10008
SECTION II
10008 8 Physiology and Testing
of Respiratory Muscles
THEODOROS VASSILAKOPOULOS
CHARIS ROUSSOS
s0010 INTRODUCTION
p0010 The respiratory muscles are the only muscles, along with the
heart, that have to work continuously, though intermittently,
to sustain life. They have to repetitively move a rather com-
plex elastic structure, the thorax, to achieve the entry of air
into the lungs and thence effect gas exchange. Their great
number mandates that they should interact properly to per-
form their task despite their different anatomic location, geo-
metric orientation, and motor innervation. They should also
be able to adapt to a variety of working conditions and respond
to many different chemical and neural stimuli. This chapter
describes some aspects of the respiratory muscles’ function
that are relevant to the understanding of the way the respira-
tory muscles accomplish the action of breathing and how their
function can be tested in the laboratory.
s0020 FUNCTIONAL ANATOMY
s0030 The Intercostal Muscles
p0020 The intercostal muscles are two thin layers of muscle fibers
occupying each of the intercostal spaces. They are termed
external and internal because of their surface relations, the
external being superficial to the internal. The muscle fibers
of the two layers run approximately at right angles to each
other: The external intercostals extend from the tubercles of
the ribs dorsally to the costochondral junctions ventrally, and
their fibers are oriented obliquely, downward, and forward,
from the rib above to the rib below. The internal intercostals
begin posteriorly, because the posterior intercostal membrane
on the inner aspect of the external intercostal muscles. From
approximately the angle of the rib, the internal intercostal
muscles run obliquely, upward, and forward from the superior
border of the rib and costal cartilage below to the floor of the
subcostal groove of the rib and the edge of the costal cartilage
above, ending at the sternocostal junctions. All the intercostal
muscles are innervated by the intercostal nerves.
p0030 The external intercostal muscles have an inspiratory action on
the rib cage, whereas the internal intercostal muscles are expira-
tory. An illustrative clinical example of the “isolated” inspiratory
action of the intercostal muscles is offered by patients who have
bilateral diaphragmatic paralysis. In these patients, inspiration is
accomplished solely by the rib cage muscles. As a result, the rib
cage expands during inspiration, and the pleural pressure falls.
Because the diaphragm is flaccid and no transdiaphragmatic pres-
sure can be developed, the fall in pleural pressure is transmitted to
the abdomen, causing an equal fall in the abdominal pressure.
Hence the abdomen moves paradoxically inward during
Albert, 978-0-323-04825-5
PHYSIOLOGY
inspiration, opposing the inflation of the lung (Figure 8-1). This
paradoxic motion is the cardinal sign of diaphragmatic paralysis
on clinical examination and is invariably present in the supine
posture, during which the abdominal muscles usually remain
relaxed during the entire respiratory cycle. However, this sign
may be absent in the erect posture.
The Diaphragm s0040
The floor of the thoracic cavity is closed by a thin musculoten- p0040
dinous sheet, the diaphragm, the most important inspiratory
muscle, accounting for approximately 70% of minute ventila-
tion in normal subjects. The diaphragm is anatomically unique
among the skeletal muscles in that its fibers radiate from a cen-
tral tendinous structure (the central tendon) to insert periph-
erally into skeletal structures. The muscle of the diaphragm
falls into two main components on the basis of its point of ori-
gin: the crural (vertebral) part and the costal (sternocostal)
part. The crural part arises from the crura (strong, tapering
tendons attached vertically to the anterolateral aspects of the
bodies and intervertebral disks of the first three lumbar verteb-
rae on the right and two on the left) and the three aponeurotic
arcuate ligaments. The costal part of the diaphragm arises from
the xiphoid process and the lower end of the sternum and the
costal cartilages of the lower six ribs. These costal fibers run
cranially so that they are directly apposed to the inner aspect
of lower rib cage, creating a zone of apposition.
The shape of the relaxed diaphragm at the end of a normal p0050
expiration (functional residual capacity, FRC) is that of two
domes joined by a saddle that runs from the sternum to the
anterior surface of the spinal column (Figure 8-2). The motor
innervation of the diaphragm is from the phrenic nerves,
which also provide a proprioceptive supply to the muscle.
When tension develops within the diaphragmatic muscle
fibers, a caudally oriented force is applied on the central ten-
don and the dome of the diaphragm descends; this descent
has two effects. First, it expands the thoracic cavity along its
craniocaudal axis and consequently the pleural pressure falls.
Second, it produces a caudal displacement of the abdominal
visceral contents and an increase in the abdominal pressure
that in turn results in an outward motion of the ventral
abdominal wall. Thus, when the diaphragm contracts, a crani-
ally oriented force is being applied by the costal diaphragmatic
fibers to the upper margins of the lower six ribs that has the
effect of lifting and rotating them outward (insertional force;
see Figure 8-2). The actions mediated by the changes in pleural
and abdominal pressures are more complex: if one assumes
that the diaphragm is the only muscle acting on the rib cage,
it seems that it has two opposing effects when it contracts.
137
 10008
138 SECTION II Physiology

Normal subject Patient with diaphragmatic paralysis

Chest Abdomen Chest Abdomen

f0010 FIGURE 8-1 Schematic demonstration of normal abdominal and rib cage movement (left panel) and the paradoxical abdominal motion of
isolated diaphragmatic paralysis (right panel). The diaphragm at resting end-expiration is shown as a solid line and after inspiration as a
dashed line. In the normal subject (left), the diaphragm moves caudally, and in the patient with diaphragmatic paralysis (right), the diaphragm
moves in a cephalad direction. The anterior abdominal wall moves inward instead of outward.

