|
|
||||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the National Adult Cystic Fibrosis Unit, Department of Respiratory Medicine, St. Vincent's University Hospital, Dublin, Ireland.
Correspondence to: C. G. Gallagher, MD, FCCP, Department of Respiratory Medicine, St. Vincent's University Hospital, Dublin 4, Republic of Ireland; e-mail: cgall{at}indigo.ie
| Abstract |
|---|
|
|
|---|
Methods: Nine subjects with CF underwent a total of three
maximal exercise tests carried out under identical circumstances over a
28-day period. Oxygen uptake
(
O2), minute
ventilation (
E), respiratory frequency (f), heart
rate (HR), and arterial oxygen saturation
(SaO2) were measured at rest, at end exercise,
and at 40% and 70% of maximum workload.
Results:
There were no significant differences in these measurements among the
three tests. Reproducibility of exercise performance was assessed using
the coefficient of variation. The mean within-subject coefficient of
variation for test variables at end exercise are as follows:
O2, 6.9%;
E, 6.2%; f, 5.8%; HR, 3.0%; and
SaO2, 1.1%. The mean within-subject
coefficient of variation for test variables at 40% and 70% of maximal
work rates are as follows:
O2, 5.2% and 4.6%;
SaO2, 0.3% and 0.9%; HR, 4.0% and 3%;
E, 5.7% and 6.5%; and f, 5.8% and 7.2%,
respectively.
Conclusions: Variables measured during clinical cycle ergometer exercise testing in adult patients with stable CF are reproducible. No learning effect was found on repeated testing.
Key Words: cystic fibrosis exercise testing reproducibility
| Introduction |
|---|
|
|
|---|
O2max), reduced
peak work rate (Wmax), and abnormal ventilatory and cardiovascular
responses to exercise.3
4
5
6
7
Despite the use of clinical
exercise testing in CF patients, the reproducibility of maximal
exercise testing has not, to our knowledge, been examined in these
patients. In order to assess the clinical significance of repeated
maximal exercise testing, we examined the reproducibility of clinical
exercise testing in our patients with CF. Both group mean and
individual responses to exercise were tested in adult patients with
stable CF undergoing repeated incremental exercise tests under
identical conditions. | Materials and Methods |
|---|
|
|
|---|
|
The study was approved by the Ethics Committee of St. Vincent's University Hospital. All subjects gave informed consent for the procedures.
Protocol
FEV1 and FVC were measured before and
after each exercise test. At least three well-coordinated
maximal efforts were obtained, and the highest value obtained for each
variable was recorded. The subjects underwent three exercise tests over
a 28-day period, with each two tests separated by at least 7 days.
Exercise was performed at the same time of day on each occasion. The
subjects were asked to avoid strenuous activity for at least 24 h
and food or caffeinated drinks for 2 h prior to exercise testing.
All patients were instructed to take all of their maintenance
medications but to avoid the use of ß-agonist inhalers for at least
2 h prior to testing.
Testing was performed on an electrically braked cycle ergometer (Excalibur; Lode BV; Groningen, The Netherlands) while the subject breathed room air. After mounting the cycle ergometer, each patient put on a nose clip, inserted the mouthpiece, and had resting measurements taken over 1 min. The initial exercise workload was 15 W and was increased by 15 W/min in a ramp fashion until exhaustion. With the use of speedometer feedback, each subject chose the pedaling rate within a range of 50 to 70 revolutions/min. All subjects were instructed in an identical manner by the same operator for all exercise studies. The subjects were told that they should continue to exercise until they could exercise no more. No other type of encouragement was offered, and no communication was made with the subjects during the testing to ensure consistency of the protocol.
ECG leads attached to the chest enabled continuous monitoring of the heart rate (HR). Arterial oxygen saturation (SaO2) was monitored by pulse oximetry (SAT-TRAK Pulse Oximeter; SensorMedics; Yorba Linda, CA). Each patient's mouthpiece was connected to a heated wire flowmeter (Mass Flow Sensor; SensorMedics). The flow signal was digitally integrated to give tidal volume (VT), and respired gases were continually analyzed by rapidly responding oxygen (paramagnetic) and carbon dioxide (infrared) analyzers. All equipment was calibrated before each exercise study using calibration syringes and precision oxygen and carbon dioxide gas mixtures. All signals were continuously displayed breath by breath on a computer screen in real time during the exercise test. Data were also stored on computer hard disk for later analysis.
