Treatment outcomes for professional voice users

MISSING IMAGE

Material Information

Title:
Treatment outcomes for professional voice users
Physical Description:
xii, 111 leaves : ill. ; 29 cm.
Language:
English
Creator:
Wingate, Judith Maige
Publication Date:

Subjects

Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 2004.
Bibliography:
Includes bibliographical references.
Statement of Responsibility:
by Judith Maige Wingate.
General Note:
Printout.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 003163662
oclc - 664556868
System ID:
AA00003588:00001

Full Text











TREATMENT OUTCOMES FOR PROFESSIONAL VOICE USERS


By

JUDITH MAIGE WINGATE
















A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2004



























Copyright 2004

by

Judith Maige Wingate
































This document is dedicated to my daughters, Lauren and Jennifer. Always remember to
follow your dreams.















ACKNOWLEDGMENTS

The journey to reach this point in my academic career has been a long one. There

have been many helpers along the journey and it is impossible to acknowledge each of

them individually. The primary person deserving the most thanks is my husband, Owen.

He has consistently believed in me and given me the support and freedom to pursue my

goals. My daughters, Lauren and Jennifer, have also been a tremendous source of

encouragement, inspiration, and humor along the way.

My mentor, Chris Sapienza, has been an inspiration for many years. She was

instrumental in encouraging me to return to graduate school and to become involved in

research. Without her support and boundless energy, my completion of this program

would not have been possible. She has provided the motivation to achieve more than I

thought possible.

Rahul Shrivastav provided invaluable technical support with computer equipment

and programming of various software programs. He has given an extra measure of time

whenever his help was needed and for that I am exceedingly grateful. My other

committee members, W.S. Brown and Paul Davenport, have also provided instruction

and support whenever needed.

Anuja Chhabra provided the therapy for all of the subjects in this study. I

appreciate her efforts and the quality of care that she provided to the participants. I also

wish to acknowledge my colleagues in the voice lab, both past and present, which have









provided encouragement and assistance whenever needed, especially Susan Baker, Bari

Hoffman-Ruddy, Erin Pearson, Jaeock Kim, Karen Wheeler, and Chris Carmichael.














TABLE OF CONTENTS

page

A C K N O W L E D G M E N T S .............................................................................................. iv

LIST OF TABLES ................ ... ......... ................. viii

LIST OF FIGURES .................... ... .... .. ... ..... .................... ........ .... ix

A B S T R A C T ............................................................... x ii

INTRODUCTION AND REVIEW OF THE LITERATURE .................................... 1

Incidence of V o ice P problem s ............................................................................ .... 4
Consequences of V oice Problem s .................... ............. ...... ........................ ... 6
Types of Vocal Pathologies Found in Professional Voice Users ...................... ... 8
Compensations for Vocal Pathologies ................................................................ 9
Assessment for Voice Problems ..................................................... .... ........ 10
Treatments for Voice Problems ............... .................................................... 13
Role of Respiratory System in Voice Production .............................................. 16
M echanism s for V ariation of V ocal Intensity......................................................... 17
Respiratory M uscle Training ................ .............. .... ............ ............. .. 18
Pressure Threshold D evice ................. .................. ...... ........... ............ .. 20
Expiratory Training Studies.................................................................................... 21
Statem ent of the Problem ...................................................... 23
Purpose of the Study ................... ............................................. .. ... .. 23
H ypotheses ............ .................. ................... .................... 24

M E T H O D O L O G Y ............................................................................ ............ 25

Experim ental D esign ...................................... ................... ...... ........... 25
Participants ................... ................... ..................................... 25
M measures ............. ............................................................ ....... 29
Screening Measures .............. .................... ................... ........... 29
Perceptual M measures ............ ..................................................... 29
Voice M measures .............. ... ...... .. ............. .............. ................ 31
Pulmonary M measures ................... ....................... .............................. 32
Aerodynamic Measures ................ ......................... .. ...................... ....... 33
A acoustic M measures ............. ................... ................................................... 34
Training Protocol ............. ................... ...... .......... ........ 36
C o m p lia n c e .................................................. .. ........ ......................... 3 8
Statistical M ethod ......... ............................................................. 38









RE SU L T S ............................................. 40

R liability ............................... ................................................................... 4 1
Perceptual Measures of Effort and Handicap ..................................................... 42
Voice M measures ................................................................. .... ........ 45
Pulmonary and Aerodynam ic M easures.................................. ........ ....... ..... 50
A acoustic M measures ........... .. .............................................................................. 51
Differences between Lesion and Non-lesion groups............ .......... .......... .. 54

DISCUSSION ................... ................... ..............................6

Perceptual M measures of Effort and H handicap .................................... ................. 56
Voice M measures ................ .............................................. .... ........ 60
Pulm onary and A erodynam ic M easures................................................. ............ 63
Acoustic M measures ............................... .. .............................. ...... ........ 66
Differences Between Lesion and Non-Lesion Groups.................. .................... 68
Combined M odality Treatment............................................. .......... 69
Strengths of the Present Study .............................. ..................... 70
Limitations of the Current Study ............ ........... ..................... .............. 70
Application of the Protocol............................... ...... ... ........... 71
Future Studies ....... .......... .. .... ............................................. 71

IN FORM ATION FLY ER .... .... ...7... ........... ......... ........ ... .. ...... .............. 73

SCREENING QUESTIONN AIRE .......... ................... ....................... ............ 74

VOICE HANDICAP INDEX ............ ................ .......... ........................... ....... 76

V O IC E R A T IN G SC A L E .................................................................. ....... ........... 79

ESTIMATE OF VOCAL EFFORT ............................... ................ ................. 81

STROBOSCOPY ASSESSMENT FORM.................................................... 82

TH E R A PY PR O T O C O L ........................... ....................... ........... 83

RESPIRATORY MUSCLE TRAINING PROGRAM ................. ............................ 99

T R A IN IN G L O G ......................................................... ............. ............ 10 1

LIST OF REFERENCES .............. .... ............... ................. .. ........ 102

BIOGRAPHICAL SKETCH ................ ..... ..... ............... 111








vii















LIST OF TABLES


Table page
Table 1-1. Summary of expiratory muscle strength training programs ....... ....... 21

Table 2-1. Demographic information for participants in the study...... ............28

T able 3-1. Intra-judge reliability......................................................... .....41

T able 3-2. Inter-judge reliability......................................................... ...42

Table 3-3. Correlations for VHI and VRS scales as well as VRS to functional and
emotional subscales of the VHI ............... .............................. ....... 45

Table 3-4. Rater evaluations of stroboscopic evaluations pre to post-treatment........49

Table 3-5. Effects of lesion on dependent variables .....................................55















LIST OF FIGURES


Figure page

Figure 2-1. Expiratory pressure threshold training device........... ............. 37

Figure 3-1. Mean Voice Handicap Index Scores before treatment, at the mid-point, and
following treatment..................................... ................... ... ......... 43

Figure 3-2. Individual VHI scores before treatment, at the mid-point, and following
treatment ...................................... .. ..................................... 44

Figure 3-3. Vocal Rating Scale scores before treatment, at the mid-point, and following
treatm ent ...................................... ....................................... 44

Figure 3-4. Individual Vocal Rating Scale scores before treatment, at the mid-point, and
following treatment........... ..................... ................... ......... 44

Figure 3-5. Changes in mean rating of voice quality pre-treatment, at the mid-point, and
follow ing treatm ent ......... ..... ............... .... ............................. ...... 46

Figure 3-6. Listener ratings for individual subjects before treatment, at the mid-point,
and follow ing treatm ent.................................................... .. ......... 46

Figure 3-7. Pre and post-treatment endoscopic image for participant 10.................48

Figure 3-8. Pre and post-treatment endoscopic image for participant 12.................48

Figure 3-9. Stroboscopic ratings pre and post-treatment for left and right vocal fold
edges ............. ............. .................... ............... 49

Figure 3-10. Stroboscopic ratings pre and post-treatment for left and right vocal fold
m ucosal w ave.................................................. ........................... 49

Figure 3-11. Maximum expiratory pressure changes before and after treatment...... 50

Figure 3-12. Individual changes in maximum expiratory pressure pre and
post-treatment................................ ................................... ........ 51

Figure 3-13. Change in phonetogram area following treatment.............. ........ 52










Figure 3-14. Changes in phonetogram area for individual participants pre, mid, and
post-treatm ent .......................................... 53

Figure 3-15. Pre-treatment phonetogram for participant 7 ........... .. ........... 53

Figure 3-16. Post-treatment phonetogram for participant 7............................ 53

Figure 3-17. Average dynamic range pre-treatment, at the mid-treatment point, and
following treatment ................... .................. ........................ .... ... 54

Figure 3-18. Changes in dynamic range for each participant pre, mid, and post-
treatm ent ................... ... .............. ....................................... 54

Figure 3-19. Estimated subglottal pressure at loud intensity for non-lesion and lesion
g ro u p s ................... .. ................. .. ... ................................... 5 5















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

TREATMENT OUTCOMES FOR PROFESSIONAL VOICE USERS


By

Judith Maige Wingate

December 2004

Chair: Christine Sapienza
Major Department: Communication Sciences and Disorders


Professional voice users comprise 25 to 35% of the United States working

population. Voice problems in this population can lead to voice quality change, lost work

days due to inability to speak, and workers' compensation claims. The estimated cost of

voice problems for teachers per year, a subset of this group, is estimated at two billion

dollars. The purpose of this study was to examine treatment outcomes of two specific

rehabilitation programs for a group of professional voice users.

Eighteen professional voice users participated in this study. Half of the participants

received five weeks of expiratory muscle strength training followed by six sessions of

traditional voice therapy. Half of the participants were diagnosed with benign vocal fold

lesions while half had complaints of throat pain or fatigue when speaking. The treatment

order was reversed for the remaining half of the group. The study was designed as a

repeated measures study. The independent variables were treatment order, laryngeal









diagnosis (lesion vs. non-lesion), gender, and time. Dependent variables were maximum

expiratory pressure (MEP), Voice Handicap Index (VHI) score, Vocal Rating Scale

(VRS) score, Voice Effort Scale score, measures of fundamental frequency and intensity

from the phonetogram, measures of irregularity and noise from the Hoarseness diagram,

phonation threshold pressure, subglottal pressure, and perceptual rankings of voice.

Results showed significant improvements in MEP, VHI scores, and VRS scores,

subglottal pressure for loud intensity, phonetogram area, and dynamic range. No

significant difference was found between laryngeal diagnosis groups. A significant

difference was not seen in treatment order. It was concluded that the combined treatment

was responsible for the improvement seen in the dependent variables. The results

indicate that a combined modality treatment may be successful in the remediation of

vocal problems for professional voice users.













CHAPTER 1
INTRODUCTION AND REVIEW OF THE LITERATURE

Speech and voice are the primary tools of trade for many occupations. Based on

conservative estimates, professional voice users comprise between 25 and 35 percent of

the working population in the United States (Titze, Lemke, & Montequin, 1997) and at

least a third of the work force in other industrialized societies (Vilkman, 2000). This

number is likely to increase with the escalating use of voice recognition systems for data

entry, word processing, and general computer control. Therefore, prevention and

treatment of voice problems in workers have social and economic significance, not only

for the professional voice user but for employers as well.

Professional voice users are individuals whose livelihoods depend partially or

wholly upon the ability to produce voice. Professional voice users may include, but are

not limited to, teachers, ministers, salesmen, telemarketers and telephone operators,

actors, singers, radio and TV announcers, and attorneys. Although the range of vocal

sophistication and quality needed may vary greatly across the range of occupations,

professional voice users share a dependence on vocal endurance (Benninger, Jacobsen, &

Johnson, 1994; Sataloff, 1997). Those professionals with persistent voice disorders may

experience deleterious effects on job performance, income and career longevity,

particularly if unable to speak.

Constant voice use by professional voice users may lead to vocal difficulties. With

sufficient recovery time, these vocal difficulties may improve but the professional voice

user typically has to return to work the next day and the cycle of repair and injury is









repeated. Other factors in the work environment may also contribute to vocal difficulties,

including high levels of background noise, poor environmental acoustics, and poor

atmospheric humidity (Carding & Wade, 2000; Ohlsson, Jarvholm, Lofqvist, Naslund, &

Stenborg, 1987; Vintturi et al., 2001).

Within the group of professional voice users, there are different levels of vocal

demand and usage as well as a wide variation in the quality of voice needed. Vilkman

(2000) has classified voice and speech professions by vocal quality and load. Vocal

loading is defined as prolonged use of the voice combined with factors such as acoustic

conditions at work and the type of communication required. Cumulative talking time is a

major factor in vocal load. Persons with high voice quality demands as well as high

vocal load may be more aware of changes in their voice production because of the

constant demand for consistent, high quality voice output. Actors and singers would be

examples of persons with high quality and high vocal load demands. These persons are

more likely to seek assistance for their voices than persons with lower quality demands

such as nurses or secretarial personnel.

The majority of professional voice users receive little, if any, instruction in voice

production. The exceptions to this are singers, actors, and broadcasters who may receive

some formal training. Those individuals who do not receive training may speak by

imitating what they hear or by trial and error (Benninger et al., 1994). However, these

methods are likely inadequate, leaving these individuals at greater risk for vocal

problems.

Professional voice users report a variety of vocal problems. These may include

changes in vocal quality as well as physical complaints. Vocal quality changes include









hoarseness (Jones et al., 2002; Sapir, Keidar, & Mathers-Schmidt, 1993; Smith, Lemke,

Taylor, Kirchner, & Hoffman, 1998; Smith, Gray, Dove, Kirchner, & Heras, 1997; Yiu,

2002), voice breaks or cracks (Jones et al., 2002; Sapir et al., 1993), voice loss (Jones et

al., 2002), and weak voice (Smith et al., 1998; Smith et al., 1997). Related physical

complaints include shortness of breath (Sapir et al., 1993; Yiu, 2002); dry throat (Jones et

al., 2002; Sapir et al., 1993; Yiu, 2002); scratchy sensation in the throat (Jones et al.,

2002; Smith et al., 1998); throat discomfort, tightness, or pain (Sapir et al., 1993; Smith

et al., 1997; Jones et al., 2002); and effortful speaking (Sapir et al., 1993; Smith et al.,

1997; Smith et al., 1998). Another common complaint is vocal fatigue (Kostyk &

Rochet, 1998; Sapir et al., 1993; Smith et al., 1997; Yiu, 2002). Vocal fatigue may be

defined as negative vocal adaptation that occurs as a consequence of prolonged voice use

(Welham & Maclagan, 2003). Vocal fatigue may be manifested as changes in vocal

quality, loudness, pitch or effort that begin to occur as the speaking day progresses.

These changes are most apparent by the end of the day and usually disappear by the

following morning (Gotaas & Starr, 1993). Further, fatigue may manifest itself as

tightness in the throat and chest, difficulty in loud talking, talking in a monopitch, and

weakening and straining of the vocal muscles (Zeine & Waltar, 2002).

Kaufman and Blalock (1988) described laryngeal tension-fatigue syndrome in a

group of professional voice users. They described a triad of typical complaints consisting

of dysphonia, vocal fatigue, and odynophonia. They further defined the tension-fatigue

syndrome as being characterized by a fluctuating voice quality, which is worse after

strain and during times of stress. They found this syndrome to be associated with a low

speaking fundamental frequency and poor breath support.








Incidence of Voice Problems

The incidence of voice problems in the general population of the United States is

estimated between 3% and 9% (Verdolini & Ramig, 2001). About 2% of that number

may be viewed as having clearly abnormal voice quality (Brindle & Morris, 1979). The

incidence of voice problems in professional voice users is higher than for the general

population.

