An Evaluation of the Breathing Strategies and

Maximum Phonation Time in Musical Theater

Performers During Controlled Performance Tasks
*Tommi Sliiden, †Sara Beck, and *Ian MacDonald, *†London, UK

Summary: Objectives. Breathing strategies for athletic dancing are quite different from those needed for legato
singing. This study considers the respiration data recorded by a telemetric heart rate monitor, mask, and gas analyzer
collected from 20 musical theater professionals performing set tasks.
Methods. Scores were taken of maximum phonation time (MPT), relative oxygen uptake, and heart rate immedi-
ately after three tasks: (1) singing only, (2) dancing only, and (3) singing while dancing. Scores were also collected of
the ability to sustain unbroken notes immediately before performing and then immediately after each of the above
tasks (1), (2), and (3). Vital capacity scores were recorded at the beginning and at the end of the testing schedule.
A questionnaire took demographic information and asked questions regarding performer perception, training,
and experience. Most were aware of breathing inconsistencies and adjustments related to changes of task and/or
direction.
Results. (Means) Tidal volume was constant for (1) singing and (2) dancing; MPT reduced by 65.2% for singing
while dancing; and minute volumes reduced by 16% but relative oxygen uptake (mL/kg/min) remained unchanged,
despite increased heart rates, and when singing while dancing compared with dancing only. The mean MPT dropped
from 20.4 seconds to 7.1 seconds (change of 65.2%) between the at rest and postsinging while dancing scores.
Conclusion. The likelihood of developing vocal dysfunction (overtime) and the risk of compromising performance
aesthetic lead the researchers to the conclusion that further study is required in this area.
Key Words: musical theater performers–respiration–maximum phonation time–relative oxygen uptake–heart rate.

BACKGROUND
The physical act of breathing forms a vital part of various tasks,
such as respiration and phonation; changes in intra-abdominal
and thoracic pressure, combined with peri-abdominal and tho-
racic muscle activities, also affect fluid dynamics, postural support,
movement, efficacy of the kinetic chain, and so on with priori-
tization of effort changing, depending on the activity being
undertaken.
During physical exercise, such as dancing, bodily process nor-
mally focuses on increasing gas exchange to a level high enough
to meet the demands of the increased muscle work. Particular-
ly taxed are the respiratory muscles simultaneously contributing
to postural support and movements of the body. In contrast to
this, singing demands that the control of the airflow takes pri-
ority. If the singing is relatively static, postural support worries
cause less concern to the performer as the need for rapid gas
exchange for larger postural movements is less of an issue.
From personal experience and from discussions (informal re-
search) over the years, combining both singing and dancing, at
the athletic end of the performance spectrum, is often fraught
with difficulty. Indeed this research has made it even more dif-
ficult to imagine how these two tasks can successfully be
combined seven (or more) days per week without such physi-
ological adaptations as to warrant concern for both the musical
Accepted for publication June 30, 2016.
From the *University College London, 74 Huntley Street, London WC1E, UK; and the
†Trinity Laban Conservatoire of Music and Dance, London.
Address correspondence and reprint requests to Ian MacDonald, 11 Falmouth House,
1 Seaton Close, London SE11 4ET, UK. E-mail: ian.macdonald11@icloud.com
Journal of Voice, Vol. 31, No. 2, pp. 253.e1–253.e11
0892-1997
© 2017 The Voice Foundation. Published by Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.jvoice.2016.06.025
aesthetic and the performer’s well-being. Yet this combined skill
is one of the basic tenants of musical theater.
Being relatively unique as a study, the aim was to design tasks
that would imitate a performance scenario as closely as possi-
ble. The hope was to take into consideration not only freedom
of movement and ease of voicing but also to somehow recreate
the heightened sense of excitement (arousal/anxiety) perform-
ers experience prior and during performance.
Reliable data on respiratory function, laryngeal function,
degrees of movement, acoustic measures, and heart rate (HR),
as well as creating (or finding) settings that would not under-
mine the performers’ psychology, were collected and these were
discussed as part of the preparation process. Realizing the spe-
cific research question, setting the hypotheses, and deciding on
the equipment and the most authentic experimental arena were
debated in the few months, preparing the research proposal with
experts from different performance and performing arts medi-
cine backgrounds.
Various types of instrumentation were considered
(Appendix S1) before deciding on the use of a portable gas ana-
lyzer that fitted snuggly on the performer like a waistcoat, a simple
recording device to monitor what was sung, a stopwatch, and
decibelometer to record and guide the participants’ voicing. This
allowed us to collect a broad spectrum of respiratory and acous-
tic data.
Questionnaire
A questionnaire collected the base demographic data and pro-
vided some information on training, experience, and perceived
difficulties and expectations regarding the participants’ abili-
ties and concerns (Appendix S1).
253.e2
Breathing during exercise and breathing
during singing
The main difference in the pedagogy between the use of breath
in singing and its use in dancing begins early on in the teach-
ing environment. Because the prosody of the art itself requires
the fuel for voicing to adapt to the fluctuations within the music,
early on it becomes part of the artistic kitbag. In dancing, fitness,
suppleness, strength, balance, and proprioception are the initial
focus. Breath and body awareness are certainly a part of the
learning-to-dance-well process, but it is not difficult to argue that
the control of breathing in singing is more crucial, particularly
in the early days.
Therefore, this piece of research focuses on the type of pro-
jected singing and athletic dancing associated with musical theater
and extensive runs. Making comparisons between the breath-
ing patterns required for singing and dancing separately and
singing while dancing proved challenging as there is sparse re-
search in this area.
Research on breathing when singing has tended to involve clas-
sically trained singers standing still, often focusing on subtle
movements of muscle activity and/or changes to sound quality,
whereas research examining breathing patterns in dancers tends
to focus on the demanding cardiovascular aspects of the activity.
This sparse research on breathing in the field of musical theater
(much, to this day, due to limitations of suitable testing meth-
odology) required the researchers to draw heavily from sources
such as sports exercises associated with running and cycling.
