66

Improved Respiratory Muscle Endurance of Highly Trained

Cyclists and the Effects on Maximal Exercise Performance
M. S. Fairbarn, K. C. Coutis, R. L. Pardy, D. C. McKenzie
Departments of Sports Science & Family Practice and Pulmonary Research Laboratory,
University of British Columbia, Vancouver, B.C., Canada

Indroduction
Abstract
Exercise scientists generally have not thought
M S. Fairbarn, K. C. Coutts, R. L. Pardy, D. that the ventilatory system was a limiting factor to peak athletic
C. McKenzie, Improved Respiratory Muscle Endurance of performance. However, respiratory muscle fatigue has been
Highly Trained Cyclists and the Effects on Maximal Exer- demonstrated following both voluntary hyperpnea (1) and

Downloaded by: University of British Columbia. Copyrighted material.

cise Performance. mt J Sports Med, Vol 12, No 1, pp 66—70, marathon running in normal humans (9). Exercise perform-
1991. ance decreased following a long-term hyperpnea regimen
which caused respiratory muscle fatigue (10), and maximal
Accepted: April 20, 1990 ventilation decreased following 60 minutes of exhausting ex-
ercise (3). Theoretically then, it is possible that the ventilatory
Insufficient respiratory muscle endurance system might contribute to exercise limitation during very in-
(RME) may be one of the factors limiting ventilation during tense exercise where high minute ventilation is maintained.
peak athletic performance. Our purpose was to determine The significance of hypoventilation during intense exercise
whether the RME of highly trained cyclists could be en- was demonstrated during a study (5) where athletes breathing
hanced and if so, to determine the effects of improved RME air while running on a treadmill developed hypoxemia, but
on their maximal exercise performance. Ten male cyclists when the exercise was repeated breathing a HE:02 mixture,
(maximal oxygen consumption (VO2max) > 60 mlkg 5 they had higher minute ventilation and the hypoxemia was
began the study by peforming 3 tests. These were VO2max, preveted. If respiratory muscle fatigue does indeed contribute
RME measured as maximal sustainable ventilatory capac- to such activities, respiratory training could hypothetically re-
ity (MSVC) and maximal exercise endurance (tlim) sult in improved exercise performance (15).
measured by an endurance cycling test to exhaustion at
90% of their maximal power output. Five subjects then There is good evidence that respiratory muscle
completed 4 weeks of isocapnic hyperpnea training (16 ses- endurance can indeed be increased in normal humans. This
sion) and 5 subjects were controls. Following this training has been achieved both by an intense exercise running pro-
interval, each subject repeated the initial tests. After the gram (13) and by specific voluntary hyperpnea training of the
RME training, the MSVC increased from 155 11 to respiratory muscles (2, 6, 8, 12). In two of these studie, the ef-
174 12 1/mm (p = 0.004) for the training subjects while fects of the increased respiratory muscle endurance on exer-
there was no change in the controls (155 26 and 150 34 cise performance was examined. In both the normal elderly (2)
1/mm). There were no changes for any of the 10 subjects in and moderately trained athletes (12), there was no effect of the
either the maximal exercise performance (VO2max = increased respiratory muscle endurance on exercise perform-
66.1 4.7 to 66.5 4.8 mlkg 5 or the maximal exercise ance. To date, there have been no studies to examine whether
endurance (tlim=335±79 to 385± 158 see). In conclu- or not the respiratory muscle endurance of highly trained ath-
sion, 4 weeks of respiratory muscle endurance training in- letes can be improved by specific respiratory muscle hyper-
creased respiratory muscle endurance but had no effect on pnea training. Furthermore, it is not known whether exercise
the maximal cycling performance of highly trained cyclists. performance in such individuals is improved by specific re-
spiratory muscle training.
Key words
Thus, the first purpose of this study was to train
respiratory muscles, endurance, maximal the respiratory muscles of highly trained endurance athletes
oxygen consumption, maximal sustained ventilatory ca- (cyclists) using a protocol that mimicked the high ventilation
pacity achieved in exercise. The second purpose of this study was to
determine the effects of respiratory muscle endurance training
on exercise performance.

