REVIEW ARTICLE Sports Med 2002; 32 (9): 567-581

0112-1642/02/0009-0567/$25.00/0

© Adis International Limited. All rights reserved.

Respiratory Muscle Training in

Healthy Individuals
Physiological Rationale and Implications for
Exercise Performance
A. William Sheel
School of Human Kinetics, University of British Columbia, Vancouver, British Columbia, Canada

Contents
Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 567
1. Respiratory Muscles (RM) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 568
1.1 Muscles of Inspiration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 569
1.2 Muscles of Expiration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 569
1.3 Breathing Pattern during Exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 569
1.4 Bioenergetics of RM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 570
1.5 Fatigue of the Diaphragm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 570
1.6 Consequences of High Levels of RM Work and Diaphragm Fatigue . . . . . . . . . . . . . . . 571
2. Effects of Whole-Body Exercise Training . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 571
2.1 Adaptations to Locomotor Muscles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 571
2.2 Adaptations to RM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 572
3. Specific RM Training . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 573
3.1 Effects on RM Endurance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 573
3.2 Effects on RM Strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 573
3.3 Effects on Exercise Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 575
3.4 Effects on Physiological Markers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 577
4. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 579

Abstract The respiratory system has traditionally been viewed to be capable of meeting
the substantial demands for ventilation and gas exchange and the cardiopulmo-
nary interactions imposed by short-term maximum exercise or long-term endur-
ance exercise. Recent studies suggest that specific respiratory muscle (RM)
training can improve the endurance and strength of the respiratory muscles in
healthy humans. The effects of RM training on exercise performance remains
controversial. When whole-body exercise performance is evaluated using sub-
maximal fixed work-rate tests, significant improvements are seen and smaller,
but significant improvements have also been reported in placebo-trained individ-
uals. When performance is measured using time-trial type performance measures
versus fixed workload tests, performance is increased to a much lesser extent with
RM training. It appears that RM training influences relevant measures of physical
performance to a limited extent at most. Interpretation of the collective literature
is difficult because most studies have utilised relatively small sample sizes and
very few studies have used appropriate control or placebo groups. Mechanisms
568 Sheel

to explain the purported improvements in exercise performance remain largely

unknown. However, possible candidates include improved ratings of breathing
perception, delay of respiratory muscle fatigue, ventilatory efficiency, or blood-
flow competition between respiratory and locomotor muscles. This review sum-
marises the current literature on the physiology of RM training in healthy
individuals and critically evaluates the possible implications for exercise perfor-
mance.

Muscular exercise places unique and multifac-
eted demands on the pulmonary system. The healthy
pulmonary system, as the body’s first and last lines
of defence for oxygen (O2) and carbon dioxide (CO2)
transport, faces three major challenges during ex-
ercise. First, increased muscle metabolism at max-
imal exercise intensities causes mixed venous O2
content to fall below <20% of resting values and
raises mixed venous PCO2 to >80mm Hg. Thus, the
demand for alveolar to arterial gas exchange is sub-
stantial, while the rising cardiac output means that
there will be considerably less time in the pulmo-
nary capillaries to accomplish this exchange. Sec-
ond, the precise regulation of alveolar oxygen par-
tial pressure (PAO2) and alveolar carbon dioxide
partial pressure (PACO2) means that alveolar ven-
.
tilation (Va) must increase more than 20-fold over
resting levels at maximal exercise. The ability of
the healthy respiratory musculature for force devel-
opment is adequate to meet this overwhelming
task, but it is equally important that the physiologic
cost of providing this ventilation not be excessive.
To this end, the medullary neural networks and
sensory reflex mechanisms that control breathing
must also have as priorities, the control of intra- and
extra-thoracic airway calibre, the synchronisation
of neural motor outputs to the upper airway and re-
spiratory muscles (RM), and the precise regulation
of breathing pattern and breath duty cycle. Finally,
since gas transport is a product of blood O2 (and
CO2) content and blood flow, it is imperative that
cardiovascular and pulmonary system responses to
exercise be precisely matched to the increase in
metabolic requirements, and that the substantial
gas transport needs of both respiratory and loco-
motor muscles be considered. Given the impor-
tance of adequate gas-exchange during exercise,
the muscles or respiration, which permit such ex-
quisite matching of the pulmonary response to the
metabolic demands, become equally important.
With exercise training, skeletal muscles un-
dergo a host of adaptations including structural,
enzymatic, neural and functional (strength and en-
durance) changes. Like all skeletal muscles, the
RM respond adaptively to an overload stimuli.
Specific RM training has been reported to result in
significant improvements in peak inspiratory pres-
sures, sustained ventilatory capacity tasks and ex-
ercise performance. The purpose of this review is
to critically examine the current literature pertain-
ing to the physiology of the RM during exercise,
specific RM training and the possible implications
of RM training for exercise performance. Studies
of RM training in patient populations such as
chronic obstructive pulmonary disease, cystic fi-
brosis, and heart failure are beyond the scope of
this manuscript and this review is limited to those
studies concerned with the healthy individual. It is
important to note that within the context of RM train-
ing the term ‘exercise’ is used to refer to whole-body
dynamic exercise (i.e. running, cycling, swimming,
and rowing etc.).
1. Respiratory Muscles (RM)
To meaningfully discuss specific RM training it
is necessary to review the basic anatomy and phys-
iology of the RM. This section briefly summarises
those muscles involved with respiration, the breath-
ing patterns that their contractions produce, the
bioenergetic cost of their contraction, and exer-
cise-induced diaphragm fatigue.

© Adis International Limited. All rights reserved. Sports Med 2002; 32 (9)
Respiratory Muscle Training
1.1 Muscles of Inspiration
Inspiration is an active process that occurs when
the inspiratory muscles contract. When these mus-
cles contract there is an expansion of the chest cav-
ity (increased volume) and pleural pressure becomes
subatmospheric. The pressure change causes alve-
olar pressure to also become subatmospheric, which
induces airflow into the lung from the atmosphere.
The primary muscle of inspiration is the diaphragm.
The diaphragm is a large dome-shaped sheet of mus-
cle that separates the abdominal and thoracic cav-
ities. When the diaphragm contracts, it pulls the cen-
tre of the muscle downwards. Additional muscles
of inspiration include the external intercostals,
which connect adjacent ribs and slope downward
and forward. When they contract, the ribs rotate
upward toward the horizontal to increase lung vol-
ume. The accessory muscles also include the ster-
nocleidomastoid and the scalene muscles, which
are attached to the first two ribs and the sternum.
When these muscles contract, they raise the rib cage,
thereby assisting inspiration. During normal quiet
breathing, approximately 50% of the active inspi-
ratory volume change is caused by diaphragm con-
traction; the external intercostals and accessory
muscles provide the rest. Knowledge of RM re-
cruitment during exercise in humans depends on
indirect techniques such as trans-diaphragmatic
pressure (Pdi) and surface electromyogram (EMG)
recordings. Measures of Pdi and EMGs show that
the diaphragm is recruited roughly in proportion to
increasing exercise hyperpnoea that occurs with
increasing work rate.[1] However, others have shown
that during exhaustive endurance exercise Pdi tends
.
to plateau at a time when minute ventilation (VE)
continues to increase markedly, suggesting that the
diaphragm is contributing less and less to the total
pressure generated by the inspiratory muscles as
exercise proceeds (i.e. accessory muscles contrib-
ute more and more).[2] Regardless of this apparent
discrepancy, the important point is that the dia-
phragm is recruited significantly more during ex-
ercise than during rest and that it, in turn, must
© Adis International Limited. All rights reserved.