On the upper rib cage, it causes a decrease in the anteroposter-
ior diameter, and this expiratory action is primarily because of
the fall in pleural pressure (see Figure 8-2). On the lower rib
cage, it causes an expansion. In fact this is the pattern of chest
wall motion observed in tetraplegic patients with transection
injury at the fifth cervical segment of the spinal cord or below,
who have complete paralysis of the inspiratory muscles except
for the diaphragm. This inspiratory action on the lower rib
cage is caused by the concomitant action of two different
forces, the “insertional” force already described and the “appo-
Au1 sitional” force c. The Neck Muscles.
s0050 The Sternocleidomastoids
p0060 The sternocleidomastoids arise from the mastoid process and
descend to the ventral surface of the manubrium sterni and
the medial third of the clavicle. Their neural supply is from
the accessory nerve. The action of the sternocleidomastoids is
to displace the sternum cranially during inspiration, to expand
the upper rib cage more in its anteroposterior diameter than in
its transverse one, and to decrease the transverse diameter of
the lower rib cage. In normal subjects breathing at rest, how-
ever, the sternocleidomastoids are inactive, being recruited
only when the inspiratory muscle pump is abnormally loaded
or when ventilation increases substantially. Therefore, they
should be considered as accessory muscles of inspiration.
The Abdominal Muscles s0070
The abdominal muscles with respiratory activity are those con- p0080
stituting the ventrolateral wall of the abdomen (i.e., the rectus
abdominis ventrally and the external oblique, internal oblique,
and transverses abdominis laterally). They are innervated by
the lower six thoracic nerves and the first lumbar nerve. As they
contract, they pull the abdominal wall inward, thus increasing
the intraabdominal pressure. This causes the diaphragm to move
cranially into the thoracic cavity, increasing the pleural pressure
and decreasing lung volume. Thus, their action is expiratory.
Expiration is usually a passive process but can become active
when minute ventilation has to be increased (e.g., during exer-
cise) or during respiratory distress. Expiratory muscle action is
also essential during cough.
TESTING RESPIRATORY MUSCLE FUNCTION s0080
Muscles have two functions: to develop force and to shorten. p0090
In the respiratory system, force is usually estimated as pressure
and shortening as lung volume change. Thus, quantitative
characterization of the respiratory muscles usually relies on
measurements of volumes and pressures.
Vital Capacity s0090
Vital capacity (VC) is an easily obtained measurement with p0100

s0060 The Scalenes
p0070 The scalenes are composed of three muscle bundles that run
from the transverse processes of the lower five cervical verteb-
rae to the upper surface of the first two ribs. They receive their
neural supply mainly from the lower five cervical segments.
Their action is to increase (slightly) the anteroposterior diam-
eter of the upper rib cage. Although initially considered as
accessory muscles of inspiration, they are invariably active dur-
ing inspiration. In fact, seated normal subjects cannot breathe
without contracting the scalenes even when they reduce the
spirometry, which, when decreased, points to respiratory mus-
cle weakness. The VC averages approximately 50 mL/kg in
normal adults. However, VC is not specific and may be
decreased because of both inspiratory and expiratory muscle
weakness and restrictive lung and chest wall diseases. A
marked fall (>30%) in VC in the supine compared with the
erect posture (which in the normal individual is 5–10%) is
associated with severe bilateral diaphragmatic weakness.
Maximal Static Mouth Pressures s0100

required inspiratory effort by reducing tidal volume. There- Measurement of the maximum static inspiratory (PI,max) or p0110
fore, scalenes in humans are primary muscles of inspiration, expiratory (PE,max) pressure that a subject can generate at the
and their contraction is an important determinant of the mouth is a simple way to estimate inspiratory and expiratory
expansion of the upper rib cage during breathing. muscle strength. These are measured at the side port of a

Albert, 978-0-323-04825-5
 10008
8 Physiology and Testing of Respiratory Muscles 139

RC
Central
Rib cage tendon
L
Costal
muscle
fibers
3
Pleural
space Lung
Insertional DI
Insertional
1
force
Diaphragm Appositional
“Zone of
apposition” Pab Pab

A B

IV IV

IV IV

V V
“Zone of
Appositional force apposition”

Pab
Position of
diaphragm in
maximal inspiration
and expiration
C D

f0020 FIGURE 8-2 Actions of the diaphragm. A, Zone of apposition and summary of diaphragm’s actions. When the diaphragm contracts, a caudally
oriented force is being applied on the central tendon and the dome of the diaphragm descends (DI). Furthermore, the costal diaphragmatic
fibers apply a cranially oriented force to the upper margins of the lower six ribs that has the effect of lifting and rotating them outward (insertional
force, arrow 1). The zone of apposition makes the lower rib cage part of the abdomen, and the changes in pressure in the pleural recess between
the apposed diaphragm and the rib cage are almost equal to the changes in abdominal pressure. Pressure in this pleural recess rises
rather than falls during inspiration because of diaphragmatic descent, and the rise in abdominal pressure is transmitted through the
apposed diaphragm to expand the lower rib cage (arrow 2). All these effects result in expansion of the lower rib cage. On the upper
rib cage, isolated contraction of the diaphragm causes a decrease in the anteroposterior diameter, and this expiratory action is primar-
ily caused by the fall in pleural pressure (arrow 3). B, Insertional force; C, appositional force; D, shape of the diaphragm and the bony
thorax at maximum inspiration and expiration.

mouthpiece that is occluded at the distal end. A small leak is
incorporated to prevent glottic closure and buccal muscle use
during inspiratory or expiratory maneuvers. The inspiratory
and expiratory pressure must be maintained, ideally for at least
1.5 sec, so that the maximum pressure sustained for 1 sec can
be recorded (Figure 8-3). The pressure measured during these
maneuvers reflects the pressure developed by the respiratory
muscles (Pmus), plus the passive elastic recoil pressure of
the respiratory system including the lung and chest wall (Prs)
(Figure 8-4). At FRC, Prs is 0 so that Pmo represents Pmus.
However, at residual volume (RV), where PI,max is usually
measured, Prs may be as much as 30 cm H2O, and thus makes
a significant contribution PI,max of up to 30% (or more if Pmus
is decreased). Similarly, PI,max is measured at total lung capacity
(TLC), where Prs can be up to 40 cm H2O. Clinical measures
and normal values of PI,max and PE,max do not conventionally
subtract the elastic recoil of the respiratory system. Normal
values are available for adults, children, and the elderly. The
tests are easy to perform and are well tolerated. However, the
measurements exhibit significant between-subject and within-
subject variability, as well as learning effect (values obtained
improve as subjects become accustomed to the maneuvers).
The normal ranges are wide, so that values in the lower quarter
of the normal range are compatible both with normal strength
and with mild or moderate weakness. However, a PI,max
of 80 cm H2O usually excludes clinically important in-
spiratory muscle weakness. Values less negative than this are
difficult to interpret, and more detailed studies are required.
A normal PE,max with a low PI,max suggests isolated diaphragmatic
weakness.