Methods and Procedures

Ten well-trained male cyclists volunteered for

tnt. J. Sports Med. 12(1991)66—70
this study. The three criteria for inclusion in the study were:
GeorgThieme Verlag Stuttgart New York
ImprovedRespiratory Muscle Endurance of Highly Trained Cyclists and the Effects Int.J.SportsMed. 12 (1991) 67

maximal oxygen consumption (VO2max) greater than 60

mlkg 1min normal values for spirometry, and an active par-
ticipation in cycling events. Since the subjects were known to
be highly fit athletes and both the experimental and control
subjects were matched for fitness, quantifying the amount of
aerobic training during this study was not included in the pro-
tocol. However, subjects were asked to report any substantial
changes in their cycling training. In addition, testing of all sub-
jects was completed while the cycling competitions were being P
held on a regular schedule. Informed consent was obtained
from each subject. Five subjects were assigned to the training A - Pump
B — Air Plow Meter
group and five to the control group. The three tests used to C — Mixing Chember
evaluate the effect of the isocapnic hyperpnea training were: D— CO2 Cylinder
H Spirometer 9 I
the maximal sustainable ventilatory capacity test (MSVC), a F — Aneathnit Bag 5 I
maximal incremental cycle ergometer test, and an endurance o — ln.pired Gag Analyzer
cycle test at 90% of the maximal power during the VO2max H • 2—way Valve & Mouthpiece
I — Pneumotachograph
test. Each of these tests was performed by all subjects both I — Expired On Analyzore
before and after the four-week training period for the experi- K — Microcomputer
mental subjects. Subjects continued their regular aerobic

Downloaded by: University of British Columbia. Copyrighted material.

training programs and were required to record the number of
kilometers per week that they cycled. Fig. 1 Ventilatory endurance apparatus for respiratory muscle en-
durance testing and training. Arrows indicate direction of airflow.

Maximal Incremental Cycle Test

other. The mean ventilation reached during the last eight
The initial maximal incremental cycle test was minutes of the highest test was the baseline measurement.
used both to ensure that subjects had a VO2max greater than 60
mlkg — min1 and to measure maximal exercise ventilation
Ventilatory Endurance Apparatus
(VEmax). The test following training was an index of whether
the aerobic fitness of the subjects had changed during the pro- The ventilatory endurance apparatus (Fig. 1)
ject as well as a measure of the effectiveness of the isocapnic was designed to allow isocapnic hyperpnea for both testing
hyperpnea training. The incremental cycle test was performed and training the endurance of the respiratory muscles. On the
on an electronically braked cycle ergometer (Mijnhardt, Hol- inspired side of the circuit, a vacuum pump (Bodine Electric,
land) using a ramp protocol beginning at 0 watts and increas- Chicago, IL) supplied a variable air flow which passed
ing by 30 watts per minute. Expired gas was measured and ana- through an air flow meter (Vacumetrics, Ventrua, CA) and
lyzed by either a Medical Graphics system with the associated then into a 13.5-litre mixing chamber. 100% CO2 was added at
2001 software package or with a separate system consisting of the rate of 3.4 to 4.25 1/mm to maintain the fractional concen-
oxygen and C02 analyzers (Beckman OM 11, LB2,FulJerton, tration of mixed expired C02 (FECO2) at each subject's pre-
CA), and a heated pneumotachograph (Fleisch, Switzerland) determined resting level. A 9-liter Respirometer (Collins, Bos-
to calculate minute ventilation. Each of the analysers were Ca- ton, MA) provided the visual reference for the target ventila-
librated with air and calibration gases before and after each tion for the subject. The subject was instructed to keep a mark
test. Data was processed on-line by an IBM microcomputer. on the bell of the respirometer below the water reservoir level.
Heart rate (HR) was recorded using direct lead ECG A 5-litre anesthesia bag was included on the inspired side of
(Lifepack 6, Physio Control Canada, Agincourt, Ontario). the system to provide a dampening effect. Distal to this bag
Criteria for attaining VO2max was a plateau in V02 with an in- was a sampling tube connected to a CO2 analyzer (Medical
creased workload, a respiratory exchange ratio greater than Graphics, St. Paul, MN) to measure the fractional concen-
1.15, and HR greater than 180 beats per minute (90% pre- tration of inspired C02 (FICO2). The subject breathed
dicted maximum heart rate). through a low-resistance 2-way valve (Hans Rudolph K.C.,
MO) and the expired gas passed through a heated pneumota-
Maximum Sustained Ventilatory Capacity chograph (Model 3 Fleisch, Switzerland) to calculate minute
(MSVC) ventilation. A temperature gauge recorded expired gas tem-
perature. Ventilation was processed on line by an IBM micro-
The MSVC test was used to measure respir- computer (Armonk, N. Y.). The gas analyzer was calibrated
atory muscle endurance; this was determined by measuring using air and calibration gas before each test. The pneumota-
the maximum ventilation the subjects could sustain for 10 chograph was calibrated at the maximum volume of 230 1/mm
minutes (2). As a warm up to the test, the subjects breathed on as measured by air flow meter. The resistance of the circuit was
the ventilatory endurance apparatus (see below) for two 0.035cm H2O/l/min.
minutes at 50% of their maximal exercise ventilation (VEmax)
followed by one minute of rest. During the first two minutes of Endurance Cycle Test
the test the air flow was gradually increased from the warm-up
level to the maximum ventilation that the subjects could Each subject performed a cycle ride to exhaus-
tolerate for the next 8 minutes. To ensure the baseline MSVC tion at a power which represented 90% of the maximum work
values were maximal, the test was peformed by each subject rate previously achieved during the incremental cycle test. The
until two tests, separated by 48 hours, were within 5% of each purpose fo the endurance cycle test was to simulate a competi-
 mt. J. Sports Med. 12 (1991) M. S. Fairbarn, K. C. Coutts, R. L. Pardy, D. C. McKenzie