569
receive a higher proportion of cardiac output dur-
ing exercise.
1.2 Muscles of Expiration
During quiet breathing at rest, expiration is
achieved passively via elastic recoil of the lung and
chest wall. During exercise the muscles of expira-
tion are recruited to forcefully exhale air and return
lung volume to normal resting values. The most
important muscles of expiration are the abdominal
muscles – rectus abdominis, internal and external
obliques, and transversus abdominis. It is experi-
mentally difficult to ascertain the exact recruit-
ment pattern of these muscles during dynamic ex-
ercise in humans. However, the general consensus
is that as exercise intensity increases, they are re-
cruited in a proportional fashion.[3]
1.3 Breathing Pattern during Exercise
Breathing during exercise is a highly coordi-
nated effort of the RM, which produces efficient
gas-exchange. The pattern of RM contractions dur-
ing exercise produces changes in tidal volume (VT),
end-inspiratory and end-expiratory lung volumes,
inspiratory and expiratory flow rates and respira-
.
tory timing. The increase in VE during exercise is
caused by increases in both VT and breathing fre-
quency (ƒb). In normal young individuals engaged
in low-intensity exercise, both increases in VT and
.
ƒb contribute to the rise in VE. VT usually plateaus
at 50 to 60% of vital capacity. During higher inten-
.
sity exercise, additional increases in VE are caused
by increases in ƒb while VT changes little. VT typ-
ically increases 3- to 5-fold from rest to maximal
exercise and ƒb increases 1- to 3-fold, although
these values are often higher for highly trained ath-
letes. The rise in VT is caused by an increase in
end inspiratory lung volume and decrease in end
expiratory lung volume. Increases in the ƒb are
caused by a fall in both inspiratory time (TI) and
expiratory time (TE), although there is a greater
decrease in TE than TI.
.
During rest, VE ranges from 6 to 8 L/min in
healthy individuals and reaches values of 120 to
Sports Med 2002; 32 (9)
570
130 L/min in untrained individuals during strenu-
ous aerobic exercise. In highly trained endurance
.
athletes VE can exceed 200 L/min. The type of
exercise is an important determinant of the ƒb, VT
.
and VE responses; at equivalent rates of oxygen
.
consumption (VO2), ƒ b will be higher and VT
lower for treadmill running than for cycle exer-
cise.[4] The influence or linking of limb movement
and breathing pattern (i.e. respiration-locomotion
entrainment) has been well described but is beyond
the scope of this review. The reader is directed to
excellent summaries on this topic.[3,5]
1.4 Bioenergetics of RM
As mentioned, the pattern of RM contraction
during exercise is significantly different than that
at rest. At rest, breathing is primarily an inspiratory
activity with little or no expiratory muscle activity.
With exercise, both inspiration and expiration are
active processes that require significant amounts
of metabolic work. Elastic work is expended for
lung expansion during inspiration and to overcome
elastic work done on the chest during expiration.
Inspiratory flow becomes more turbulent and expir-
atory work increases disproportionally with increas-
ing work rates. At rest, the O2 cost of breathing is
approximately 2% of total body O2 consumption.
During moderate exercise, the O2 cost of hyperpnoea
.
increases to 3 to 5% of total VO2.[6] The metabolic
and circulatory costs of hyperventilation during
heavy exercise are substantial, amounting to 10%
. .
of maximal VO2 (VO2max) in the untrained person
.
and up to 15 to 16% of VO2max and of maximal
cardiac output in highly fit humans.[6,7] Interest-
ingly, the reader should note that in these studies,
those individuals with the highest cost of breathing
.
(% VO2max) at maximal exercise had the highest
degree of expiratory flow limitation.
Blood flow to the RM during exercise in hu-
mans has not been directly quantified. However,
animal preparations show substantial increases in
blood flow to the diaphragm and inspiratory and
expiratory muscles.[8-12] Microsphere studies in an
equine model showed large increases in blood flow
to both inspiratory and expiratory muscles which
© Adis International Limited. All rights reserved.
Sheel
totalled about 16% of total cardiac output at max-
imal exercise.[8] The highest values for diaphragm
blood flow were equal to, or slightly higher than,
blood flow per 100g to the primary locomotor mus-
cles (gluteus medius and biceps femoris). Collec-
tively, these animal studies clearly demonstrated
that the diaphragm and other RM receive a substan-
tial percentage of cardiac output and achieve a high
O2 extraction rate (80% at maximal exercise). The
implication of this is that the RM certainly receive
their ‘share’ of cardiac output during exercise to per-
form respiratory work. Importantly, how well the
RM ‘compete’ with locomotor muscles for blood
flow is addressed in section 1.6.
1.5 Fatigue of the Diaphragm
The generalisation that the healthy lung and
chest wall are usually ‘overbuilt’, with respect to
the demands of exercise, is true for the most part.
One exception to this near-perfection, which can
occur during short-term and long-term exercise, is
exercise-induced diaphragm fatigue. The diaphragm
is a highly oxidative, densely capillarised muscle
sheet, which has traditionally been thought to be
highly resistant to fatigue. Indeed, during exercise
.
to exhaustion at intensities ≤80% of VO2max the
diaphragm does not fatigue. However, at higher in-
tensities of sustained exercise, the diaphragm con-
sistently shows fatigue at end-exercise, as demon-
strated using the objective technique called bilateral
phrenic nerve stimulation.[2] This fatigue occurred
even though the force output of the diaphragm dur-
ing exercise was well below the fatigue threshold
of this muscle determined during voluntary hyper-
pnoea at rest.[13] This ‘fatigue-sensitising’ effect of
prolonged heavy exercise may be caused by the
redistribution of cardiac output during exercise to
working limb muscles, thereby depriving the dia-
phragm of adequate blood flow in the face of very
high, sustained metabolic requirements. It is pos-
sible that circulating metabolic by-products of ex-
ercise may alter the contractile properties of the
diaphragm and make it susceptible to fatigue. To
date, it is unknown if specific RM training attenu-
ates exercise-induced diaphragmatic fatigue.
Sports Med 2002; 32 (9)
Respiratory Muscle Training
1.6 Consequences of High Levels of RM
Work and Diaphragm Fatigue
An important consequence of high levels of RM
work and diaphragm fatigue is a vasoconstriction
and reduction in blood flow to working locomotor
muscles, accompanied by changes in noradrena-
line spillover.[14] These effects were demonstrated
by mechanically loading, with added resistance, or
unloading, with a pressure-assist ventilator, the RM
at near-maximum exercise. Harms et al.[14] showed
changes in leg blood flow indicating a competitive
relationship between locomotor and RM for a lim-
ited cardiac output. When this study was repeated
at submaximal exercise, changes in limb vascular
resistance did not occur, even though changes in
. The regular application of a specific exercise
RM work were still sufficient to alter VO2 and car-
diac output.[15] With regard to this relationship, it
is interesting to note that exercise-induced dia-
phragm fatigue at the end of exhaustive exercise
does not occur until relative exercise intensity ex-
. cise.[23] In this section, the effects of whole-body
ceeds 80% VO2max.[2] Recent evidence suggests
that when the diaphragm is subjected to fatiguing
levels of work, there is a marked increase in sym-
pathetic activation and a decrease in resting leg
blood flow,[16,17] and that this response is not pres-
ent during non-fatiguing levels of respiratory work
(see Dempsey et al.[18] for a review of respiratory
influences on sympathetic vasomotor outflow in
humans).