Much of the available data regarding voice problems in professional voice users

pertain to teachers. According to the United States Bureau of Labor Statistics (2002),

teachers held 7.65 million jobs in 2001, about 6 percent of the U.S. workforce. The cited

percentage of teachers with vocal difficulties ranges from 38% (Smith et al., 1998) to as

high as 80% (Gotaas & Starr, 1993). Teachers, especially in the primary grades, may

spend 90 minutes or more a day talking, with about half of the total talking time at greater

than 80 decibels (Masuda, Ikeda, Manako, & Komiyama, 1993). Teachers of singing are

also at-risk for vocal problems, with 64% of 125 voice teachers surveyed indicating past

vocal problems compared to 33% of a control group (Miller & Verdolini, 1995). Daycare

workers are also at risk for voice problems. In a study comparing Finnish daycare

workers to nurses, daycare workers demonstrated significantly more voice problems with

54% of the daycare workers reporting at least one voice symptom (Sala, Laine, Simberg,

Pentti, & Suonpaa, 2001).

There are limited data available regarding the incidence of voice problems in other

occupational groups of professional voice users. In a sample of 304 telemarketers, Jones

and colleagues (2002) found that 68% of this group had vocal complaints and were twice

as likely to have vocal problems when compared to an age-matched control group.

Similarly, aerobics instructors are at high-risk for voice problems. In one study, 44%






5


percent of a sample of 54 aerobics instructors experienced voice loss associated with

class instruction (Long, Williford, Olson, & Wolfe, 1998).

Persons using voice recognition software to complete their jobs may also be at risk

for vocal problems. At least one case of vocal fatigue and strain has been documented in

a graduate student lecturer following the introduction of voice recognition software. In

this particular case, the individual developed tendonitis and synovitis in the hands, wrists,

arms, and shoulders as a result of repetitive stress and turned to voice recognition

software for her writing. Subsequently, she experienced vocal fatigue, strain, and

difficulty projecting her voice in the classroom. Prior to this, the individual had

experienced no vocal problems (Haxer, Guinn, & Hogikyan, 2001). The Voice and

Speech Laboratory of the Massachusetts Eye and Ear infirmary also reported four cases

of voice recognition product users who developed vocal problems. In each case, the

individuals complained of vocal fatigue, changes in vocal quality, and throat tightness.

Endoscopy confirmed edema and/or vocal hyperfunction in all four cases (Williams,

2003).

While persons in other occupations are also at-risk for voice problems, specific data

regarding the incidence of persons in these occupations that experience vocal complaints

are minimal. The information that is available is based on surveys from voice clinics

indicating the percentage of professional voice users seeking medical treatment,

therefore, this is not a true representation of all the professional voice users experiencing

vocal difficulty. In an examination of over 8,000 patient records from two clinics over a

two-year period, the most frequently represented occupations were executive manager

(7.8%), teacher (5.2%), clerical worker (4.6%), factory worker (4.4%), sales (2.3%),









retail (2.2%), and singer (1.5%) (Coyle, Weinrich, & Stemple, 2001). Other occupations

that are frequently identified as being at-risk for voice problems include ministers,

telephone operators, public relations specialists, speech pathologists, and broadcasters

(Titze, Lemke, & Montequin, 1997).

Consequences of Voice Problems

Professional voice users. Professional voice users may experience a number of

work-related problems due to vocal complaints. Reduced productivity may result from

physical discomfort associated with speaking or from vocal fatigue (Jones et al., 2002;

Smith et al., 1997). Chronic hoarseness, roughness, or other vocal quality changes may

result in reduced quality of work, especially for persons in jobs requiring moderate to

high vocal quality. This, in turn, may lead to problems with low self-esteem. The voice

problems may adversely impact daily activities and social function. The limitation of, or

restricted participation in, daily activities may result in deterioration of the quality of an

individual's life (Ma & Yiu, 2001). Individuals with vocal problems have significantly

lower scores on quality of life scales than do matched controls (Wilson, Deary, Millar, &

Mackenzie, 2002).

Voice problems in professional voice users may be considered a form of repetitive

strain injury (Verdolini & Ramig, 2001) caused by the repeated contact of the vocal folds

during voice production. The repeated movement and resulting collision of the vocal

folds during speaking may occur over a million times per day on the job (Vilkman,

2000). Over time, chronic voice problems and laryngeal pathology may result. In a study

that followed education majors through the first years of professional teaching, 50% of

the subjects reported voice problems after two years of professional activity (Debodt,

Wuyts, Van de Heyning, Lambrechts, & Abeele, 1998).









Voice loss, even if temporary, is likely to result in missed workdays (Yiu, 2002).

For those persons not in permanent employment situations, such as actors and singers,

this can have a negative impact on earning potential. Frequent cancellation or

postponement of jobs may result in fewer job offers and could eventually lead to

dismissal from a particular performance venue.

Career decisions may also be impacted by voice problems. Smith and colleagues

(1997) found that 40% of teachers reported that their voice problems made them

reconsider future career plans. Four percent of teachers in their survey considered a job

change. Data regarding career changes related to voice problems are not available for

other occupations.

Employers. Employers may see the impact of voice problems on their employees

in several ways. Occasional voice loss or throat pain may result in absenteeism.

Employer job satisfaction may drop, leading to strained relations between management

and employees. When voice problems become chronic, employee turnover may increase.

Verdolini and Ramig (2001) estimate the annual treatment costs for teachers with voice

problems to be over 1.4 billion dollars. Further, they speculate that if each teacher with

vocal difficulty missed three workdays annually requiring the hiring of a substitute

teacher at approximately $60 per day, the cost in the United States would be over 372

million dollars. The total cost for teachers annually would amount to nearly 2 billion

dollars with most of this cost being absorbed by the employer.

For those occupational voice users who are unable to perform their job duties,

workers compensation claims may be filed. For 2001, the total cost to American

businesses for disabling injuries and lost workdays was 132 billion dollars (Chubb









Group, 2004). The National Safety Council estimates the average per case cost of wage

and productivity losses, medical expense, and administrative expense of workers'

compensation claims to be $33,000 (2002). Specific figures for voice-related disability

complaints are not available.

Types of Vocal Pathologies Found in Professional Voice Users

Professional voice users are susceptible to the same vocal pathologies as other

voice users. Pathologies may occur as a result of repeated trauma to the vocal folds or in

response to environmental triggers, such as toxins, molds, and dust. The most common

disorder of the vocal folds in an occupational setting is laryngitis, although the

development of vocal fold nodules, vocal fold polyps, and contact ulcers may also be

seen (Williams, 2002; Sala et al., 2001).

In a study with high-risk vocal performers, laryngoscopic and stroboscopic

examination showed performers to have vocal fold edema, increased vascularity of the

vocal folds, and irregular vocal fold edges (Hoffman, Lehman, Crandell, Ingram &

Sapienza, 2001). In another study of 20 singing teachers who volunteered at a national

conference for laryngoscopic examination, 61% were found to have subepithelial vocal

fold cysts. The cysts were accompanied by varicosities at their bases, suggesting

traumatic etiology. Another 8% of the teachers in the sample had vocal fold polyps

(Heman-Ackah, Dean, & Sataloff, 2002).

Another common pathology of professional voice users is vocal fold nodules.

Vocal fold nodules typically form on the edges of the vocal folds resulting in hoarseness,

breathiness, loss of vocal range and reduction in vocal flexibility (Benninger et al., 1994;

Colton & Casper, 1990). Nodules typically occur from misuse or abuse of the voice

resulting in vocal fold trauma (Sataloff, 1991). Other lesions of the vocal fold, such as








hemorrhage or cysts, may occur with forceful voice use and are often seen in professional

voice users (Benninger et al., 1994; Lin, Stern, & Gould, 1991).

Extra-esophageal reflux may also lead to voice difficulties for professional voice

users. The recurring irritation to the laryngeal tissues may cause vocal fold edema and/or

compensatory muscle tension dysphonia (Ross, Noordzji, & Woo, 1998; Sataloff, 1997).

Professional voice users, particularly those who travel frequently or eat late at night, may

be especially prone to reflux (Benninger et al., 1994). In fact, Sataloff(1997) found

reflux laryngitis in 45% of consecutive professional voice users seen in a voice clinic.

Compensations for Vocal Pathologies

Professional voice users perform under a number of adverse conditions that may

lead to vocal compensations. A common condition is increased background noise, such

as that occurring in a classroom, factory, day care center, or busy office. A study done

with welders in Sweden showed a mean workplace noise level of 95 dBA (Ohlsson et al.,

1987). Background noise levels in day care centers were found to be 76 dB on average

(Sodersten, Granqvist, Hammarberg, & Szabo, 2002). Speakers may increase their vocal

intensity two to five decibels for every ten-decibel increase in noise level (Garber, Siegel,

& Pick, 1981). If the increased vocal loudness is sustained for long periods of time,

speakers may increase vocal fold adduction and this may lead to irritation or

inflammation of the vocal folds (Ohlsson et al., 1987; Sapienza, Crandell, & Curtis,

1999). A related issue is poor room acoustics that may lead to poor acoustic feedback for

the speaker. This situation may also lead to increased vocal strain (Caldwell et al., 2001).

Anxiety, stress, and stage fright may lead to increased musculoskeletal tension and

altered breathing patterns. Stress may also be manifested as dry mouth, fatigue,

headaches, insomnia, and heartburn (Sataloff, 1991). In turn, this may lead to increased








strain of the laryngeal muscles to produce an acceptable vocal quality. Similar

compensatory muscle tension may occur in response to the quality changes created by

throat infection as the speaker attempts to maintain a normal tone (Caldwell et al., 2001).

Heavy costumes and makeup may create special problems for singers and actors.

Costumes may alter respiratory function and posture. Makeup, especially the use of

prosthetics or masks, may alter face and jaw movement and make projection and

articulation more difficult (Raphael, 1991; Hoffman-Ruddy et al., 2001). The performer

may compensate for these altered movements by changing laryngeal dynamics.

Frequent travel may lead to alterations in diet and sleep patterns for professional

voice users. These changes may lead to dehydration, fatigue, and digestive problems

(Caldwell et al., 2001). The professional voice user's response to each of these problems

may lead to compensatory behaviors that adversely affect vocal production.

As compensatory behaviors lead to increases in musculoskeletal tension, reduced

respiratory function, and altered laryngeal dynamics, the professional voice user may find

it increasingly difficult to produce voice consistently with the desired quality. When this

occurs, the professional voice user is likely to seek medical treatment for the vocal

problem in order to regain and maintain desirable vocal quality.

Assessment for Voice Problems

Assessment of voice production may be accomplished in a variety of ways. A

medical examination of the larynx is typically accomplished using visualization via an

endoscope. This allows for assessment of laryngeal structures and mobility. With the

addition of stroboscopy, a more detailed look at vocal fold movement can be obtained.

Self-perception of vocal quality and its impact on a person's quality of life may be

assessed using any number of self-assessment scales. The Voice Handicap Index









(Jacobson et al., 1997) was developed for this purpose and was chosen for use in this

study. It consists of thirty items that are ranked by the subject on a five-point ordinal

scale with a maximum score of 120. Higher scores indicate a higher degree of vocal

disability. The developers of the VHI indicate that an 18-point difference is necessary to

determine meaningful change between administrations of the test that is not due to

inherent variability.

The voice may also be assessed using perceptual measures. Perceptual assessment

typically involves the use of a rating scale that may be completed by an individual

regarding their voice or by another listener. Several types of scales have been used for

this purpose including equal-appearing interval scales and direct magnitude estimation.

For this study, a visual analog scale was used. The visual analog scale has been shown to

be a reliable, valid, and sensitive method for self-report of subjective experiences (Gift,

1989). The visual analog scale uses a line, usually 100 millimeters long, with descriptors

at each end. The subject is asked to make a mark along the line that measures his/her

perception of the parameter at the present moment. The data yielded are of a ratio level

and are useful for parametric statistical analysis (Cannito et al., 1997).

The physical characteristics of the voice may be assessed using various measures.

The phonetogram, or voice range profile, is one such measure that has been used to track

changes in both frequency and intensity over time. The phonetogram is a display of

vocal intensity range plotted with fundamental frequency (Titze, 1994). Coleman (1993)

describes the phonetogram as a descriptor of total vocal tract output. He also states that

the covariance of pitch and loudness may be used to track changes in the pathological

conditions of the vocal system.









Akerlund (1993) used the phonetogram to determine if treatment affected mean

sound levels and frequencies in a group of subjects with non-organic dysphonia. Results

showed that both sexes extended their vocal range an average of 1.5 to 2.1 semitones.

Frequency has been shown to have little variation, of about 1 semitone, between trials

with normal subjects (Coleman, Mabis, and Hinson, 1977). Intensity has a short-term

mean variation of 2-3 dB between administrations of the phonetogram (Gramming,

Sundberg, and Akerlund, 1991).

Another method for describing physical characteristics of the voice is the

hoarseness diagram. Typically, analysis of the dysphonic voice has been difficult to

accomplish because of the aperiodicity of the signal. The hoarseness diagram seeks to

minimize this difficulty. The method, devised by Frolich, Michaelis, Strube, and Kruse

(1997, 2000), is based on four acoustic measures that allow a quantification of voice

quality with regard to roughness and breathiness. The roughness, or irregularity

component, is calculated using jitter, shimmer, and the mean correlation coefficient

between consecutive glottal cycles. The breathiness, or noise, component is based on the

glottal-to-noise excitation ratio (GNE). The GNE is applicable even for highly irregular

glottal oscillation. Any narrow-band noise generated by turbulence does not influence

the degree of turbulence measured (Michaelis, Gramms, & Strube, 1997). The GNE is

based on correlations between Hilbert envelopes calculated for different frequency

channels of an inverse-filtered speech wave. Values close to one reflect a pulse-like

excitation. Conversely, values close to 0 result from a noise-like excitation (Frolich et al,

2000). The values obtained from the hoarseness diagram may be plotted on a graph with

the irregularity component on the x-axis and the breathiness, or noise component, plotted









on the y-axis. Voices that are normal cluster in the lower left region of the graph and

voices that are more pathologic are undifferentiated in their placement (Michaelis,

Frolich, & Strube, 1998).

Physiologic measures may also be useful in voice analysis. These may include

aerodynamic measures, electromyography, and electroglottography. For this study,

aerodynamic measures of subglottal and intraoral pressure were obtained.

Treatments for Voice Problems

Treatments for voice problems may be surgical, medical, or behavioral. Surgical

treatment is indicated for problems that are not amenable to therapy alone. These

problems would include, but are not limited to, vascular lesions, structural abnormalities,

hemorrhagic polyps, and vocal fold cysts (Benninger et al., 1994; Davies & Jahn, 1998).

Management with medications may be indicated for cases of allergy, edema, and

gastroesophageal reflux. Behavioral therapy is often the treatment of choice for voice

disorders as surgical management may result in vocal fold scarring, resulting in an

adynamic segment along the vibratory margin of the vocal fold (Sataloff, 1991). There

are four primary approaches to behavioral voice therapy in common use at the present

time: hygienic, symptomatic, psychogenic, and physiologic. Physiologic therapy is the

type used for this study.

Physiologic voice therapy. Physiologic voice therapy is based on improving the

balance between the vocal subsystems of respiratory support, laryngeal muscle strength,

and resonance. The objective of this type of therapy is to promote a healthy vocal fold

cover (Stemple, Glaze, & Gerdeman, 1993). The physiologic condition of the vocal folds

is considered in developing direct exercise or manipulation to improve the patient's vocal

condition. Several approaches fit under the physiologic voice therapy heading. These









include vocal function exercises (Stemple, Lee D'Amico & Pickup, 1994; Sabol, Lee &

Stemple, 1995), resonant voice therapy (Verdolini, Druker, Palmer, & Samawi, 1998),

and the accent method of voice therapy (Kotby, El-Sady, Basiouny, Abou-Rass, &

Hegazi, 1991).