Table 1 Anthropometric Data for 10 Subjects 200

(Mean SD)
Experimental Control 80
(n=5) (n=5) 'I
Height (cm) 174 178
E 160
7.7 3.8
Weight (kg) 69 73 0
6.2 6.8 so
140
Age (yr) 22 23 0s
>
2.7 4.0
FVC (liters) 5.5 5.7
120
0.9 0.7
FEy1 (liters) 4.4 4.8
0.3 0.6 A

MVV12 (1/mm) 206 215 2 4 6 8 10

15 20 time (mm)

FVC = Forced vital capacity

FEV1 = Forced expiratory volume in one second Pre-Exp. — — Pre-Con. Post.-Exp. -. A Post-Con.
(n—SI (n..5( (n5) (n5(
MW12 = Maximal voluntary ventilation in 12 seconds.
Fig. 2 Mean ventilation for each minute during the MSVC test for

Downloaded by: University of British Columbia. Copyrighted material.

both the experimental and control groups pre- and post-training pe-
tive cycle ride while standardizing the external environment. riod. The mean of the last 8 minutes of the MSVC was greater for the
Previous experience in this laboratory with similar subjects experimental group (—A—) post training.
has shown that the approximate time for a test at this work level
is from 6 to 10 minutes, a duration similar to the MSVC test and
isocapnic hyperpnea training times. Following a 3-minute the eighth training session, the duration of both the work and
warm-up, resistance was increased over 10 seconds until the rest intervals was increased to 10 minutes each.
predetermined work rate was attained. From this point, ex-
pired minute ventilation (VEtlim) was recorded as well as the Statistical Analysis
time to exhaustion (tlim). The criterion for the end of the test
was the inability to maintain the minimum pedal frequency of Analysis of variance was used to test the simi-
60 rpm for 3 consecutive revolutions. Subjects were not aware larity of the experimental and control groups prior to respir-
of the elapsed time. atory muscle endurance training for the following variables:
age, height, weight, FYC, FEVi, MVV12, MSVC, VO2max,
Pulmonary Function Tests VEmax, tlim, and VEthm.