Data Analysis
Minute ventilation (
E), VT,
respiratory frequency (f), HR, oxygen uptake
(
O2), and carbon dioxide
output (
CO2) were measured
breath by breath using standard formulas.10
11
E and VT were expressed at body
temperature and pressure, saturated with water vapor;
O2 and
CO2 were expressed at standard
temperature pressure, dry. Predicted
O2max during exercise was
calculated as follows10
:
![]() |
![]() |
![]() |
Statistical Analysis
Data collected at rest, at maximal exercise, and at two matched
submaximal work rates were used in the analysis. Submaximal work rates
of 40% and 70% of the highest Wmax achieved during the three tests
were chosen for each patient. Comparisons were then made at matched
work rates for the three studies.
Statistical significance of group mean data from the three experiment days were determined by repeated-measures analysis of variance.14 The variability of subject results for the three experiments was assessed using the coefficient of variation. The coefficient of variation was derived by dividing the SD by the mean.15 Analysis of the Borg scale was performed using Wilcoxon's signed rank test; p value < 0.05 was considered significant. The results are shown as mean (± SD).
Analysis of our data showed that the sample size was sufficient to
detect a 12% increase in both
O2max (0.18 L/min) and
Wmax (16 W) throughout the three exercise tests with 90%
power.16
| Results |
|---|
|
|
|---|
Resting Results
Analysis of group mean data collected before exercise showed no
significant difference in baseline spirometric values. Mean
FEV1 was 56% predicted (range, 30 to 82%). Mean
FVC was 76% predicted (range, 55 to 100%). The mean coefficient of
variation for FEV1 was 4.7% (range, 1.0 to
8.8%), and for FVC, 5.2% (range, 2.0 to 9.4%). The mean resting
SaO2 was 95% (range, 93 to 97%).
There was no significant difference in baseline values of
O2,
CO2,
E,
VT, f, SaO2, and HR in
the three separate studies.
Exercise Results
All patients had evidence of impaired exercise tolerance. Mean
(± SD) values for
O2max as
percent predicted was 62 ± 8% (range, 51 to 71%). The mean maximal
heart rate was 86 ± 6% (range, 78 to 98%) of predicted. Mean
oxygen desaturation was 4 ± 2.7% (range, 1 to 9%). The three
subjects with the lowest FEV1 had significant
desaturation of > 5%. The
E/MVV ratio was
98 ± 27% (range, 57 to 124%). Reasons for discontinuing exercise
were dyspnea (three patients), leg discomfort (four patients), or both
(two patients). This did not vary from test to test.
Eight of the nine patients reached their anaerobic threshold as calculated using the modified V-slope method.17
Table 2
lists the group mean data collected at end of exercise. There was no
significant difference at end of exercise in
O2,
CO2,
E,
VT, f, HR, SaO2, exercise
time, or work rate. In addition, there was no significant difference in
any variables at 40% Wmax or 70% Wmax.
|
|
O2max was 6.9% (range, 1 to
13% throughout exercise); for SaO2,
1.1% (range, 0 to 5% throughout exercise); and HR, 3.0% (range, 1.7
to 8.0% throughout exercise).
The mean within-subject coefficient of variation for
E at end exercise was 6.2% (range, 3.0 to
12.3% throughout exercise); for VT, 3.8% (range, 1 to
15% throughout exercise); and for f, 5.8% (range, 2 to 15%
throughout exercise).
The mean within-subject coefficient of variation for exercise duration was 4.7% (range, 1.9 to 13.2%) and for Wmax, 6.0% (range, 0.6 to 14.0%). At end exercise, the mean coefficient of variation for Borg scale leg discomfort was 8.4% (range, 0 to 25%) and for Borg scale dyspnea, 11.3% (range, 0 to 25%).
