Dysphagia (2012) 27:452–459
DOI 10.1007/s00455-011-9389-2
ORIGINAL ARTICLE
Deglutitive Subglottic Air Pressure and Respiratory System Recoil
Roxann Diez Gross • Ricardo L. Carrau •
William A. Slivka • Ronit G. Gisser •
Libby J. Smith • David J. Zajac • Frank C. Sciurba
Received: 19 April 2011 / Accepted: 19 December 2011 / Published online: 24 January 2012
Ó Springer Science+Business Media, LLC 2012
Abstract The purpose of this experiment was to confirm
the presence of positive subglottic air pressure during
swallowing, known as deglutitive subglottic air pressure
(DPsub), in a group of healthy individuals. We also sought
to determine if respiratory system recoil is responsible for
generating the pressure. Ten healthy volunteers underwent
direct DPsub measurement via percutaneous puncture of
the cricothyroid membrane. Simultaneous DPsub and nasal
airflow volumes were recorded while participants swal-
lowed calibrated boluses over a wide range of lung vol-
umes. Body plethysmography was used to determine
functional residual capacity and residual volume. A custom
respiratory recoil measurement system was used to mea-
sure recoil pressures. Regression analysis of lung volume
on DPsub and lung volume on recoil pressure yielded
strong linear relationships (P \ 0.0001, R2 = 0.71 and
P \ 0.0001, R2 = 0.69, respectively). A mixed-model
analysis of the effect of method (direct puncture or recoil)
R. D. Gross
R. L. Carrau
R. G. Gisser
L. J. Smith
Department of Otolaryngology, Eye and Ear Institute,
School of Medicine, University of Pittsburgh,
200 Lothrop St., Pittsburgh, PA 15213, USA
W. A. Slivka
F. C. Sciurba
Division of Pulmonary, Allergy and CCM, UPMC
Comprehensive Lung Center, School of Medicine,
University of Pittsburgh, 3601 Fifth Ave., 4th Floor,
Pittsburgh, PA 15213, USA
D. J. Zajac
Craniofacial Center, University of North Carolina,
Chapel Hill, NC 27599, USA
R. D. Gross (&)
The Children’s Institute of Pittsburgh, 1405 Shady Ave,
Pittsburgh, PA 15217, USA
e-mail: RGR@the-institute.org
123
on pressure showed that there was no effect (F = 0.63;
P = 0.43). By confirming the presence of DPsub in healthy
adults and showing that respiratory system recoil is the
most likely mechanism that generates DPsub, treatment of
persons with dysphagia has even greater potential to be
expanded to include consideration of factors that affect
respiratory control and recoil forces.
Keywords Subglottic air pressure
Lung volume

Recoil
Dysphagia
Swallowing
Deglutition

Deglutition disorders
The association between the generation of subglottic air
pressure during the pharyngeal phase of swallowing, called
deglutitive subglottic air pressure (DPsub), and improved
swallowing function was first described nearly 20 years
ago by studies that compared swallowing physiology in
patients with open versus closed tracheostomy tubes.
Investigators reported that when the tube was closed during
swallowing, significant physiologic changes occurred,
including the reduction or elimination of prandial aspira-
tion (food and drink entering below the true vocal folds and
into the trachea) [1–5]. Although not measured directly,
researchers and clinicians speculated that closing the tube
enabled the generation of DPsub and that a relationship
between the subglottic pressure and swallowing function
seemed to exist. The resulting subglottic pressure theory
for swallowing states that subglottic mechanoreceptors
provide respiratory-related afferent input to the brainstem
central pattern generator for swallowing and that efferent
output is modified accordingly [6]. The viability of this
theory is entirely dependent upon confirmation of the
presence of positive (above atmospheric) air pressure
below the closed vocal folds at the time of each swallow.
R. D. Gross et al.: Deglutitive Subglottic Air Pressure
In patients with indwelling tracheostomy tubes, sub-
glottic airway pressure measurements during swallowing
are readily obtainable because the tubes are positioned
within the trachea below the vocal folds. The location of
the tube creates open communication between the trachea
and the atmosphere; thus, DPsub is equal to atmospheric
pressure (zero) when the tube is open during swallowing.
In 1994, Gross et al. [7] described a single patient who was
observed to aspirate under fluoroscopy when her trache-
ostomy tube was open. A direct measurement of DPsub in
the aspirating condition showed that zero DPsub was
present during the swallow. In that same patient, aspiration
could be prevented by occluding the tube with a one-way
airflow valve that closed spontaneously at the onset of
deglutitive apnea (Passy-Muir, Inc.). The DPsub that was
measured in the occluded condition was 9.3 cmH2O [7, 8].
In another study, DPsub values ranging from 7 to 10
cmH2O were reported in nonaspirating patients with
indwelling tracheostomy tubes [9].