Running, in particular, is similar to many dance movements.
During exercise, gas exchange, particularly the efficient elim-
ination of the biproduct carbon dioxide, is crucial. Once athletic
activity reaches a certain intensity, lactic acid enters the blood
stream and needs to be broken down. To maintain a healthy blood
pH, sodium bicarbonate buffers the increase of acid by break-
ing it down into water and carbon dioxide. It is essential that
this carbon dioxide elimination process remains effective and
consistent. Too high a level of pH will have a negative effect
on a wide range of chemical processes in the body.
A healthy body adapts well to these increased demands, spon-
taneously increasing minute volume by increasing tidal volumes
and/or respiratory rates (RR). Although there are many factors
involved in regulating this automatic response, it can be over-
ridden consciously when needed, such as in swimming under
water or singing long phrases.
TABLE 1.
A Comparative List of Patterns of Breathing for Singing and for Dancing
Breathing During Singing
1. Consciously controlled breathing
2. Varied minute volume
3. Controlled expiration
4. Lower air flow
5. Airway—some impedance most of the time
6. Adducted vocal folds most of the time
7. Valsalva maneuver discouraged
8. Breathing pattern regulated by musical structure
9. Changes likely to depend on musical prosody
Journal of Voice, Vol. 31, No. 2, 2017
At rest, a highly coordinated diaphragm, abdominal, and in-
tercostal muscle recruitment keeps the shape of the rib cage
preferentially undistorted.1 During exercise (dancing), the in-
crease in minute volume comes predominately from increases
in tidal volume (which can more than triple in some athletes).
This increased volume is gained by recruiting peri-abdominal
musculature to contract, pulling the ribs inferiorly and medi-
ally, which in turn increases the volume of expired breath utilizing
expiratory reserve volume within the lungs. This action enables
a greater diffusion rate as the distal part of the lungs allows for
greater gas exchange. During times of high ventilation, in-
creased RR is preferred.1–3
Rapid breathing and increased airflow during high intensity
exercise are facilitated by an unconstricted airway, and this can
be helped by keeping the breathing constant, or if necessary
making only slow and gradual changes. Problems in this area
can include exercise-induced asthma or bronchoconstriction,
exercise-induced laryngeal obstruction, vocal cord dysfunc-
tion, paradoxical vocal fold motion, adduction or collapse of
supraglottic structures, and inappropriate closure of the glottis.4,5
Differences between breathing during dancing
(athletic movement/exercise) and breathing during
singing (athletic voicing/musical theater singing)
The following comparative table (Table 1) bullet-points the main
points of differences between the strategies employed for breath-
ing during singing and those needed for dancing. These are the
perceived norms taught in many institutions and are well rec-
ognized within the literature (Table 1). The point of labeling them
clearly here next to each other is to highlight the differences likely
to cause physiological and ergonomic challenges for the musical
theater performer when attempting to sing while dancing.
It is not difficult to see the potential for respiratory difficulty
when we examine this table. It is true that some items in table
“a” could appear in table “b” and vice versa. But we do not need
to concern ourselves with the vagaries of stylistics for the pur-
poses of this study. In the more athletic arenas of dance and
singing, having to sing while dancing will likely create some-
thing of a challenge.
Certainly when it comes to artistry (the non-subjective type
such as sustaining long high pitches as required within the score,
or performing a jete and singing an accented note, again as
Breathing During Dancing
1. Spontaneous breathing
2. High minute volume
3. Free expiration
4. Higher airflow
5. Airway—low impedance most of the time
6. Abducted vocal fold most of the time
7. Valsalva—reasonably common
8. Breathing patterns regulated by respiratory need
9. Slow progressive changes related to respiratory need
Tommi Sliiden et al Evaluation of Breathing Strategies and Maximum Phonation Time
demanded of the performance), it is easy to appreciate how the
musical theater performer may run into difficulty.
Thus, demanding cardiovascular work involves spontaneous
breathing with high minute volumes and high airflow, open
airways and glottis with vocal folds abducted, and as regular
breathing rate and tidal volumes as possible.
Singing, on the other hand, requires controlled breathing pat-
terns, with focus on low and controlled airflow through the partly
adducted vocal folds during expiration, with varied and often
sudden changes to rate, air volume, and airflow depending on
the prosody of the piece.
Among singers, there are many schools of thought with regard
to optimum breathing for any given role and vocal quality. But
despite a greater body awareness in this group of performers,
it has been demonstrated that there is often a discrepancy between
perceived and actual actions, and even among highly trained
singers there can be a lack of accurate knowledge of the mecha-
nisms associated with their singing performance.6,7
Control of rib cage movements, in particular, is considered
to be of great relevance in singing technique. Some studies suggest
that breathing when singing resembles more that of rest than
of athletic activity, with less differentiation between rib cage
and abdominal wall excursion, with the rib cage and abdomi-
nal muscles jointly activated so that the respiratory system acts
as a single compartment.8,9 Other studies, most of them em-
ploying trained opera singers, indicate that there are many
individual variations in muscle recruitment patterns from the rib
cage, abdominal wall, and accessory respiratory muscles, sug-
gesting that professional opera singing does not require strictly
uniform breathing strategies.10,11
Respiratory muscles also play an important role in postural
support. The great variations in intra-abdominal pressure and
muscle vector tension (diaphragm, pelvic floor, and peri-
abdominal musculature) assist stability and poise. This stability
is an important factor of balance, delicate control of airflow during
phonation, as well as larger movements of limbs and trunk as-
sisting smooth transference of weight and changes to the center
of gravity. The more stability that is needed, the more the breath-
ing musculature is recruited into this process, with increased
contraction reducing lower rib excursions and creating a flatter
diaphragm resulting in smaller excursions for breathing.
Maximum phonation time
When singing or playing wind instruments, the ability to control
the airflow is vital, especially when attempting long, sustained
phrases. This ability is more challenging during the dynamic task
of singing and dancing simultaneously.