There are few experimental studies utilizing prospective, randomized designs that

have addressed the issue of outcomes for voice therapy. One study compared the effect

of intensive hydration therapy to a control period for six subjects with vocal fold nodules

(Verdolini-Marston, Sandage, & Titze, 1994). A reduction in perceived vocal effort,

improvement in laryngeal appearance, and reduced phonation threshold pressure were

found under the hydration condition. In another study examining subjects with vocal fold

nodules, the effect of confidential voice therapy was compared to resonant voice therapy

over a 12-day period. Findings suggested improvements in voice quality with both

therapy methods as long as subjects were compliant with treatment (Verdolini-Marston,

Burke, Lessac, Glaze, & Caldwell, 1995). Basiouny (1998) compared use of the accent

method to instruction in vocal hygiene. Significant improvements were found for vocal

function and aerodynamic measures as well as improvement in laryngeal appearance for

the group receiving the accent method of treatment.

A study of voice therapy outcomes for women with vocal fold nodules was

-'undertaken at the Massachusetts Eye and Ear Infirmary in Boston. The treatment

protocol consisted of five basic behaviorally based phases: vocal hygiene, direct

facilitation, respiration, relaxation, and carry-over (Holmberg, Doyle, Perkell,

Hammarberg, & Hillman, 2003). They found that significant changes in perceptual

parameters occurred after completion of the direct facilitation and/or respiration treatment









approaches. For 9 of the 10 subjects, the nodules decreased in size post treatment and

accompanying edema was reduced.

In a study by Carding, Horsley, and Docherty (1999) direct and indirect treatment

groups were compared to a control group. The direct treatment group received individual

therapy personalized to meet their specific needs. The indirect treatment group received

lifestyle management and vocal hygiene suggestions only. After an eight-week therapy

trial, the direct treatment group showed significant improvement in voice severity. When

results for both treatment groups were combined, 92% of the subjects showed

improvement.

In the largest comparative treatment study utilizing 204 patients with functional

dysphonia, subjects received six sessions of direct voice treatment compared to a non-

treatment group. Significant improvements were found in ratings of voice quality, both

self-rated by the subjects and by a group of expert listeners. Improvement was also found

in shimmer values (MacKenzie, Millar, Wilson, Sellars, & Deary, 2001). None of the

studies examined were specific to professional voice users. In fact, this population was

excluded from most of the studies reviewed.

While no existing therapy programs specifically target the professional voice user,

any of the behavioral therapy programs may be utilized as appropriate for an individual's

particular problems and complaints with some degree of success. However, as previously

discussed, professional voice users may have significant difficulty with vocal endurance

and may have demands for sustained loudness due to working conditions. No existing

therapy programs target these requirements. Two of the physiologic types of therapy,

Vocal Function Exercises and the Accent Method, specifically target the respiratory









component of voice production. However, there is no program that focuses on

strengthening the respiratory muscles used for voice production. It is believed that

strengthening the respiratory drive used for voice production will assist professional

voice users in meeting loudness and endurance demands by increasing the respiratory

system's ability to generate respiratory pressures.

Role of Respiratory System in Voice Production

The respiratory system provides the power for vocal fold vibration. Ample air

must be moved in an outward direction through the glottis in order for voice to be

established and maintained (Titze, 1994). This occurs through a combination of thoracic

and abdominal muscle function coupled with adjustments to the closure of the vocal folds

themselves. As demands for loudness and/or duration increase, the abdominal muscles

are responsible for increasing expiratory force.

At rest, air exchange occurs as the thorax is enlarged. During inspiration, the

diaphragm descends, increasing the vertical dimension of the chest, allowing the lungs to

expand and fill with air. At the same time, the external intercostal muscles pull the ribs

up and forward to further increase the dimensions of the chest cavity. The system then

recoils to its resting state during passive exhalation. Speech is typically produced in the

midrange of vital capacity, near the resting expiratory level, or at 50% to 65% of vital

capacity (Mitchell, Hoit, & Watson, 1996).

During more active breathing for exercise or for the demands of singing or public

speaking, the muscles of exhalation become more active. These muscles include the

internal and external obliques, rectus abdominis, transverse abdominis, and the quadratus

lumborum (Zemlin, 1998). They serve to drive the abdominal contents inward and create

a piston-like effect on the diaphragm. The contraction of the internal intercostals and the








serratus posterior inferiors assists in depressing the rib cage. These muscle contractions

decrease lung volume, increase lung pressure, and increase the velocity of airflow out of

the lungs (West, 2000).

Mechanisms for Variation of Vocal Intensity

Lung pressures generated during the expiratory phase vary with the amount of

muscular force applied to the lung and chest wall unit. An increase in expiratory muscle

force will increase lung pressure and, in turn, increase the pressure below the vocal folds

(subglottal pressure). This is the primary mechanism for increasing vocal intensity

(Isshiki, 1964). The required subglottal pressure for generating normal conversational

speech is 4 to 6 cmH20. The elastic recoil forces of the lung and chest wall usually

produce this pressure level at lung volume initiations between 40 and 60% lung volume.

Men have higher static recoil pressures than females which suggests that women need to

initiate speech at higher lung volumes than men to achieve the same static recoil

pressures (Stathopoulous & Sapienza, 1993). When greater pressures are required for

loud speech or singing, expiratory muscle force plays a critical role in sustaining

appropriate subglottal pressure (Sundberg, 1987). Subglottal pressure may increase to 10

to 15 cmH20 for loud conversation and has been measured as high as 20 to 30 cmH20 for

shouting (Hixon, 1987). In singing, subglottal pressures for loud singing also fall in the

20 to 30 cmH20 range (Sundberg, 1987).

Sound intensity may be influenced by other factors as well. Physiologically, there

is a strong relationship between the shut-off of glottal airflow and the sound pressure

level of the voice. The sound pressure level, or intensity, increases as the rate of glottal

shut-off of airflow increases (Scherer, 1991). Degree of glottal adduction is also linked

to sound intensity. Increased vocal fold adduction decreases the ratio of the intensity of









the fundamental frequency to the overall intensity. The greatest increase in intensity

appears to occur when the vocal processes are almost touching during the closed phase of

vibration (Scherer, 1991). Acoustic properties of vocal production may also increase

intensity. As fundamental frequency increases, there is a 6 dB increase for every octave

change that occurs (Titze, 1994).

Professional voice users frequently have high demand for increased loudness due to

increased background noise or poor environmental acoustics in the workplace. As

previously discussed, background noise in some work settings may range from 76 to 95

dBA (Ohlsson et al., 1987; Sodersten et al., 2002). Therefore, the demands on the

respiratory system must be increased to accommodate the increased background noise.

Without an adequate increase in subglottal pressure for loudness demands, the

professional voice user may increase laryngeal hyperfunction which may lead to vocal

fatigue and increased risk of tissue trauma (Sataloff, 1997; Titze, 1994). Therefore,

strengthening the expiratory muscle system should prove beneficial to professional voice

users; both in fulfilling the demands for higher respiratory drive and in minimizing

potential vocal fold damage from trauma.

Respiratory Muscle Training

The training of respiratory muscle is a concept that has been applied to both normal

and diseased populations. The training is designed to strengthen either inspiratory or

expiratory muscles to increase the ability to generate increased muscle pressure. The

underlying rationale for expiratory muscle training is that respiratory muscle, like other

skeletal muscle, responds to strengthening and conditioning (Leith & Bradley, 1976;

Powers, Coombes, & Demeril, 1997). Muscle strength can be increased through weight

or resistance training. Resistance training can be accomplished through isometric








training (resistance without movement), isotonic training (resistance with movement) or

isokinetic training (resistance at a constant speed) (Saxon & Schneider, 1995). The goal

of strength training is to increase the maximum amount of force that can be generated by

a muscle or muscle group. High intensity training is required in order to increase muscle

strength (Powers & Howley, 2001). Resistance training has been demonstrated to

increase both muscle size and muscle strength, and to increase both Type I and Type II

fibers (Powers & Howley, 2001).

The overload principle states that muscle must be exercised beyond a level that it is

accustomed to in order to achieve a training, or strength, effect (Powers & Howley,

2001). Since the muscle, or muscle system, gradually adapts to this overload through

adaptation, increased load must be incorporated into the training program (Saxon &

Schneider, 1995). This increased load must be progressive to prevent musculoskeletal

injury (Powers & Howley, 2001; Saxon & Schneider, 1995). Further, training must be

specific to the muscle fibers involved in the activity and to the metabolic energy system

used for the activity (Powers & Howley, 2001). The training program must train the

muscles actually used for a specific skill. For example, a runner must practice running in

order to strengthen the energy and muscle systems used for running.

Muscle strength can be increased through weight or resistance training. The most

common devices used for respiratory muscle strength training are resistance-based

trainers or pressure threshold trainers. Resistance training requires that a subject inspire

or expire through a mouthpiece that provides resistance to the flow of air. With pressure

threshold training devices, the subject must produce a specific amount of mouth pressure

to open a one-way valve and allow air to pass through the mouthpiece. Pressure









threshold training is not dependent on airflow and allows training to occur only when the

respiratory muscles produce a sufficient lung pressure. The respiratory muscles are

known to compensate during resistance-based training by decreasing flow, thereby

reducing the amount of muscle contraction needed to complete the task (Powers et al.,

1997). Pressure threshold training requires higher levels of muscle force and therefore,

most often produces a greater training effect.

Pressure Threshold Device

Pressure threshold devices have a spring that presses a cover (valve) over a tube

opening. The subject must produce a respiratory pressure sufficient to overcome the

spring compression strength to open the valve. Airflow through the device occurs only

when the valve is open. Since lung volume changes simultaneously with airflow, the

expiratory muscles must produce a constant minimum force throughout the breath. The

muscle activation increases as lung volume decreases so the training effect increases late

in the expiratory cycle. This results in greater force stress on the expiratory muscles and

a greater strengthening effect, thus allowing more specific training for the expiratory

muscles (Kellerman, Martin, & Davenport, 2000).

A pressure threshold trainer, designed by University of Florida investigators (Paul

Davenport, Daniel Martin, and Christine Sapienza) is available for experimental use.

This device allows an individual to train the expiratory muscles at near maximum level.

Most commercially available trainers have threshold settings from 4 to 20 cmH2O that do

not provide enough overload to the muscle to increase strength. The current experimental

trainer used in this study allows the pressure threshold to be set at levels of up to 155

cmH20, much higher than previously available on commercial pressure threshold

trainers, thus meeting the required overload principle. This trainer can be calibrated to









allow for quantifiable increases in the pressure threshold level over time, a feature not

available on some commercially available trainers.

Expiratory Training Studies

Several studies have explored the use of expiratory muscle training. The results of

these studies are summarized in Table 1-1. In a study of six healthy subjects who

performed expiratory muscle training over a four-week period, twice daily for 15

minutes, a significant increase in the strength of expiratory muscle as well as a reduced

sensation of respiratory effort during exercise was noted. All subjects showed an

increase in maximum expiratory pressure following training with the group showing a

25% increase in pressure overall (Suzuki, Sato, & Okubo, 1995).

Table 1-1. Summary of Expiratory Muscle Strength Training Programs.
Authors Subject # Training # sets Training % increase
Group subjects Duration threshold MEP
Suzuki et Healthy 10 4 weeks 15 min/ 30% of 25%
al. men twice a day MEP
Smeltzer Multiple 20 3 months 3 sets/15 Not 36%
et al. sclerosis breaths twice reported
a day
Gosselink Multiple 28 3 months 3 sets/15 60% of 35%
et al. sclerosis breaths twice MEP
a day
Sapienza Band 40 2 weeks 4 sets/6 75% of 47% for
et al. students breaths MEP females;
53% for
males
Hoffman- High-risk 8 4 weeks 4 sets of 6 80% of 84%
Ruddy performers breaths/twice MEP
daily
Roy et al. Teachers 27 6 weeks 5 sets of 5 80% of 21%
breaths/twice MEP
daily
Baker Healthy 32 4 weeks 5 sets of 5 75% of 41% for 4
males and and 8 breaths/5 MEP wk. Group;
females weeks days a week 50% for 8
Swk. group








Expiratory muscle training was completed with high school wind instrumentalists,

five days a week for a two-week period using four sets of six training breaths at loads

between 14 and 80 cmH20. The valve was set at 75% of each subject's maximum

expiratory pressure. With training, females showed an increase of 43 cmH20 or a 47%

increase from baseline. Males had an increase of 53 cmH20 post-training, or an increase

of 48% from baseline. This study found no change in maximum expiratory pressures for

a control group (Sapienza, Davenport, & Martin, 2002).

In a study of 20 severely disabled multiple sclerosis patients, ten subjects

completed three months of expiratory training using a threshold training device and were

compared to five subjects who completed three months of sham training using the same

device without an expiratory training threshold load (Smeltzer, Lavietes, & Cook, 1996).

A significant increase in expiratory muscle strength was found in the training group. In

another study with severely disabled multiple sclerosis patients, 28 subjects completed

three months of training. Patients completed three series of 15 contractions at 60% of

maximum expiratory pressure twice daily. The patients achieved a 35% increase in

maximum expiratory pressure (Gosselink, Kovacs, Ketelaer, Catron, & Decramer, 2000).

Another study of expiratory muscle training was completed with a group of eight

high-risk vocal performers with a high degree of vocal load. These performers trained

daily for a period of four weeks. Mean maximum expiratory pressures for the

experimental group post-training increased from 70.5 cmH20 to 130 cmH20 (84%).

Subjects reported a decrease in breathlessness during performances and an increase in

duration of sustained phrases. A control group experienced no substantial increase in









maximum expiratory pressure (Hoffman-Ruddy, 2001; Sapienza, Hoffman-Ruddy,

Davenport, Martin, & Lehman, 2001).

Expiratory muscle training was also explored in a group of teachers as part of a

larger study comparing three voice therapy treatments. Twenty-seven full-time teachers

with vocal problems underwent expiratory muscle training. Training was performed five

days a week, with five sets of five repetitions performed twice daily for six weeks. The

expiratory muscle-training group showed a significant improvement in maximum

expiratory pressure at the conclusion of the study. However, this group did not

demonstrate significant change on self-ratings of voice severity or vocal disability (Roy

et al., 2003). The training paradigm used with the teachers was not as carefully

controlled as in previous studies as the load was not measured or changed weekly.

Baker (2003) examined changes in healthy individuals. Some subjects participated

in 4 weeks of training while others completed 8 weeks of expiratory muscle training. The

two groups showed increases in MEP of 41% and 50% respectively.

Statement of the Problem

While expiratory muscle training has been found to increase expiratory muscle

strength, the benefits of expiratory muscle strength training for professional voice users

have not been fully demonstrated across a range of occupational settings. Further, no

studies have combined expiratory muscle strength with any other therapeutic

intervention.

Purpose of the Study

The purpose of this study was to examine the use of expiratory muscle strength

training combined with voice therapy in a group of professional voice users. Physiologic

and acoustic changes occurring in the voice pre to post treatment will document outcome.









The changes in patient perception of vocal difficulty following treatment will also be

compared.

Hypotheses

Central Hypothesis: It is hypothesized that expiratory muscle strength training

combined with voice therapy will produce greater improvements in voice production than

will voice therapy alone.

Hypothesis 1. Both groups will report a greater reduction in vocal symptoms and

vocal effort following EMST-training plus voice therapy as compared to the voice

therapy condition alone.

Hypothesis 2: Both groups will produce greater improvements in perceptual ratings

post-EMST training plus voice therapy than with voice therapy alone.

Hypothesis 3: Both groups will produce greater improvement in respiratory

pressure post-EMST training plus voice therapy than following voice therapy alone.

Hypothesis 4: Both groups will produce greater improvements in frequency and

time domain measures, as measured by the phonetogram, post-EMST training plus voice

therapy than following voice therapy alone.

Hypothesis 5. The group with benign vocal fold lesions will demonstrate greater

improvements overall than the non-lesion group following the combined treatments of

EMST and voice therapy than following voice therapy alone.














CHAPTER 2
METHODOLOGY

This study examined the use of expiratory muscle strength training combined with

voice therapy for professional voice users currently experiencing complaints of vocal

problems. Participants were grouped by laryngeal diagnosis. Both groups received voice

therapy and expiratory muscle strength training. Half of the participants received voice

therapy first and half received expiratory muscle strength training first in order to avoid

an order effect.

Experimental Design

The design of this project was a prospective, complete repeated measures design.

The independent variables were laryngeal diagnosis (lesion/non-lesion), gender, and

treatment order The dependent variables were: Voice Handicap Index (VHI) score,

Vocal Rating Scale score, Voice Effort Scale score, irregularity and noise measures from

the hoarseness diagram, total area, and the range of intensity and fundamental frequency

obtained from the phonetogram, phonation threshold pressure, subglottal pressure,

maximum expiratory pressure, and listener ratings of voice quality.

Participants

Sample size determination. Calculation of the sample size was performed using

one response variable as suggested by Marks (2001). The VHI score was used as the

variable to determine sample size since this measure is an important measure for

determining the impact of voice problems as reported by individuals. The VHI is a

measure of a person's perception of their vocal difficulty on their quality of life. The








common standard deviation, denoted as a, was determined from the range of VHI scores

obtained prior to treatment. The range was 77, which yielded a common standard

deviation of 19.25. The critical relative difference of the VHI (or bound on error) was set

at 18 points as suggested by the developers of the scale (Jacobsen et al., 1997). This is

the minimum significant difference between administrations of the scale that is not due to

a different test administration. Using the two obtained values as described above, Delta

was calculated using the formula: Delta=bound/standard deviation and was determined to

be 0.93. The significance level (a), or probability of committing a Type I error, was

predetermined at 0.05. The power of the test, or the ability to reject the null hypothesis if

the null is false, was set at 80%. Using Marks' sample size table (2001) for the obtained

values, the number of participants needed was 9 per group. Therefore, at least 9

participants per group were needed to detect an average difference of eighteen points in

VHI scores, pre to post-treatment. Thus, a total of 20 participants (10 participants x 2

groups) were recruited for the study.

Recruitment and selection. Approval for the study was obtained from the

University of Florida Health Science Center Institutional Review Board prior to enrolling

participants. Participants were recruited from the Gainesville, Florida area. Flyers

containing information about the study along with contact information were posted at

various locations across campus (Appendix A). Additional copies of the flyer were

distributed to schools, businesses, and organizations with members who were

professional voice users.








Inclusion criteria. Participants were included based on the following criteria:

1. Age between 21 and 65 years,

2. Professional voice users who speak at least 4 hours a day on the job,

3. Complaints of vocal problems including fatigue, hoarseness, throat pain, and voice
loss,

4. Normal hearing (25dB HL at 500 Hz; 20dB HL for 1000, 2000, and 4000 Hz.),

5. Persons who were able to maintain their current level of physical activity
(including both aerobic exercise and weightlifting) throughout the entire training
period. Participants were asked to report any changes in their level of activity
throughout their participation in the study. Participants were excluded from the
study if they made a significant change in activity level as described above.

Exclusion criteria. Participants were excluded from the study if they reported any

of the following:

1. History of chronic and acute cardiac disease including hypertension, pulmonary
disease, neuromuscular disease, and/or immune system disease as reported on a
health questionnaire (Appendix B),

2. Upper respiratory infection at the time of baseline measurements as reported on the
health questionnaire or during the training period,

3. Pulmonary function test values that were below 75% of the predicted normative
value,

4. Vocal fold lesion requiring immediate medical attention,

5. History of smoking or tobacco use within the last 5 years.

Assignment of participants to groups. Participants meeting inclusion criteria

were assigned to one of two groups, lesion versus non-lesion, based on their laryngeal

examination. Participants with laryngeal irritation or edema were assigned to Group 1.

Participants with evidence of benign vocal fold lesions were assigned to Group 2.

Participant demographics. Twenty participants, 12 females and 8 males, were

enrolled in the study. One female participant (number 19) completed only the initial

measures and withdrew from the study due to illness. In order to keep the two treatment








groups equal, one randomly selected participant's data were removed from the analysis.

The final number of participants included in the study is 10 females and 8 males.

Demographic data for the participants are listed in Table 2-1. A total of 9 participants (2

males, 7 females) had vocal fold lesions. The lesions included vocal fold nodules (5

females), unilateral vocal fold polyp (1 female), bilateral sulcus vocalis (1 male), and

petechial hemorrhage (1 male, 1 female). The average age of the participants was 41

years with a range of 22 to 59 years. The average number of years on the job was 12.

Participants reported on a written questionnaire that they spent an average of 61% of their

time at work talking.

Table 2-1. Demographic information for participants in the study.
Partici- Gender Occupation Age Years Hours Estimated Lesion
pant on worked/week Talking
Job Time
1 M Manager 27 1 20 70% No
2 M Professor 34 3 20 50% No
3 F Professor 37 5 35 38% No
4 M Attorney 49 7 40 50% No
5 F Teacher 25 4 40 68% No
6 M Chaplain 55 27 43 25% Yes
7 F Surgery 38 8 40 90% Yes
Scheduler
9 F Corrections 29 0.5 25 75% Yes
Officer
10 F Teacher 46 5 50 50% Yes
11 M Librarian/ 48 7 40 40% No
Teacher
12 F Teacher 59 38 55 55% Yes
13 M Professor 44 9 35 60% Yes
14 F Professor 53 32 55 75% Yes
15 F Computer 41 18 50 75% No
Coordinator
16 F Marketing 36 9 37.5 60% Yes
17 M Minister 59 18 50 25% No
18 M Counselor 49 17 40 65% No
20 F Medical 35 10 40 85% Yes
Technologist








Measures

Measures of voice production, subglottal pressure, and perceived vocal effort were

obtained for each participant. Laryngoscopic evaluation and lung function measures

were used as screening tools prior to the study. All other measures were taken before and

after each of the training phases, for a total of three measurement points.

Screening Measures

An endoscopic examination of the vocal folds was performed on each participant

prior to acceptance into the study. Examinations were completed using a rigid oral

endoscope or a flexible nasal endoscope under medical supervision. As per previously

cited exclusion criteria, participants were not included in the study if they had a vocal

fold lesion requiring immediate medical attention. Participants were asked to fill out a

health questionnaire (Appendix A) to ensure that all other criteria were met.

Measures of forced expiratory volume at one second (FEV1) and forced vital

capacity (FVC) were taken for each participant. Participants were asked to inhale to their

total lung capacity and then blow out as forcefully as possible into a computerized

spirometer (Spirovision 3+, Futuremed of America). FEV1 is a measure of expiratory

volume during the first second of expiration. FVC represents the total volume exhaled

(West, 2000). The values of FEVI and FVC had to be at 75% or greater of expected

values for participants to be eligible for the study.

Perceptual Measures

Self-rating scales. Each participant completed three self-evaluation scales at each

of the three measurement points. The first was the Voice Handicap Index (Appendix C).

This thirty-item scale, developed by Jacobsen and colleagues (i997) was designed to

measure the impact of voice problems on a person's quality of life. The participant









responds to a series of statements and ranks each statement on a scale from 0 to 4

yielding a possible score of 120. It is divided into three scales to assess physical,

emotional, and functional domains. The authors consider a minimum difference of

eighteen points between administrations to be indicative of a statistically significant

change in perceived handicap due to the voice problem.

The second perceptual scale, the Voice Rating Scale (Appendix D) was developed

by the investigator for the present study to explore its use for further detailing work-

related voice problems. This scale simply represents a pilot development to evaluate

whether it has potential use for ascertaining problems that occur with the voice in the

workplace. The Voice Rating Scale is a series of statements that required the participant

to rank the magnitude of work-specific vocal problems. Statements concerned the degree

of difficulty speaking and being heard, the degree of vocal fatigue and voice loss, and the

degree to which the voice problem affected job performance. It employed a visual analog

scale in which the participant was asked to rank their response to each question on a 100-

millimeter line. The response was then converted to a percentage. Participants were not

shown their previous responses on the rating scales. For data analysis, the percentages

obtained for each question were totaled, yielding a possible total score of 1000 points.

The third scale, also developed by the investigator, was the Vocal Effort Scale

(Appendix E). This also employed a 100-millimeter visual analog scale on which

participants rated their vocal effort at work and in social situations during the previous

week. Responses were converted to percentages. Again, participants were blinded to

responses from any previous measurement sessions.








Voice Measures

Listener ratings. Voice recordings were collected for each participant at each of

the three measurement points. Each participant produced three /a/ and three /i/ vowels as

well as seven sentences taken from the Consensus Auditory Perceptual Evaluation of

Voice (CAPE-V). The CAPE-V was developed by the American Speech-Language-

Hearing Association's Special Interest Division 3 for Voice and Voice Disorders in 2002.

The voice samples were recorded in a sound-treated room using a Shure SM48 cardioid

microphone with a frequency response of 55 to 14,000 Hz. Mouth-to-microphone

distance was kept constant at six inches as recommended by Kay Elemetrics. The

samples were digitized directly in the Computerized Speech Lab Model 4300B (Kay

Elemetrics). A sampling rate of 44,100 Hz was used for later analysis with the

hoarseness diagram software. One sentence ("We were away a year ago") from each

measurement point for each speaker was selected for perceptual testing. Therefore, a

total of 54 stimuli were used for perceptual tests (18 speakers X 3 measurement points).

A group of 10 listeners, all speech pathologists with special training in voice disorders,

were recruited to rate the voice quality for the sentence samples. The raters were blinded

to subject group and were unaware of whether the recordings were pre, mid, or post

treatment conditions. The ratings for each listener were averaged together for analysis.

The perceptual experiment was carried out using the EcosWin software (Avaaz

Innovations, Inc.) The 54 sentences were used to form 10 blocks of stimuli. Each block

consisted of one occurrence of each sentence in random order. Therefore, each listener

rated a total of 540 sentences (54 sentences X 10 blocks). The listening tests were carried

out in a single-walled sound-treated booth. Stimuli were presented through a RP2

processor (Tucker-Davis Technologies, Inc.) over TDH-39 headphones at a comfortable








loudness level. Listeners made their response using a computer monitor and keyboard.

Each listener heard all 10 blocks of stimuli once, but the order of the blocks was

randomized across listeners. The listeners were asked to indicate the severity of voice

quality for each sentence using a five-point rating scale with "one" representing normal

voice quality and "five" representing severe voice quality. To minimize fatigue, listeners

were provided with at least two breaks during the test session. Listeners required

approximately 50 minutes to rate all 10 blocks of stimuli. In order to minimize the

variability within and across listeners, data from the multiple observations was averaged

to obtain a single judgment for each stimulus (Shrivastav & Sapienza, 2004).

Stroboscopic ratings. A group made up of seven speech pathologists and one

medical doctor rated the pre and post-treatment stroboscopic examinations for the

participants in the lesion group only. Stroboscopic assessments were completed using a

stroboscopic rating form (Appendix F) to rate vocal fold edge, mucosal wave, amplitude

of mucosal wave, glottic closure, and phase closure. The examinations were randomized

on videotape and raters were not provided with information on whether the examination

was completed pre or post-treatment. Data for the multiple observations of each of the

parameters was averaged across raters for analysis.

Pulmonary Measures

Measures of expiratory muscle strength were taken weekly during the respiratory

training condition. Maximum expiratory pressure (MEP) measured at the mouth, was

used as the indirect measure of expiratory muscle strength. The measurement apparatus

consisted of a mouthpiece connected to a Smart 350 series pressure manometer (Meriam

Instruments, USA) by 46 cm of 6 mm inner diameter tubing with a 14-gauge needle air-

leak. MEP was measured with the participant standing and their nose occluded with a









nose clip. After inhaling to total lung capacity, the participant placed their lips around

the mouthpiece and blew out as forcefully as possible. Repeated measures were taken

with a one to two minute rest between trials, until three measurements within five percent

of each other were obtained. The average of these three values was recorded. The

percentage of change in MEP from baseline to the end of respiratory training served as

the primary index for documenting changes in expiratory muscle strength as used in other

studies (Gosselink et al., 2000; Smeltzer et al., 1996; Suzuki et al., 1995).

Aerodynamic Measures

Subglottal pressure was defined as the minimum lung pressure necessary for the

initiation and maintenance of vocal fold vibration for voice production. Subglottal

pressures may be estimated from measures of intraoral pressure as described by

Smitheran and Hixon, (1981). Specifically, the pressure between two oral air pressure

peaks was linearly interpolated during a syllable train of alternating voiceless plosives

and voiced vowels, making it possible to estimate the subglottal pressure during the

voiced vowel segment. The middle four /pa/ syllables produced during the syllable train

were used to measure subglottal pressure. The syllable train was produced at both

"comfortable" and "loud" intensity levels, as determined by the participants themselves.

Recordings were made of three trials from each participant. Intraoral pressure was

collected by inserting a small pitot tube (2 mm. diameter) into the oral cavity between the

lips and behind the front teeth. A pressure transducer (Glottal Enterprises, PTL-1)

recorded the air pressure. The pressure transducer was calibrated at 5 cmHz0 for each

participant with a dedicated calibration unit (Glottal Enterprises MCU-4). The recordings

for pressure were made in a quiet room.









Estimates of phonation threshold pressure were made in a similar manner.

Phonation threshold pressure was defined as the minimal pressure required for initiating

vocal fold vibration (Titze, 1994). The same syllable train used for measuring intraoral

pressure was utilized for estimating phonation threshold pressure. Participants were

instructed to initiate voice at the lowest possible intensity level without whispering. They

were allowed to practice until they were able to barely produce voice. Three successful

attempts were recorded for each participant.

All pressure measures were recorded using PowerLab 8SP with Chart 4 for

Windows software (AD Instruments). The sampling rate was set at 10,000 samples per

second for all pressure measures.

Acoustic Measures

Sound pressure levels. Root-mean-square (RMS) intensity levels were calculated

for intraoral pressure tasks using Matlab, (Mathworks, Inc., Version 6.5.1.) An ATM75

headset cardioid microphone (Audiotechnica), connected to an ART Tube MP

preamplifier, was set at a two-centimeter distance from the participant's mouth to record

intensity level simultaneously with the pressure measures. The intensity of the

microphone signal was calibrated in using 180 Hz pure tone signal at an intensity of

80dB SPL.

Phonetogram. An individual phonetogram, or voice range profile, a display of

vocal intensity range versus fundamental frequency (Titze, 1994) was obtained at each of

the three measurement points for each participant. The phonetogram was collected using

the Voice Range Profile (VRP) software (Kay Elemetrics, Model 4326, Version 2.3). A

Shure SM 48 cardioid microphone, mounted on a microphone stand, was used at a

constant six-inch mouth to microphone distance as recommended by Kay Elemetrics.