Extremes in lung volumes at swallow onset have also
been used as a means to manipulate DPsub [6]. In that study
by Gross et al. [6], 28 healthy volunteers swallowed stan-
dardized boluses under fluoroscopy at three targeted lung
volumes: total lung capacity (TLC), functional residual
capacity (FRC), and residual volume (RV). Significant
durational differences in pharyngeal swallowing physiology
were measured when the lung volumes were compared. The
authors attributed the changes in swallowing physiology to
alterations in DPsub generation; however, subglottic pres-
sures were not measured directly.
As previously stated, DPsub is easily measured in tra-
cheostomy patients because there is direct access to the
subglottic air space. Subglottic pressure readings have been
taken in healthy persons using topical anesthesia on the
pharynx and larynx to place pressure-measuring catheters
through the vocal folds and into the subglottic air space [10,
11]. Due to the methods used, only dry swallows could be
studied. To determine if DPsub is also present in healthy
persons during bolus swallowing and to avoid using anes-
thesia or placing instrumentation between the vocal folds,
Gross et al. [12] used a direct cricothyroid puncture tech-
nique to detect and measure DPsub across a variety of lung
volumes in a healthy person. Their findings provided addi-
tional evidence to show that DPsub is also generated while
swallowing food boluses (pudding) and is a likely compo-
nent of normal swallowing physiology. In addition, the
investigation revealed a linear relationship between the
estimated lung volume at the time of the swallow, the
amount of air pressure generated, and polarity of the pres-
sure recordings. The authors postulated that respiratory
system recoil was the most plausible physiologic mecha-
nism that could be responsible for generating DPsub. Nev-
ertheless, their experiment had two important limitations.
453
First, there was only one subject, and consequently, gener-
alization of the findings has been limited. Second, the lung
volumes were estimated and not measured directly.
Accordingly, the purpose of the present experiment was to
use the same method to confirm the presence of DPsub
generation in a larger group of healthy individuals and to
correlate the subglottic pressures with known lung volumes.
The second objective of our experiment was to obtain
within-subject comparisons of DPsub values and respiratory
system recoil pressures to determine if lung-thoracic unit
recoil is responsible for generating DPsub. There were three
hypotheses: (1) DPsub would be detected and measured in
all subjects. (2) DPsub pressure would be dependent upon
lung volume at swallow onset. (3) Respiratory system recoil
pressures would be highly correlated with DPsub pressures.
Methods
To achieve our objectives and to test our hypotheses, a
prospective, repeated-measures design was used. The
experiment consisted of three separate steps that all took
place on the same day.
Participants
This investigation was approved by the University of
Pittsburgh Institutional Review Board. Each participant
signed a consent form prior to participation in any portion
of this experiment. A total of ten nonsmoking volunteers
consisting of 6 males and 4 females of mixed ethnicity
completed the experiment. The average age was 32.5
years (range = 27–37). The average height was 67.3 in.
(range = 61–71), average weight was 152 lbs. (range =
104–172), and average body mass index was 23.52
(range = 20.39–29.62).
Four different screening procedures were completed on
each participant prior to data collection:
(1) Cognitive screening was completed using the Folstein
Mini Mental Exam (MME) [13]. All participants
scored 29 or 30 on the MME indicating that cognitive
functioning was intact.
(2) Each participant also completed a screening ques-
tionnaire [14] to assure that none had a history of
swallowing difficulty, closed head injury, neurologic
disease, muscle disease, or respiratory disease.
(3) All subjects passed a respiratory screening using the
best of three forced exhalations into a hand-held
spirometer. Criterion for inclusion stated that the ratio
of the amount of air exhaled in 1 sec (FEV1) over the
total amount of air exhaled (forced vital capacity or
FVC) had to be greater than 70%.
123
R. D. Gross et al.: Deglutitive Subglottic Air Pressure
In patients with indwelling tracheostomy tubes, sub-
glottic airway pressure measurements during swallowing
are readily obtainable because the tubes are positioned
within the trachea below the vocal folds. The location of
the tube creates open communication between the trachea
and the atmosphere; thus, DPsub is equal to atmospheric
pressure (zero) when the tube is open during swallowing.
In 1994, Gross et al. [7] described a single patient who was
observed to aspirate under fluoroscopy when her trache-
ostomy tube was open. A direct measurement of DPsub in
the aspirating condition showed that zero DPsub was
present during the swallow. In that same patient, aspiration
could be prevented by occluding the tube with a one-way
airflow valve that closed spontaneously at the onset of
deglutitive apnea (Passy-Muir, Inc.). The DPsub that was
measured in the occluded condition was 9.3 cmH2O [7, 8].
In another study, DPsub values ranging from 7 to 10
cmH2O were reported in nonaspirating patients with
indwelling tracheostomy tubes [9].
Extremes in lung volumes at swallow onset have also
been used as a means to manipulate DPsub [6]. In that study
by Gross et al. [6], 28 healthy volunteers swallowed stan-
dardized boluses under fluoroscopy at three targeted lung
volumes: total lung capacity (TLC), functional residual
capacity (FRC), and residual volume (RV). Significant
durational differences in pharyngeal swallowing physiology
were measured when the lung volumes were compared. The
authors attributed the changes in swallowing physiology to
alterations in DPsub generation; however, subglottic pres-
sures were not measured directly.