Albert, 978-0-323-04825-5
 10008
140 SECTION II Physiology

1s Peak pressure
1s
0 20 Pe,max

–2
Pressure (kPa) 15

Pressure (kPa)
–4

–6 10

–8
5
Pi,max
–10
0
Peak pressure

0 1 2 3 1 2 3 4 5
A Time (s) B Time (s)

f0030 FIGURE 8-3 A, Pressure tracing from a subject performing a maximum inspiratory maneuver (PImax). A peak pressure is seen, and the 1-sec
average is determined by calculating the shaded area. B, Typical pressure tracing from a subject performing a maximum expiratory maneuver
Au 6 (PEmax). (Adapted from ATS/ERS: Statement on Respiratory Muscle Testing. Am J Respir Crit Care Med 2002; 166:518–624.)

s0110 Transdiaphragmatic Pressure rapid, fully coordinated recruitment of the diaphragm and
p0120 When inspiratory muscle weakness is confirmed, the next diag- other inspiratory muscles. The nose acts as a Starling resistor,
nostic step is to unravel whether this is due to diaphragmatic so that nasal flow is low and largely independent of the driving
weakness, because the diaphragm is the most important in- pressure that is the esophageal pressure. Pdi measured during a
spiratory muscle. This is accomplished by the measurement of sniff (Pdi,sn,max) reflects diaphragm strength, and Pes reflects
maximum transdiaphragmatic pressure (Pdi,max). Pdi,max is the the integrated pressure of the inspiratory muscles on the lungs
difference between gastric pressure (reflecting abdominal pressure) (Figure 8-5). Pressures measured in the mouth, nasopharynx,
and esophageal pressure (reflecting intrapleural pressure) on or one nostril give a clinically useful approximation to eso-
a maximum inspiratory effort after the insertion of appropri- phageal pressure during sniffs without the need to insert
ate balloon catheters in the esophagus and the stomach, esophageal balloons, especially in the absence of significant
respectively. obstructive airway disease. To be useful as a test of respiratory
muscle strength, sniffs need to be maximal, which is relatively
s0120 Sniff Pressures easy for most willing subjects, but may require some practice.
p0130 A sniff is a short, sharp voluntary inspiratory maneuver per- The nasal sniff pressure is the easiest measurement for the sub-
formed through one or both unoccluded nostrils. It achieves ject. Pressure is measured by wedging a catheter in one nostril
by use of foam, rubber bungs, or dental impression molding (Fig-
ure 8-6). The subject sniffs through the contralateral unob-
structed nostril. There is a wide range of normal values,
reflecting the wide range of normal muscle strength in different
100
individuals. In clinical practice, Pdi,sn,max values greater than
100 cm H2O in males and 80 cm H2O in females are unlikely
80 to be associated with clinically significant diaphragm weakness. Au2
Prs Values of maximal sniff esophageal or nasal pressure numeri-
60 Pmus + Prs cally greater than 70 cm H2O (males) or 60 cm H2O (females)
% VC

Max.
expiration are also unlikely to be associated with significant inspiratory
40 Max.
inspiration muscle weakness. However, these reflect the integrated pres-
Pmus + Prs sure of all the inspiratory muscles, and it is possible that there
20
Pmus Pmus
could be a degree of weakness of one or more of these muscle
0 groups that would not be detected at this level. In chronic
–150 –100 –50 0 50 100 150 200 250 obstructive pulmonary disease, nasal sniff pressure tends to
cm H2O underestimate sniff esophageal pressure because of dampened
pressure transmission from the alveoli to the upper airway but
can complement PI,max in excluding weakness clinically.
f0040 FIGURE 8-4 Relationship of muscle and respiratory pressures at dif-
ferent lung volumes. Vertical axis, lung volume as a percentage of vital
capacity (%VC). Horizontal axis, alveolar pressure in cm H2O. The bro- Electrophysiologic Testing s0130
ken lines indicate the pressure contributed by the muscles. Pmus, The next diagnostic step consists of determining whether p0140
pressure developed by the respiratory muscles; Prs, pressure of the weakness is due to muscle, nerve, or neuromuscular transmis-
respiratory system. (Adapted from Agostoni E, Mead J: Statics of sion impairment. This requires the measurement of Pdi in
the respiratory system. In Fenn WO, Rahn H, eds: Handbook of response to bilateral supramaximal phrenic nerve electrical or
physiology: respiration, Vol. 1, Section 3. Washington, DC: American
magnetic stimulation, with concurrent recording of the elicited
Physiology Society; 1964: 387–409.)

Albert, 978-0-323-04825-5
 10008
8 Physiology and Testing of Respiratory Muscles 141

Normal subject Diaphragm paralysis

Pressure 100 100

(cm H2O)

FIGURE 8-5 Examples of the f0050

Sniff Pdi
50 50 sniff maneuver. Left panel shows
a recording from a healthy sub-
ject. Note that the esophageal
(pleural) pressure change is
subatmospheric, whereas the
intraabdominal pressure be-
0 0 comes more positive. Measure-
ment conventions for the sniff
esophageal (Sn Pes) and sniff
transdiaphragmatic pressures
Sniff Pes (Sn Pdi) are illustrated. The right
panel shows an example from
-50
a patient with bilateral diaphrag-
matic paralysis. Note that
there is now a negative pres-
500 ms 500 ms sure change in the abdominal
compartment, because the
diaphragm fails to prevent
Pdi Pgas Pes
pressure transmission from
the thorax.

electromyogram (EMG) of the diaphragm (called the com-
pound muscle action potential, CMAP) with either surface
or esophageal electrodes (Figure 8-7). If the phrenic nerve is
stimulated, the diaphragm contracts. This contraction is called
a twitch. If the stimulus is intense enough, all phrenic fibers
are activated synchronously giving reproducible results. The
intensity of the twitch increases with the frequency of stimula-
tion. If multiple impulses stimulate the phrenic nerve, the con-
tractions summate to cause a tetanic contraction. Thus, if both
phrenic nerves are stimulated with various frequencies (1, 10,
20, 50, and 100 Hz) at the same lung volume with closed air-
way (to prevent entry of air and thus changes in lung volume
and initial length of the diaphragm), the isometric force-
frequency curve of the diaphragm is obtained (Figure 8-8).
(It should be noted that the usual rate of motor nerve dis-
charge during voluntary muscle contraction in humans is
between 5 and 15 Hz, and, because of the steep shape of the
force-frequency curve in this range, small alterations in the
discharge rate cause significant changes in the force produced.
Maximum voluntary contractions, such as the PI,max are
achieved with discharge rates higher than 50 Hz, but cannot
be sustained for long. Stimulation of the phrenic nerve with
high frequencies is technically difficult to achieve (because of
displacement of the stimulating electrode by local contraction
of the scalene muscles and movement of the arm and shoulder
because of activation of the brachial plexus). Therefore, the
transdiaphragmatic pressure developed in response to single
supramaximal phrenic nerve stimulations at 1 Hz, called the
twitch Pdi, is commonly measured.
Although technically demanding, this approach has the p0150
great advantage of being independent of patient effort/motiva-
tion. This also allows for the measurement of phrenic nerve con-
duction time or phrenic latency (i.e., the time between the onset
of the stimulus and the onset of CMAP [Mwave] on the dia-
phragmatic EMG) (Figure 8-7, B). A prolonged conduction
time suggests nerve involvement.

f0060
FIGURE 8-6 The sniff maneuver, which makes use of
a nasal bung and an adapted pressure meter. A, Mea-
surement setup. The meter returns a numerical value
0 that is the amplitude of the pressure swing between
atmospheric (0) pressure and the nadir. B, The trace
produced. The meter returns a numerical value that
Sniff nasal
inspiratory is the amplitude of the pressure swing between
pressure atmospheric (0) pressure and the nadir.
(SNIP)

Albert, 978-0-323-04825-5
 10008
142 SECTION II Physiology

FIGURE 8-7 Electrophysiologic testing: The “twitch Pdi.” f0070

A, The twitch transdiaphragmatic pressure (Pdi) after mag-
netic/electrical stimulation. B, A detailed (enlarged) view of 15

cm H2O
a compound muscle action potential (M wave), where the

Pes
latency (time from stimulus to muscle depolarization), 0
the duration, and the amplitude of the electromyogram are –15
evident. a.u., Arbitrary units; L-CMAP, left compound motor
action potential; Pes, esophageal pressure; Pga, gastric pres-
40
sure; R-CMAP, right compound motor action potential.

cm H2O
(Adapted from Vassilakopoulos T, Roussos C: Neuromuscular

Pga
20
respiratory failure. In Clinical Critical Care Medicine, Slutsky A,
Au7 Takala R, Torres R, eds. St. Louis: Mosby.) 0

40

cm H2O
Pdi
20

10

R-CMAP
a.u.
0

–10

10

L-CMAP
a.u.
0

–10
0 0.2 0.4
A Time (s)
Amplitude 2 mV

Amplitude

Latency Duration

p0160 However, electrophysiologic testing also has shortcomings.
Although the conduction time or latency is prolonged in neu-
ropathies that are predominantly demyelinating, it may be pre-
served in neuropathies that are predominantly axonal despite
substantial diaphragm weakness. Moreover, when the preceding
technique is used, it is important that costimulation of the bra-
chial plexus be avoided, otherwise the action potential recorded
from surface electrodes may originate from muscles other than
the diaphragm. This problem is compounded if the phrenic
nerve is stimulated by use of a magnetic technique. Classically,
an axonal neuropathy is characterized by the finding of pre-
served latencies with diminished CMAP. Lack of CMAP after
nerve stimulation is an indication of paralysis with the lesion
located proximal to or at the neuromuscular junction. De-
creased twitch Pdi in the face of normal CMAP is characteristic
of contractile dysfunction that resides within the muscle.
PHYSIOLOGY: THE ABILITY TO BREATH: s0140
THE LOAD/CAPACITY BALANCE
For a human to take a spontaneous breath, the inspiratory mus- p0170
cles must generate sufficient force to overcome the elastance of
the lungs and chest wall (lung and chest wall elastic loads), as
well as the airway and tissue resistance (resistive load). This
requires an adequate output of the centers controlling the mus-
cles, anatomic and functional nerve integrity, unimpaired neu-
romuscular transmission, an intact chest wall, and adequate
muscle strength. This can be schematically represented by con-
sidering the ability to take a breath as a balance between inspira-
tory load and neuromuscular competence (Figure 8-9). Under
normal conditions, this system is polarized in favor of
neuromuscular competence (i.e., there are reserves that permit
considerable increases in load). However, for a human to

Albert, 978-0-323-04825-5
 10008
8 Physiology and Testing of Respiratory Muscles 143

100
Force
Energy supplies
(% maximum) Blood substrate
80 concentration
Arterial oxygen
content Energy demands
60
Energy stores/ Efficiency
nutrition (VT/tl)
40
Ability to extract V’E
energy sources
20 (tl/ttot)
Inspiratory muscle
blood flow PI/PI,max
0
0 20 40 60 80 60
Stimulation frequency
Fatigued muscle Fresh muscle

f0080 FIGURE 8-8 Force-frequency relationship of in vivo human respira-

Au4 tory muscles. The force-frequency curve of the fresh muscle is shown
in red, and the relationship after a fatiguing task is shown in blue; a
disproportionate force loss at low stimulation frequencies is
FIGURE 8-10 Balance between energy supplies and energy f0100
observed. (Adapted from Moxham J, Wiles CM, Newham D,
demands. Respiratory muscle endurance is determined by the bal-
Edwards RHT: Ciba Found Symp 1981; 82:197–212.)
ance between energy supplies and demands. Normally, the supplies
meet the demands, and a large reserve exists. Whenever this bal-
ance weighs in favor of demands, the respiratory muscles ultimately
become fatigued, leading to inability to sustain spontaneous Au8
breathing decreased. VT/tI, Mean inspiratory flow (tidal volume/
inspiratory time); tI/ttot, duty cycle (fraction of inspiration to total
breathing cycle duration); PI/PI, max, inspiratory pressure/maximum
inspiratory pressure ratio; V 0 E, minute ventilation. (Adapted from
Vassilakopoulos T, Roussos C: Neuromuscular respiratory failure.
In Slutsky A, Takala R, Torres, eds. Clinical critical care medicine.
St. Louis: Mosby.)

Lung elastic loads Central drive

Chest wall
Neural and breathe spontaneously, the inspiratory muscles should be able to
neuromuscular sustain the aforementioned load over time as well as adjust the
elastic loads
transmission
minute ventilation in such a way that there is adequate gas
Resistive loads Muscle strength exchange. The ability of the respiratory muscles to sustain this
Load (per breath) Neuromuscular load without the appearance of fatigue is called endurance and
(PI) competence is determined by the balance between energy supplies and
(PI, max) energy demands (Figure 8-10).
Energy supplies depend on the inspiratory muscle blood p0180
flow, the blood substrate (fuel) concentration and arterial oxy-
gen content, the muscle’s ability to extract and use energy
sources, and the muscle’s energy stores. Under normal circum-
stances, energy supplies are adequate to meet the demands,
and a large recruitable reserve exists (see Figure 8-10). Energy
f0090 FIGURE 8-9 Balance between inspiratory load and neuromuscular demands increase proportionally with the mean pressure
competence. The ability to take a spontaneous breath is deter- developed by the inspiratory muscles per breath (PI) expressed
mined by the balance between the load imposed on the respiratory as a fraction of maximum pressure that the respiratory muscles
system (pressure developed by the inspiratory muscles; PI) and the can voluntarily develop (PI/PI,max), the minute ventilation
neuromuscular competence of the ventilatory pump (maximum
(VE), the inspiratory duty cycle (TI/TTOT), and the mean
inspiratory pressure; PI, max). Normally, this balance weighs in favor
of competence, permitting significant increases in load. However, if
inspiratory flow rate (VT/TI) and are inversely related to the
the competence is, for whatever reason, reduced below a critical efficiency of the muscles. Fatigue develops when the mean rate
point (e.g., drug overdose, myasthenia gravis), the balance may of energy demands exceeds the mean rate of energy supply
then weigh in favor of load, rendering the ventilatory pump insuffi- (i.e., when the balance is polarized in favor of demands).
cient to inflate the lungs and chest wall. (Adapted from Vassilakopou- The product of TI/TTOT and the mean transdiaphragmatic p0190
los T, Roussos C: Neuromuscular respiratory failure. In Slutsky A, pressure expressed as a fraction of maximal (Pdi/Pdi,max)
Takala R, Torres, eds. Clinical critical care medicine. St. Louis: Mosby.) defines a useful “tension-time index” (TTIdi) that is related to
the endurance time (i.e., the time that the diaphragm can

Albert, 978-0-323-04825-5
 10008
144 SECTION II Physiology

Energy supplies Energy demands

Blood substrate ↓ Efficiency
concentration
(VT, tl)
Arterial oxygen
content VE
Energy stores/ (tl/tlot)
nutrition
Pl/Pt, max
Ability to extract
energy sources
Inspiratory muscle
blood flow

Lung elastic loads

Chest wall
elastic loads Central drive
Resistive loads Neural and neuromuscular
transmission
Load
Muscle strength

Competence

f0110 FIGURE 8-11 System of two balances: load and competence, energy supplies and demands. The system of two balances, incorporating the
various determinants of load, competence, energy supplies, and demands is represented schematically. The PI/PI, max, one of the determi-
nants of energy demands (see Figure 8-10) is replaced by its equivalent: the balance between load and neuromuscular competence (see Figure 8-
9). In fact, this is the reason the two balances are linked. When the central hinge of the system moves upward or is at least at the horizontal level, a
balance exists between ventilatory needs and neurorespiratory capacity, and spontaneous ventilation can be sustained. In healthy persons, the
hinge moves far upward, creating a large reserve. For abbreviations, see legends to Figures 8-9 and 8-10. (Adapted from Vassilakopoulos T,
Roussos C: Neuromuscular respiratory failure. In Slutsky A, Takala R, Torres R, eds. Clinical critical care medicine. St. Louis: Mosby.)

sustain the load imposed on it). Whenever TTIdi is smaller
than the critical value of 0.15, the load can be sustained indefi-
nitely; but when TTIdi exceeds the critical zone of 0.15–0.18,
the load can be sustained only for a limited time period, in other
words, the endurance time. This was found to be inversely
related to TTIdi. The TTI concept is assumed to be applicable
not only to the diaphragm but also to the respiratory muscles
as a whole:
TTI ¼ PI =PI;maxw TI =TTOT
p0210 Because endurance is determined by the balance between
energy supply and demand, TTI of the inspiratory muscles
system (Figure 8-11). Schematically, when the central hinge
of the system moves upward, or is at least at the horizontal
level, spontaneous ventilation can be sustained indefinitely
(see Figure 8-11). The ability of a subject to breathe spontane-
ously depends on the fine interplay of many different factors.
Normally, this interplay moves the central hinge far upward
and creates a great ventilatory reserve for the healthy individ-
ual. When the central hinge of the system, for whatever reason,
moves downward, spontaneous ventilation cannot be sustained,
and ventilatory failure ensues.
Hyperinflation s0150

has to be in accordance with the energy balance view. In fact,
as Figure 8-4 demonstrates, PI/PI,max and TI/TTOT, which con-
stitute the TTI, are among the determinants of energy
demands; an increase in either that will increase the TTI value
will also increase the demands. But what determines the ratio
PI/PI,max? The nominator, the mean inspiratory pressure devel-
oped per breath, is determined by the elastic and resistive
loads imposed on the inspiratory muscles. The denominator,
the maximum inspiratory pressure, is determined by the neuro-
muscular competence (i.e., the maximum inspiratory muscle
activation that can be voluntarily achieved). It follows, then,
that the value of PI/PI,max is determined by the balance between
load and competence (see Figure 8-9). But PI /PI,max is also one
of the determinants of energy demands (see Figure 8-10);
therefore, the two balances (i.e., between load and competence
and energy supply and demand) are in essence linked, creating a
Hyperinflation (frequently observed in obstructive airway dis- p0220
eases) compromises the force-generating capacity of the dia-
phragm for a variety of reasons: First, the respiratory muscles,
like other skeletal muscles, obey the length-tension relation-
ship. At any given level of activation, changes in muscle
fiber length alter tension development. This is because the
force-tension developed by a muscle depends on the interaction
between actin and myosin fibrils (i.e., the number of myosin
heads attaching and thus pulling the actin fibrils closer within
each sarcomere). The optimal fiber length (Lo) where
tension is maximal is the length at which all myosin heads
attach and pull the actin fibrils. Below this length (as with
hyperinflation, which shortens the diaphragm), actin-myosin
interaction becomes suboptimal, and tension development
declines. Second, as lung volume increases, the zone of apposi-
tion of the diaphragm decreases in size, and a larger fraction of

Albert, 978-0-323-04825-5
 10008
8 Physiology and Testing of Respiratory Muscles 145

Thorax Thorax

Upper
3
rib cage

Ppl Ppl
Diaphragm Diaphragm
Shortened (hyperinflated)
DI muscle fibers
2 1
Lower Decreased
rib cage Pap Insertional curvature
force
Zone of Decreased zone
apposition of apposition

Appositional Medial orientation of

force diaphragmatic fibers
A B
FIGURE 8-12 Consequences of hyperinflation on the diaphragm. A, Normal actions of the diaphragm as in Figure 8-2. B, Deleterious effects f0120
of hyperinflation on the diaphragm.

the rib cage becomes exposed to pleural pressure. Hence, the
diaphragm’s inspiratory action on the rib cage diminishes.
When lung volume approaches total lung capacity, the zone
of apposition all but disappears (Figure 8-12), and the dia-
phragmatic muscle fibers become oriented horizontally inter-
nally (Figure 8-12, B). The insertional force of the diaphragm
is then expiratory, rather than inspiratory, in direction. This
explains the inspiratory decrease in the transverse diameter of
the lower rib cage in subjects with emphysema and severe
hyperinflation (Hoover’s sign). Third, the resultant flattening
of the diaphragm increases its radius of curvature (Rdi) (see
Figure 8-12, B) and, according to Laplace’s law, Pdi ¼ 2Tdi/
Rdi, diminishes its pressure-generating capacity (Pdi) for the
same tension development (Tdi).
s0160 RESPIRATORY MUSCLE RESPONSES
TO CHANGES IN LOAD
s0170 Acute Responses
s0180 Increased Load
p0230 Respiratory Muscle Fatigue. Fatigue is defined as the loss of
s0190 capacity to develop force and/or velocity in response to a load
that is reversible by rest. Thus, fatigue may be present before
the point at which a muscle is unable to continue to perform
a particular task (task failure). In applying this concept to the
inspiratory muscles, one could conclude that they may be
fatigued before there is hypercapnia because of their inability
to continue to generate sufficient pressure to maintain alveolar
ventilation.
p0240 Fatigue should be distinguished from weakness in which
reduced force generation is fixed and not reversed by rest,
although the presence of weakness may itself predispose a
muscle to fatigue. The site and mechanisms of fatigue remain
controversial. Theoretically, the site of fatigue may be located
at any link in the long chain of events involved in voluntary
muscle contraction leading from the brain to the contractile
machinery. A widely used convention is to classify fatigue as
central fatigue, peripheral high-frequency fatigue, or peripheral
low-frequency fatigue.
Central Fatigue. Central fatigue is present when a maximal p0250
voluntary contraction generates less force than does maximal s0200
electrical stimulation. If maximal electrical stimulation super-
imposed on a maximal voluntary contraction can potentiate
the force generated by a muscle, a component of central
fatigue exists. This procedure applied to the diaphragm con-
sists of the twitch occlusion test that may separate central from
peripheral fatigue. This test examines the transdiaphragmatic
pressure (Pdi) response to bilateral phrenic nerve stimulation
superimposed on graded voluntary contractions of the dia-
phragm. Normally, the amplitude of the Pdi twitches in
response to phrenic nerve stimulation decreases as the volun-
tary Pdi increases. During Pdi,max, no superimposed twitches
can be detected. When diaphragmatic fatigue is present, super-
imposed twitches can be demonstrated. A number of experi-
ments have suggested that a form of central diaphragmatic
“fatigue” may develop during respiratory loading so that, at
the limits of diaphragmatic endurance, a significant portion
of the reduction in force production is due to failure of the
central nervous system to completely activate the diaphragm.
Central fatigue may be caused by a reduction in the number
of motor units that can be recruited by the motor drive or by
a decrease in motor unit discharge rates or both. The observed
decreased central firing rate during fatigue may, in fact, be a
beneficial adaptive response preventing the muscle’s self-
destruction by excessive activation.
Peripheral Fatigue. Peripheral fatigue refers to failure at the p0260
neuromuscular junction or distal to this structure and is pres- s0210
ent when muscle force output falls in response to direct electri-
cal stimulation. This type of fatigue may occur because of failure
of impulse propagation across the neuromuscular junction,
the muscle surface membrane or the t tubules (transmission
fatigue), impaired excitation—contraction coupling, or failure
of the contractile apparatus of the muscle fibers (because of
alterations in metabolism or changes in contractile proteins).