Figure 1
shows each subject's
O2max
and exercise duration for each of the three exercise tests. Figure 2 is a graphic representation of
O2 against
E throughout the exercise tests for three patients
with varying degrees of lung dysfunction. The relationship between
O2 and
E is
similar in each subject's three tests.
|
|
| Discussion |
|---|
|
|
|---|
O2max.
Our patients had increased ventilatory requirements with a high
ventilation (
E), an elevated ventilatory equivalent
for oxygen
(
E/
O2)
and an increased
E/MVV at end exercise. Two patients
reached their predicted maximum HR. The three patients with the lowest
FEV1 showed significant oxygen desaturation
during exercise. We looked at the reproducibility and within-subject variability of measured parameters during maximal incremental exercise testing in our adult CF population. This variability, as measured by the coefficient of variation, is similar to that measured in studies carried out on healthy subjects and in patients with chronic airflow limitation (CAL), interstitial lung disease (ILD), and cardiac failure.
Previous studies have examined the reproducibility of maximal exercise
testing. Garrard and Emmons18
found some diurnal variation
in their healthy subjects, with a coefficient of variation during
maximal exercise testing for
O2max of 8.4%;
E, 12.0%; and HR, 3.8%. Nordrehaug et
al19
examined the reproducibility of maximal treadmill
exercise testing in healthy subjects and found a coefficient of
variation for
O2max of 5.0%;
E, 7.0%; and HR, 3.0%. They also found that the
variability at end exercise was less than that at submaximal levels of
exercise.
The reproducibility of repeated exercise testing has also been
extensively examined in patients with CAL. Swinburn et
al20
examined the repeatability of walking, step, and
cycle ergometer tests and found a significant learning effect on
repeated exercise testing, as well as significant variability in
O2 and
E,
depending on which exercise test was performed. Noseda et
al21
examined repeated cycle ergometer exercise tests in
patients with CAL at intervals of 1 month. Their coefficient of
variation for FEV1 was 10.2%;
O2, 9.0%;
E, 8.1%; and HR, 5.0%. Owens et al22
found similar results with maximal ergometer exercise tests also
separated by 1 month, with a coefficient of variation for
FEV1 of 7.5%;
O2, 6.6%;
E, 6.3%; and HR, 3.5%. Cox et al23
examined reproducibility of exercise testing carried out on consecutive
days and calculated a relative duplicate error of 3.5% for
O2 and 6.6% for
E.
In patients with ILD, Marciniuk et al24
examined the
reproducibility of maximal ergometer exercise testing and found a
coefficient of variation for FEV1 of 3.5%;
O2, 5.3%;
E, 5.5%; and HR, 4.0%. This is similar to
findings in patients with cardiac failure where Janicki et
al25
have shown good reproducibility, although Elborn et
al26
showed significant increases in exercise duration and
workload with repeated exercise testing.
Our patients with CF had a coefficient of variation for
FEV1 of 4.7%;
O2max, 6.9%; and
E, 6.2%. This is similar to the findings of
Nordrehaug et al19
in healthy subjects. Comparison with
patients with respiratory disease shows adult patients with CF to have
variability slightly greater than ILD patients but less variability
than CAL patients.
All of our patients stated that they made a maximal effort and
exercised until exhaustion. The Borg scores at end exercise (dyspnea,
4.1 ± 1.1; leg discomfort, 3.8 ± 1.1) are similar to those in
other studies assessing breathlessness at end exercise in other
diseases. Marciniuk et al24
found that patients with ILD
had mean Borg scores of 4.5 at end exercise. In their study of added
dead space during maximal incremental exercise testing in ILD, the
control group of the study (with no dead space) had mean scores of 5.0
at end exercise.27
Studies looking at the
reproducibility of Borg scoring after maximal exercise testing in
patients with CAL have also found a mean of 5.0 after repeated
testing.28
Other factors favoring a maximal effort by our
patients is that the majority of the patients either reached their
maximal predicted HR at end exercise or showed a high (> 90%)
E/MVV ratio at end exercise. Patients with low Borg
scores (
3) for dyspnea all gave higher scores (> 4) for leg
discomfort and vice versa. Finally, all our exercise tests were
observed by a physician who felt that, at end exercise, all the
patients had given maximal effort.