As previously stated, DPsub is easily measured in tra-
cheostomy patients because there is direct access to the
subglottic air space. Subglottic pressure readings have been
taken in healthy persons using topical anesthesia on the
pharynx and larynx to place pressure-measuring catheters
through the vocal folds and into the subglottic air space [10,
11]. Due to the methods used, only dry swallows could be
studied. To determine if DPsub is also present in healthy
persons during bolus swallowing and to avoid using anes-
thesia or placing instrumentation between the vocal folds,
Gross et al. [12] used a direct cricothyroid puncture tech-
nique to detect and measure DPsub across a variety of lung
volumes in a healthy person. Their findings provided addi-
tional evidence to show that DPsub is also generated while
swallowing food boluses (pudding) and is a likely compo-
nent of normal swallowing physiology. In addition, the
investigation revealed a linear relationship between the
estimated lung volume at the time of the swallow, the
amount of air pressure generated, and polarity of the pres-
sure recordings. The authors postulated that respiratory
system recoil was the most plausible physiologic mecha-
nism that could be responsible for generating DPsub. Nev-
ertheless, their experiment had two important limitations.
453
First, there was only one subject, and consequently, gener-
alization of the findings has been limited. Second, the lung
volumes were estimated and not measured directly.
Accordingly, the purpose of the present experiment was to
use the same method to confirm the presence of DPsub
generation in a larger group of healthy individuals and to
correlate the subglottic pressures with known lung volumes.
The second objective of our experiment was to obtain
within-subject comparisons of DPsub values and respiratory
system recoil pressures to determine if lung-thoracic unit
recoil is responsible for generating DPsub. There were three
hypotheses: (1) DPsub would be detected and measured in
all subjects. (2) DPsub pressure would be dependent upon
lung volume at swallow onset. (3) Respiratory system recoil
pressures would be highly correlated with DPsub pressures.
Methods
To achieve our objectives and to test our hypotheses, a
prospective, repeated-measures design was used. The
experiment consisted of three separate steps that all took
place on the same day.
Participants
This investigation was approved by the University of
Pittsburgh Institutional Review Board. Each participant
signed a consent form prior to participation in any portion
of this experiment. A total of ten nonsmoking volunteers
consisting of 6 males and 4 females of mixed ethnicity
completed the experiment. The average age was 32.5
years (range = 27–37). The average height was 67.3 in.
(range = 61–71), average weight was 152 lbs. (range =
104–172), and average body mass index was 23.52
(range = 20.39–29.62).
Four different screening procedures were completed on
each participant prior to data collection:
(1) Cognitive screening was completed using the Folstein
Mini Mental Exam (MME) [13]. All participants
scored 29 or 30 on the MME indicating that cognitive
functioning was intact.
(2) Each participant also completed a screening ques-
tionnaire [14] to assure that none had a history of
swallowing difficulty, closed head injury, neurologic
disease, muscle disease, or respiratory disease.
(3) All subjects passed a respiratory screening using the
best of three forced exhalations into a hand-held
spirometer. Criterion for inclusion stated that the ratio
of the amount of air exhaled in 1 sec (FEV1) over the
total amount of air exhaled (forced vital capacity or
FVC) had to be greater than 70%.
123
454
(4) To ensure that the true vocal folds could fully adduct,
transnasal endoscopic examination of laryngeal func-
tion was completed approximately 10 min before
proceeding to step 1 of the experiment. For this
screening, the nasal passage was decongested using
AfrinÒ nasal spray. Topical anesthetic (2% viscous
lidocaine) was then applied with a cotton swab to the
surfaces of the inferior meatus of the nares. The
flexible endoscope was then passed along the floor of
the nose and positioned above the larynx. Vocal fold
motion was observed as each subject completed
breath-holding and coughing tasks. The endoscopic
examination was recorded and complete vocal fold
adduction was reverified by reviewing the images.
Step 1: Measurement of Deglutitive Subglottic Air
Pressure
Simultaneous nasal airflow and subglottic pressure mea-
surements were made during bolus swallowing. For each
subject, air flow and air pressure transducers were calibrated
using previously described methods [12]. Nasal airflow was
detected by means of a snuggly fitted nasal mask with a flow
tube that was connected to a heated pneumotachograph
(Fleisch, #1, Switzerland). To reduce dead space to less than
30 cc and limit the rebreathing of CO2, the pneumotacho-
graph was positioned directly to the nasal mask by using a
custom headgear. Direct measurement of DPsub was
obtained via a percutaneous puncture of the cricothyroid
membrane by an otolaryngologist. Approximately 2 cc of a
lidocaine and epinephrine mixture was injected in the area
over the cricothyroid membrane to anesthetize the skin and
control bleeding. Next, a small needle (19–21 gauge) that
was housed within a polyethylene catheter was inserted
through the skin and cricothyroid membrane. Following the
insertion, the needle was withdrawn leaving the catheter in
the air space below the true vocal folds. The distal end of the
catheter was then connected to the pressure transducer
which was referenced to atmosphere. During data collec-
tion, the catheter’s position within the airway was stabilized
manually. No anesthesia was applied to the larynx or
pharynx to assure that swallowing function would not be
affected by sensory loss. The simultaneous subglottic
pressure and nasal airflow signals were digitized to a
computer at a rate of 500 samples per second with 16-bit
resolution using WinDaq/Pro model 720 (DATAQ Instru-
ments, Akron, OH), a computer-based signal acquisition
and processing system. The computer was used to display,
record, and analyze the data.