Pulmonary function tests were performed to To determine the effects of the respiratory
ensure normal spirometry of the subjects. Forced vital capac- muscle endurance training, the mean group pre- and post-
ity (FVC), forced expiratory volume in one second (FEVi), training differences were tested by multiple analysis of vari-
and maximal voluntary ventilation in 12 seconds (M\TV12) ance for the variables in the following groups:
were measured and analyzed using the Medical Graphics 1. Pulmonary functions: FVC, FEY1, MYV
computerized spirometer system (St. Paul, MN) with the as- 2. Maximal aerobic capacity test: VO2max, VEmax
sociated 1070 software package. 3. Endurance cycle test: turn, VEtlim.

Respiratory Muscle Endurance Training Analysis of variance was used to determine the
Protocol significance of the difference in the MSVC test between the
groups. The data analysis was performed using the statistical
To improve the endurance performance of the package, SYSTAT, version 4 (16). The level of significance for
respiratory muscles, a volume overload technique, isocapnic each test was P <0.05. Data are expressed as mean SD.
hyperpnea, was selected. The subjects attended three or four
training sessions per week for a total of 16 sessions. Each ses- Results
sion consisted of three 8-minute work intervals of isocapnic
hyperpnea alternated with 8-minute intervals of rest. The All 10 cyclists completed the study and main-
training overload was a combination of increasing both venti- tained the same average number of kilometers cycled per week
lation and duration of the work intervals. Initially, the target (range 125—500) throughout the study. Prior to the isocapnic
ventilation for each work interval was the ventilation each sub- hyperpnea training period, there were no differences between
ject achieved during the initail MSVC test with the remainder the training and control groups for any of the variables studied
of each training session at the maximum tolerable ventilation. (Tables 1 & 2).
Progressively, the subjects were able to maintain this target
ventilation for each of the three work intervals. To provide a Following 16 training sessions, the experimen-
training stimulus, the target ventilation was then increased to a tal group demonstrated a significant increase in respiratory
level that could only be maintained for the first work interval muscle endurance. The ventilation during the MSVC test in-
of a training session and the subject's new goal was to maintain creased from 155.4± 11.2 to 173.9± 11.6 1/mm, p=0.OO4.
this larger ventilation for all three work intervals. Following The control groups values were not different (155.1 and
ImprovedRespiratory Muscle Endurance of Highly Trained Cyclists and the Effects mt. J. Sports Med. 12(1991) 69

Table 2 Pre- and Post-Training Exercise Variables

(Mean SD)
Control Experimental
(n=5) (n=5)
Pre-train Post-train Pre-train Post-train

VO2max 68.0 67.1 64.2 65.8

(mI'kg 1min) 6.6 5.8 1.9 4.8
VEmax 171.4 167.5 177.0 177.1
(I/mm) 35.5 21.6 12.6
VEtlim 174.0 176.3 176.0 178.5
(I/mm) 29.2 26.8 19.3
turn 328.6 342.4 427.8
(seconds) 99.0 79.6 74.9

VO2max = Maximal oxygen consumption

MSVC flaX VEttim VEmax Maximal ventilation during VO2max test
FnaCon, EJ PoatCon. VEtlim = Maximal ventilation during endurance cycle test
Froop. Poat•Ex9.
tlim=Time to exhaustion (endurance cycle test).

Fig. 3 Ventilation for each of the three tests: MSVC (maximal sus-
tainable ventilatory capacity test), VEmax (maximal incremental

Downloaded by: University of British Columbia. Copyrighted material.