There are limitations to the current assessment of factors that limit
exercise,29
but allowing for these, one patient appeared
primarily limited by cardiac factors, seven patients appeared limited
by respiratory factors, and one appeared limited by a combination of
both. The patients with the most severe CF were limited primarily by
respiratory factors, with
E/MVV ratios > 90% and
low predicted HR, while the two patients with the mildest disease were
primarily limited by cardiac and/or respiratory factors, reaching
> 97% predicted maximum HR.
In clinical practice, changes in individual patients are more useful
than changes within groups of patients. We looked at within-subject
variation and how this could be applied in a clinical setting. Our
results indicate that changes of 13% in exercise duration, 19% in
O2max, and at least 17% in
peak
E are unlikely to occur by
chance.14
In conclusion, we have found that parameters measured in repeated cycle ergometer exercise testing are reproducible; this variability is similar to that seen in healthy subjects and in patients with CAL and ILD. There was no obvious learning effect between the tests. We also noted that the reproducibility of spirometric measurements (FEV1 and FVC) before and after exercise was similar to that of healthy subjects and, although slightly less than that measured in ILD, is more reproducible than that seen in CAL. In addition, we looked at exercise testing at submaximal workloads (40% and 70% Wmax) and also found this to be reproducible.
Our patients showed no learning or training effect with repeated exercise testing, indicating that practice testing is not necessary in patients with stable CF who have not previously used a cycle ergometer.
| Footnotes |
|---|
Abbreviations:
CAL = chronic airflow limitation; CF = cystic fibrosis;
f = respiratory frequency; HR = heart rate; ILD = interstitial
lung disease; MVV = maximum voluntary ventilation;
SaO2 = arterial oxygen saturation;
CO2 = carbon dioxide output;
E = minute ventilation;
O2 = oxygen uptake;
O2max = peak oxygen uptake;
VT = tidal volume; Wmax = peak work rate
Received for publication May 27, 1998. Accepted for publication February 24, 1999.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
J. D. Dodd, S. C. Barry, R. B. M. Barry, C. G. Gallagher, S. J. Skehan, and J. B. Masterson Thin-Section CT in Patients with Cystic Fibrosis: Correlation with Peak Exercise Capacity and Body Mass Index. Radiology, July 1, 2006; 240(1): 236 - 245. [Abstract] [Full Text] [PDF] |
||||
![]() |
E. F. McKone, S. C. Barry, M. X. FitzGerald, and C. G. Gallagher Role of arterial hypoxemia and pulmonary mechanics in exercise limitation in adults with cystic fibrosis J Appl Physiol, September 1, 2005; 99(3): 1012 - 1018. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. I. Parekh, J. A. Blumenthal, M. A. Babyak, R. LaCaille, S. Rowe, L. Dancel, R. M. Carney, R. D. Davis, S. Palmer, and for the INSPIRE Investigators Gas Exchange and Exercise Capacity Affect Neurocognitive Performance in Patients With Lung Disease Psychosom Med, May 1, 2005; 67(3): 425 - 432. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. E. Hansen, X.-G. Sun, Y. Yasunobu, R. P. Garafano, G. Gates, R. J. Barst, and K. Wasserman Reproducibility of Cardiopulmonary Exercise Measurements in Patients With Pulmonary Arterial Hypertension Chest, September 1, 2004; 126(3): 816 - 824. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. G. Thin, S. J. Linnane, E. F. McKone, R. Freaney, M. X. FitzGerald, C. G. Gallagher, and P. McLoughlin Use of the Gas Exchange Threshold to Noninvasively Determine the Lactate Threshold in Patients With Cystic Fibrosis* Chest, June 1, 2002; 121(6): 1761 - 1770. [Abstract] [Full Text] [PDF] |
||||
![]() |
H. Blau, H. Mussaffi-Georgy, G. Fink, C. Kaye, A. Szeinberg, S. A. Spitzer, and J. Yahav Effects of an Intensive 4-Week Summer Camp on Cystic Fibrosis* : Pulmonary Function, Exercise Tolerance, and Nutrition Chest, April 1, 2002; 121(4): 1117 - 1122. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |