline levels. Further study could consider how SOVTEs might be optimized in individuals with vocal fatigue and whether certain SOVTEs are more beneficial than others.

Speakers reporting vocal fatigue experienced significant benefits in LVT and vocal effort with recovery strategies. Following SOVTE and vocal rest, these individuals initiated and terminated speech at lung volumes above baseline (although only the change in LVT reached statistical significance). Initiating speech at a higher lung volume is a more efficient respiratory strategy since greater recoil forces are available to support increased subglottic pressure.^49,50 This adaptation may have helped participants avoid compensatory laryngeal strain, thereby contributing to the improvement in self-perceived vocal effort as well. LVI and LVT declined again after the second vocal exertion task, but not to baseline levels, likely because they started at a higher lung volume post recovery. This finding supports the protective nature of vocal rest and SOVTE. This is plausible as 10-minute straw phonation has been shown to produce carryover effects on subsequent phonation.³⁷ Vocal rest was probably effective in part due to the nature of vocal fatigue, a condition ameliorated by rest.^2,5,63 These findings are promising for individuals experiencing frequent vocal fatigue; however, neither strategy protected against changes in voice measures.

CONCLUSION

Acute vocal exertion worsened PTP, maximum F0, and vocal effort and resulted in lower LVT in individuals reporting frequent vocal fatigue. SOVTE and vocal rest returned vocal effort to baseline and increased LVT, but PTP and maximum F0 were only partially improved. SOVTE and vocal rest had a protective effect on LVT once vocal exertion was resumed by raising LVT above baseline. Additional study is needed to consider the implications of the respiratory adjustments observed in this study and the manner in which SOVTE and vocal rest can best be employed with this at-risk population.