What are the effects of high levels of RM work
on exercise performance? Recently, Harms and co-
workers[19] elegantly addressed this question by
subjecting highly trained cyclists to 9 to 12 ran-
domly assigned exercise performance trials, con-
sisting of cycling to exhaustion, beginning at 90%
.
VO2max, under control conditions and during RM
resistive loading and RM unloading. Unloading in-
creased endurance time in 76% of performance tri-
als by a mean of 14 ± 5% and RM loading reduced
performance time in 83% of trials by 15 ± 3% (p <
.
0.05). The rate of rise of VO2 over exercise time
was statistically reduced with unloading and in-
creased with loading, most likely in response to the
accompanying changes in the work of breathing.
© Adis International Limited. All rights reserved.
571
Similarly, the rate of rise of the individuals’ per-
ception of both respiratory and limb discomfort were
markedly reduced with unloading and increased
with RM loading. These significant effects of RM
unloading on exercise performance are consistent
with the deleterious effects of fatiguing the RM on
subsequent exercise performance,[20] but differ from
the negative results from unloading previously re-
ported.[21,22] It is important to note that these pre-
vious studies were conducted at lower intensity
work rates in less fit individuals and did not find
.
significant changes in VO2 with RM unloading.
2. Effects of Whole-Body
Exercise Training
overload enhances physiological function to bring
about a training response. Exercise training induces
a variety of highly specific adaptations that enable
the body to function more efficiently during exer-
endurance training on locomotor and RM are re-
viewed.
2.1 Adaptations to Locomotor Muscles
The collective changes observed in skeletal
muscle in response to exercise training represent
an adaptive ‘strategy’ to enhance the functional
capacity of the organism. All living organisms pos-
ses the inherent capacity to alter the structural and
functional properties of their organ systems in ac-
cordance with the environmental conditions im-
posed on a particular system. The ability of muscle
cells to alter the amount and composition of their
subcellular components (such as contractile ma-
chinery, enzyme levels, etc.) in response to exer-
cise is an area of study rich in history. Over the last
75 to 100 years the adaptations that occur in limb
locomotor muscle in response to exercise training
have been well documented. For a detailed sum-
mary of muscle plasticity, the reader is directed to
the excellent writing of Booth and Baldwin.[24] Ta-
ble I briefly summarises known adaptations in lo-
comotor muscle to whole-body endurance train-
Sports Med 2002; 32 (9)
572
ing. These changes have clearly been shown to im-
prove the functional capacity of skeletal muscle.
2.2 Adaptations to RM
While the adaptations to human locomotor mus-
cles have been studied extensively and the struc-
tural, biochemical and functional consequences
have been well established (see table I), studies ex-
amining the effects of whole-body training on the
RM are relatively few. Like all vertebrate skeletal
muscle, it seems intuitive that the RM can respond
adaptively to strength and endurance training (see
section 3). However, the question remains: how do
the muscles of respiration adapt to whole-body ex-
ercise training? To answer this question, the animal
literature must be reviewed.
The majority of the animal literature has focussed
on the rat diaphragm. It appears that endurance
training elicits small (20 to 30%) but significant
increases in rat diaphragm mitochondrial enzyme
activity and antioxidant enzyme activity, resting
glycogen levels,[25-34] which is consistent with the
changes described for limb muscle in table I. Re-
cently, Vrabas et al.[35] had rats complete treadmill
exercise training for 5 days/week, 60 min/day at
. were observed for citrate synthase or phosphofruc-
70% of VO2max, for 10 weeks. Trained diaphragms
had 10% higher citrate synthase and 12.1% higher
superoxide dismutase activities compared with the
control diaphragms. This was a consistent finding,
Table I. Overview of metabolic and structural changes in limb
locomotor muscles associated with whole-body endurance training
Glycogen
Number of mitochondria
Mitochondrial volume
Adenosine triphosphate
Phosphocreatine
Creatine
Glycolytic enzymes
Aerobic enzymes
Maximal lactate
Type I fibres
Type II fibres
Oxygen extraction
Capillarisation
↑ = increase; ↓ = decrease.
© Adis International Limited. All rights reserved.
Sheel
namely that endurance training of sufficient inten-
sity and duration can increase the oxidative capac-
ity of the rat diaphragm. An up-regulation of anti-
oxidant enzyme activity (i.e. superoxide dismutase)
in the diaphragm is important because it points to
an increased ability of the diaphragm to eliminate
free radicals during exercise. Consistent with the
limb locomotor data, these authors[35] also observed
a significant decrease in type IIb and an increase in
type I myosin heavy chains in trained diaphragms. It
is presumed that this results in a functional im-
provement in the endurance properties of the mus-
cle.
It appears that endurance training results in
small but significant biochemical alterations of rat
RM; however, do these changes improve RM per-
formance? The effect of exercise training on dia-
phragmatic contractile properties has not been stud-
ied extensively. In one study, Metzger and Fitts[25]
studied the effects of endurance treadmill running
on the contractile, biochemical and fatigue proper-
ties of the rat diaphragm. The exercise-training regi-
men did not alter maximal force production, the
maximal speed of shortening, or diaphragmatic en-
durance. It is important to note that no differences
tokinase in the diaphragm post-training implying
that the training stimulus may not have been suffi-
cient to elicit biochemical and/or structural changes.
More recently, enzyme and fibre type changes
have indeed been observed in the diaphragm of an
exercising rat.[35] Using an in vitro fatigue proto-
col, control diaphragms produced 12% less force
↑
compared with trained diaphragms, suggesting that
↑
↑
exercise training produced an increased resistance
↑ to fatigue. Despite these significant changes, the
↑ maximal shortening velocity in the diaphragm was
↑ not different between trained and untrained states.
↑ These findings suggest that endurance training re-
↑
duces the rate of diaphragm fatigue in vitro, but has
↑
no effect on maximal shortening velocity or spe-
↑
↓ cific force.
↑ Direct studies examining the effects of RM
↑ training on critical bioenergetic enzymes and fa-
tigue resistance in humans are lacking. Studies of
Sports Med 2002; 32 (9)
Respiratory Muscle Training
human RM adaptation to exercise training have
shown improvements in ventilatory performance
as evidenced by increases in maximal sustainable
ventilatory capacity (MSVC) and maximal volun-
tary ventilation (MVV).[36-38] Comparisons of ath-
letes and nonathletes have also shown differences
in ventilatory performance.[39,40]
Martin and Stager[39] showed that athletes could
sustain 80% of their 12-second MVV for 11 min-
utes; whereas age, gender, vital-capacity and body-
size matched nonathletes could sustain this venti-
latory load for only 3 minutes (p < 0.05). Despite
these results, it is important to note that not all
studies have reported similar findings. Following
6 weeks of endurance training in the hypoxia of high
altitude (3600m), maximal mouth pressure [maxi-
mal inspiratory pressure (PImax)], MSVC and 12-
second MVV were unchanged.[41] It was hypo-
thesised that exercise training coupled with the
well known adaptations that occur with high alti-
tude may have had an additive effect on RM per-
. formed 63 to 83% of initial MVV for 30 min/day,
formance. In this study, VO2max significantly in-
creased by +19% and yet there were no detectable
differences in RM performance. Clearly, further
study is needed to objectively evaluate the effects
of whole-body endurance training on human ven-
tilatory performance.
3. Specific RM Training
Thus far, this review has briefly summarised the
actions of the RM during exercise, the bioenerget-
ics of the RM, and the effects of endurance-exercise
training on the RM. It is clear that during whole-
body endurance type exercise, the RM perform sig-
nificant amounts of metabolic work. Furthermore,
there is evidence to suggest that the RM, as with
limb skeletal muscles, respond adaptively to exer-
cise training. In this section, the effects of specific
RM training on RM strength, RM endurance, ef-
fects on exercise performance, and effects on phys-
iological markers are discussed. Most RM-training
studies have employed two modes of training: (i)
voluntary isocapnic hyperpnoea to improve RM
endurance; or (ii) inspiratory resistive loading to
improve RM strength.
© Adis International Limited. All rights reserved.
573
3.1 Effects on RM Endurance
During isocapnic hyperpnoea training individ-
uals maintain a given level of ventilation for up to
30 minutes. This regimen is carried out typically
for 3 to 5 times per week for 4 to 5 weeks (see table
II). Using this type of training, several investiga-
tors have shown improvements in RM endurance.
Table II summarises the training protocols utilised
and the subsequent change in RM endurance. In a
recent well controlled study, individuals performed
30 minutes of RM work equal to 50 to 60% of
15-second MVV using a ƒb of 50 to 60 breaths/min
for 5 weeks.[42] Weekly in-laboratory tests were
conducted to ensure that end-tidal CO2 values were
within ±2mm Hg of eupnoeic values. This protocol
failed to improve 15-second MVV (pre = 213 vs
post = 214 L/min, p > 0.05) or endurance breathing
performance at 90% of 15-second MVV (pre =
0.66 min vs post 1.15 min, p > 0.05). Conversely,
in an equally well controlled study, participants per-
5 days/week for 4 weeks.[43] In this study, a signif-
icant increase in breathing endurance (VT = 60%
of vital capacity, ƒb = 40 to 50 breaths/min) was
observed (pre = 4.6 vs post = 29.1 min, p < 0.05).
It is not clear why the two studies reported differ-
ent results. Despite the discrepancy between these
two studies, it should be pointed out that most stud-
ies that utilise a percentage of MVV to exhaustion
have indeed demonstrated improvements in breath-
ing endurance (see table II).
3.2 Effects on RM Strength
Inspiratory resistive loading training usually
consists of loads [approximately 15 to 50% maxi-
mal inspiratory pressure (PImax)] applied to the in-
spiratory circuit 3 to 5 times per week for a dura-
tion of 5 to 20 minutes while at rest. This is typically
accomplished by using a custom-built apparatus or
a commercially available device. As correctly point-
ed out by Dempsey and colleagues,[3] few studies
have controlled for lung volume or breathing strat-
egy during pre- and post-training evaluation mak-
ing interpretation of the training effects difficult.
Sports Med 2002; 32 (9)
574 Sheel

Table II. Summary of reported changes in exercise performance after specific respiratory muscle training
Studya Participants RM training Change in resting RM Exercise load Change in
function (%) performance (%)
Leith and Untrained: E1: static inspiration and E1: no change breathing Exercise not NA
Bradley[44] n = 4 E1, expiration at 20% intervals endurance performed
‘strength’ over vital capacity,
30 min/d, 5 d/wk, 5wk
n = 4 E2, E2: ventilate to ‘exhaustion’, E2: +19% breathing
‘endurance’ 3-5 × 5 d/wk, 5wk endurance (S)
n=4C C: no RM training C: no change breathing
endurance
.
Morgan et al.[45] Trained: E: 85% MVV increasing E: +14% MVV (S), 95% VO2max cycling E: –6% (NS)
n=4E duration, 5 d/wk, 3wk +1575% endurance to exhaustion
breathing time (S)
n=5C C: no RM training C: 0% MVV (NS), C: –8% (NS)
0% endurance breathing
time (NS)
Hanel and Untrained: E: 50% PImax, 10 min, E: +10% PImax (S) Cooper’s 12 min run E: +8% (NS)
Secher[46] n = 10 E 2/d, 27.5d test
n = 10 C C: 0% PImax, 10 min, 2/d, C: +4% PImax (NS) C: +6% (NS)
27.5d
Fairbarn et al.[47] Trained: E: ≥MSVC 3 sessions of 8 E: +12% breathing 90% Wmax cycling to E: +25% (NS)
n=5E min, 3-4 d/wk, 4wk endurance (S) exhaustion
n=5C C: no RM training C: –4% breathing C: +4% (NS)
endurance (NS)
.
Boutellier and Untrained: 58-63% MVV 20-30 min, +268% breathing 64% VO2peak cycling +50% (S)
Piwko[48] n=4 5 d/wk, 4wk endurance (S) to exhaustion
.
Boutellier et Trained: 55-68% MVV 30 min, +555% breathing 77% VO2peak cycling +38% (S)
al.[49] n=8 5 d/wk, 4wk endurance (S) to exhaustion
Suzuki et al.[50] Untrained: E: 30% PImax, 15 min, 2/d, E: PImax +30% (S), Incremental treadmill Performance not
n=6E 4wk MVV +12% (S) to exhaustion or for measured. No
.
n=6C C: no RM training C: PImax 0% (NS), a maximum of 10 min change in VE or
MVV +4.8% (NS) ‘respiratory
sensation’ at any
stage of exercise
Spengler et Trained: 60-85% MVV, 30 min, +532% breathing 85% Wmax cycling +28% (S)
al.[43] n = 20 5 d/wk, 4wk endurance (S)
Inbar et al.[51] Trained: E: 30-80%, PImax, 30 min, E: +25% PImax (S) Incremental running Performance not
n = 10 E 6 d/wk, 10wks and +27% breathing test to exhaustion measured. No
.
endurance (S) change in VO2max,
.
n = 10 C C: same training, C: +1% PImax (NS) VEmax
no resistance and +1% breathing
endurance (NS)
Stuessi et al.[52] Untrained: E: 40 (15wk) sessions of E: +632% breathing 70% Wmax cycling to E: +24% (S)
n = 13 E 30 min 65-70% MVV endurance (S) exhaustion
n = 15 C C: no RM training C: –2% breathing C: –4% (S)
endurance (NS)
Sonetti et al.[42] Trained: E: 5wk, 5 d/wk, 30 min/d, E: +8% PImax (S), Cycling: 8km time E: +1.8% 8km time
n=9E 50-60% MVV and 4-5 min no change 15 sec MVV, trial, and 80-85% trial (S), +26% fixed
50% PImax no change breathing Wmax to exhaustion work-rate (S)
endurance
n=8C C: 5wk, 5 d/wk, 30 min/d C: +3.7% PImax (NS) C: –0.3% 8km time
placebo, ‘hypoxic trainer’ no change 15 sec MVV, trial (NS), +16%
no change breathing fixed work-rate (S)
endurance
Continued over page

© Adis International Limited. All rights reserved. Sports Med 2002; 32 (9)
Respiratory Muscle Training
Table II. Contd
Studya Participants RM training
Volianitis et Trained: E: 50% PImax 30 breaths,
al.[53] n=7E 2 × day, 7 d/wk, 11wk
n=7C C: 15% PImax, 60 breaths,
1 × day, 11wk
Markov et al.[54] Untrained: E: 60% MVV, 30 min,
n = 13 E 40 sessions over 15wk
n = 9 C1 C1: 30 min cycling/running,
40 sessions over 15wk
n = 15 C2 C2: no RM training
a Studies listed in chronological order.
C = control; E = experimental; MSVC = maximal sustainable ventilatory capacity; MVV = maximal voluntary ventilation; NA = not applicable;
NS = not statistically significant (p > 0.05); PImax = maximal inspiratory pressure; RM = respiratory muscle; S = statistically significant (p <
. .
0.05); VE = minute ventilation; VEmax = maximal minute ventilation; VO2max = maximal oxygen consumption; VO2peak = peak oxygen
consumption; Wmax = maximal power output.