For each participant, 20 single boluses of 5 ml of
ambient-temperature pudding (Hunt’s Snack PackÒ,
ConAgra Foods) were premeasured onto individual plastic
123
R. D. Gross et al.: Deglutitive Subglottic Air Pressure
teaspoons using a syringe. The viscosity of the pudding was
approximately 43,950 cps (spindle #4, 20 rpm; Brookfield
Viscometer, Middleboro, MA). Once the bolus was placed
within the mouth, the subjects were instructed to keep their
lips closed and to swallow the entire bolus with one swal-
low. Three target lung volumes were randomly assigned: (1)
total lung capacity (TLC), (2) tidal volume (TV), and (3)
functional residual capacity (FRC). For the TLC condition,
the participant was told to ‘‘fill your lungs with air and then
swallow’’; for TV, ‘‘swallow at the top of a normal breath’’;
and for FRC, ‘‘swallow at the bottom of a normal breath.’’
To signal that the bolus had been swallowed and to place a
mark within the raw data, an EMG surface electrode was
placed over the muscles on the palmar surface at the base of
the subject’s thumb. Each subject was instructed to move
their thumb toward the palm after they swallowed the bolus.
This motion resulted in an increased EMG signal when
compared to the baseline and allowed bolus swallows to be
easily distinguished from saliva swallows. All swallows
were easily identified by the combined presence of the
EMG signal and deglutitive apnea where respiratory airflow
equals zero (Fig. 1). The goal was to record five swallows
per lung volume for a total of 15 swallows per subject.
Step 2: Pulmonary Function Testing
Within 1 h of completing step 1 and without additional
food intake, pulmonary function testing was completed so
that lung volumes could be calculated for DPsub pressures
that were obtained in step 1 and recoil pressures obtained in
TM
step 3. A body plethysmograph (Model MasterScreen
Body, Jaeger/CareFusion, Yorba Linda, CA) was cali-
brated prior to testing according to the manufacturer’s
guidelines. FRC was determined via thoracic gas volume
technique where three to five separate, acceptable trials
yielded values within 5% of the mean reported FRC. TLC
and residual volume (RV) were determined to ensure that
no subject was restricted or hyperinflated. Forced vital
capacity testing was also performed in standard fashion to
confirm normal forced spirometry results.
Step 3: Measurement of Lung-Thoracic Unit Recoil
Subjects walked approximately 15 ft from the body ple-
thysmograph to the custom lung mechanics system used to
assess respiratory system recoil. They sat for a minimum of
10 min prior to further testing. No significant change in
FRC was assumed as a result of this relocation. Plethys-
mographically determined FRC was used to reference
absolute lung volume against quasistatic respiratory system
recoil measurement. The custom respiratory recoil mea-
surement system consisted of a pneumotachograph (model
3700, Hans Rudolph, Kansas City, MO), differential
R. D. Gross et al.: Deglutitive Subglottic Air Pressure
Fig. 1 Example of the raw data from one swallow. The tracing at the top is from the flow transducer. The tracing underneath flow is from the
pressure transducer. The next line below is from the palmar surface EMG electrode. The bottom tracing is the integrated flow signal
pressure transducers (model MP45-871, Validyne Engi-
neering, Northridge, CA), and a pneumatic occlusion valve
(model 9340, Hans Rudolph). Transduced signals were
processed (CD19 Carrier Demodulator, Validyne), digi-
tized, and displayed in real time during testing.
To simulate the central inhibition of respiratory muscles
that occurs during deglutitive apnea, each subject was
instructed to inhale until their lungs were nearly full and
then to passively exhale without using respiratory muscle
effort. Surface electromyography (sEMG) recordings were
used to detect possible respiratory muscle activity during
the passive exhalations. Two standard 2-cm-diameter Ag/
AgCl EMG electrodes were coated with electrolyte gel and
then secured to the skin which was first cleaned using an
alcohol pad. The electrodes were placed on the abdomen
parallel with the fiber alignment of the upper and lower
rectus abdominis muscles. Two electrodes were also posi-
tioned horizontally on the chest over the pectoralis muscle.