cycle test), VEtlim (endurance cycle test). Experimental subjects in- mean age of the subjects of the Belman and Gaesser (2) study
creased MSVC post training ( p=O.OO4). No other changes. was 67 years and the ventialtions during the MSVC were only
62—64%oftheMVVt 2, while the pre-training value for the sub-
jects in the present study was 75% of their MVVi 2. Morgan (12)
149.5 34 liters) following the four week interval (Fig. 2). reported a 1 7-fold increase in respiratory muscle endurance in
There were no differences in the maximal minute ventilations 4 subjects whose VO2max was the nearest to that of our sub-
in the incremental and endurance cycle tests either between jects (51 vs 66 mFkg min , but the test of respiratory
the 2 groups of subjects before or after training (Fig. 3). muscle endurance they used was not comparable to the MSVC
used in the present study.
The results of the post-training test showed that
there were no differences in any of the variables measured Although the MSVC increased with specific
during the exercise tests for either the experimental or control respiratory muscle training, there was no change in maximal
ventilation during either the incremental or the endurance
groups (Table 2).
cycle tests in our study. One potential explanation for the lack
Discussion
of a carryover effect for the voluntary hyperpnea of training to
the involuntary hyperventilation of exercise was that the
This study demonstrates that 16 training ses- breathing patterns were different. Klass and Dempsey (7)
sions of isocapnic hyperpnea increased the respiratory muscle showed that the end-expiratory lung volume was higher
endurance of highly trained cyclists. The five experimental during the voluntary hyperpnea tests compared to exercise at a
subjects were able to increase their maximal sustained ventila- similar ventilation. We did not measure end-expiratory lung
tory capacity by 12%, demonstrating that their respiratory volume, but the breathing pattern seen at VEmax during the
muscle endurance was not maximal as a result of their regular voluntary hyperpnea (tidal volume [VT] 2,56 0.2 1, respi-
cardiovascular training programs. Since the control subjects ratory rate [RR] 62±3 breathjmin) in the pretraining
did not increase their respiratory muscle endurance, we can as- MSVC test was different from the breathing pattern seen at
sume the increase for the experimental subjects was due to the VEmax during the exercise tests (VT = 3.46 0.6 1,

respiratory muscle training. It has previously been established RR 54 7 breaths/mm). At the end of the training sessions
that athletes involved in sports requiring a high degree of aero- we found that the breathing pattern during the MSVC
bic fitness have greater respiratory muscle endurance than (VT = 3.4 0.21, RR = 53 3 breaths/mm) had become al-
non-athletes (11), but the present study illustrates that the res- most identical to that found at VEmax in the incremental exer-
piratory muscle endurance of highly trained athletes may still cise test. We therefore speculate that it is unlikely that the dif-
be influenced by training programs designed specifically for ferences in breathing pattern during the training compared to
these muscles. the exercise tests would explain the fact that maximum ventila-
tion did not change after training in either the incremental or
The 12% increase in MSVC in this study is not endurance cycle tests. Our study would support the notion that
as large an increase as demonstrated by other studies (2, 6, 8, ventilatory factors do not contribute to exercise limitation in
12). Some of this difference can be explained by the higher in- highly trained athletes.
itial MSVC of our athletic subjects. Leith and Bradley (8) re-
ported an increase of 21 %, but the endurance breathing test The VO2max of our subjects did not change
(SVC) they used was a composite of 8 to 10 points of different following hyperpnea training, and these results are consistent
levels of ventilation against time and therefore not strictly with the studies of Morgan et al (12) and Belman and Gaesser
comparable to the MSVC used in this study. The greater im- (2). To increase the VO2max, it would be necessary to increase
provement reported in the study by Keens et al (6) was possibly the amount of oxygen available to the working muscles, and
because their subjects (age 28.3 2.6 yr) had 30 training this can be accomplished by increasing the cardiac output
sessions compared to 16 for the subjects in this study. The and/or the systemic arteriovenous oxygen difference. The
70 mt. J. Sports Med. 12 (1991) M. S. Fairbarn, K. C. Coutts, R. L. Pardy, D. C. McKenzie