Nonetheless, studies using resistive loading have
reported increases in PImax in the range of 8 to 45%
(see table II). It is interesting to note that those
studies that have utilised the highest percentage
PImax and longest duration of training also report
the highest change with training in PImax
(+45%);[53] whereas studies which utilise a lower
percentage PImax or training for a shorter duration
show less of an improvement in PImax.[42] It is
tempting to speculate that, as with other skeletal
muscles, the RM respond in a dose-response fash-
ion to a given training stimulus. However, the lim-
ited available data are too few to make such a con-
clusion. In summary, it is reasonable to conclude
that inspiratory muscle loading can improve the
strength of the RM as reflected by PImax (table II),
and comparable changes to Pdi during inspiratory
efforts have also been observed.[50]
3.3 Effects on Exercise Performance
A critical review of the literature revealed that
the effects of RM training on exercise performance
are controversial. The reason for much of the con-
troversy surrounds the fact that studies have used
different RM-training regimens, different labora-
tory tests of exercise ‘performance’, and differ-
ences in the training status of individuals. Further-
more, most studies have utilised relatively small
© Adis International Limited. All rights reserved.
575
Change in resting RM Exercise load Change in
function (%) performance (%)
E: +45% PImax (S) 6 min ‘all out’ rowing +4% (S)
C: +5% PImax (NS) +2% (NS)
Breathing endurance. 70% Wmax cycling to E: +24% (S)
E: +770% (S) exhaustion
C1: +45% (S) C1: +41% (S)
C2: –25% (NS) C2: –0.05% (NS)
. .
sample sizes (typically <8 to 10 and as low as 4),
although more recent studies have employed larger
samples.
Of the eight studies listed in table II that mea-
sured exercise performance, six showed statisti-
cally significant improvements. In these studies,
performance was determined using a fixed work
.
rate [i.e. submaximal percentage of VO2max or
maximal power output (Wmax)]. The reported per-
formance improvements ranged from 24 to 50%.
While the improvements are seemingly large, it is
important to note that in the only study to use a true
placebo group (‘hypoxic’ trainer), a 16% improve-
ment was seen in the placebo group compared with
a 26% improvement in the RM-trained individuals.
In those studies that demonstrated an improvement
in performance, the fixed work-rate ranged from
.
64 to 80% of VO2max or Wmax. In the two remaining
studies which did not show improvements, the
work rate was near maximal (90% Wmax and 95%
.
VO2max). This remains a curious observation, as at
.
intensities above 85% VO2max diaphragm fatigue
has been observed (see section 1.5) and there may
be an inter-relationship between RM training, dia-
phragm fatigue, recruitment of accessory RM, and
exercise intensity. Potentially, RM training may
reduce diaphragm fatigue for a given workload;
however, to date, no data are available to address
Sports Med 2002; 32 (9)
576
this observation. From the available data it is dif-
ficult to distinguish between the exercise perfor-
mance effects of hyperpnoea training versus inspi-
ratory-load training.
From the above summary, it appears from the
limited data that RM training can improve time to
volitional exhaustion at fixed-rate submaximal
workloads (although this is not a completely con-
sistent finding) and not at near-maximal exercise
intensities. Indeed, the most commonly used labo-
ratory protocols have employed exercise time to ex-
haustion as a measure of performance.[55-59] Con-
troversy surrounds the value of fixed work-rate
laboratory tests for two reasons. Firstly, these tests
are not true measures of athletic performance, as
they do not mimic competitive situations. Athletes
often change velocity for strategic or environmen-
tal reasons (i.e. hill, wind, etc.). Secondly, those
studies which evaluated performance using fixed
work-rate type performance tests are reported to
have poor reliability.[59] Most data suggest that lab-
oratory tests that require individuals to complete a
fixed amount of work or to cover a set distance in
the shortest possible time are much more reliable
than constant-load tests.[57,60-64] Coefficient of
variations (CV) range from 1.0 to 3.1% for cyclists
performing time trials,[61,62] and 2.7 to 3.4% when
cyclists perform as much work as possible in 1
hour.[57,60] This contrasts sharply with fixed work-
rate tests with CV values of 17 to 40%,[57] and
25%.[55]
Taking the comparison of fixed work-rate tests
and time-trial type tests into account, an examina-
tion of the effects of RM training on exercise per-
formance yielded remarkably different results.
During an 8km laboratory time-trial, cycling per-
formance was increased significantly (+1.8 ± 1.2%)
in well trained athletes who performed 5 weeks of
RM training.[42] In this study, individuals who per-
formed placebo RM training did not show an im-
provement in performance; however, the improve-
ment in the RM-training group was not significantly
different than the placebo group. Figure 1 shows
the individual responses for both RM-training and
placebo participants. Improved performance was
© Adis International Limited. All rights reserved.
Sheel
seen in 8 of 9 RM-training participants and 5 of 8
placebo participants. The lack of a consistent effect
of RM on exercise performance can be interpreted
in two ways: (i) improvements in 8km time trials
were caused by a familiarisation effect; or (ii) the
changes can be explained by the placebo effect.
Further study is required to fully discern the effects
of RM training, familiarisation and placebo effects
on exercise performance. In a different study, fol-
lowing 11 weeks of RM training, female rowers
improved rowing performance (as measured by
distance covered in 6 minutes), by +3.5 ± 1.2%.[53]
The control group in this study performed ‘sham’
RM training (15% PImax compared with 50% PImax
in the training group) and also significantly im-
proved performance (+1.6 ± 1.0%). The +1.9% im-
provement in the RM-training group over and
above the improvement of the sham-training group
suggests that the RM training did indeed have an
effect on the rower’s performance and may indeed
be relevant within the context of competitive ath-
letics.
It is difficult to interpret the differences in ob-
served performance changes between Sonetti et
al.[42] and Volianitis et al.[53] One explanation is
related to the duration of the RM-training period
(5 vs 11 weeks). This may be reflected in the dif-
ferences in changes to PImax (+8 vs +45%). It may
also be that because of the locomotion-breathing
coupling (i.e. entrainment) that is frequently ob-
served during rowing that improvements in PImax
are more important than during cycling. This highly
speculative assessment requires further study, and
it remains controversial whether RM training im-
proves exercise performance.
In summary, when exercise performance is
evaluated using submaximal fixed work-rate tests,
significant improvements are seen in RM-training
participants. Smaller, but significant improve-
ments have also been reported in placebo-trained
individuals. When performance is measured using
time-trial type performance measures, perfor-
mance is increased to a much lesser extent with RM
training (range: +1.8 to 3.5%) than during fixed
work-rate tests (range: +24 to 50%). The concept
Sports Med 2002; 32 (9)
Respiratory Muscle Training
that RM training improves performance is dimin-
ished by the observation that studies with well de-
a RMT group
15
14
13
Mid- or post-treatment time (min)
12
b Placebo group
15
14
13
12
12 13 14
Pre-treatment time (min)
Fig. 1. Effects of 3 weeks (mid-test) and 5 weeks (post-test) of:
(a) respiratory muscle training (RMT); or (b) placebo, on 8km
cycling time-trial finishing times. Points lying below the line of
identity (diagonal line) indicate improvement from initial test.