The ground electrode was placed at the level of the ster-
num. The sEMG signal was preamplified (Grass Instru-
ments model P511 high-performance AC preamplifier,
Astro-Med, Inc., West Warwick, RI) and bandpass filtered
between 100 and 1 kHz. A notch filter was set at 60 Hz.
The EMG signal was displayed on a computer screen
viewable by the subject as biofeedback to facilitate
455
relaxation of the abdominal muscles. Each subject took
practice breaths by inhaling to just below total lung
capacity and then exhaling passively. Once the subject
could perform the maneuver without evidence of abdomi-
nal or chest muscle activity, he/she was instructed to
breathe through a low-resistance rubber mouthpiece with
nose clips in place. Each participant supported their cheeks
with both hands to prevent the buccal spaces from filling
with pulmonary air. Following four or more resting tidal
breaths to establish FRC, the subject was instructed to
inspire to near TLC and exhale passively to FRC.
Simulation of true vocal fold closure during swallowing
was achieved by setting an occlusion valve to close for
650 ms at random times during each passive exhalation.
The 650-ms duration was selected because it is consistent
with the mean deglutitive apnea duration measured in our
previous investigations that used normal subjects [14, 15]. It
is also very close to the 623-ms mean duration of vocal fold
closure that was reported by Kawasaki et al. [16] and the
average true vocal fold closure time of 700 ms (±80)
reported by Shaker et al. [17]. The occlusion valve closure
was activated by the test technician and expiratory airflow
was interrupted two to four times in stepwise fashion. At
least 10 acceptable interrupter-derived pressure and volume
exhalation data points were obtained from each participant.
123
456 R. D. Gross et al.: Deglutitive Subglottic Air Pressure

The valve closure was controlled in such a way that subjects

were not able to anticipate the pressure measurements.
Acceptable quasistatic airway pressure measurement was
characterized by a short plateau of the airway pressure
tracing and without evidence of abdominal or chest muscle
EMG activity. Every time that the occlusion valve closed,
the airway pressure was measured and recorded by the
computer. Each passive breath generated between one to
three pressure measurements at random lung volumes.

Data Analysis

A total of 90 of a potential 150 swallows were recorded.

Subglottic pressure data were lost when the tube became Fig. 2 Direct pressure method showing the linear relationship
occluded or ‘‘kinked’’ due to the upward laryngeal motion between lung volume in l and DPsub in cmH2O. Each filled circle
under the skin that is associated with swallowing. Also, if represents a single swallow
an ‘‘EMG marked swallow’’ was followed by additional
unmarked swallows, the DPsub data were not included in
the final analysis because the pattern indicated that the
entire bolus may not have been swallowed at one time.
Using this criterion, a total of 71 of 90 data points (swal-
lows) were analyzed from the step 1 experiment. To
determine tidal volume for each swallow, the nasal airflow
signal was integrated using a computer program within
WinDaq called advanced CODAS. Absolute lung volume
was then calculated by adding the FRC to the tidal volume
that was measured during deglutitive apnea.
For the interrupter method in step 3, occasional outliers
were easily identified by the presence of very high pressure
(contraction of expiratory muscles during exhalation),
negative pressure (brief inhalation), or zero pressure (if Fig. 3 Interrupter method showing a significant linear relationship
they came off of the mouthpiece). These pressures were not between lung volume in l and pressure in cmH2O measured during
passive exhalation
included in the analysis of the recoil pressures.
All final pressure and volume data were statistically yielded a significant linear relationship between volume on
analyzed with SAS v9.2 (SAS Institute, Cary, NC). A pressure (P \ 0.0001, R2 = 0.69). After adjusting for sub-
significance level of 0.05 was preset prior to statistical ject, 69% of variance in recoil pressure was explained by
analysis. Data were analyzed using linear regression to volume and subject (38% by volume and 31% by subject)
examine the relationship between pressure and volume. A (Fig. 3).
linear mixed-model analysis accessed the effect of method A mixed model was used to analyze the effect of method
(direct puncture or interrupter) using compound symmetry (direct puncture and interrupter) on pressure, adjusting for
correlation matrix. volume and allowing a repeated-subject effect. There was
no effect of method (F = 0.63; P = 0.43). Volume was a
significant covariate (F = 252.26; P \ 0.0001). The
Results interaction between volume and method was not significant
(F = 0.07; P = 0.79) (Fig. 4; Table 1).
Linear regression of volume on pressure for step 1 (direct
measurement of DPsub) yielded a significant linear rela-
tionship between volume and pressure (P \ 0.0001, Discussion
R2 = 0.71). After adjusting for subject, 71% of variance in
pressure measurements can be explained by volume and Deglutitive Subglottic Air Pressure
subject (53% by volume and 18% by subject) (Fig. 2).
Linear regression of volume on pressure using the In all participants, air pressures above atmospheric were
interrupter method to measure lung-thoracic recoil also consistently detected in the subglottic airspace during the

123
R. D. Gross et al.: Deglutitive Subglottic Air Pressure
Fig. 4 Combined data is graphed to show that both methods produce
similar pressures. The filled circles represent DPsub and the triangles
represent recoil pressures in cmH2O
Table 1 Statistical table
Effect Num Denum F value P value
DF DF
Method (puncture or 1 6 0.63 0.43
recoil)
Lung volume 1 189 252.26 \0.0001
Volume*method 1 189 0.07 0.79
Num DF degrees of freedom in the numerator; Denum DF degrees of
freedom in the denominator
pharyngeal phase of swallows that occurred above FRC.
The pressures that were measured (DPsub) replicate and
support those that were previously reported by showing a
significant linear relationship between lung volume and
DPsub [6, 12].
Subglottal pressure generation during dry (saliva) swal-
lowing at unknown lung volumes was studied previously by
Shaker et al. [10] who made the measurements by placing a
solid-state strain gauge catheter between the vocal folds.
They reported a ‘‘biphasic’’ pattern with pressures initially
falling below atmospheric then rising above atmospheric.
Their findings were consistent with Shin et al. [11] who also
placed a pressure transducer through the vocal folds into the
trachea of at least one human subject; however, Shin also
took intramuscular EMG recordings of the thyrohyoid
muscle. Using this combined method, they noted that the
subglottic pressures initially elevated with swallow onset,
dropped with laryngeal elevation, and then rose again while
the larynx was in descent [11].
Similar biphasic air pressure recordings, obtained by
cricothyroid puncture, were reported by Gross et al. [12],
but only for swallows that occurred near FRC. The same
pattern was observed during this experiment, but again,
only for swallows occurring at or near FRC. Similarities
and differences between the experiments may be related to
457
the type of instrumentation used, dry versus bolus swal-
lows, and/or to lung volume at the time of the swallow. A
possible explanation for biphasic pressures may be dis-
covered when the action of the diaphragm during swal-
lowing apnea is considered.
Hardemark Cedborg and her research team [18] were the
first to use intramuscular EMG to study diaphragmatic
muscle activity during swallowing apnea. Their study
revealed that during passive exhalation, diaphragmatic
muscle activity was initially quiescent, but that a consistent
increase in diaphragmatic muscle activity occurred just prior
to the pharyngeal swallow. The authors postulated that the
diaphragm becomes active prior to swallow onset to main-
tain lung volume during the swallow to assist with DPsub
generation and to assure that there will be sufficient air for
postswallow exhalation. The onset of diaphragmatic muscle
contraction may be more easily detected in the experiments
that placed catheters between the vocal folds because the
presence of the catheter may prevent complete closure and
permit a slight draw of air when the diaphragm becomes
tonic. It is also possible that an initial drop in tracheal air
pressure may be detected only at lower lung volumes (near
or at FRC) where the lung-thoracic unit is closer to
homeostasis.
Wheeler Hegland et al. [19] were the first to examine
lung volumes at the time of swallow initiation in healthy
young subjects. Consistent with previous reports, they
measured a predominant breathing and swallowing pattern
of exhale-swallow-exhale and a second most frequent
pattern of inhale-swallow-exhale [18, 20–23]. A new and
significant finding of theirs was that lung volume at swal-
low onset was not associated with the respiratory pattern.
Subjects exhaled down to or inhaled up to a specific range
of 43–64% of vital capacity (VC), where 95% of all
swallows occurred. The authors speculated that this lung
volume range may enable the safest and most efficient
swallowing motor output. To examine the possible DPsub
range that might be associated with the lung volume range,
we converted our lung volumes to VC. Wheeler Hegland
and colleagues reported that paste consistency was swal-
lowed at a mean VC of 51.16% and thin-liquid VC was
56.27%. In our subjects, this range of VC was found to
have produced DPsub pressure that ranged from 5.5 to 9.5
cmH2O (Fig. 5). This pressure range is similar to a previ-
ously proposed optimal DPsub range of 7–10 cmH2O that
was based upon data from nonaspirating tracheostomy
patients. The range is also consistent with the average
subglottic pressure of 6.0 ± 1.8 mmHg (8.1 cmH2O,
5.7–10.6 cmH2O), that was reported by Shaker et al. [10].
More recently, Wheeler Hegland et al. [24] found that
healthy individuals timed swallows to occur at significantly
higher lung volumes when drinking from a cup compared
to single-bolus swallows.
123
458
Fig. 5 Lung volumes converted
to % vital capacity and the
associated DPsub values. The
darkest area highlights the
mean VC range reported by
Wheeler-Hegland et al. [19].
The vertical lines surround our
swallows that occurred within
their mean VC range and also
bracket the associated DPsub
range. The large shaded area is
where 95% of all swallows
occurred in the Wheeler-
Hegland et al. study
Respiratory System Recoil
This experiment has also provided direct evidence to
confirm that in healthy individuals, DPsub is generated by
respiratory system recoil pressure. The experimental
method used passive exhalation to simulate the inhibition
of respiratory muscles that occurs during swallowing, and a
brief occlusion of 650 ms simulated closure of the true
vocal cords that also occurs when swallowing. This tech-
nique replicated DPsub because when the glottis is open, a
continuous column of air exists from the mouth to the distal
lung alveoli. Mouth pressure reflects the subglottic pressure
since the subglottis is in open communication with the
mouth. The method also took advantage of the fact that
compliance of the total respiratory system is basically
linear over the midrange of the respiratory system’s recoil
from 75% of TLC to FRC. The basic interrupter technique
that we employed has been shown to be a valid method for
measuring tracheal pressure during passive expiratory air-
flow in anesthetized and anesthetized/paralyzed adults [25,
26]; prior to this experiment it had not been used in alert
individuals. Our results are consistent with those of Shin
et al. [11] who used direct airway punctures to simulta-
neously measure subglottal and intrathoracic pressures
during water swallows in felines. In their experiment the
two pressures were described as ‘‘nearly identical.’’
Summary
By confirming and characterizing the presence of DPsub
generation in healthy adults and showing that respiratory
123
R. D. Gross et al.: Deglutitive Subglottic Air Pressure
system recoil is the mechanism that generates DPsub, our
experiment supports the subglottic pressure theory for
swallowing. Our findings suggest that factors affecting
respiratory control and respiratory system mechanics may
need to be considered and evaluated when treating persons
with dysphagia. For example, in addition to addressing
impaired coordination of the respiratory cycle with swal-
lowing in patients with chronic obstructive pulmonary
disease [15, 27], they may need to be instructed to time the
onset of their swallows to occur at higher tidal volumes in
order to compensate for poor recoil and reduce their risk of
prandial aspiration. Dysphagic patients with Parkinson’s
disease, who also have disordered breathing and swallow-
ing coordination [14, 28], may also have impaired recoil
and may benefit from therapies that target lung volume at
swallow onset, in addition to traditional methods that focus
on strengthening swallowing musculature. Lastly, any
factors that affect lung volume and recoil, such as the
position of the body during deglutition, may need to be
taken into consideration when swallowing function is being
evaluated and treated.
Study Limitations
Due to the invasive nature of direct airway puncturing
and the technical difficulties with maintaining a patent
subglottic catheter during laryngeal movement, we elec-
ted to study 10 subjects and not a larger group. The
generalization of the study findings may be limited by the
sample size and our use of a relatively young, healthy
group.
R. D. Gross et al.: Deglutitive Subglottic Air Pressure
Conclusion
During normal swallowing, respiratory system recoil gen-
erates subglottic air pressure. The amount of pressure that
is generated is dependent upon the volume of air within the
lungs at swallow onset. The ‘‘modified’’ interrupter tech-
nique used for this experiment can be used as a noninvasive
method to estimate deglutitive subglottic air pressures.
Acknowledgments The authors acknowledge Marike Vuga, PhD,
for statistical support. This investigation was supported by the UPMC
Competitive Medical Research Fund and the Parkinson Foundation of
Western Pennsylvania. Sheryl Ross, MA, CCC/SLP for assisting with
data collection.
References
1. Logemann JA, Pauloski BR, Colangelo L. Light digital occlusion
of the tracheostomy tube: a pilot study of effects on aspiration
and biomechanics of the swallow. Head Neck. 1998;20(1):52–7.
2. Stachler RJ, Hamlet SL, Choi J, Fleming S. Scintigraphic quan-
tification of aspiration reduction with the Passy-Muir valve.
Laryngoscope. 1996;106(2 Pt 1):231–4.
3. Elpern EH, Borkgren Okonek M, Bacon M, Gerstung C,
Skrzynski M. Effect of the Passy-Muir tracheostomy speaking
valve on pulmonary aspiration in adults. Heart Lung. 2000;29(4):
287–93.
4. Suiter DM, McCullough GH, Powell PW. Effects of cuff defla-
tion and one-way tracheostomy speaking valve placement on
swallow physiology. Dysphagia. 2003;18(4):284–92.
5. Dettelbach MA, Gross RD, Mahlmann J, Eibling DE. Effect of
the Passy-Muir valve on aspiration in patients with tracheostomy.
Head Neck. 1995;17(4):297–302.
6. Gross RD, Atwood CW Jr, Grayhack JP, Shaiman S. Lung vol-
ume effects on pharyngeal swallowing physiology. J Appl
Physiol. 2003;95(6):2211–7.
7. Gross RD, Dettelbach MA, Eibling DE, Zajac DJ. Measurement
of subglottic air pressure during swallowing in a patient with
tracheostomy. Otolaryngol Head Neck Surg. 1994;111(2):133.
8. Eibling DE, Gross RD. Subglottic air pressure: a key component
of swallowing efficiency. Ann Otol Rhinol Laryngol. 1996;
105(4):253–8.
9. Gross RD, Mahlmann J, Grayhack JP. Physiologic effects of open
and closed tracheostomy tubes on the pharyngeal swallow. Ann
Otol Rhinol Laryngol. 2003;112(2):143–52.
10. Shaker R, Dua KS, Ren J, Xie P, Funahashi A, Schapira RM.
Vocal cord closure pressure during volitional swallow and other
voluntary tasks. Dysphagia. 2002;17(1):13–8.
11. Shin T, Maeyama T, Morikawa I, Umezaki T. Laryngeal reflex
mechanism during deglutition–observation of subglottal pressure
and afferent discharge. Otolaryngol Head Neck Surg. 1988;99(5):
465–71.
12. Gross RD, Steinhauer KM, Zajac DJ, Weissler MC. Direct
measurement of subglottic air pressure while swallowing.
Laryngoscope. 2006;116(5):753–61.
13. Folstein MF, Folstein SE, McHugh PR. ‘‘Mini-mental state’’. A
practical method for grading the cognitive state of patients for the
clinician. J Psychiatr Res. 1975;12(3):189–98.
14. Gross RD, Atwood CW Jr, Ross SB, Eichhorn KA, Olszewski
JW, Doyle PJ. The coordination of breathing and swallowing in
Parkinson’s disease. Dysphagia. 2008;23(2):136–45.
459
15. Gross RD, Atwood CW Jr, Ross SB, Olszewski JW, Eichhorn
KA. The coordination of breathing and swallowing in chronic
obstructive pulmonary disease. Am J Respir Crit Care Med.
2009;179(7):559–65.
16. Kawasaki A, Fukuda H, Shiotani A, Kanzaki J. Study of move-
ments of individual structures of the larynx during swallowing.
Auris Nasus Larynx. 2001;28(1):75–84.
17. Shaker R, Milbrath M, Ren J, Campbell B, Toohill R, Hogan W.
Deglutitive aspiration in patients with tracheostomy: effect of
tracheostomy on the duration of vocal cord closure. Gastroen-
terology. 1995;108(5):1357–60.
18. Hardemark Cedborg AI, Sundman E, Boden K, Witt Hedstrom H,
Kuylenstierna R, Ekberg O, et al. Coordination of spontaneous
swallowing with respiratory airflow and diaphragmatic and
abdominal muscle activity in healthy adults. Exp Physiol. 2009;
94(4):459–68.
19. Wheeler Hegland KM, Huber JE, Pitts T, Sapienza CM. Lung
volume during swallowing: single bolus swallows in healthy
young adults. J Speech Lang Hear Res. 2009;52(1):178–87.
20. Hardemark Cedborg AI, Boden K, Witt Hedstrom H, Kuylen-
stierna R, Ekberg O, Eriksson LI et al. Breathing and swallowing
in normal man: effects of changes in body position, bolus types,
and respiratory drive. Neurogastroenterol Motil. 2010;22(11):
1201–8, e316.
21. Boden K, Cedborg AI, Eriksson LI, Hedstrom HW, Kuylenstierna
R, Sundman E, et al. Swallowing and respiratory pattern in young
healthy individuals recorded with high temporal resolution.
Neurogastroenterol Motil. 2009;21(11):e101–1163.
22. Martin-Harris B, Brodsky MB, Michel Y, Ford CL, Walters B,
Heffner J. Breathing and swallowing dynamics across the
adult lifespan. Arch Otolaryngol Head Neck Surg. 2005;131(9):
762–70.
23. Klahn MS, Perlman AL. Temporal and durational patterns asso-
ciating respiration and swallowing. Dysphagia. 1999;14(3):
131–8.
24. Wheeler Hegland K, Huber JE, Pitts T, Davenport PW, Sapienza
CM. Lung volume measured during sequential swallowing in
healthy young adults. J Speech Lang Hear Res. 2011;54(3):
777–86.
25. Gottfried SB, Higgs BD, Rossi A, Carli F, Mengeot PM, Calverly
PM, et al. Interrupter technique for measurement of respiratory
mechanics in anesthetized humans. J Appl Physiol. 1985;59(2):
647–52.
26. Pesenti A, Pelosi P, Foti G, D’Andrea L, Rossi N. An interrupter
technique for measuring respiratory mechanics and the pressure
generated by respiratory muscles during partial ventilatory sup-
port. Chest. 1992;102(3):918–23.
27. Shaker R, Li Q, Ren J, Townsend WF, Dodds WJ, Martin BJ,
et al. Coordination of deglutition and phases of respiration: effect
of aging, tachypnea, bolus volume, and chronic obstructive pul-
monary disease. Am J Physiol. 1992;263(5 Pt 1):G750–5.
28. Pinnington LL, Muhiddin KA, Ellis RE, Playford ED. Non-
invasive assessment of swallowing and respiration in Parkinson’s
disease. J Neurol. 2000;247(10):773–7.
Roxann Diez Gross PhD
Ricardo L. Carrau MD
Ronit G. Gisser MSc
Libby J. Smith DO
David J. Zajac PhD
Frank C. Sciurba MD
123