Maximum phonation time (MPT) (or maximum phonation du-
ration) measures the ability to sustain a long note (voicing) and
is a simple and non-invasive method often used as a rough es-
timate of breath control and vocal proficiency. Procedures of
testing MPT are not standardized, but often the vowel sound /a:/
or consonants /s/ and /z/ on an habitual pitch and with normal
amplitude are used. Various factors affect MPT, but reported
normal values in non-singers include an MPT of 30.0 seconds
(standard deviation [SD]: 10) for men and 22.6 seconds (6.0)
for women.
253.e3
MPT has been shown to have a correlation to various types
of voice disorders and can easily be used to help in assessing
progress of treatment, surgery, or voice therapy.12–18 None of this
literature or previous research, however, included the possible
impact of increased cardiovascular demands from physical ac-
tivity as a cofactor.
Although dysfunction of the respiratory muscles does occur,
the issue for singers often lies in the imbalances in muscle func-
tion or commonly hyperfunction of musculature, leading to
irregularities at the level of the vocal folds (such as insufficient
approximation). This can lead to increased airflow and reduced
breath efficiency and is associated with shorter MPT, and often,
but not always, a breathier sound.14,16,19 Breathiness does not nec-
essarily indicate pathology, and is often used as a vocal quality
by jazz singers to create a more intimate sound, or as one study
suggests be an indication of a hypofunctional voice production
and vocal fatigue.20
Vocal training has regularly been shown to affect MPT; one study
demonstrated classically trained female singers having higher MPT
than non-trained women, whereas in classically trained men this
difference only pertained to the higher pitches and greater volumes
(in other words, sustained voicing outside of their normal speak-
ing range). Results showed a declining MPT with an increasing
sound pressure (3 seconds per 10 dB).21 This was not factored into
the analysis as the number of the participants was too low.
Vital capacity (VC, the maximum volume of air that can be
expelled following a maximum inhalation) in normal healthy
adults is usually between 4 and 5 liters, depending on chest size,
rib cage, physical fitness and health, age, gender, and the testing
posture. One study found no significant correlation between VC
and MPT,19 whereas in another there was a correlation between
the classically trained female scores.21
Aims
This study investigates the breathing patterns of musical theater
performers during some common performance tasks by care-
fully considering the measuring instrumentation. It evaluates the
data with respect to the respiratory effort required for pro-
jected voicing and athletic dancing, and considers the impact upon
performance ability and artistic aesthetic.
Hypotheses
The following are the study hypotheses:
(1) The equipment chosen for collecting real-time data from
musical theater performers would yield reliable and useful
information.
(2) The singing while dancing task would produce the great-
est respiratory measures for all compared with the singing
only and dancing only tasks.
(3) There would be significant differences for respiratory
scores taken for singing only compared with dancing only.
(4) The MPT for the four sets of MPT scores would de-
crease incrementally as the testing sequence progressed
and this will show an inverse with HR.
(5) Decreasing MPT was a result of increased efforts in car-
diovascular activity in performers.
253.e4 Journal of Voice, Vol. 31, No. 2, 2017

TABLE 2.
Demographic Data
Female Male All
Age (mean) 25.3 24.7 25.00 (SD: 03.45)
Height (m) (mean) 1.63 1.78 01.71 (SD: 00.11)
Weight (kg) (mean) 57.01 70.26 63.60 (SD: 10.32)
Body mass index (kg/m2) (mean) 21.20 22.17 20.36 (SD: 05.33)
Years as professional performer 05.40 (SD: 03.70)
Abbreviation: SD, standard deviation.

(6) VC should remain the same at the beginning as at the
end for these healthy performers.
METHODS
A protocol in 10 sections was designed to collect respiratory and
acoustic data from the participants (Table 2).
The final methodology selected for data collection was a result
of several small pilot experiments (Appendix S1: Pilot infor-
mation) addressing ease of performance movement and continuity
of vocalization for the performers. The protocol that follows is
the result of the most optimal outcome from apparatus and process
according to testimonials from the performers and consistency
of the readings (Figure 1).
Demographics
Twenty healthy participants took part: 10 female and 10 male
(age range: 21–34 years), all currently working professionals in
London’s West End theaters.
Inclusion and exclusion criteria
Inclusion was straightforward in so much as the self-reported
vocal and physical (dance) health of the performer must meet
three specific points.
(1) No reported illness within the past month
(2) That they had performed yesterday and are expected to
perform tomorrow
(3) That they would take part in the study within a few hours
of performing that evening
Instrumentation
A telemetric HR monitor, mask, and gas analyzer (METAMAX
3B, CORTEX, 04229 Leipzig, Germany) (Figure 1), as used in
studies of athletes and contemporary dancers,22–25 collected the
majority of data on respiration. An independent decibelometer
(Precision Gold, Model N05CC, fast response) (Maplin Elec-
tronics Ltd, Brookfields Way, Manvers, Rotherham S63 5DL)

FIGURE 1. Performer with instrumentation (gas analyzer, fitted demonstrating various athletic dance moves unhindered).
Tommi Sliiden et al Evaluation of Breathing Strategies and Maximum Phonation Time 253.e5

TABLE 3.
Musical Excerpts
Test Pieces
Show (Bold) Song From the Show Extracted Segments (3 min)
Prestudy ONLY “The opening,” Act 1 London cast live show recording, 1994: 1’45”–3’35” + 4’25”–5’35”
(musical Cats)
Wicked (all) “Loathing” (“What is this Original Broadway cast recording 2003, track 4: 1’56”–3’28,”
feeling?”), Act 1 repeated
Singin’ in the “All I do,” Act 1 London cast rehearsal accompaniment recording, 2012: first 40 s,
Rain (women) repeated
Singin’ in the “Singin’ in the rain,” Finale, London cast live show recording, 2012: Last 1’15”–2’25,” repeated
Rain (men) Act 2
Top Hat (all) “Let’s face the music and London cast live show recording, 2012: whole number, 2 min,
dance/Top Hat,” Finale, repeated (repeating lyrics from previous phrase twice during
Act 2 dance break)

(A-weighted, low setting [30–100 dB]) recorded the phonation
intensity levels. A pitch pipe kept the performers on the allo-
cated pitches, and a portable recording device (iPod Nano with
built in microphone) was secured to the performer so as not to
interfere with either the singing or dancing tasks. The prere-
corded accompaniment for performance was played on a portable
sound system.