Participants were highly trained male cyclists. Experimental
participants (n = 9) performed RMT (50 to 60% maximal volun-
tary ventilation for 30 min/d and 4 to 5 min breathing at 50%
maximal inspiratory pressure) 5 d/wk for 5 weeks. Placebo par-
ticipants (n = 8) used a placebo breathing device for 30 min/d,
5 d/wk for 5 weeks. Participants were told the device reduced
the oxygen content of each breath, mimicking the effects of high
altitude. The time-trial performance increased in the respiratory
training group with 8 of 9 individuals increasing; the placebo
group showed a nonsignificant change in 5 of 8 participants (p
= 0.07) [reprinted from Respiration Physiology, 127, Sonetti DA,
Wetter TJ, Pegelow DF, et al.,[42] Effects of respiratory muscle
training versus placebo on endurance-exercise performance,
185-99, Copyright 2001, with permission from Elsevier Sci-
ence].
© Adis International Limited. All rights reserved.
577
signed control groups and sham-training groups
also show variable degrees of performance im-
provements. It appears that RM training influences
relevant measures of physical performance in highly
Line of identity fit individuals to a limited extent at most. As a
Mid-test caveat, this conclusion is based on the available
Post-test
studies which have utilised relatively small sample
sizes and very few studies have used appropriate
control or placebo groups (see table II). For exam-
ple, most studies have utilised: (i) no control group;
(ii) a control group that performs ‘less’ training
than the experimental group (i.e. sham training);
or (iii) the control performs no intervention.
3.4 Effects on Physiological Markers
If we operate on the premise that RM training
can increase fixed work-rate exercise performance,
what physiological mechanisms might be respon-
sible? Usually, increased performance is attributed
to improvements of the cardiorespiratory system
or to improvements in skeletal muscles. From the
available data, it does not appear that RM training
. .
has any systematic effect on maximal VE, VO2, heart
rate, stroke volume, blood gas concentrations or
oxyhaemoglobin saturation during incremental
exercise to exhaustion.[42,43,45-47,51,52,54] However,
some studies reported modest reductions in blood
lactate concentrations of ~2 mmol/L,[43] while oth-
15 ers reported no change as a result of RM train-
ing.[42] A physiological mechanism must reside
elsewhere.
The respiratory system can become fatigued
with exercise.[2,13] Could RM training delay fa-
tigue of the respiratory musculature? This is an
attractive possibility, but objective measures of di-
aphragm fatigue (Pdi, assessed with bilateral phre-
nic nerve stimulation) following exhaustive exer-
cise pre– and post–RM training have not been
performed. One study has attempted to examine
RM fatigue following RM training. Volianitis et
al.[53] measured PImax pre- and post-exercise (6
minutes ‘all-out’ rowing) before and after RM
training. At the initial session, PImax was decreased
11.2 and 11.1% for the RM-training group and the
sham-training group, respectively. Following 4
Sports Med 2002; 32 (9)
578
weeks, these values were 3.1 and 10.7%, respec-
tively, and after 11 weeks were 4.5 and 10.7%, re-
spectively. Between-group differences were statis-
tically significant at both 4 and 11 weeks (p <
0.05). Based on these data, it seems that RM train-
ing can prevent RM fatigue and that the effects
reach a plateau at approximately 4 weeks of train-
ing. However, the use of PImax as a surrogate mea-
sure of RM fatigue has been questioned. Decramer
and Macklem[65] suggested that a problem with the
measurement of PImax is that these measurements
are clearly effort dependent and, consequently,
may be related to poor coordination or submaximal
effort. This may especially hold true following ex-
haustive exercise where individuals may not be
willing, or able, to produce a maximal voluntary
inspiratory effort for reasons unrelated to RM fa-
tigue. While the findings of Volianitis et al.[53] are
certainly provocative, electrical or magnetic stim-
ulation of the phrenic nerves with simultaneous
measures of Pdi pre- and post-exercise are required
to objectively evaluate the effect of RM training on
diaphragm fatigue.
The regulation and distribution of cardiac out-
put during exercise is a highly complex process. It
has been hypothesised that during exercise the RM
‘compete’ with limb locomotor muscles for their
‘share’ of cardiac output.[7,14,19,66] When the work
of breathing (measured by inspiratory and expir-
atory oesophageal pressure) is reduced by 50% us-
ing a proportional-assist ventilator, blood flow to
exercising legs is increased by 5 to 7%[14] and en-
durance performance (fixed work-rate) is increased
by +15% in highly trained cyclists.[19] It is difficult
to explain the dramatic improvements seen in fixed
work-rate exercise performance (range: +25 to
50%) with RM training when reducing the work of
breathing by 50% yields significantly less im-
provements. Sonetti et al.[42] pointed out that it
seems inconceivable that the effects of RM train-
ing alone could surpass those seen with substantial
mechanical unloading – unless RM training has ad-
ditional influences on locomotor muscles, which
are not realised via substantial RM unloading and
the prevention of diaphragm fatigue. It remains un-
© Adis International Limited. All rights reserved.
Sheel
known if RM training can ‘free’ cardiac output
from the RM to the locomotor muscles as these
measures have not been performed. Further study
examining these relationships is warranted.
Dyspnoea is defined as discomfort associated
with breathing, and dyspnoeic sensations are com-
monly cited as factors limiting exercise (for review
see Killian and Campbell[67]). Despite the frequent
occurrence of dyspnoea during exercise, causal
mechanisms remain elusive and dyspnoea remains
poorly understood. In healthy individuals, a signif-
icant conscious perception of the ‘effort’ to breathe
usually does not occur until heavy-intensity exer-
cise is achieved and hyperventilation begins to de-
velop. Recent evidence suggests that dyspnoea in
heavy endurance exercise, may contribute to exer-
cise limitation. Harms et al.[19] determined the im-
pact of altering the work of breathing on exercise
performance and sensations of dyspnoea and limb
ratings of perceived exertion (RPE) in highly trained
cyclists. At isotime points during constant load cy-
.
cling at high-intensity (~90% VO2max), inspiratory
loading resulted in significantly higher leg RPE
and dyspnoea ratings than control, whereas un-
loaded breathing had the opposite effect. Unloaded
breathing also resulted in a significant increase in
time to exhaustion relative to control or loaded
breathing. The change in exercise performance
was significantly correlated with the changes in
limb RPE and the changes in dyspnoea ratings. The
results from this study suggest that exercise perfor-
mance is related to the amount and perception of
respiratory effort. Furthermore, it has been hypo-
thesised that RM training may decrease perception
of respiratory exertion and contribute to the ob-
served increase in constant-load exercise perfor-
mance. A critical examination of those studies that
have sought to determine the possible interplay be-
tween RM training, dyspnoea, and exercise reveal
that supporting data are few and conflicting.[42,49,50]
Boutellier et al.[49] have commented that their
study participants reported that RM training elim-
inated the feeling of breathlessness even during up-
hill running or cycling. Interpretation of this is dif-
ficult as no data were presented and it is unknown
Sports Med 2002; 32 (9)
Respiratory Muscle Training
how this was evaluated. In contrast to this prelim-
inary observation, no significant changes have
been reported in perception of limb or respiratory
effort during incremental, fixed work-rate, or time-
trial tests following RM-training or placebo peri-
ods.[42,50] To date, only one study has shown sig-
nificant improvements in the perception of RM
effort during exercise throughout an 11 week RM-
training regimen.[53] Differences were observed
during submaximal intensities and absent during
higher intensities of rowing. Given the importance
of dyspnoea during exercise, these observations
warrant further study and await to be incorporated
into the context of understanding exercise-related
dyspnoea, RM training, and exercise performance.