best exercise necessary to effect these changes would be activ- References

ities which require rhythmic contractions of large muscle
groups, and VO2max would most likely improve as a result of 1 Bai, T. R., Rabinovitch B. J., Pardy R. L.: Near-maximal voluntary
increased cardiac output (4). Since respiratory muscle training hyperpnea and ventilatory muscle function. JAppi Physiol 57 (6):
1742—1748,1984.
involves a small group of muscles, there is no reason to antici- 2
Belman M. J., Gaesser G. A.: Ventilatory muscle training in the
pate an adaptation in the central circulatory response to exer- elderly. JApplPhysiol64 (3): 889—905, 1988.
cise unless the respiratory muscles become so efficient after Bender P. R., Martin B. J.: Maximal ventilation after exhausting
training that their oxygen consumption at any level of ventila- ' exercise. Med Sci Sports Exerc 17(1): 164—167, 1985.
tion decreases substantially. Under these circumstances, there Clausen J. P.: Effect of physical training on cardiovascular adjust-
might be more oxygen available to the exercising limb muscles ments to exercise in man. PhysiolRev 57:770—815, 1977.
(1, 15). It is apparent from this study that such theoretical con- Dempsey J., Hanson P., Pegelow D., Claremont A., Rankin J.:
Limitations to exercise capacity and endurance: Pulmonary sys-
siderations did not apply since VEmax and VO2max in exer- tem. Can JSportSci7 (1): 4—13, 1982.
cise did not change with training. Also, these results support 6
Keens T. G., Krastins I. R. B., Wannamaker E. M., Levison H.,
Weibel and Kayar's (14) concept that each component in the Crozier D. N., Bryan A. C.: Ventilatory muscle endurance training
system responsible for oxygen delivery or oxygen utilization in normal subjects and patients with cystic fibrosis. Am Rev Respir
develops simultaneously. Consequently, adaptation of the Disll6:853—860, 1977.
respiratory muscles is ineffective without the concomitent Kias J. V., Dempsey J. A.: Voluntary versus reflex regulation of
maximal exercise flow: volume loops. Am Rev Respir Dis 139:
adaptations in other components of the oxygen delivery sys- 150— 156,1989.
tem.

Downloaded by: University of British Columbia. Copyrighted material.

Leith D. E., Bradley M.: Ventilatory muscle strength and en-
durance training.JApplPhysiol4 1(4)508—516, 1976.
There were no significant changes in the en- Joke J., Mahler D. A., Virgulto J. A.: Respiratory muscle fatigue
durance cycling test measurements and again these results are 10
after marathon running. JApplPhysiolS2 (4): 821—824, 1982.
similar to the study of Morgan et al (12). Although the tlim ap- Martin B., Heintzelman M., Chen H.: Exercise performance after
peared to increase by 86 seconds (from 342 to 428 226 ventilatorywork.JApplPhysjol52(6): 1581—1585,1982.
Martin B. J., Stager J. M.: Ventilatory endurance in athletes and
seconds) in the experimental subjects in this study, these re- non-athletes. Med Sci Sports Exerc 13(1): 21—26, 1981.
sults were affected by one subject whose dim increased for 459 12
Morgan D. W., Kohrt W. M., Bates B. J., Skinner J. S.: Effects of
to 801 seconds. Since this subject's V02 and VEtlim during the respiratory muscle endurance training on ventilatory and en-
endurance cycle test were very similar to the VO2max and durance performance of moderately trained cyclists. mt J Sports
VEmax during the maximal incremental test, perhaps other 13
Med8 (2): 88—93,1987.
variables such as motivation or buffering capacity may have Robinson E. P., Kjeldgaard J. M.: Improvement in ventilatory
been responsible for this individual's variability in the tlim. muscle function with finning. JApp Physiol 52 (6): 1400—1406,
1982.
14
Wiebel E., Kayar J.: Matching 02 delivery to 02 demand in
Neither oxygen consumption of the respir-
atory muscles nor respiratory muscle fatigue were measured
during this study. Under ideal circumstances, measuring these
' muscle: I. Adaptive variation. Adv Exp Med Biol 227: 159—169,
1987.
Whipp B. J., Pardy R. L.: Breathing during exercise. Handbook of
Physiology: The Respiratory System III. Chap. 34: 605—629,
two variables during a hyperpnea training protocol would pro-
vide information on the intensity and duration of training ne- 1986.
16
Wilkinson, L.: SYSTAT: The System for Statistics, Evanston, IL:
cessary to effect metabolic adaptation of the respiratory SYSTAT, Inc. 1988.
muscles.

In conclusion, four weeks of isocapnic hyper- M S. Fairbarn

pnea exercise increased the respiratory muscle endurance of Pulmonary Research Laboratory
five highlytrained male cyclists, but did not effect a change in St. Paul's Hospital
maximal exercise ventilation or performance. 1081 BurrardSt.
Vancouver, B. C., Canada
V6K 1Y6
Acknowledgement
The authors wish to thank the subjects for their time
and effort and also to thank Joe Comeau for his technical assistance.