Performance music
Table 3 details the music chosen for a prestudy pilot and the final
pieces selected, as well as timings showing the content used from
the final edits.
Prerecorded excerpts from suitable musical numbers were
chosen on the advice of the current or resident musical direc-
tors. All of the musical excerpts had been slightly edited to provide
a more homogenous section, in which both singing and dancing
were required more or less continuously throughout. Simply put,
this was achieved by adding singing to any “dance breaks” or
repeating the extract, if too short, to make up an exactly 3-minute
test piece.
Pitch selection for sustained pitch exercise to
demonstrate MPT (Table 4)
Literature has stipulated “average intensity” being 60–70 dB for
speech and average habitual speaking pitch being 128Hz for men
and 225Hz for women.
recorded on the iPod Nano placed exactly 25 cm directly in front
of them.
The HRs of all the participants were noted before they per-
formed and were labeled as “subject resting heart rate.” A return
to this resting heart rate was required for each performance (length
of time it took varied slightly for each) before the next task could
commence. During the three performances (singing only, dancing
only, and singing while dancing), continuous measurements were
taken of HR, relative oxygen uptake volume, minute volume,
tidal volume, and RR.
Protocol sequence (lasting between 30 and
45 minutes)
(1) Resting HR was recorded and labeled resting HR.
(2) MPT was taken voicing a steady “Ah” at 80 dB
(+/−5 dB).
(3) The task singing only was performed and scored (1).
(4) MPT was taken immediately following the task.
(5) Resting HR returned to baseline.
(6) The task dancing only was performed and scored (2).
(7) MPT was taken immediately following the task.
(8) Resting HR returned to baseline.
(9) The task singing while dancing was performed and
scored (3).
(10) MPT was taken immediately following the task.

Setup for sustained phonation VC was taken prior and post protocol sequence. Participants
Performers were instructed to phonate for as long as possible were asked to empty their lungs fully after a full inspiration (in
in their normal voice (no effect or qualities) on the vowel “Ah” their own time) with three consecutive tries with no extra breaths
(as in car) and at a steady volume of 80 dB (+/−5 dB), which in between. Although as stated earlier the correlation between
was guided by a decibelometer that they held. The sound was MPT and VC is unclear, measures were obtained from the par-
ticipants and these were compared with similar measures from
a bicycle exercise study.3
TABLE 4.
Pitches for Male and Female Participants
Pitch Sustained on a “Ah” [a:/] Questionnaire
A short questionnaire (17 questions) gathered additional infor-
Female A3/220.0 Hz
mation from participants regarding their technique, the length
Male C3/130.8 Hz
and type of training received, and their knowledge of the science
253.e6
and physiology of singing and dancing (Appendix S1: Partici-
pant questionnaire). The researchers also hoped to gain some
insight, from these answers, as to which direction future re-
search should take.
Statistical analysis
The statistical package SPSS 17.00 (SPSS Inc. Released 2008.
SPSS Statistics for Windows, Version 17.0. Chicago) was used.
Analysis of variance analyzed the differences between group
means. Analysis tested for normality, comparisons of means, dif-
ferences of variables between tasks, groups, gender, and
any correlation between MPT and HR. Significance was set at
P < 0.05.
RESULTS
Table 5 shows the results from the instrumentation.
Differences due to gender and musical pieces
Overall, the analysis showed few significant differences (P < 00.5)
for gender or musical excerpts. Scores that were significant were
related to demographic and physical characteristics. The fol-
lowing are the differences between the genders: mean height of
men were on average 15 cm taller and 13 kg heavier. One show
piece (“All I do”) required performers to need lower tidal volumes
(35% less) and lower minute volume (44% less) ONLY with the
singing only task. However, these differences did not signifi-
cantly influence VC scores.
Differences in heart rate (1/min)
The average resting heart rate was surprisingly high at 80.7 beats
per minute, but with no significance between gender or for the
musical pieces. Significant scores for HR readings were taken
after each of the three performance tasks.
Differences in relative oxygen uptake (VO2 max)
(mL/kg/min)
Some significance for most scores between groups (men/
women) was observed. The greatest increase and the greatest
significance were found between singing and dancing, and
between singing and the singing while dancing task, with
P > 0.001 for all. There was no significance between the tasks
involving dancing.
Minute volume (liter/min)
Significance for comparisons of all tasks lies between P < 0.001
and P < 0.026. (Table 5). Interestingly, for the dance task com-
pared with the singing while dancing task, minute ventilation
(MV) of men decreased by 14.82% (P < 0.026) and that of women
decreased by 17.59% (P < 0.001).
Changes to respiratory rate
The rate at which the performers needed to respire necessarily
increased when the activity they were being asked to do
Journal of Voice, Vol. 31, No. 2, 2017
required much more energy. The results reflect these antici-
pated changes (Table 5). No significant differences were seen
between groups; however, significant differences between
tasks were observed. So comparing changes between singing and
dancing, men increased their RR on average 24.13 units (SD:
5.94), an increase of 120.46% (P < 0.001), and women in-
creased their RR by 20.01 units (SD: 15.25), an increase of
100.91% (P = 0.002).
When comparing singing task with the singing while dancing
task, the results were as follows: Men’s RR increased at an
average of 9.33 units (SD: 4.95), a change of 46.58% (P < 0.001),
whereas women’s RR increased at an average of 9.82 units (SD:
6.60), a change of 49.52% (P = 0.001).
Tidal volume (liters)
There were significant changes for this parameter except between
the singing task and the dancing task.
Differences in MPT
The average MPT was 20.4 seconds, widely spread with an SD
of 5.5 seconds. At a higher HR, after dancing, it went down to
as low as 7.1 seconds. There were no significant differences for
the four MPT readings for gender or musical pieces. There were
significant differences related to the tasks being performed, as
clearly represented in Figure 2 (mean MPT).
Figure 2 clearly describes the changes in the ability to sustain
voicing in relation to the task being performed. There is a marked
difference in phonation time once the much more physical task
of dancing is included, with a reduction efficacy of more than
50%.
Vital capacity (liter)
VC was generally unaffected in our healthy professional per-
formers, although it is interesting to note that for men the percent
of VC used during the MPT reduced from 90.15% (SD: 18.02)
to 75.11% (SD: 17.69), a reduction of 15.03% (SD: 13.87), which
at 16.68% gave a significance score of P = 0.008.
Results from the questionnaire
The results from the questionnaire showed consensus between
our group of performers around issues of preparation and train-
ing (Table 6).
Apart from the information that may be of potential use in
future studies, this table shows a clear pattern in the issues per-
formers find difficult in musical theaters. These included a 100%
agreement that dancing had the greatest potential to negatively
impact upon prosody and phraseology of singing because the
control of the breathing apparatus diminished rapidly over time.
The majority of them also agreed that the cardiovascular (80%)
demands of dancing and the types of movement (jumping, spin-
ning, running, twisting, etc) (75%) had the most negative effects
upon singing.
Most (70%) were aware that repertoire or direction required
constant modification of breathing strategies for singing while
dancing.
 Tommi Sliiden et al
TABLE 5.
Results of Respiration From the Gas Analyzer and From MPT Readings
During 3 Min Singing and
Rest/Activity Rest (R) Singing (S) Dancing (D) Dancing (S + D) Statistically Significant Differences (P < 0.05)
Heart rate M: 75.75 (11.98) M: 91.88 (11.10) M: 143.46 (11.75) M: 146.59 (12.64) Statistically significant differences when comparing any task
(1/min) W: 85.63 (10.18) W: 101.35 (19.18) W: 139.06 (30.09) W: 156.03 (10.13 combination (all between P = 0 < 0.001–P = 0.004), apart from
d: no difference d: no difference d: no difference d: no difference) D versus S + D for women (P = 0.063)
Relative oxygen M = 5.94 (1.43) M = 9.02 (1.55) M = 37.39 (8.56) M = 36.60 (11.25) Statistically significant differences when comparing any task
combination (all P < 0.001), apart from D versus S + D; no

Evaluation of Breathing Strategies and Maximum Phonation Time

uptake W = 5.43 (0.75) W = 7.55 (1.07) W = 28.98 (3.22) W = 28.31 (3.35)
(mL/kg/min) d: no difference d = 1.48; P = 0.024 d = 8.41; P = 0.009 d = 8.29; P = 0.038 statistically significant difference for either gender
Minute volume M = 11.54 (2.95) M = 24.43 (2.61) M = 64.84 (13.58) M = 55.23 (9.48) Statistically significant differences when comparing any task
(L/min) W = 9.18 (1.06) W = 16.75 (4.00) W = 45.08 (6.43) W = 37.15 (6.54) combination. (all between P < 0.001 and P < 0.026)
d: no difference d = 7.68; P < 0.001 d = 19.75; P = 0.001 d = 18.07; P < 0.001 Special interest: D versus S + D
Men: 9.61 L/min (11.44) [14.82%], P < 0.026
Women: 7.93 L/min (2.58) [17.59%], P < 0.001
Respiratory rate M = 13.52 (4.76) M = 20.03 (1.24) M = 44.17 (6.24) M = 29.36 (5.03) Statistically significant differences when comparing any task
(1/min) W = 15.59 (3.11) W = 19.83 (3.64) W = 39.85 (15.04) W = 29.65 (7.70) combination (all between P < 0.001 and P < 0.0461)
d: no difference d: no difference d: no difference d: no difference Special interest: S versus D
Men: increased by 24.13 (5.94) units [120.46%], P < 0.001
Women: increased by 20.01 (15.25) units [100.91%], P = 0.002
S versus S + D
Men: increased by 9.33 (4.95) units [46.58%], P < 0.001
Women: increased by 9.82 (6.60) units [49.52%], P = 0.001
Tidal volume (L) M: 0.96 (0.38) M: 1.32 (0.13) M = 1.51 (0.28) M = 2.03 (0.32) Statistically significant differences when comparing any task
W: 0.63 (0.14) W: 0.96 (0.27) W = 1.08 (0.20) W = 1.38 (0.39) combination apart from S versus D; no statistically significant
d: no difference d = 0.33l, P = 0.018 d = 0.431, P = 0.001 d: no difference difference for either gender
% of VC used (TV/ M = 22.87 (9.50) M = 32.22 (7.15) M = 36.89 (11.15) M = 49.33 (11.19) Statistically significant differences when comparing any task
VC) W = 22.57 (77.96) W = 35.93 (16.07) W = 38.84 (12.50) W = 51.89 (22.72) combination (all between P < 0.001 and P = 0.011), apart from S
d: no difference d: no difference d: no difference d: no difference versus D; no statistically significant difference for either gender

Standing Still
Following
3 Min Rest/
Activity
Vital capacity (L) M = 4.24 (0.85) M = 3.76 (0.84) No statistically significant difference between the tasks
W = 3.05 (1.21) W = 3.29 (1.21)
d: no difference d: no difference
Maximum M = 20.23 (5.10) M = 7.37 (1.80) Statistically significant differences between the tasks
phonation W = 20.49 (5.81) W = 6.92 (1.59) M: reduced by 12.84 s (4.21) [63.47%], P < 0.001
time, MPT (s) d: no difference d: no difference W: reduced by 13.57 s (13.49) [65.84%], P < 0.001
% of VC used M = 90.15 (18.02) M = 75.11 (17.69) W: no statistically significant difference between the tasks
during W = 79.57 (26.07) W = 76.79 (23.72) M: reduced by 15.03 (13.87) units [16.68%], P = 0.008
MPT test d: no difference d: no difference
(TV/VC)
Notes: Baseline values. Mean (standard deviation) [% difference] for men (M) and women (W). Differences of mean values between genders (d) are only specified when statistically significant (P < 0.05).

253.e7
Abbreviations: MPT, maximum phonation time; TV, tidal volume; VC, vital capacity.
253.e8
FIGURE 2. Mean maximum phonation time.
Another interesting piece of information related to feeling com-
pletely prepared for public performance, with the majority
admitting to not feeling ready for opening night in most
situations.
DISCUSSION
This study was designed to understand the impact on the human
respiratory system of an athletic musical theater performance.
The overriding concern was that we and the performers felt that
the performances being tested were as close to actual real-life
performances as possible.
As one of the first studies to attempt to understand the complex
transitions within human respiratory physiology during change-
able athletic performance (musical theater), we would accept that
one of the major limitations, apart from time and financial re-
straints, was the inability to look back through previous literature
to fine-tune the processes. The relevant literature, in the main,
only partly furnished us with clues to methodology pitfalls and
confounding variables.
TABLE 6.
Scores of the Time Taken for These Performers to Feel
Fully Able to Demonstrate Their Full Potential, Sing All
Notes, and Articulate All Words While Dancing
Preparedness of the Performers for Performing
By opening night 45%
After first week of the run 50%
By week 5 75%
By week 8 85%
Notes: 15% (n = 3) still not 100% confident at 11 (n = 1), 108 (n = 1), and
200 (n = 1).
Journal of Voice, Vol. 31, No. 2, 2017
That being said, we are pleased that we have a project that
satisfied the research objectives, was enthusiastically taken up
by our professional participants, and that has yielded some
interesting data that could perhaps be a good starting place for
several qualitative and quantitative studies in the future.
Thoughts on the differences of scores between the various parts
of the study follow and some notes on the pilots studies under-
taken in advance.
Music selected for the performance tasks
It is more usual in voicing research to give participant the same
vocal exercise and to make comparisons between the execu-
tion of these exercises. However, it was more important to us
that the participants had already performed the pieces profes-
sionally and that we were able to include performers from
different shows.
In the end, we opted for a different approach. The reasons are
that from those who volunteered, from a very thorough and intense
advertising campaign, getting them to learn a new song and new
steps would have:
• meant most would not have thoroughly embodied the piece
to performance standard;
• would have therefore moved the project away from its over-
riding concern.
Research into dance specific fitness and singing fitness has
shown the importance of the familiarity of tasks for test
subjects24,26 as positively relating to greater efficacy of the uti-
lization of lung volumes and rib cage movements.
Respiratory and acoustic measures were included as between-
subject control measures in the four MPT tasks between each
performance piece.
Environment for testing
Pilots were run in a dance studio and a sports lab, with a dis-
cussion concerning the differences and suitability of these
environments raised.
All agreed that:
• the studio, although suitable, was somewhat alien that the
“atmosphere” of the actual show and testing would have
to be done at unsuitable times; and
• the lab while boasting good equipment interfaces and with
its obvious suitability for experimentation induced an ad-
ditional layer of stress (perhaps akin to “White Coat
Syndrome” [conjecture]).
Hypothesis 1: The equipment chosen for collecting real-time data
from musical theater performers would yield reliable and useful
information.
Equipment used in the experiment
The two alternative pieces of equipment we considered but
decided in the end not to use were the following:
Tommi Sliiden et al Evaluation of Breathing Strategies and Maximum Phonation Time
EGG—electroglottography—laryngoscope—
(Laryngograph Microprocessor EGG-D200 with
interface for EGG electrodes, analog output of
speech and EGG/Lx + omnidirectional microphone,
20 Hz–20,000 Hz response, −65 dBA, 3.5 mm
jack + medium electrodes and elongated electrodes
for singing)
This piece of equipment would have given clear and reliable data
for numerous acoustic parameters for voicing but not for res-
piration. The EGG is very well regarding for identifying
perturbations of amplitude and frequency within speech and
singing, and can be used as predictive of vocal pathology. As
our subject caused perturbations not related (necessarily) to vocal
pathology but rather to the necessary movement needed for dance,
the researchers considered it useful for only some of the readings.
• The performers found the piece equipment to be the most
physically restrictive.
• The researchers had concerns about the position of the elec-
trodes (level of the thyroid cartilage) remaining stable
throughout the testing.
• The equipment was less easy to set up and remove than
the gas analyzer (for this cohort).
Previous research using the first two methods required a near
static positioning for reliable readings.11,26–28
Spirometry: Micro 1 Spirometer—using the Micro I™
PC configuration software
The spirometry method of respiratory data collection was con-
sidered as a simple and effective process; however, the method
of securing and standardizing the positioning was much better
achieved (for this cohort) with the gas analyzer.
In addition to this, the performers having tried the other equip-
ment found the gas analyzer easier to move around when reporting
about no impediment to articulatory accuracy, although unfor-
tunately the mask did prevent any high-quality recordings of the
sound made. High-quality singing, however, was not a re-
quired parameter for this study.
The order of the tasks
We surmise that a more rigorous set of results would likely have
transpired if we could have repeated the study with a different
ordering of tasks and compared both sets of results. Although
care was taken to assure all performers had returned to their resting
RR before proceeding on the next task, this one order is a lim-
itation (conjecture) of the study.
Order 1: singing only followed by dancing only followed by
singing while dancing.
Order 2 (potential): dancing only followed by singing only
followed by singing while dancing.
We concluded that the first order allowed the performers a
gentler and more uniform start to the testing protocol. It was also
the order they all preferred. This order also seemed the most
253.e9
logical for warming up the body and voice in the most health-
conscious way (conjecture).
Hypothesis 2: The singing while dancing task would produce
the greatest respiratory measures for all compared with the singing
only and dancing only tasks.
This hypothesis was not met. Although there seemed to be a
logic to this assumption, the readings reveal a more complex ap-
proach to the singing while dancing task than for the separate
tasks. We concluded from the data that aerobic and anaerobic
respirations were employed to generate the extra effort re-
quired, giving smaller readings for VO2 for the combined task.
Reductions in liters per minute (MV) were significant. We con-
clude that in experienced performers, a more complex approach
to combined task meant that breathing was regulated, necessar-
ily, to accommodate the prosody and sense of the text, but despite
attempts to be faithful to the singing phraseology breathing was
compromised by the need to fuel the “dance muscles” with
oxygen.
As no one confessed to ever having been taught how to modify
breathing to accommodate both tasks, we assume a posteriori
that each individual chose his or her own adaptations to their
breathing. Further, we assumed it is likely these adaptive pat-
terns changed regularly (and perhaps unconsciously) throughout
any long theater run. This randomness of execution, in the absence
of any proffered strategy, during high levels of athleticism could
be causative of stress- and strain-related injury. However, on a
positive note, if adaptation can be made unconsciously, it is pos-
sible that they could also be taught, although the researchers accept
lesson plans for such tutoring may prove difficult.
Hypothesis 3: There would be significant differences for respi-
ratory scores taken for singing only compared with dancing only.
With this hypothesis, we assumed correctly that athletic
dancing (for this cohort) would eclipse singing in its respirato-
ry requirements.
Hypothesis 4: The MPT for the four sets of MPT scores would
decrease incrementally as the testing sequence progressed and
this will show an inverse with HR.
This assumption was met with a dramatic fall of phonation
time as HR practically doubled. These changes were consis-
tent across the group. A slight increase in phonation time for the
singing only task may be due to differences in the effort re-
quired to sing pitches other than the controlled lower sustained
pitches selected for this part of the study (women: 220 Hz/
men: 130.8 Hz). The ability of these performers to sustain pitch
once the body has assumed the posturing and energizing re-
quired for (athletic) dancing is clearly affected by the whole range
of supportive, stabilizing, torsion, acrobatic, and other varia-
tions in biotensegrity the body is required to undergo.
Hypothesis 5: Decreasing MPT was a result of increased efforts
in cardiovascular activity in performers.
The significance is difficult to call because of the small numbers
involved here, but it does seem like a clearly inverse relation-
ship between the increased HR and the reduced MPT, and opens
253.e10
the door for further investigation on the subject of the perfor-
mance impact for these parameters.
Hypothesis 6: VC should remain the same at the beginning as
at the end for these healthy performers.
We assumed that in our healthy, experienced professionals,
their ability to utilize as much capacity as was required of them
during athletic performance would remain high and consistent
across all tasks. This hypothesis was met.
Furthermore, resting HR seemed high in this group even after
they had rested and calmed down between tasks. As the testing
was performed at their venues within a few hours prior to curtain
up (and with some people milling around and watching in the
background), we assume a posteriori that some level of perfor-
mance anxiety was responsible (conjecture).
As at the higher HR (after dancing) MPT was reduced to as
little as 7.1 seconds, it is interesting to mention that the final
note sung in the Singin’ in the Rain finale is 7 seconds long,
begging the question how many could actually achieve this night
after.
It would also be interesting to examine the “overtime fatigue
effect” in musical theater as in this study only half of the sub-
jects who scored above average MPT at rest were able to attain
above average at higher HRs.
Oxygen uptake
Singing produced higher HR and relative oxygen uptake scores
compared with the at rest scores but much less than the other
two tasks. The slight drop-off when singing while dancing likely
indicates an increased use of anabolic energy sources or a much
modified employment of effort overall for this simultaneous task.
The normal oxygen uptake requirements at rest according to
the literature are 250 mL/min, during light and moderate exer-
cise (for everyday tasks such as dressing, washing, walking) 300–
1000 mL/min, during heavy work (laborious jobs in heavy
industry, building, mining, etc) 1000–2000 mL/min, and during
severe work >2000 mL/min.29 Our performers’ mean (men/
women) was, at rest, 417/310 mL/min, during singing 634/
430 mL/min, during dancing 2627/1652 mL/min, and during
singing while dancing 2572/1614 mL/min.
The literature goes on to say that tidal volume can increase
up to 300% during exercise 1; the increase in our subject group
was only 68% comparing rest and dancing, or 125% compar-
ing rest with singing while dancing. Maximum tidal volume
during maximal exercise has been quoted for men as 50–60%
of VC3; our male dancers used 37% during dancing and 49%
during singing while dancing (at submaximal levels). When the
spontaneously chosen tidal volume and RR were measured in
a group of cyclists, cycling so that the set minute volume of 72
L/min was achieved, their mean tidal volume was 2.5 L and their
RR was 30 breaths/min2. Our male dancers when dancing with
spontaneous breathing had a mean minute volume of 65 L/min,
with a tidal volume of 1.5 L and RR of 44 breaths/min.
Questionnaire
Findings showed that the ability to even support and sustain notes
was the biggest challenge perceived by this group when having
Journal of Voice, Vol. 31, No. 2, 2017
to sing while dancing. They also indicated little training in breath-
ing functions and technique during their initial training, with only
45% of them feeling able to perform the combined singing while
dancing to their full potential by opening night.
SUMMARY AND FURTHER LIMITATIONS
Apart from these limitations related to time frames, resources,
and the sparseness of examples within the literature, the main
limitations were the low numbers of participants, the neces-
sary imposition of the instrumentation (although minimal),
potential differences in skill and effort for the tasks, and our in-
ability to ratify anxiety into the analysis.