Another possibility to explain the effects of RM
training is altered ventilatory efficiency. It is pos-
.
sible that VE is reduced following RM training
during submaximal exercise and could contribute
to the observed improvements in fixed work-rate
. The author received no funding for the prepartion of this
exercise tests. Reduced VE for a given workload
would reduce the metabolic requirements of the
RM and result in a diminished competition for
blood flow requirements between the RM and lo-
comotor muscles. This remains an interesting pos-
sibility that awaits experimental evidence. Fur-
thermore, measures of RM oxygen consumption
. 2. Johnson BD, Babcock MA, Suman OE, et al. Exercise-induced
(VO2resp) have not been performed in conjunction
. diaphragmatic fatigue in healthy humans. J Physiol 1993;
with RM training. It is well known that VO2resp
increases significantly from rest to moderate and
maximal exercise.[6,7]
4. Conclusion
Specific RM training has been shown to im-
prove the endurance and strength of the RM in
healthy humans. The effects of RM training on ex-
ercise performance remain controversial. When
exercise performance is evaluated using submaxi-
mal fixed work-rate tests, significant improvements
are seen and smaller, but significant improvements
have also been reported in placebo-trained partic-
ipants. When performance is measured using time-
trial type performance measures, performance is
increased to a much lesser extent with RM training.
It appears that RM training influences relevant
© Adis International Limited. All rights reserved.
579
measures of physical performance to a limited ex-
tent at most. Interpretation of the collective litera-
ture is difficult because most studies have utilised
relatively small sample sizes and very few studies
have used appropriate control or placebo groups.
Mechanisms to explain the purported improve-
ments in exercise performance remain largely un-
known. However, candidates include improved rat-
ings of breathing perception, delay of RM fatigue,
ventilatory efficiency, or blood-flow competition
between respiratory and locomotor muscles. Fu-
ture well controlled studies with larger sample sizes
are warranted to ascertain whether delaying or at-
tenuating RM fatigue does indeed improve athletic
performance and what physiological mechanisms
might be responsible.
Acknowledgements
article and has no conflicts of interest relavant to the contents
of this review.
References
1. Bye PTP, Esau SA, Walley KR, et al. Ventilatory muscles dur-
ing exercise in air and oxygen in normal men. J Appl Physiol
1984; 56: 464-71
460: 385-405
3. Dempsey JA, Adams L, Ainsworth DM, et al. Airway, lung,
and respiratory function during exercise. In: Rowell LB,
Shepherd JT, editors. Exercise: regulation and integration of
multiple systems. New York (NY): Oxford University Press,
1996: 448-514
4. Henke KG, Sharratt M, Pegelow D, et al. Regulation of end-
expiratory lung volume during exercise. J Appl Physiol 1988;
64 (1): 135-46
5. Dempsey JA, Forster HV, Ainsworth DM. Regulation of hy-
perpnea, hyperventilation, and respiratory muscle recruit-
ment during exercise. In: Dempsey JA, Pack AI, editors.
Regulation of breathing. New York (NY): Marcel Dekker,
1995: 1065-134
6. Aaron EA, Seow KC, Johnson BD, et al. Oxygen cost of exer-
cise hyperpnea: implications for performance. J Appl Physiol
1992; 72: 1818-25
7. Harms CA, Wetter TJ, McClaran SR, et al. Effects of respira-
tory muscle work on cardiac output and its distribution during
maximal exercise. J Appl Physiol 1998; 85: 609-18
8. Manohar M. Blood flow to the respiratory and limb muscles
and to abdominal organs during exertion in ponies. J Physiol
1986; 377: 25-35
Sports Med 2002; 32 (9)
580
9. Musch TI, Friedman DB, Pitetti KH, et al. Regional distribution
of blood flow of dogs during graded dynamic exercise. J Appl
Physiol 1987; 63 (6): 2269-77
10. Manohar M. Costal vs crural diaphragmatic blood flow during
submaximal and near-maximal exercise in ponies. J Appl
Physiol 1988; 65 (4): 1514-9
11. Manohar M, Hassan AS. Diaphragm does not produce ammonia
or lactate during high-intensity short-term exercise. Am J
Physiol 1990; 259 (4 Pt 2): H1185-9
12. Fixler DE, Atkins JM, Mitchell JH, et al. Blood flow to respi-
ratory, cardiac, and limb muscles in dogs during graded exer-
cise. Am J Physiol 1976; 231 (5 Pt 1): 1515-9
13. Babcock MA, Pegelow DF, McLaran SR, et al. Contribution of
diaphragmatic power output to exercise-induced diaphragm
fatigue. J Appl Physiol 1995; 78: 1710-9
14. Harms CA, Babcock MA, McClaran SR, et al. Respiratory mus-
cle work compromises leg blood flow during maximal exer-
cise. J Appl Physiol 1997; 82 (5): 1573-83
15. Wetter TJ, Harms CA, Nelson WB, et al. Influence of respira-
. training on myosin heavy-chain isoforms and enzyme activity
tory muscle work on V O2 and leg blood flow during submaxi-
mal exercise. J Appl Physiol 1999; 87 (2): 643-51
16. Sheel AW, Derchak PA, Morgan BJ, et al. Fatiguing inspiratory
muscle work causes reflex reduction in resting leg blood flow.
J Physiol (Lond) 2001; 537 (1): 277-289
17. St Croix CM, Morgan BJ, Wetter TJ, et al. Fatiguing inspiratory
muscle work causes reflex sympathetic activation in humans.
J Physiol 2000; 529 Pt 2: 493-504
18. Dempsey JA, Sheel AW, St Croix CM, et al. Respiratory influ-
ences on sympathetic vasomotor outflow in humans. Respir
Physiol Neurobiol 2002; 130: 3-20
19. Harms CA, Wetter TJ, St Croix CM, et al. Effects of respiratory
muscle work on exercise performance. J Appl Physiol 2000;
89 (1): 131-8
20. Sliwinski P, Yan S, Gauthier AP, et al. Influence of global in-
spiratory muscle fatigue on breathing during exercise. J Appl
Physiol 1996; 80: 1270-8
21. Krishnan B, Zintel T, McParland C, et al. Lack of importance
of respiratory muscle load in ventilatory regulation during
heavy exercise in humans. J Appl Physiol 1996; 490: 537-50
22. Gallagher CG, Younes M. Effect of pressure assist on ventila-
tion and respiratory mechanics in heavy exercise. J Appl
Physiol 1989; 66: 1824-37
23. McArdle WD, Katch FI, Katch VL. Exercise physiology: en-
ergy, nutrition, and human performance. 5th ed. Baltimore
(MD): Lippincott, Williams and Wilkins, 2001
24. Booth FW, Baldwin BJ. Muscle plasticity: energy demand and
supply processes. In: Rowell LB, Shepherd JT, editors. Exer-
cise: regulation and integration of multiple systems. New
York (NY): Oxford University Press, 1996: 1075-123
25. Metzger JM, Fitts RH. Contractile and biochemical properties
of diaphragm: effects of exercise training and fatigue. J Appl
Physiol 1986; 60 (5): 1752-8
26. Gosselin LE, Betlach M, Vailas AC, et al. Training-induced
alterations in young and senescent rat diaphragm muscle. J
Appl Physiol 1992; 72 (4): 1506-11
27. Grinton S, Powers SK, Lawler J, et al. Endurance training-induced
increases in expiratory muscle oxidative capacity. Med Sci
Sports Exerc 1992; 24 (5): 551-5
© Adis International Limited. All rights reserved.