Participants were asked to produce maximum variations in fundamental frequency at

minimum and maximum intensity levels while producing an /a/ vowel in ascending and

descending pitch glides. Intensity range was plotted vertically on a graph and frequency

was plotted horizontally in linear units. The participant's productions were plotted in real

time and served as visual feedback for the participant. The participant was instructed to

repeat the maneuver multiple times until he/she felt that they had produced a maximum

plot of intensity and frequency. The total area of the phonetogram was calculated using

Matlab Version 6.5.1. Minimum and maximum intensity and minimum and maximum

fundamental frequency produced by each participant were also recorded.

The Voice Range Profile hardware is internally calibrated. However, the accuracy

of the module was verified at the beginning of the study. Three tones, C4 (261.63 Hz),

C5 (532.25 Hz), and C6 (1046.4 Hz), an octave apart, and encompassing the speaking

range were generated using a BK Precision 5 megahertz function generator attached to a

JBL Pro III speaker. Sound levels were verified using a Radio Shack Sound Level Meter

set on the C Scale. The greatest error was noted at the lowest frequency with an input of

50dB. The Voice Range Profile system read this tone as 60dB. Errors ranging from +5

to +7 dB were noted from 60 to 80dB at this frequency. Minimal errors of 1 to 3 dB were

noted at C5 from 60 to 90dB. The VRP module was most accurate at the highest

frequency tested, C6.

Hoarseness diagram. Measures were made to calculate the coordinates of the

vowel /a/ produced by all speakers on the hoarseness diagram. The hoarseness diagram

is a relatively new measure, which consists of four acoustic measures that are plotted

together (Frolich et al., 2000). The measures, jitter, shimmer, and mean period









correlation, form the horizontal axis of the diagram, which is labeled as the irregularity

component. The fourth measure, the glottal-to-noise excitation ratio, is plotted on the

vertical axis and is labeled as the noise component. This glottal-to-noise excitation ratio

indicates the extent to which the voice excitation is due to a pulse train or due to noise.

Samples analyzed for the hoarseness diagram were obtained from the middle one

second of sustained /a/ productions. Freeware available from the authors' website,

www. physik3.gwdg. de/micha/english/hd. html, was used for determining the irregularity

and noise components.

Training Protocol

Each participant was assigned to one of two groups based on laryngeal diagnosis.

Group 1 consisted of persons with a diagnosis of muscle tension dysphonia or laryngeal

irritation. Group 2 consisted of persons with a diagnosis of benign vocal fold lesions.

Both groups received twice weekly sessions of voice therapy for a period of three weeks

for a total of six sessions. Both groups also received five weeks of expiratory muscle

strength training. Four of the participants with lesions and five of the participants from

the non-lesion group completed the respiratory training first. Conversely, five of the

participants with lesions and four of those without lesions completed the voice therapy

component first.

The voice therapy sessions lasted approximately 45 minutes each and were

conducted by a speech-language pathologist with specialized training in voice therapy. A

script was followed for each session to minimize variability between participants. Topics

covered included vocal and respiratory anatomy and physiology, vocal hygiene, increased

awareness and activation of abdominal musculature during voice production, vocal








projection and resonance exercises, and daily homework activities. The voice therapy

protocol is contained in Appendix G.

Expiratory pressure threshold trainer. The expiratory pressure threshold trainer

used to complete the expiratory muscle-training program was a cylindrical device that

consisted of a mouthpiece and a one-way spring-loaded valve (Figure 2-1). The valve

blocked expiratory airflow until a sufficient threshold pressure was reached to overcome

the spring force. To achieve this threshold pressure, the participant breathed out with an

increased expiratory effort. As long as the threshold pressure was maintained, air flowed

through the device. The device contains an adjustable spring, which allows the required

threshold pressure to be increased. As stated previously, the participants' MEP was

measured at the initiation of the study and at the beginning of each subsequent training

week. The threshold pressure was set at 75% of the participant's MEP at the time of

measurement (pre-training and at the beginning of weeks 1-5). Each training breath

lasted 3 to 4 seconds. Participants performed the exercise five times per set and

completed five sets for five days of the week as reported by other investigators (Baker,

2003; Roy et al., 2003; Hoffman-Ruddy, 2001).


Figure 2-1. Expiratory Pressure Threshold Training Device








Both groups received five weeks of expiratory muscle strength training. MEP was

measured weekly at approximately the same time of day. Participants were provided

with written and verbal instructions for the completion of the training protocol (Appendix

H).

Compliance

Participant compliance during the voice therapy phase was documented by the

participant's practice record (see Appendix G). Likewise, participants completed a

training log daily during the expiratory muscle-training phase (Appendix I). Participants

were provided with written instructions for the therapy component and for the use of the

device during the expiratory muscle-training phase. Participants were instructed to call

their voice therapist or the author at any time if they had questions or if problems arose in

their practice regimen.

Statistical Method

The primary statistical method that was used to examine treatment differences with

respect to the change from baseline scores across the two treatment groups for subglottal

and phonation threshold pressures, voice ratings, and acoustic measures, was the analysis

of variance for repeated measures (ANOVA). The between-subject factors tested were

lesion group, gender, and treatment group. The within-subject factor was the number of

weeks of treatment. Paired sample t-tests were used to analyze MEP since only pre and

post-treatment measures were taken. The Wilcoxon Signed Ranks test was used to

analyze rater evaluations of stroboscopic examinations. All analyses were carried out

using SPSS software version 11.5.

Inter- and intra-rater reliability was carried out on 10 % of the data that was

measured by hand. To test the inter-judge reliability of the dependent variables, a





39


different examiner, a student trained in scoring the various measures, re-analyzed the

data. The student was blinded to the purpose of the study. Pearson r correlations were

used to compare the results between examiners. To test intra-rater reliability, the author

re-analyzed 10% of the data and compared the first set of measures against the second

using Pearson r correlations.













CHAPTER 3
RESULTS

This study determined the effects of expiratory muscle strength training combined

with voice therapy on voice production in two groups of professional voices users, those

with and without vocal fold lesions. The central hypothesis stated that expiratory muscle

strength training combined with voice therapy would produce greater improvements in

voice production than would voice therapy alone. Specific hypotheses were also

proposed that a) both groups of participants would produce greater improvements in self-

ratings of voice symptoms and effort following the combined modality treatment; b) that

both groups would produce greater improvements in perceptual ratings of voice

following the combined treatment modality; c) that both groups would produce greater

improvement in respiratory pressures following the combined modality treatment; d) that

both groups would produce greater improvements in acoustic measures, specifically the

phonetogram and hoarseness diagrams, reflecting improvement in voice production,

following the combined treatment approach; and e) that there would be a significantly

greater response to treatment for the lesion group as compared to the non-lesion group.

In order to test these hypotheses, a repeated measures analysis was used. Main effects

were tested and comparisons of the pre-treatment, mid-treatment, and post-treatment

conditions were completed. The mid-treatment condition indicated the time where each

of the independent intervention methods was examined. At the mid-treatment condition,

the effects of EMST or voice therapy were interpreted. The post-treatment condition

reflected the effect of the combined modality treatment. The independent variables for









the study were treatment order, laryngeal diagnosis, and gender. The dependent variables

were the scores on the VHI, the VRS, the Voice Effort Scale, measures from the

hoarseness diagram, measures made from the phonetogram, respiratory pressures, and

perceptual ratings of voice. Main effects were found for the Voice Handicap Index, Vocal

Rating Scales, MEPs, subglottal pressure produced at loud intensity, phonetogram area,

and dynamic range.

Reliability

High correlations were found between the two sets of measurements made by the

experimenter, suggesting high intra-judge reliability. High correlations were also found

between ratings across multiple judges suggesting high inter-judge reliability. The

Pearson r between the first and second set of measurement (intra-judge reliability) ranged

from 0.81 to 1.00 (Table 3-1). Likewise, for inter-judge measures, there was a strong

positive correlation between the two measurers for the dependent variables listed (Table

3-2) with a range of 0.406 to 1.00. The reliability for the strobe ratings, although

statistically significant, raises a question regarding their consistency and utility. Given

this data, the reliability of the majority of the dependent variables was considered

adequate for the purpose of the present experiment.

Table 3-1: Intra-judge reliability

Dependent Variable r p
MEPs 1.000 <.001
Strobe ratings 0.81 .015
Subglottal pressure comfortable 1.000 <.01
Subglottal pressure loud 1.000 <.01
Phonation threshold pressure 1.000 <.01











Table 3-2: Inter-judge reliability
Dependent Variable r p
MEPs 1.000 <.001
Strobe ratings .406 .026
Subglottal pressure-comfortable .979 <.001
Subglottal pressure loud .959 <.001
Phonation threshold pressure .931 <.001


Perceptual Measures of Effort and Handicap

The first hypothesis stated that both treatment groups would report a greater

reduction in vocal symptoms and vocal effort following EMST-training plus voice

therapy as compared to the voice therapy condition alone. The results of the present

study suggest this to be true, thus the first hypothesis was accepted. The results of the

repeated measures ANOVA showed a significant main effect for the VHI, F (2,20) =

5.593, p = 0.012 and the VRS, F (2,20) = 3.703, p = 0.043. The mean VHI scores were

found to be significantly reduced by 9 points between pre and mid-treatment, F (1, 10) =

8.416, p = 0.016, and 4 points between the mid and post-treatments which was not

significantly different, F (1,10) = 3.68, p = 0.084. Similarly, the VRS scores were found

to reduce by 83 points between pre and mid-treatment, F (1, 10) = 2.48, p = 0.146 and 43

points between the mid and post-treatments, F (1, 10) = 5.493, p = 0.041. Results for the

group averages for the VHI are illustrated in Figure 3-1 and individual VHI scores are

detailed in Figure 3-2. After a single treatment, either EMST training or voice therapy,

83% of the participants had an improved VHI score with the range of score decreasing

from 1 to 32 points. Following the combined modality treatment, 39% of the participants

either had no change in VHI score or an increase of 1 to 14 points. Average VRS scores

are in Figure 3-3 and individual VRS scores are in Figure 3-4. After a single treatment at











the mid-point of the study, 77% of the participants demonstrated a decrease in VRS


scores (range = 3-304 points) while 23% had an increase in their scores (range = 28 to


358 points). Following the combined modality treatment, 61% of the participants


indicated a decreased in VRS scores (range = 7 -235 points) while 39 % of the


participants had an increase in their VRS scores (range = 14 to 136 points).


60



50



S40



30
30 -


20
Pre Mid Post



Figure 3-1. Mean Voice Handicap Index scores before treatment, at the mid-point and
following treatment.
120


100


80 O


60 I


40 -


20 Mid

0 Post
1 3 5 7 10 12 14 16 18
2 4 6 9 11 13 15 17 20

Participant



Figure 3-2. Individual VHI scores before treatment, at the mid-point, and following
treatment.











700


600


S500


S400


300


200




Figure 3-3.
treatment.


Vocal Rating Scale scores before treatment, at the mid-point and following


1000


800


SPre

200 A Mid


0 Post
1 3 5 7 10 12 14 16 18
2 4 6 9 11 13 15 17 20

Participant

Figure 3-4. Individual Vocal Rating Scale scores before treatment, at the mid-point, and
following treatment.


The VHI and VRS were highly correlated across treatment conditions across


different speakers with correlations ranging from 0.751 to 0.894 (Table 3-3).


Correlations for the functional and emotional subscales of the VHI to the VRS were also


calculated. The only significant correlation was found between the functional subscale of


1









the VHI and the VRS prior to the initiation of treatment (r = 0.668, p = 0.002). No

significant main effect was found for the vocal effort scales, F (1, 30) = 0.930, p = 0.343.


Table 3-3. Correlations for VHI and VRS scales as well as VRS to functional and
emotional subscales of VHI.

Treatment Total p Functional p Emotional p
Point Scale r subscale subscale
Pre .894 <.001 .668 .002 .326 .187
Mid .751 <.001 .483 .042 .241 .334
Post .783 <.001 .193 .442 -.003 .991


Voice Measures

The second hypothesis stated that both groups would produce greater

improvements in perceptual ratings of voice post-EMST training plus voice therapy than

with voice therapy alone. The listener ratings for voice quality showed no significant

main effect in the severity of voice quality ratings F (1.544, 24.712) = 3.233, p = 0.068

(Figure 3-5). No significant differences were seen between the pre to mid treatment, F

(1.544, 24.712) = 3.777, p = 0.07) and the mid to post-treatment conditions, F (1.544,

24.712) = 3.603, p = 0.076). The sphericity assumption was not met; therefore, the

Huynh- Feldt correction was used to adjust the degrees of freedom for the averaged tests

of listener ratings. The listener ratings of voice quality used a 5-point rating scale, with a

range from 1 (normal voice quality) to 5 (severe voice quality). The ratings for individual

participants are illustrated in Figure 3-6. After the single therapy was received, 78 % of

the participants were judged to have increased severity of voice quality rating (range =

0.1-2.48 points) while 11% had no change and 11% had an improvement in voice quality

(range = 0.1-0.74 points). Following the combined modality treatment, 72% of the

participants had an improvement in voice quality rating (range = 0.12-2.61 points) with





46


11% of the participants' ratings remaining the same and an additional 17% of the

participants showing increased severity of voice quality ratings post-treatments (range =

0.07-0.1 points).


Figure 3-5. Changes in mean ratings of voice quality pre-treatment, at the mid-point, and
following treatment.


5

4

S3


O Pre

v Mid

SPost


1 3 5 7 10 12 14 16 18
2 4 6 9 11 13 15 17 20

Participant
Figure 3-6. Listener ratings for individual subjects before treatment, at the mid-point
point, and following treatment.


W 9w1


4 -L


I tl








Stroboscopic ratings. Evaluations ofvideostroboscopic examinations were

completed only for the lesion group because these participants were the only ones

expected to have any change in their physical laryngeal examination post-treatment. One

participant did not have a follow-up stroboscopic examination due to medical problems

precluding the use of the nasal endoscope as well as difficulty tolerating the oral

endoscope. Therefore, subjective ratings of the stroboscopic examinations were obtained

for only 8 of the 9 participants. Four of the 8 participants showed partial resolution of

the pathology on post-examination as evidenced by visual examination. For the purpose

of illustration, Figures 3-7 and 3-8 show the pre and post-treatment endoscopic images

for two of the participants with resolution of their lesions. The Wilcoxon signed ranks

test was used to analyze the stroboscopy ratings. A significant improvement was found

for the left vocal fold edge from pre to post-treatment. The change in other ratings

obtained was not statistically significant. The results for all the stroboscopic ratings are

detailed in Table 3-4 and individual ratings for vocal edge and mucosal wave are

illustrated in Figure 3-9 and Figure 3-10 respectively. Ratings of "0" represent normal

with ratings of"6" representing severe abnormality. Pre to post-treatment, 75% of the

subjects showed a decrease in rating for the left vocal fold edge (range = 0.13-1.42

points) while 1 participant (12.5%) showed no change in rating and 1 participant (12.5%)

showed an increased rating. For the right vocal fold edge, 63% of the participants

showed decreased ratings post-treatment (range = 0.125-0.875 points), 1 participant

(12%) had no change and 25% had an increased rating (range = 0.02-0.875 points). For

mucosal wave, 1 participant did not receive ratings due to an inadequate stroboscopic

examination so data are reported for only 7 participants. For the left mucosal wave, 43%








of the participants had an increase in rating (range = 0.25-1.37 points), 1 participant

(0.05%) had no change, and 43% had a decrease in rating (range = 0.125-1 point). For

right mucosal wave, 57% had decreased ratings (range = 0.375-1.25 points), 1 (14%) had

no change, and another 29% had increased ratings pre to post-treatment (range = 0.25-

0.75 points). The second hypothesis was rejected based on the results obtained.