CONCLUSIONS
Singing (projected) while dancing (athletically) will lead to some
compromises of both tasks. The results showed drastically reduced
MPT at high exertion levels with singing, restricting the natural
response of breathing when dancing and resulting in reduced
minute volume. Despite the increase in HR, there was no in-
crease in relative oxygen uptake, indicating a likely increase in
anaerobic energy yield when combining the two tasks.
In our study, dancing did not incur a shallower breathing than
when singing, mean tidal volume stayed the same, and mean RR
only increased when dancing.
Our questionnaire revealed that the specific training and prep-
aration for successfully combining these athletic tasks repeatedly
is not satisfactory, even for performers of some ability, with seem-
ingly little training in breathing outside singing classes. Most
also felt that rehearsal periods were too short for them to achieve
their perceived full potential by opening night.
This study found the equipment, previously used on athletes
and dancers, helpful in studying aspects of singing and breath-
ing under performance-like conditions, with little negligible impact
upon normal dancing or singing presentation.
More research in the field is needed to support the possible
development of training methods that could better equip those
performers required to sing while dancing, play wind instru-
ments while moving (marching bands, etc), and for those involved
in leading aerobic circuit training classes.
SUPPLEMENTARY DATA
Supplementary data related to this article can be found online
at doi:10.1016/j.jvoice.2016.06.025.
REFERENCES
1. Aliverti A. Lung and chest wall mechanics during exercise: effects of
expiratory flow limitation. Respir Physiol Neurobiol 2008;163:90–99.
2. Mallios VJ, Hodgson JL. Imposed breathing pattern alters respiratory work
during exercise. Eur J Appl Physiol. 1994;69:262–267.
3. Younes M, Kivinen G. Respiratory mechanics and breathing pattern during
and following maximal exercise. J Appl Physiol. 1984;57:1773–1782.
4. Christensen PM, Thomsen SF, Rasmussen N, et al. Exercise-induced
laryngeal obstructions: prevalence and symptoms in the general public. Eur
Arch Otorhinolaryngol. 2011;268:1313–1319.
5. Wilson JJ, Theis ST, Wilson EM. Evaluation and management of vocal cord
dysfunction in the athlete. Curr Sports Med Rep. 2009;8:65–70.
6. McCoy S. Breath management: gender-based differences in classical singers.
Folia Phoniatr Logop. 2005;57:246–254.
Tommi Sliiden et al Evaluation of Breathing Strategies and Maximum Phonation Time
7. Watson PJ, Hixon TJ. Respiratory kinematics in classical (opera) singers.
J Speech Hear Res. 1985;28:104–122.
8. Binazzi B, Lanini B, Bianci R, et al. Breathing patterns and kinematics in
normal subject during speech, singing and loud whispering. Acta Physiol
(Oxf) 2006;186:233–246.
9. Thorpe WT, Cala SJ, Chapman J, et al. Patterns of breath support in
projection of the singing voice. J Voice. 2001;15:86–104.
10. Thomasson M, Sundberg J. Consistency of phonatory breathing patterns
in professional operatic singers. J Voice. 1999;13:529–541.
11. Watson AH, Williams C, James BV. Activity patterns in latissimus dorsi
and sternocleidomastoid in classical singers. J Voice. 2012;26:e95–e105.
12. Gelfer MP, Pazera JF. Maximum duration of sustained /s/ and /z/ and the
s/z ratio with controlled intensity. J Voice. 2006;20:369–379.
13. Hirano M. Clinical Examination of Voice. Austria: Springer-Verlag/Wien;
1981.
14. Hirano M, Koike Y, von Leden H. Maximum phonation time and air usage
during phonation. Folia Phoniatr Logop. 1968;20:185–201.
15. Hixon T, Hoit JD. Evaluation and Management of Speech Breathing
Disorders. Tucson, Arizona, USA: Redington Brown LLC; 2005.
16. Iwata S, von Leden H, Williams D. Air flow measurement during phonation.
J Commun Disord. 1972;5:67–79.
17. Larson GW, Mueller PB, Summers PA. The effect of procedural variations
on the s/z ratio of adults. J Commun Disord. 1991;24:135–140.
18. Mathieson L. Greene and Mathieson the Voice and its Disorders. 6th ed.
London: Whurr; 2006.
253.e11
19. Solomon NP, Shannon JG, Milbrath RL. Respiratory and laryngeal
contributions to maximum phonation duration. J Voice. 2000;14:331–340.
20. Leino T, Laukkanen A-M, Ilomäki I, et al. Assessment of vocal capacity
of Finnish university students. Folia Phoniatr Logop. 2008;60:199–209.
21. Schmidt P, Klingholz F, Martin F. Influence of pitch, voice sound pressure
and vowel quality on the maximum phonation time. J Voice. 1988;2:245–249.
22. Redding E, Weller P, Ehrenberg S, et al. The development of a high intensity
dance performance fitness test. J Dance Med Sci. 2009;13:3–9.
23. Wyon M, Redding E, Abt G, et al. Development, reliability and validity
of a multistage dance specific aerobic fitness test (DAFT). J Dance Med
Sci. 2003;7:80–84.
24. Wyon M, Head A, Sharp C, et al. Oxygen uptake during modern dance class,
rehearsal and performance. J Strength Cond Res. 2004;18:646–649.
25. Pocock G, Richards CD. Human Physiology—The Basis of Medicine. 2nd
ed. Oxford University Press; 2004:358–361.
26. Thomasson M, Sundberg J. Lung volume levels in professional classical
singing. Logoped Phoniatr Vocol. 1997;22:61–70.
27. Pettersen V. Muscular patterns and activation levels of auxiliary breathing
muscles and thorax movement in classical singing. Folia Phoniatr Logop.
2005;57:255–277.
28. Iwarsson J, Sundberg J. Effects of lung volume on vertical larynx position
during phonation. J Voice. 1998;12:159–165.
29. Wyon M, Redding E, Head A, et al. Physiological monitoring of
cardiorespiratory adaptations during rehearsal and performance of
contemporary dance. J Strength Cond Res. 2005;19:611–614.