Albert, 978-0-323-04825-5
 10008
146 SECTION II Physiology
Peripheral fatigue can be further classified into high-frequency
and low-frequency fatigue on the basis of the shape of the muscle
force-frequency curve (Figure 8-8). High-frequency fatigue
results in depression of the forces generated by a muscle in
response to high-frequency electrical stimulation (50–100 Hz),
whereas low-frequency fatigue results in depression of force
generation in response to low-frequency stimuli (1–20 Hz).
Low-frequency fatigue can occur in isolation, but high-
frequency fatigue is invariably associated with some alterations
in muscle force generation at lower frequencies.
p0270 High-Frequency Fatigue. During artificial stimulation of a
s0220 motor neuron, especially at high frequencies, muscle force
declines rapidly in association with the decline in CMAP
amplitude. This response, known as “high-frequency fatigue”
(see Figure 8-8), is attributed to transmission fatigue. The site
of this type of fatigue may be located postsynaptically (from a
decrease in sarcolemmal membrane endplate excitability or a
reduction in action potential propagation into the t tubular
system) or presynaptically (probably in fine terminal filaments
of the motor nerve or less frequently from depletion of synap-
tic transmitter substance). Teleologically, transmission block
could be beneficial in some instances by protecting the muscle
against excessive depletion of its ATP stores, which would lead
to rigor mortis. Normal subjects breathing against high-intensity
inspiratory resistive loads develop high-frequency fatigue,
which resolves very quickly after cessation of the strenuous
diaphragmatic contractions.
p0280 Low-Frequency Fatigue. All processes that link the electrical
s0230 activation of the muscle fiber and the various metabolic and
enzymatic processes providing energy to the contractile
machinery are called excitation-contraction coupling pro-
cesses. Impaired excitation contraction coupling is thought to
be responsible when the loss of force is not accompanied by
a parallel decline in the electrical activity. This type of fatigue
is characterized by a selective loss of force at low frequencies
of stimulation (low-frequency fatigue) (see Figure 8-8) despite
maintenance of the force generated at high frequencies of
stimulation, indicating that the contractile proteins continue
to generate force provided that sufficient calcium is released
by the sarcoplasmic reticulum. The mechanism of this type
of fatigue is not understood. It may occur because of a reduced
level of calcium availability caused by alterations in sarcoplas-
mic reticulum function or a reduction in the calcium sensitiv-
ity of the myofilaments (troponin binding site) at submaximal
calcium concentrations (because of high hydrogen and phos-
phate ion concentration). These defects would reduce the
force developed at low-frequency stimulation. In contrast, at
higher stimulation frequencies, a relatively normal force can
be generated when the interior of the fiber is saturated with
calcium.
p0290 This type of fatigue is characteristically long lasting, taking sev-
eral hours to recover. Low-frequency fatigue occurs during high-
force contractions and is less likely to develop when the forces
generated are smaller, even if these are maintained for longer time
periods, thereby achieving the same total work. As previously
stated, fatigue develops when the mean rate of energy demands
exceeds the mean rate of energy supply to the muscle (see Fig-
ure 8-10), resulting in depletion of muscle energy stores, pH
changes from lactate accumulation, and excessive production of
oxygen-derived free radicals. However, the exact interplay of all
these factors is not yet identified in either the diaphragm or the
Albert, 978-0-323-04825-5
other skeletal muscles. Accordingly, low-frequency fatigue occurs
in the diaphragm of experimental animals during cardiogenic or
septic shock. Low-frequency fatigue has also been found in the
diaphragm and sternocleidomastoid of normal subjects after they
breathed against very high inspiratory resistance or after sustain-
ing maximum voluntary ventilation (for 2 min) (Figure 8-13).
The clinical relevance of respiratory muscle fatigue is diffi- p0300
cult to figure out, because the measurements that are required
for fatigue detection are hard to apply in situations where
fatigue is likely present (such as during acute hypercapnic
respiratory failure).
Inflammation and Injury. Strenuous diaphragmatic contrac- p0310
tions (induced by resistive breathing, which accompanies s0240
many disease states such as COPD and asthma) initiate an
inflammatory response consisting of elevation of plasma cyto-
kines and recruitment and activation of white blood cell sub-
populations. These cytokines are produced within the
diaphragm secondary to the increased muscle activation.
Strenuous resistive breathing results in diaphragmatic ultra-
structural injury (such as sarcomere disruption, necrotic fibers,
flocculent degeneration, and influx of inflammatory cells) in
both animals and humans. The mechanisms involved are not
definitely established but may involve intradiaphragmatic
cytokine induction, adhesion molecule upregulation, calpain
activation, and reactive oxygen species formation. Cytokines
are also essential in orchestrating muscle recovery after injury
by enhancing proteolytic removal of damaged proteins and
cells (through recruitment and activation of phagocytes) and
by activating satellite cells. Satellite cells are quiescent cells
of embryonic origin that reside in the muscle and are
Twitch Pdi
(% of baseline 100
value)
90
80
0
0 10 20 40 60 80 100
Minutes after maximal hyperventilation
Intense diaphragm contraction Sham run
FIGURE 8-13 Demonstration of low-frequency fatigue of the dia- f0130
phragm in normal subjects. The mean twitch tension (twitch trans-
diaphragmatic pressure; Tw Pdi) is shown before and at intervals
up to 90 min after intense diaphragmatic contraction (in this case,
a 2-min period of maximal normocapnic hyperventilation) in nine
healthy adults (blue symbols). Data in the same subjects after a
sham (“normal breathing”) run are shown in red. A significant
decline in Tw Pdi is observed that has only partially recovered at
90 min, which confirms the presence of low-frequency diaphragmatic
fatigue. (Data from Hamnegård C-H, Wragg SD, Kyroussis D, et al:
Diaphragm fatigue following maximal ventilation in man. Eur Respir
J 1996; 9:241–247.)
 10008
8 Physiology and Testing of Respiratory Muscles 147