Sheel
28. Ianuzzo CD, Noble EG, Hamilton N, et al. Effects of strepto-
zotocin diabetes, insulin treatment, and training on the dia-
phragm. J Appl Physiol 1982; 52 (6): 1471-5
29. Powers SK, Criswell D, Lawler J, et al. Regional training-induced
alterations in diaphragmatic oxidative and antioxidant en-
zymes. Respir Physiol 1994; 95 (2): 227-37
30. Powers SK, Criswell D, Lieu FK, et al. Exercise-induced cellu-
lar alterations in the diaphragm. Am J Physiol 1992; 263 (5
Pt 2): R1093-8
31. Powers SK, Lawler J, Criswell D, et al. Aging and respiratory
muscle metabolic plasticity: effects of endurance training. J
Appl Physiol 1992; 72 (3): 1068-73
32. Powers SK, Grinton S, Lawler J, et al. High-intensity exercise
training-induced metabolic alterations in respiratory muscles.
Respir Physiol 1992; 89 (2): 169-77
33. Powers SK, Lawler J, Criswell D, et al. Endurance-training-
induced cellular adaptations in respiratory muscles. J Appl
Physiol 1990; 68 (5): 2114-8
34. Sugiura T, Morimoto A, Murakami N. Effects of endurance
in the rat diaphragm. Pflugers Arch 1992; 421 (1): 77-81
35. Vrabas IS, Dodd SL, Powers SK, et al. Endurance training re-
duces the rate of diaphragm fatigue in vitro. Med Sci Sports
Exerc 1999; 31 (11): 1605-12
36. Clanton TL, Dixon GF, Drake J, et al. Effects of swim training
on lung volumes and inspiratory muscle conditioning. J Appl
Physiol 1987; 62 (1): 39-46
37. O’Kroy JA, Coast JR. Effects of flow and resistive training on
respiratory muscle endurance and strength. Respiration 1993;
60 (5): 279-83
38. Robinson EP, Kjeldgaard JM. Improvement in ventilatory mus-
cle function with running. J Appl Physiol 1982; 52 (6): 1400-6
39. Martin BJ, Stager JM. Ventilatory endurance in athletes and
non-athletes. Med Sci Sports Exerc 1981; 13 (1): 21-6
40. Coast JR, Clifford PS, Henrich TW, et al. Maximal inspiratory
pressure following maximal exercise in trained and untrained
subjects. Med Sci Sports Exerc 1990; 22 (6): 811-5
41. Thomas RG, LaStayo PC, Hoppeler H, et al. Exercise training
in chronic hypoxia has no effect on ventilatory muscle func-
tion in humans. Respir Physiol 1998; 112 (2): 195-202
42. Sonetti DA, Wetter TJ, Pegelow DF, et al. Effects of respiratory
muscle training versus placebo on endurance exercise perfor-
mance. Respir Physiol 2001; 127 (2-3): 185-99
43. Spengler CM, Roos M, Laube SM, et al. Decreased exercise
blood lactate concentrations after respiratory endurance train-
ing in humans. Eur J Appl Physiol Occup Physiol 1999; 79
(4): 299-305
44. Leith DE, Bradley M. Ventilatory muscle strength and endur-
ance training. J Appl Physiol 1976; 41 (4): 508-16
45. Morgan D, Kohrt W, Bates B, et al. Effects of respiratory mus-
cle endurance training on ventilatory and endurance perfor-
mance of moderately trained cyclists. Int J Sports Med 1987;
8: 88-93
46. Hanel B, Secher NH. Maximal oxygen uptake and work capac-
ity after inspiratory muscle training: a controlled study. J
Sports Sci 1991; 9 (1): 43-52
47. Fairbarn MS, Coutts KC, Pardy RL, et al. Improved respiratory
muscle endurance of highly trained cyclists and the effects on
maximal exercise performance. Int J Sports Med 1991; 12 (1):
66-70
Sports Med 2002; 32 (9)
Respiratory Muscle Training
48. Boutellier U, Piwko P. The respiratory system as an exercise
limiting factor in normal sedentary subjects. Eur J Appl Phys-
iol 1992; 64: 145-52
49. Boutellier U, Buchel R, Kundert A, et al. The respiratory system
as an exercise limiting factor in normal trained subjects. Eur
J Appl Physiol Occup Physiol 1992; 65 (4): 347-53
50. Suzuki S, Yoshiike Y, Suzuki M, et al. Inspiratory muscle train-
ing and respiratory sensation during treadmill exercise. Chest
1993; 104 (1): 197-202
51. Inbar O, Weiner P, Azgad Y, et al. Specific inspiratory muscle
training in well-trained endurance athletes. Med Sci Sports
Exerc 2000; 32 (7): 1233-7
52. Stuessi C, Spengler CM, Knopfli-Lenzin C, et al. Respiratory
muscle endurance training in humans increases cycling en-
durance without affecting blood gas concentrations. Eur J
Appl Physiol 2001; 84 (6): 582-6
53. Volianitis S, McConnell AK, Koutedakis Y, et al. Inspiratory
muscle training improves rowing performance. Med Sci
Sports Exerc 2001; 33 (5): 803-9
54. Markov G, Spengler CM, Knopfli-Lenzin C, et al. Respiratory
muscle training increases cycling endurance without affect-
ing cardiovascular responses to exercise. Eur J Appl Physiol
2001; 85 (3-4): 233-9
55. Billat V, Renoux JC, Pinoteau J, et al. Reproducibility of run-
.
ning time to exhaustion at V O2max in subelite runners. Med
Sci Sports Exerc 1994; 26 (2): 254-7
56. Coggan AR, Costill DL. Biological and technological variabil-
ity of three anaerobic ergometer tests. Int J Sports Med 1984;
5 (3): 142-5
57. Jeukendrup A, Saris WH, Brouns F, et al. A new validated
endurance performance test. Med Sci Sports Exerc 1996; 28
(2): 266-70
58. Kuipers H, Verstappen FT, Keizer HA, et al. Variability of
aerobic performance in the laboratory and its physiologic
correlates. Int J Sports Med 1985; 6 (4): 197-201
© Adis International Limited. All rights reserved.
581
59. McLellan TM, Cheung SS, Jacobs I. Variability of time to ex-
haustion during submaximal exercise. Can J Appl Physiol
1995; 20 (1): 39-51
60. Bishop D. Reliability of a 1-h endurance performance test in
trained female cyclists. Med Sci Sports Exerc 1997; 29 (4):
554-9
61. Hickey MS, Costill DL, McConell GK, et al. Day to day vari-
ation in time trial cycling performance. Int J Sports Med
1992; 13 (6): 467-70
62. Palmer GS, Dennis SC, Noakes TD, et al. Assessment of the
reproducibility of performance testing on an air-braked cycle
ergometer. Int J Sports Med 1996; 17 (4): 293-8
63. Schabort EJ, Hawley JA, Hopkins WG, et al. High reliability
of performance of well-trained rowers on a rowing ergometer.
J Sports Sci 1999; 17 (8): 627-32
64. Schabort EJ, Hawley JA, Hopkins WG, et al. A new reliable
laboratory test of endurance performance for road cyclists.
Med Sci Sports Exerc 1998; 30 (12): 1744-50
65. Decramer M, Macklem PT. Pressures developed by the respi-
ratory muscles. In: Roussos C, editor. The thorax. New York
(NY): Marcel Dekker, 1995: 1099-126
66. Harms CA, Dempsey JA. Cardiovascular consequences of ex-
ercise hyperpnea. Exercise Sport Sci Rev 1999; 27: 37-62
67. Killian KJ, Campbell EJM. Dypsnea. In: Roussos C, editor. The
thorax. New York (NY): Marcel Dekker, 1995: 1709-47
Correspondence and offprints: Dr A. William Sheel, School
of Human Kinetics, University of British Columbia, 210-
6081 University Blvd, Vancouver, BC V6T-1Z1, Canada.
E-mail: bill.sheel@ubc.ca
Sports Med 2002; 32 (9)