Figure 3-7. Pre and post treatment endoscopic image for participant 10.














Figure 3-8. Pre and post endoscopic images for participant 12.











Table 3-4. Rater evaluations of stroboscopic evaluations pre to post-treatment.
Stroboscopic Parameter Z p Mean S.D. Mean Post S.D.
pre

Vocal fold edge left -2.201 .028 1.05 .722 .639 .54
right -1.103 .270 .89 .55 .676 .46

Mucosalwave left -.135 .892 1.58 1.55 1.16 .60
right -.677 .498 1.03 .66 .91 .41
Amplitude left -.933 .351 1.17 .68 1.07 .52
right -.734 .463 1.17 .64 1.07 .39

Closure -.700 .484 2.99 .89 3.36 .8
Phase -1.344 .176 2.15 .6 2.32 .24


6


5 -


4


3
as
2


1*

0


L edge pre

Left edge post

Right edge pre

Right edge post


Participant


Figure 3-9. Stroboscopic ratings pre to post-treatment for left and right vocal fold edges.


6


5


Sj 4


1 3-

.-1


6 12 1i 1 20
6 7 9 10 12 13 16 20


L mucosal wave pre

L mucosal wave post

R mucosal wave pre

R mucosal wave post


Participant


Figure 3-10. Stroboscopic ratings pre to post-treatment for left and right vocal fold
mucosal wave.


6 7 9 10 12 13 16 20








Pulmonary and Aerodynamic Measures

The third hypothesis theorized that both groups would produce greater

improvement in respiratory pressure post-EMST training plus voice therapy than

following voice therapy alone. This was true for maximum expiratory pressures that

showed a significant main effect from pre to post-treatment t(17) = -8.063, p = <.001.

Mean MEP pre-treatment was 85.92 cmH20 (s.d.=26.14) and mean MEP post-treatment

was 147.87 cmH20 (s.d.=42.27). Increases in MEP post-treatment ranged from 17.88%

to 130% with an average increase of 76.94%. Results for MEP are shown in Figure 3-11

with individual measures of MEP illustrated in Figure 3-12.

A significant main effect in estimated subglottal pressure for loud phonation was

observed increasing with treatment, F (2, 20) = 5.234, p = 0.015. The greatest increase

occurred after the first treatment, F (1,10) = 5.847, p = 0.036. No main effects were

obtained for estimated subglottal pressures at comfortable loudness, F (2, 20) = 0.406,

p =. 672 or for phonation threshold pressures, F (2, 20) = 0.297, p = 0.746: No

statistically significant differences were indicated from pre- to mid- or mid- to post-

treatment for subglottal pressure at comfortable loudness or for phonation threshold

pressure. Based on this data and the data for MEPs, the third hypothesis was accepted.


Figure 3-11. Maximum expiratory pressure changes before and after treatment.









300



200



00 Pre


0 Post
1 3 5 7 10 12 14 16 18
2 4 6 9 11 13 15 17 20

Participant


Figure 3-12. Individual changes in maximum expiratory pressure pre to post-treatment.

Acoustic Measures

The fourth hypothesis predicted that both groups would produce greater

improvements in frequency and time domain measures post-EMST training plus voice

therapy than following voice therapy alone. This proved to be true for the phonetogram

(Figure 3-13). The area of the phonetogram increased significantly from pre to post-

treatment, F (2, 20) = 21.667, p = <0.001. After the first treatment condition, the increase

was not statistically significant, F (1, 10) = 0.859, p = 0.376. Between mid and post-

treatment, the area increased significantly, F (1, 10) = 48.74, p = <0.001. Changes for

each participant for phonetogram area are detailed in Figure 3-14 with examples of pre

and post-treatment phonetograms shown in Figures 3-15 and 3-16. After the first

treatment condition, 14 of 18 subjects (77%) had an increase in phonetogram area (range

= 60-435 units). The remaining 4 subjects (23%) had a decrease in area (range = 157-293

units). After the combined modality treatment, 16 of 18 subjects (89%) showed an

increase in phonetogram area (range = 8-596 units) while area for 2 of the 18 (11%)

remained the same or decreased (range = 0-84 units). Dynamic range increased









significantly pre to post-treatment F (2, 20) =7.153, p = 0.007 (Figure 3-17) with the

greatest change occurring between the mid and post- treatment conditions, F (1, 10) =

20.793, p = 0.001. No significant increase occurred between the pre and mid-treatment

conditions, F (1, 10) = 0.443, p = 0.521). Individual data for change in dynamic range is

detailed in Figure 3-18. After either EMST or voice therapy, 10 of the 18 participants

(56%) demonstrated an increased dynamic range (range 1-19 dB). For 8 of the 18

participants (44%), dynamic range did not change or decreased (range = 0-18 dB). After

the combined modality treatment, 13 of the 18 participants (72%) demonstrated an

increase in dynamic range (range = 3-22 dB) while 5 of the 18 (28%) stayed the same or

had a decrease in range (range = 0-6 dB). Frequency range did not demonstrate a

significant main effect, F (2, 20) = 1.830, p = 0.186. A significant increase did not occur

between the mid and post-treatment conditions, F = (1, 10) 4.344, p = 0.064. There was a

significant difference noted for frequency range between males and females. This

suggests that the increase in the phonetogram area following treatment was primarily a

consequence of a change in the intensity dynamic range.

1000


900


800


< 700


600


500


rre


Yost


Figure 3-13. Change in phonetogram area following treatment.











1400

1200

1000

800

600

400

200
0


TII Y"'Y
ai 4





1 7 10 12 14 16 18
2 4 6 9 11 13 15 17 20

Participant


0 Pre

SMid

w Post


Figure 3-14. Changes in phonetogram area for individual participants pre, mid, and post-
treatment.



















-tra
--- -- -- --- -- -- -- -- ------ -- ---- -

Figure 3-15. Pre-treatment phonetogram for participant 7.

.......... ........ .. .. ... ....... ..
-. ................................... .............. .............---------------------_ -------


Figure 3-16. Post-treatment phonetogram for participant 7.


~ "
PII m
XQ PZ
:::

I











50

48

464




42

40

38
Pre Mid Post


Figure 3-17. Average dynamic range pre-treatment, at the mid-treatment point, and
following treatment..
70

60 -

50 TT

vA 40 TA

30 0 Pre

20 0 Mid

10 __Post
1 3 5 7 9 11 13 15 17
2 4 6 8 10 12 14 16 18

Participant

Figure 3-18. Change in dynamic range for each participant pre, mid, and post-treatment.


The measures of irregularity, F (2, 20) = 0.637, p = 0.539 and noise, F (2, 20) =


1.883, p = 0.178, in the voice, taken from the hoarseness diagram, did not show a


significant main effect for either treatment group. Therefore, the fourth hypothesis was


accepted based on the improvements in the phonetogram.


Differences between Lesion and Non-lesion groups


The fifth hypothesis proposed that the group with benign vocal fold lesions would


demonstrate greater improvements overall than the non-lesion group following the


combined treatments of EMST and voice therapy compared to voice therapy alone. Only










one dependent variable, estimated subglottal pressure at loud intensity levels, showed a

statistically significant difference between the lesion and non-lesion groups F (1,30) =

26.543, p = <. 001). This is highlighted in Figure 3-19. Data for the remaining

dependent variables is shown in Table 3-5. Therefore, the fifth hypothesis was rejected.

Table 3-5. Effect of lesion on dependent variables.

Dependent Variable F p
Voice Handicap Index .004 .950
Vocal Rating Scale 1.020 .321
Speaking effort at work .602 .444
Speaking effort socially .029 .866
Subglottal pressure, comfortable intensity 3.681 .065
Subglottal pressure, loud intensity 26.543 <.001
Phonation threshold pressure .041 .841
Phonetogram area .090 .766
Dynamic range 3.59 .068
Frequency range .597 .446
Irregularity .510 .481
Noise .725 .401


16-
14
12 T--
0



Pre Md Post
4


Pre Mid Post


Non-lesion


Pre Mid Post


Lesion


Figure 3-19. Estimated subglottal pressure at loud intensity for non-lesion and lesion
groups.













CHAPTER 4
DISCUSSION

The central hypothesis for this study was that expiratory muscle strength training

combined with voice therapy would produce greater improvements in voice production

than would voice therapy alone. In general, this was the case. Main effects were found

for more than half of the dependent variables examined. Thus, it seems that the

combination of EMST training and voice therapy is a beneficial treatment paradigm for

professional voice users. While there was some significant effect found when comparing

the pre to mid-treatment conditions, the number of dependent variables that responded to

the independent treatments were fewer. Certainly it could be argued that it was simply

the greater duration of treatment that influenced the dependent variables as opposed to

the combination of EMST and voice therapy this argument cannot be resolved from the

study. The findings of this study do indicate that EMST plus voice therapy when

combined did result in different and more improved function than EMST or voice therapy

alone. Future study of additional research arms is suggested.

Perceptual Measures of Effort and Handicap

Voice Handicap Index. The first hypothesis was that both therapy groups would

report greater reduction in vocal symptoms and vocal effort following the combined

treatments as compared to the voice therapy condition alone. Significant improvements

from pre to mid-treatment were found for the VHI but not for the mid to post-treatment

condition, suggesting that the individual treatments (EMST or voice therapy) resulted in a

decrease in vocal handicap as perceived by the participants involved in those treatments.








The individual subject ratings support the pre to mid post treatment contrast showing that

83% of the patients responded positively on the VHI following the individual treatments.

The finding of no significant effect from mid to post-treatment indicates that

combining treatments did not further benefit the participant by decreasing vocal handicap

to a greater extent. It may be that once the participant perceived a change in the status of

their voice, they responded more immediately to the change. Typically, changes in the

physical condition of the vocal folds and/or the voice quality occur relatively quickly

when a patient is exposed to treatment. Most therapeutic protocols can result in a

positive effect within 4 to 6 weeks. Furthermore, when a person receives attention from a

clinician, one of the more immediate effects can be a change in the patient's perception of

how they are responding to the treatment. This early change may have to do with

common factors that are present across therapies, with the likely component being the

client-therapist relationship (Haas, Hill, Lambert, & Morrell, 2002). This effect tends to

lessen as the treatment continues and the patient habituates to the newfound voice quality.

The focus of outcomes research is often placed on the patient's perspective about the

impact of a disease or disability, and its treatment (Rosen & Murry, 2000). Therefore,

the patient's perception of improvement following treatment is an important factor to

indicate the success of treatment (Rosen et al., 2000). The reduction in VHI scores

observed in this study suggests that a treatment of EMST or voice therapy can be

successful and is consistent with the decreased scores reported by various researchers

using this tool across a variety of populations (Roy et al., 2003; Roy et al., 2001; Rosen,

Murry, Zinn, Sullo & Sonbolian, 2000).








Vocal Rating Scale. The statements on the VRS were pertinent to the problems of

professional voice users on the job, with content focusing on specific situations and needs

in the workplace. The majority of the statements were of a functional nature. This scale

was developed by the author to complement the VHI that has functional (physical),

emotional, and social subscales related to overall functioning including home and social

settings. The VRS was found to be highly correlated with overall VHI score but not

highly correlated with the functional or emotional subscales. So while there is some

redundant information being gathered from the use of the VHI and VRS it is also

apparent that there is information on the VRS that is distinct from the VHI.

Interestingly, the VRS scores did not show significant change pre to mid-treatment

but did change mid to post-treatment. The individual participant data showed that 61% of

the participants decreased their VRS scores following the combined modality treatments.

This suggests that the combined modality treatment served to decrease the participant's

perception of how their voice difficulty was impacting functional activities in daily

living. There must be a reason that the participants felt like there was a greater

improvement with the combined modality treatment. First, it could be that the VRS was

more specific to issues surrounding work and as the participants spent more time in

therapy, they were able to learn how to modify their voice in the workplace, therefore

impacting their rating of their vocal difficulties. And while the VHI and VRS were

highly correlated, correlation does not equate with specificity. While items may have a

relationship with each other, certain questions are able to specifically target a

participant's perception of how well they are performing in a particular situation better

than others. For example, participants were asked to rank the degree to which their job








performance was affected by their voice problem or the degree to which they lost their

voice after prolonged talking. These statements are much more highly focused and

specific than the statements on the VHI. For example, the VHI asks the participant to

rank how their voice difficulty restricts their personal and social life, a much more

general question that does not target specifically work performance. The professions of

these participants were all very reliant on the use of the voice. Any degradation of the

voice in these professional voice users would highly impact their job performance. So, it

appears that the VHI and the VRS are complementary but yet different in their specificity

with regard to how voice disorders handicap individual situations.

Vocal effort scales. The ratings of effort to speak, both at work and socially, did

not show a statistical main effect. However, notably 13 of the 18 participants (72%)

indicated a decrease in effort to speak at work and 10 of the 18 (55%) reported decreased

effort when speaking socially. Accordingly, the effort scores at work decreased from a

mean of 48 to 36 points and the social effort scores decreased from a mean of 43 to 34

points. While these scores were not statistically significant, the decrease in effort scores

appears to have some clinical significance. As previously discussed, patient perception

of a problem and its subsequent improvement has always been an important indicator of

clinical improvement. If the patient perceives decreased effort to speak, especially if this

was a complaint prior to treatment, as was the case for the current participants, then

decreased effort represents significant clinical change whether or not statistical

significance is reached. It is possible that the minority of participants who did not report a

reduction in perceived effort might have interpreted the increased awareness of how they

produced voice as an increase in effort. Participants were required to learn new









behaviors in the course of therapy. These new behaviors require time to become

habituated and the 8 to 9 week time period during treatment may have not been adequate

in the case of these participants for habituation to take place. Habits appear to develop as

positive reinforcement is repeated over time (Aarts, Paulussen & Shaalma, 1997). The

exact time frame required for establishment of habit vary among individuals but may take

six months or more to become fully established (Vallis, Ruggiero, Greene, Jones, Zinman

et al., 2003). So, it is reasonable to assume that some participants may have an increased

awareness of their previous vocal behaviors but may not have completely mastered the

new vocal behaviors, thereby resulting in no significant change in their perception of

vocal effort.

Voice Measures

Listener ratings. The second hypothesis predicted that both groups would produce

greater improvements in voice quality as judged by perceptual ratings made by qualified

listeners post EMST training plus voice therapy than with voice therapy alone. This

hypothesis was supported by listener evaluation of voice quality for all of the

participants. The listener ratings between the pre and mid-treatment conditions and the

mid and post conditions were not significant.

There were 2 subjects that had fairly high ratings when evaluated in the mid-

treatment condition, which may have skewed the results to some degree. Participant 12,

a middle school classroom teacher, showed a large increase in voice quality rating at the

mid-treatment test point. She reported that she might have been developing a throat

infection at the time. Additionally, she was in a period of high personal stress due to

illness in her family. Also, participant 18 may have an emerging diagnosis of spasmodic

dysphonia, in the opinion of the author and the clinician who provided the treatment for








the study. His voice quality was typically inconsistent throughout the study. However,

when the results were re-analyzed with these 2 participants removed from the database,

the contrast of the mid to post-treatment condition remained non-significant. Finally, it is

important to recall that the majority of participants reported improvement by self-rating

and that other quantitative measures showed improvements as well.

The main reason that the listener ratings did not result in significant effects with

either the individual or combined modality treatments is most likely due to the large

number of voice qualities that were in the mildly disordered range. Because so many of

the voices, regardless of the presence of lesion, were rated as mild pre-treatment,

obtaining a change in quality as a function of treatment was difficult.

The individual subject data does positively describe however that 72% of the

participants improved their voice quality with the combined modality treatments,

although the differences in ratings were small.

Stroboscopic ratings.