FF FR S

IIb IIa I
A MHC2B MHC2A MHCSlow
Force

Force

Force
B Time Time Time
Size

Fat
igu
e re
sist
anc
e

d
ee
Sp I
IIb IIa

FIGURE 8-14 Properties of skeletal muscle fiber types. Different fiber types in the diaphragm muscle are distinguished by size, myosin heavy f0140
chain content, contractile characteristics (force and speed of contraction), and fatigue resistance (type S, slow; type FR, fast-twitch, fatigue
resistant; and type FF motor units, fast-twitch, fatigable), as well as myosin heavy chain (MHC) isoform expression (MHCSlow, MHC2A, and
MHC2B). A, Size; B, force; C, size, speed of contraction, fatigue resistance. (Adapted from Mantilla CB, Sieck GS: Mechanisms underlying
motor unit plasticity in the respiratory system. J Appl Physiol 2003; 94:1230–1241; and Jones DA: Skeletal muscle physiology. In Roussos C,
ed: The Thorax, 2nd ed. New York: Marcel Dekker; 3–32).

transformed into myocytes, when the muscle becomes injured, COPD is the paradigm of a disease with chronically p0330
to replace damaged myocytes. increased respiratory muscle load. A major adaptation of the
respiratory muscles is fiber type transformation. The myosin
s0250 Chronic Responses heavy chain component of the myosin molecule constitutes
s0260 Increased Load the basis for the classification of muscle fibers as either (type I)
p0320 Plasticity and Adaptation. The respiratory muscles are plas- or (type II) (Figure 8-14, A). Myosin heavy chain exists in
s0270 tic organs that respond to chronic changes of the load they various isoforms, which in increasing order of maximum
are facing and thus of their activity with structural and func- shortening velocity are myosin heavy chain (MHC) I, IIa,
tional changes/adaptations. and IIb, the latter being the faster (Figure 8-14, C ). The

Albert, 978-0-323-04825-5
 10008
148 SECTION II Physiology
diaphragm in healthy humans is composed of approximately
50% type I fatigue-resistant fibers, 25% type IIa, and 25% type
IIb. There are two ways in which muscles can modify their over-
all MHC phenotype: preferential atrophy/hypertrophy of fibers
containing a specific MHC isoform and actual transformation
from one fiber type to another. In COPD, there is a transforma-
tion of type II to type I fibers, resulting in a great predominance
of type I fatigue-resistant fibers. This increases the resistance
of the diaphragm to fatigue development, but at the same time
compromises the force-generating capacity, because type I
fibers can generate less force than type II fibers.
p0340 Adaptation is not only restricted to fiber type transforma-
tion. In an animal model of COPD (emphysematous ham-
sters), the number and the length of sarcomeres decrease,
resulting in a leftward shift of the length-tension curve, so that
the muscle adapts to the shorter operating length induced by
hyperinflation. These alterations may help restore the mechan-
ical advantage of the diaphragm in chronically hyperinflated
states. In humans, this adaptation seems to occur by sarcomere
length shortening.
s0280 Inactivity-Unloading
p0350 Respiratory muscles do not only adapt when they function
against increased load but also when they become inactive, as
happens during denervation or when a mechanical ventilator
undertakes their role as force generator to create the driving
pressure permitting airflow into the lungs. Inactivity and
unloading of the diaphragm caused by mechanical ventilation
is harmful, resulting in decreased diaphragmatic force-
Albert, 978-0-323-04825-5
generating capacity, diaphragmatic atrophy, and diaphragmatic
injury, which are described by the term ventilator-induced dia-
phragmatic dysfunction (VIDD). The mechanisms are not fully
explained, but muscle atrophy, oxidative stress, structural
injury, and muscle fiber remodeling contribute to various
extents in the development of VIDD.
SUGGESTED READINGS
ATS/ERS: Statement on Respiratory Muscle Testing. Am J Respir Crit
Care Med 2002; 166:518–624.
Laghi F, Tobin M: Disorders of the respiratory muscles. Am J Respir Crit
Care Med 2003; 168:10–48.
Mantilla CB, Sieck GS: Invited review: Mechanisms underlying motor unit
plasticity in the respiratory system. J Appl Physiol 2003; 94:1230–1241.
Orozco-Levi M: Structure and function of the respiratory muscles in
patients with COPD: impairment or adaptation? Eur Respir J 2003;
22:Suppl.46, 41s–51s.
Roussos C, Zakynthinos S: Fatigue of the respiratory muscles. Intensive
Care Med 1996; 22:134–55. Au5
Vassilakopoulos T: Ventilator-induced diaphragmatic dysfunction: The
clinical relevance of animal models. Intensive Care Med 2007 (in press). Au3
Vassilakopoulos T, Roussos C: Neuromuscular respiratory failure. In
Slutsky A, Takala R, Torres R, eds. Clinical Critical Care Medicine.
St. Louis: Mosby: 275–282.
Vassilakopoulos T, Roussos C, Zakynthinos S: The immune response to
resistive breathing. Eur Respir J 2004; 24:1033–1043.
Vassilakopoulos T, Zakynthinos S, Roussos C: Respiratory muscles and
weaning failure. Eur Respir 1996; J9:2383–2400.
Vassilakopoulos T, Zakynthinos S, Roussos C: Muscle function: Basic con-
cepts. In Marini JJ, Slutsky A, eds. Physiologic Basis of Ventilator
Support. New York: Marcel Dekker; 1998: 103–152.