Lesion status. Half of the participants in the lesion group had some

resolution of their benign vocal fold lesions from mid to post-treatment. The other half

did not show measurable improvement on stroboscopic examination post-treatment. This

is similar to results obtained by Holmberg and colleagues (2001) with a group of women

with vocal fold nodules. Following voice therapy, 80% of their participants demonstrated

a decrease in nodule size and edema. Resolution of lesion is an important clinical

indicator of improvement, reflecting a return to a more normal physiologic status. This

further supports the positive outcome from EMST training combined with voice therapy.

Certainly it could be argued that the change in physical status of the vocal folds was not









related to either of the specific treatments but rather to time in that the lesions, following

an 8 to 9 week time course of rehabilitation, would respond better than at a 4 week time

frame. Obviously, in order to determine this potential criticism, one would have to study

multiple treatment techniques while manipulating duration of treatment. Of note, the

participants whose lesions resolved all had reductions in VHI from pre to post-treatment

and all but one had reductions in VRS scores pre to post-treatment.

Other physiologic parameters. Only one parameter, the left vocal fold

edge, showed a statistically significant improvement pre to post-treatment. The mid to

post-condition was not compared with stroboscopic examination because it was a time-

consuming process requiring the participant to go off-site. Given the participants'

already full work-schedule and large time commitment to the study, only the pre-post

condition was examined. In performing the stroboscopic examinations, 3 different

clinicians were involved using 2 different stroboscopy systems. This created differences

in technique as well as light and color variations between systems that may have affected

the ratings. A few of the raters did not rate all of the strobe parameters for each

examination as they felt the samples were inadequate for evaluation. Furthermore, in our

study there were 8 raters evaluating a total of 16 video samples. It is possible that the

smaller number of raters and samples yielded variable results. In contrast, Poburka and

Bless (1998) found high agreement among raters for ratings of vocal fold edge, mucosal

wave, and amplitude with lower agreement for glottal closure and phase symmetry. They

had a total of 39 raters with ratings of 45 video samples. Rather than drawing specific

conclusions about the changes that occurred in the laryngeal dimensions in the

stroboscopic examinations, it appears that more careful control of the stroboscopic








system requirements as well as the number of clinicians who are rating the examinations

must be considered. These concerns were evident based on the poor inter-rater reliability

found for the stroboscopic parameters.

Pulmonary and Aerodynamic Measures

The third hypothesis stated that both groups would produce greater improvement in

respiratory pressure post-EMST training plus voice therapy than following voice therapy

alone. This was the case for MEP and for subglottal pressure produced at loud intensity

level.

The average increase in MEP of nearly 77% was consistent with increases in MEP

found in previous studies with healthy individuals, performers, and high-school band

students (Baker, 2003; Hoffman-Ruddy, 2001; Sapienza et al., 2001). All participants

increased MEP from pre to post treatment.

It is reasonable to conclude that the increased MEP could aid in improving the

overall voice quality produced by the participants. For the group that initiated the therapy

program with EMST first, the increase in MEP was associated with a positive change in

VHI score. For the group that initiated EMST training second with regard to treatment

order, there was a significant decrease in VRS scores. Given the correlation between the

VHI and VRS, it is reasonable to conclude that both groups exposed to the EMST

training paradigm experienced reductions in vocal handicap and/or vocal difficulties. A

greater demand for increased respiratory drive exists during long speaking tasks (Hixon,

1987). This necessitates the active use of expiratory muscles to increase the drive for

phonation, as passive lung recoil is not able to meet the pressure demands. The increased

strength of the respiratory muscles may enhance the individual's ability to generate and








maintain the required pressure because the driving force of the respiratory system is

increased (Baker, 2003), thereby improving the physiological function of the vocal folds.

Subglottal pressure for normal conversation typically ranges from 4 to 6 cmHzO

(Baken & Orlikoff, 2000). Subglottal pressures for participants without lesions were

within the normative range pre (3.6-8.7 cmH20) and post-treatment (5.3-8.6 cmH20). The

participants identified as having vocal fold lesions demonstrated an average higher

subglottal pressure both pre (5.1-16.4 cmH20) and post-treatment (4.3-14.8 cmH20). This

is consistent with reports of higher transglottal air pressure in patients with benign

lesions. In a study by Holmberg and colleagues (2003), transglottal pressures were 2 to 6

standard deviations higher for patients with lesions as compared to normal subjects. This

may reflect the increased mass and stiffness of the vocal fold and/or hyperfunctional

voice production. No significant main effect was seen in subglottal pressure produced at

comfortable intensity post-treatment. This makes sense for the non-lesion group because

their subglottal pressures were already within normal limits. The lack of change in

subglottal pressure for the lesion group was disappointing although there was a drop of

approximately 1.5 cmH20 following the combined treatment.

Subglottal pressures associated with increased vocal intensity typically range from

8 to 20 cmH20 (Baken & Orlikoff, 2000; Hixon, 1987). The non-lesion group showed a

range of 5.2-11 cmH20 pre-treatment, which is within normal limits but on the lower end

of the normal range (Isshiki, 1964). Following the combined modality treatment, the

range increased from 6.5-14, a 1 to 3-cmH20 increase. As a group, the pressures moved

closer toward the normal range. The ability to increase pressure for loud talking is likely

an important contributing element to the decreased VRS scores discussed previously.









The capability to increase the subglottal pressures used to produce a louder voice

certainly would impact professional voice users function in the workplace, particularly

the teachers, ministers, and attorney involved in this study. Additionally, the ability to

increase pressure may be important when talking in conditions of increased noise, a

situation often encountered by professional voice users. In increased noise, individuals

may have difficulty monitoring their intensity levels and, as a result, may increase their

effort by increasing vocal fold tension and medial vocal fold compression (Case, 1991;

Titze, 1994). By increasing subglottal pressure with the increased expiratory muscle

strength, the need for increased medial compression should be reduced and laryngeal

tissue trauma from excess compression is minimized (Titze, 1994).

Phonation threshold pressures did not show significant main effect. Phonation

threshold pressure is reported to be between 3 to 4 cmH20 (Baken & Orlikoff, 2000;

Titze, 1994). The non-lesion group demonstrated a range of 1.97 to 10.6 cmH20 pre-

treatment with a range of 2.7-5.4 cmH20 post-treatment. This indicates some move

toward more normalized phonation threshold pressures for the non-lesion group. The

lesion group did not show a similar trend (pre-treatment: 2.25-7.24 cmH20; post-

treatment: 1.96-9.90 cmH20).

Admittedly, producing phonation threshold pressures was a difficult task for the

participants to perform, as is evident from the wide range of values acquired. During

completion of the phonation threshold task it appeared as if the participants were

performing the task correctly, yet the pressure values remained high. With cases of voice

disorders, obtaining the softest phonation could have resulted in a glottal configuration

that was more likened to a whisper rather than a more abducted state for soft phonation.








In fact, while not perceptually notable, the participants could have been producing the

task with a certain degree of hyperfunction thus resulting in a higher phonation threshold

pressure. Colton and Casper (1990) discussed how patients can obtain a targeted voice

quality with a variety of glottal configurations, therefore concluding that what you hear is

not always related to what is expected in glottal configuration.

Acoustic Measures

The fourth hypothesis predicted that both groups would produce greater

improvements in acoustic measures taken from the phonetogram and hoarseness diagram,

post-EMST training plus voice therapy than following voice therapy alone. Although

significant increases in the phonetogram area and the dynamic range were observed,

measures of irregularity and noise did not show any significant main effect.

Phonetograms are representative of the output of the entire phonatory mechanism

(Coleman, 1993) and have been considered as a measure of voice coordination (Ikeda,

Masuda, Manako, Yamashita, Yamamoto, et al., 1999). The increased area shown mid to

post-treatment is considered a positive outcome of the combined modality treatment

(Speyer, Weineke, VanWijck-Warnaar, & Dejonckere, 2003). Eighty-nine percent of the

individual participants increased phonetogram area with the combined treatment as

opposed to 77% with the individual treatments. So, there was some slight benefit, on the

order of 10% improvement by combining the treatment techniques. Increased respiratory

drive, coupled with the participants' ability to efficiently control airflow as learned from

the voice therapy exercises, likely contributed to the increased dynamic range that, in

turn, contributed to the increased phonetogram area.

Frequency range and dynamic range were not targeted in the treatment protocol.

The treatment focused on coordination of breath stream and phonation, or voice onset,








increased resonance, and projection of the voice. The combined modality treatment may

have resulted in the improvements seen in both of these parameters. A practice effect

could be responsible for the changes but is not likely. Previous research has shown the

vocal range to extend only slightly (1.5 to 2.1 semitones) across repeated administrations

of the phonetogram (Akerlund, 1993; Gelfer, 1986). On average, the frequency range of

participants in the current study increased 6.5 semitones. Exercises targeting the

extension of the frequency range were not included in the therapy protocol so this

expansion likely reflects improved phonatory output as a result of the combined treatment

modality rather than change as a result of a practice effect. Variability of dynamic range

has also been investigated previously, and while it may differ up to 10dB across

administrations, on average the variation is about 3dB (Gramming et al., 1991). The

participants in this study demonstrated a significant change mid to post-treatment (i.e.

combined modality) with an average increase of 10dB. Individual participants increased

dynamic range from 3 to 22 dB above and beyond the 1 to19 dB range acquired with the

individual treatments. This change may also be attributed to the combined treatment

modality and reflects increased power.

The hoarseness diagram, particularly its noise component, is applicable for highly

irregular oscillations (Michaelis et al., 1997). The measures of irregularity and noise did

not demonstrate a significant main effect. The group mean for the irregularity component

prior to treatment was 4.304 and decreased to 4.084 post-treatment. The group mean for

noise was 1.128 pre-treatment and decreased to 0.867 post-treatment. While this is

generally a sensitive measure, it may not be sensitive enough to detect differences in

moderately or mildly impaired voices. According to Frolich and colleagues (2000), voice








disturbances not primarily affecting the degree of glottal closure or the regularity of vocal

fold vibration could not be expected to lead to significant differences from those of

normal voices (p. 716). The stroboscopic ratings for the participants in this study also did

not indicate a disturbance to glottal closure or irregularity of vocal fold vibration. This

explains why no significant improvement was seen for this measure. This measure was

selected prior to enrolling participants in the study. It was unanticipated that so many

voices would be normal or mildly impaired as ranked by listeners. Participants

volunteered for the study based on their vocal complaints and symptoms and voice

quality ratings were not a part of the inclusion criteria.

Differences Between Lesion and Non-Lesion Groups

The fifth hypothesis stated that the group of participants with benign vocal fold

lesions would show greater improvements in all measures following the combined EMST

training and voice therapy. The only difference between the two groups was for

subglottal pressure at loud intensity level. No significant differences were found between

the lesion and non-lesion groups for any of the other dependent variables. The original

inclusion criterion was to accept all participants with benign lesions. In retrospect, this

may have been in error. Benign lesions vary in size, type, and position on the vocal fold.

The benign lesions involved in this study ranged from those that altered vocal fold mass

to those that had little impact on vocal fold vibration (i.e., small vocal nodules). Future

studies of the impact of therapy on benign lesions needs to be more specific by focusing

on one type of benign lesion controlling for the most specific details of the lesion such as

size, impact on vibration, and impact on glottal closure.








Combined Modality Treatment

Most research in voice therapy has examined the effect of a single variable, such as

hydration or vocal hygiene, on treatment outcomes. Only a few studies have investigated

the outcome of combining two or more treatment approaches for patients with voice

disorders. Basiouny (1998) studied the efficacy of the Accent Method of voice therapy,

which combines respiratory and phonatory exercise. The therapy protocol consisted of

ten sessions over a five-week period. Like the present study, he found improvements in

vocal and aerodynamic parameters. Murry and Woodson (1995) compared the effect of

Botox treatment alone to Botox treatment combined with 5 sessions of voice therapy for

patients with adductor spasmodic dysphonia. Therapy targeted muscle hyperfunction and

regulation of airflow during phonation They found a prolonged effect of Botox when it

was combined with voice therapy. Holmberg et al. (2001) found that significant changes

in voice parameters occurred after direct facilitation of voice and respiration phases of

treatment. The direct facilitation phase utilized reduction of loudness and yawn-sigh

exercises. The respiration phase focused on decreased effort for speech breathing and

improved breath management. Holmberg's experiment was similar to the present study

but was spread over 15 sessions occurring in a 4 to 6 month time frame. The current

study is the first to combine specific expiratory muscle training with voice therapy. The

problem with studying these two techniques is that there is not a great deal of information

about the effects of either EMST or voice therapy with voice disordered patients when

they are exposed to these individual treatments. Each of the studies references above

examined one particular patient group (those with nodules and spasmodic dysphonia).

The lack of data on EMST and how it effects patients with voice disorders makes the

current data a bit more difficult to interpret and gives room for one to argue that it is









merely time in therapy that resulted in the positive findings for the combined modality

treatments. Therefore, following the results of this study, it is necessary to examine the

impact of combined modality treatments with a more complicated design. For example,

in order to determine whether it was truly the combination of EMST plus voice therapy

as opposed to simply the time spent in therapy, additional arms would have to be added

to the design. The 4 researched arms would include participants enrolled in EMST plus

voice therapy, voice therapy plus EMST, EMST only, and voice therapy only across an

8-week period.

Strengths of the Present Study

The present study is an easily followed program with strong applicability to

professional voice users. The time frame is reasonable and results occur within a

relatively short period of time (8 to 9 weeks). The EMST component is device-driven

and serves to illustrate quantifiable results to participants immediately, providing positive

reinforcement for continued practice. Many participants commented that they felt this

aspect of the program made a discernible difference in their daily respiratory patterns for

speech. The voice therapy component is easy to follow and allows for daily practice by

the participant in a reasonable amount of time. The emphasis is on speaking activities

that can be readily incorporated into routine speaking activities for the professional voice

user. Further, this program requires minimal training for voice clinicians and could

easily be incorporated into their clinical practice.

Limitations of the Current Study

The therapy protocol for this study was scripted so that all participants received the

same instruction. However, this did not allow for individual differences in speed of

learning for each phase of the treatment. Likewise, there was no provision for dealing









with specific problems that are commonly encountered in voice therapy, such as

excessive laryngeal tension. Additionally, some participants may have benefited from

additional practice with the resonance and projection exercises which were covered in

two sessions. These concepts were difficult for some participants to fully master within

the limited time frame.

The study did not address maintenance of the behaviors learned. This is an area

that is lacking in most therapy programs. As previously noted, habituation of learning

may take up to 6 months to occur. Therefore, a provision for intermittent follow-up may

be of use in the future.

Application of the Protocol

The current protocol demonstrated success in affecting change in many voice

parameters that are important for professional voice users. The reasonable time frame is

attractive to the larger population of professional voice users. The time demands should

be easily incorporated into most work schedules. The activities utilized can be quickly

applied in the employment setting, minimizing time away from work. The ultimate result

of the voice improvements seen should be a reduction in missed workdays due to vocal

problems.

Future Studies

Future studies should add additional research arms as stated above as well as

comparing longer treatment protocols to the current protocol to fully determine the

impact of time on these vocal changes. Additionally, the development of a maintenance

component to study carryover of behavior is needed. Comparison of participants

receiving maintenance therapy to those receiving no follow-up should be explored.








To determine if changes in the phonetogram were a result of the combined

modality treatment or resulted from a practice effect, a study of changes in the

phonetogram over time, with no treatment offered, should be undertaken for a sample of

professional voice users.

Other types of voice treatment, specifically Vocal Function Exercises (Stemple et

al., 1994) and the Lee Silverman Voice Treatment (Ramig, Countryman, Thompson, &

Horii, 1995) should be combined with EMST training. By comparing different

combinations of treatments, recommendations for designing a highly effective program

for occupational voice users could be made.

The VHI and VRS should be utilized in future studies as they effectively

demonstrate clinically significant improvements in self-perception. Additionally, the

phonetogram appears to be a sensitive and efficient means for quantifying vocal change.

As previously mentioned, other acoustic analysis may be more sensitive for documenting

vocal improvement in professional voice users than the measures used in the current

study.

Further, the prevention of voice disorders is an area lacking in research. Future

research, of a longitudinal nature, could be undertaken with professional speakers to

determine if the combination of EMST and voice therapy might help delay or prevent the

development of vocal problems in occupational voice users.











APPENDIX A
INFORMATION FLYER



DO YOU HAVE

PROBLEMS WITH

YOUR VOICE?




To be eligible you must:
*have complaints of voice problems
*use your voice at work for 4 or more hours a day
*be between 21 and 65 years of age
*have no history of cardiac, lung, neuromuscular, or
immune system disease, or hypertension
have no history of smoking or tobacco use in the last
five years

As part of this study you will receive voice therapy. For more
information contact Judy Wingate at (352)392-2046, ext. 221,
Department of Communication Sciences and Disorders,
University of Florida.

This study is supervised by Christine Sapienza, (352) 392-2046,
ext. 221














APPENDIX B
SCREENING QUESTIONNAIRE


WRITTEN QUESTIONNAIRE


I. Demographics

Name
Address


Sex


City
Birthdate
Tel: (H)
Occupation
# of hours worked per w


State Zip Code

(W)
# years in occupation


'eek


II. Physical Characteristics

Height Weight

1. Rate your health on this scale compared to others your age.


1= very good 1 2 3
2= good
3= fair
4= poor
5= very poor

2. List the major surgeries you have had within the last 5 years.


3. Are you being treated at the present time for any medical conditions? If yes, please
specify.


I I I I









4. Please list your medications.



5. Do you have asthma or other conditions that affect your breathing? _Y N
6. Y N Have you had a recent cold or flu?
7. Y N Do you have allergies?
8. Y _N Have you ever smoked tobacco products?
If you used to smoke, for how many years and when did you stop?


III. Past Medical History (Have you had a past medical history for any of the
following? Circle as many as apply. )


high blood pressure
kidney disease
urinary tract infections
diabetes
gastritis
diverticulosis
gall bladder disease
lung disease
hormonal problems
ovarian or uterine problems
prostate problems
spastic bowel syndrome
cardiac problems
gastroesophageal reflux disease


thyroid problems
liver diseases
bleeding problems
peptic ulcer
hiatal hernia
skin conditions
colitis
pancreatitis
arthritis
neurological problems
history of cancer
hernias
incontinence


10. Do you plan to increase or decrease your physical activity over the next several
months? Yes/No If yes, how do you plan to change?

11. How much water do you drink each day?

12. How much caffeine do you drink each day?


13. Do you use an amplifier or microphone when you are at work?


Y N


14. Please estimate the percentage of time you spend talking while at work.


15. Do you participate in any of the following when you are not at work?

Singing Coaching Other activities requiring voice use __ (Please
describe)














APPENDIX C
VOICE HANDICAP INDEX


Instructions: These are statements that many people have used to describe their voices
and the effects of their voices on their lives. Circle the response that indicates how
frequently you have the same experience.

0 = Never 1 = Almost Never 2 = Sometimes 3 = Almost Always 4 = Always
Part I-F
1. My voice makes it difficult for people to hear me.
0 1 2 3 4
2. People have difficulty understanding me in a noisy room.
0 1 2 3 4
3. My family has difficulty hearing me when I call them throughout the house.
0 1 2 3 4
4. I use the phone less often than I would like to.
0 1 2 3 4
5. I tend to avoid groups of people because of my voice.
0 1 2 3 4
6. I speak with friends, neighbors, or relatives less often because of my voice.
0 1 2 3 4
7. People ask me to repeat myself when speaking face-to-face.
0 1 2 3 4
8. My voice difficulties restrict personal and social life.
0 1 2 3 4
9. I feel left out of conversations because of my voice.
0 1 2 3 4
10. My voice problem causes me to lose income.
0 1 2 3 4










Part II-P
1. I run out of air when I talk.
0 1 2 3 4
2. The sound of my voice varies throughout the day.
0 1 2 3 4
3. People ask, "What is wrong with your voice?"
0 1 2 3 4
4. My voice sounds creaky and dry.
0 1 2 3 4
5. I feel as thought I have to strain to produce voice.
0 1 2 3 4
6. The clarity of my voice is unpredictable.
0 1 2 3 4
7. I try to change my voice to sound different.
0 1 2 3 4
8. I use a great deal of effort to speak.
0 1 2 3 4
9. My voice sounds worse in the evening.
0 1 2 3 4
10. My voice "gives out" on me in the middle of speaking.
0 1 2 3 4


Part III-E
1. I am tense when talking to others because of my voice.
0 1 2 3 4
2. People seem irritated with my voice.
0 1 2 3 4
3. I find that other people don't understand my voice problem.
0 1 2 3 4









4. My voice problem upsets me.
0 1 2 3


5. I am less outgoing because of my voice problem.
0 1 2 3 4
6. My voice makes me feel handicapped.
0 1 2 3 4
7. I feel annoyed when people ask me to repeat.
0 1 2 3 4
8. I feel embarrassed when people ask me to repeat
0 1 2 3 4
9. My voice makes me feel incompetent.
0 1 2 3 4
10. I am ashamed of my voice problem.
0 1 2 3 4
















APPENDIX D
VOICE RATING SCALE



Name Date

Please rank the severity of each statement by marking an "x" on the line provided.




1. I have problems with my voice.


mild


moderate


severe


2. The degree to which my job performance is affected by my voice problem is:


mild moderate severe

3. The degree to which I have thought about changing my job because of my voice

problem is:



mild moderate severe

4. The degree to which people have difficulty understanding me on the phone is:



mild moderate severe

5. The degree to which I have difficulty being understood in noisy environments is:



mild moderate severe

6. The degree to which I have difficulty projecting my voice is:







80



mild moderate severe

7. The degree to which my throat feels sore after prolonged talking is:


moderate


severe


8. The degree to which my voice quality changes after prolonged talking is:


moderate


severe


9. The degree to which my voice tires after prolonged talking is:




mild moderate severe

10. The degree to which I lose my voice after prolonged talking is:


moderate


severe

















APPENDIX E
ESTIMATE OF VOCAL EFFORT


Please rank the severity of each statement by marking an "x" on the line provided.




The degree of effort I have needed to speak at work this week has been:


moderate


severe


The degree of effort I have needed to speak socially this week has been:


moderate


severe
















APPENDIX F
STROBOSCOPY ASSESSMENT FORM



Exam #:

Please rank the following by circling your choice:


Vocal Fold Edge normal
Left 0
Right 0


Mucosal Wave




Amplitude


normal
Left 0
Right 0


mod.decrease
2 3
2 3

mod decrease


normal


rough/irregular
5
5


absent
5
5


no movt


Left 0
Right 0


Glottic Closure complete posterior irreg. Spindle
0 1 2 3


Phase Closure


whisper


anterior
4


normal


hourglass incomplete
5 6


hyperadduction


1 2 3 4 5


Adapted from University of Wisconsin, Stroboscopic Assessment Form













APPENDIX G
THERAPY PROTOCOL


Session 1: Controlled Breathing (Clinician Script)

Your voice is produced using a power source and a sound source. The power source for
your voice is air coming from your lungs. The sound source is the larynx. The lungs
take in air. As you exhale and close your vocal folds, the vocal folds vibrate and produce
sound. The sound is then shaped into speech by the tongue, teeth and lips.

In order to sing or to produce long utterances, it is useful to prolong exhalation. The
abdominal muscles help control exhalation. Good posture allows the abdominal muscles
to work best. Your spine should be straight with shoulders relaxed. Your head should be
in a neutral position with the chin parallel to the floor. Ideally, you should be standing
and your knees should be relaxed so that you can shift your weight easily.

You should do each of these exercises daily until your next session and record your
practice on the exercise sheet. We will go through these together and make sure
that you understand each exercise and can perform it correctly.

Exercise 1:

1. Correct your posture as described above.
2. Blow out all of your air.
3. Now, breathe in through an open, relaxed mouth. Concentrate on the movement of
your abdominal muscles. They will move in an outward direction as air comes in.
You should not experience movement in your chest or shoulders as you breathe in.
4. Next, exhale slowly on "f' and concentrate on keeping the flow of air steady. You
should feel the abdominal muscles begin to pull in as you near the end of your breath.
5. Repeat slowly until you feel comfortable with this exercise.

Exercise 2:

1. Place yourself in correct posture.
2. Blow out all of your air.
3. Pant like a dog for 5 seconds. Feel the abdominal muscles contract and release. Be
sure to let your tongue hang out of your mouth as you pant.
4. Repeat 5 times.








Exercise 3:

1. Place yourself in correct posture.
2. Blow out all of your air then inhale.
3. Blow out 5 times on "sh" as if you are blowing out candles. Stop and start the airflow.
4. Repeat 4 times. Feel the movement of your abdominal muscles.

Session 1,Exercise 1 (Subject Practice Worksheet):

1. Place yourself in correct posture.
2. Blow out all of your air.
3. Now, breathe in through an open, relaxed mouth. Concentrate on the movement of
your abdominal muscles. They will move in an outward direction as air comes in.
You should not experience movement in your chest or shoulders as you breathe in.
4. Next, exhale slowly on "f' and concentrate on keeping the flow of air steady. You
should feel the abdominal muscles begin to pull in as you near the end of your breath.
5. Repeat slowly until you feel comfortable with this exercise.

Session 1, Exercise 2:

1. Place yourself in correct posture.
2. Blow out all of your air.
3. Pant like a dog for 5 seconds. Feel the abdominal muscles contract and release. Be
sure to let your tongue hang out of your mouth as you pant.
4. Repeat 5 times.

Session 1, Exercise 3:

1. Place yourself in correct posture.
2. Blow out all of your air then inhale.
3. Blow out 5 times on "sh" as if you are blowing out candles. Stop and start the airflow.
4. Repeat 5 times. Feel the movement of your abdominal muscles.

Practice Record


Date Time Ex. 1-# reps. Ex. 2 # reps Ex. 3 # reps.









Session 2: Voice Onset (Clinician Script)

Today's session will focus on how you turn your voice on. First we will review the
exercises from the previous session. Please let me know if you have any questions.
(Review exercises from session 1 on previous sheet.)

As you speak, it is important that you use exhalation. Remember, the air flowing out of
your body is the power supply for your voice. With the following exercises, you will
learn to coordinate airflow and your voice. To avoid any excessive force when turning
on your voice, you will begin by letting out a small amount of air and then turning on
your voice. When speaking, you will take a short inhalation followed by a prolonged
exhalation. The drawing below illustrates this idea. Your voice will continue as long as
you have air available.


voice


exhalation



inhalation
We will go over the next set of exercises. Again, you will practice these exercises
daily until your next session. Record your practice in the table provided. You may
also use these exercises as a "warm-up" before speaking on the job.


Exercise 1:

1. Take a short breath in. Begin to release air and prolong the sound "ha". Feel the air at
your mouth as you begin to exhale. Sustain the "ha" for as long as you can.
2. Repeat 4 times. Remember to always begin with air.

Exercise 2:

1. Repeat the exercise above with each of the following vowels: "he", "ho", "who".
Be sure to do 4 repetitions of each, sustaining as long as you can.


Session 2, Exercise 1 (Subject Practice Worksheet)

1. Take a short breath in. Begin to release air and prolong the sound "ha". Feel the air at
your mouth as you begin to exhale. Sustain the "ha" for as long as you can.
2. Repeat 5 times. Remember to always begin with air.









Session 2, Exercise 2:

1. Repeat the exercise above with each of the following vowels: "he", "ho", "who".
Be sure to do 5 repetitions of each, sustaining as long as you can.


Date Time "ha" # "he" #reps "ho" # reps "who" #
reps. reps.









Session 3: Breathing for Speech (Clinician Script)

Review the exercises from Session 2 and ask if there are any questions or concerns.

So far, you have practiced breathing and coordinating the breath and your voice. Today,
you will begin applying what you have learned to speech. You will continue to use a
short inhalation followed by a prolonged exhalation. Remember to let air out as you
speak. You will practice first with sentences and then with short paragraphs.

You will practice these exercises daily until your next session. We will go through
them together to make sure you understand how to do them.

Exercise 1:

1. Take a breath.
2. As you exhale, count for as long as possible. Try to count at a normal speed. See if
you can count a little further on each breath as you practice.

Exercise 2:

1. Take a breath.
2. As you exhale, read a sentence from the list provided.
3. Continue with 10 sentences. Take a breath in the middle of the sentence if you feel
you are running low on air.

Exercise 3:

1. Select a paragraph to read. (Two are provided for you.)
2. Read aloud, just as you read the sentences in exercise 2. Take a breath at each
punctuation mark and whenever you feel the need to breathe.









3. Read aloud for 5 minutes each day.

Session 3, Exercise 1 (Subject Practice Worksheet)

1. Take a breath.
2. As you exhale, count for as long as possible. Try to count at a normal speed. See if
you can count a little further on each breath as you practice.

Session 3, Exercise 2:

1. Take a breath.
2. As you exhale, read a sentence from the list provided.
3. Continue with 10 sentences. Take a breath in the middle of the sentence if you feel
you are running low on air.

Session 3, Exercise 3:

1. Select a paragraph to read. (Several are provided for you.)
2. Read aloud, just as you read the sentences in exercise 2. Take a breath at each
punctuation mark and whenever you feel the need to breathe.
3. Read aloud for 5 minutes each day.


Date Time Exercise 1 Exercise 2 Exercise 3










Session 4: Resonance and Projection (Clinician Script)

Review exercises from Session 3 and answer any questions or concerns.

As sound waves travel through the vocal tract, they are modified by the oral and nasal
cavities, including the palate and teeth. The sound waves may be dampened or enhanced
as they travel through the tract. In general, the more open the tract is, the more the voice
can resonate and achieve a rich, full sound. Your vocal tract functions as an inverted
megaphone that helps to move sound out. As you perform the following exercises, try to
keep this megaphone position.

Think about voices you have heard in places where people are making announcements on
the street (newspaper vendor or fair barker) or those speaking without microphones. The
words are stretched out slightly, the mouth is open, the quality is slightly nasal, the
volume is slightly louder, and the pitch varies more than in normal speaking. All of these
combined help the voice to carry or "project". Remember the demonstration provided for
you and practice phrases produced in this way during the week.

As usual, we will practice your exercises together. Remember to record your
practice each day on your practice sheet.


Exercise 1:

1. Keep your mouth and vocal tract in an open, relaxed position.
2. Hum on "m". Feel the buzz created around your nose and sinus cavities.
3. Now say the following phrases using an exaggerated nasal quality. Continue to feel
the buzz in your face.

Oh my Oh me Oh no Oh my no Oh me oh my

3. Say the phrases again at a slow rate. Repeat at a fast rate.

Exercise 2:

1. Keep your mouth and vocal tract in an open, relaxed position.
2. Say the following phrases. Concentrate on keeping an open mouth position. Prolong
the vowels and emphasize the accented syllables. Be sure to use breath.

Fore! Hey! Go away Hang on
Over here Help me You may go Where were you?
Hurry up Come here Hey you Who is it?
Oh no Hello there Go and see You may go
Hooray Up, up and away Hold your horses
Read all about it Take it away Places everybody