Peak exercise capacity estimated from incremental shuttle walking test in patients with COPD: a methodological study
© Arnardóttir et al. 2006
Received: 31 May 2006
Accepted: 17 October 2006
Published: 17 October 2006
In patients with COPD, both laboratory exercise tests and field walking tests are used to assess physical performance. In laboratory tests, peak exercise capacity in watts (W peak) and/or peak oxygen uptake (VO2 peak) are assessed, whereas the performance on walking tests usually is expressed as distance walked. The aim of the study was to investigate the relationship between an incremental shuttle walking test (ISWT) and two laboratory cycle tests in order to assess whether W peak could be estimated from an ISWT.
Ninety-three patients with moderate or severe COPD performed an ISWT, an incremental cycle test (ICT) to measure W peak and a semi-steady-state cycle test with breath-by-breath gas exchange analysis (CPET) to measure VO2 peak. Routine equations for conversion between cycle tests were used to estimate W peak from measured VO2 peak (CPET). Conversion equation for estimation of W peak from ISWT was found by univariate regression.
There was a significant correlation between W peak and distance walked on ISWT × body weight (r = 0.88, p < 0.0001). The agreement between W peak measured by ICT and estimated from ISWT was similar to the agreement between measured W peak (ICT) and W peak estimated from measured VO2 peak by CPET.
Peak exercise capacity measured by an incremental cycle test could be estimated from an ISWT with similar accuracy as when estimated from peak oxygen uptake in patients with COPD.
Measurements of exercise capacity are important and widely used in rehabilitation of patients with chronic obstructive pulmonary disease (COPD). Exercise testing in COPD varies from maximal laboratory tests, requiring advanced technical equipment, to simple field tests. Maximal laboratory tests are mostly constructed to measure peak exercise capacity (W peak), and/or peak oxygen uptake (VO2 peak) whereas field tests have been considered to reflect functional capacity [1–3]. The incremental shuttle walking test (ISWT) is a field test which is similar to the laboratory tests as it is externally paced and progressive . During ISWT there is a linear relationship between VO2 and walking speed, similar to the relationship between VO2 and work rate in incremental laboratory testing [5, 6]. Singh et al found, when comparing incremental laboratory treadmill test and ISWT, that VO2 peak could be estimated from distance walked on ISWT .
Singh et al compared two different walking tests (treadmill and ISWT) and it is unclear whether such a strong relationship would also be found between ISWT and cycle performance. In laboratory testing, the treadmill test evokes slightly higher ratings of VO2 peak than the cycle tests [7, 8], whereas different protocols of cycle tests usually result in similar VO2 peak but different W peak, depending on the slope of increased load during the test [9–11]. Body weight is an important contributor to the work load during walking, whereas it is of minor importance during cycling. Thus, the correlation between performance on timed walking tests and VO2 peak from a cycle test becomes stronger if distance walked is multiplied by body weight (distance × weight = work of walking at horizontal level) [12, 13]. Recent findings indicate that metabolic and ventilatory responses to walking may differ from the responses to cycling in patients with COPD [8, 14, 15]. In pulmonary rehabilitation many exercise programmes are conducted on ergometer cycles. Target training intensity is often expressed as a percent of W peak, which in turn usually is defined as the highest work rate tolerated during an incremental cycle test with 1-minute steps or ramp increments. From a known VO2 peak it is possible to estimate W peak [16, 17], and as VO2 peak can be estimated from an ISWT  it seems reasonable to assume that W peak could be estimated from ISWT through the estimated VO2 peak. However, using two conversion equations would make the results less reliable. Therefore, an estimation of W peak directly from the performance on ISWT would be preferable. This would be of clinical interest when expensive laboratory tests are not available.
The aim of the present study was to investigate whether W peak (assessed on a cycle ergometer) could be estimated from an ISWT in patients with moderate or severe COPD. For this purpose, comparisons of ISWT and two different cycle tests were made.
Ninety-three subjects with moderate or severe COPD according to the British Thoracic Society  were consecutively invited to take part in the study when being referred for training to the Physiotherapy Unit of the Pulmonary Section at the Akademiska Hospital, Uppsala, Sweden during 2001–2004. All were smokers or ex-smokers. The study was approved by the Medical Ethics Committee of Uppsala University and all subjects gave informed consent.
Inclusion criteria were COPD with forced expiratory volume in one second (FEV1) < 60% of predicted value and FEV1/VC (vital capacity) < 0.7 after bronchodilatation . Exclusion criteria were other diseases interfering with exercise such as ischemic coronary disease and musculo-skeletal problems.
Lung function was measured with a Masterlab Trans spirometer, Masterlab Body Plethysmograph and Masterlab Transfer (Erich Jaeger AG, Würzburg, Germany) in accordance with the ATS guidelines . Swedish reference values were used [20, 21].
Incremental shuttle walking test (ISWT) was performed in a level corridor. Two cones were placed 9 m apart comprising a 10 m track as described by Singh et al . Instructions to the subject and the pace of the test were played from a tape recorder. The test commenced at a speed of 30 m/min which then was increased by 10 m/min every minute. The subjects continued walking until they were not able to reach the next cone in time for the signal or got too exhausted to continue. The total distance walked was the main outcome of the test. Before and directly after walking, oxygen saturation (SpO2), heart rate, peak expiratory flow (PEF), perceived exertion (Borg RPE-scale)  and dyspnoea (Borg CR-10-scale)  were registered. The patients carried a pulse oximeter (Optovent Respons, Optovent, Linköping, Sweden) by a shoulder strap throughout the test. In 52 of the patients the test was identically repeated within a week.
Symptom-limited incremental cycle ergometer test (ICT) (Case 8000 Exercise Testing System, GE Medical Systems, Milwaukee, USA) with continuous ECG-registration was conducted to measure peak work load (W peak). The subjects started pedalling at 20 W and the load was then increased by 10 W every fulfilled minute until exhaustion. Heart rate, breathing frequency and SpO2 (Optovent Respons) were registered every minute during exercise. Systolic blood pressure, subjective ratings of perceived exertion and dyspnoea were recorded every other minute [22, 23]. All variables were measured before and 1, 2, 4 and 10 minutes after exercise.
A semi-steady-state cardiopulmonary exercise test with breath-by-breath gas exchange analysis (CPET) was performed, according to routines at our clinic, after resting for at least 30 min after the incremental cycle test (longer if needed for all resting parameters to be stable at pre-exercise levels). Measurements of heart rate, SpO2 and ratings of perceived exertion and dyspnoea were made identically to the incremental cycle test. To enable measurement of VO2 peak, the subjects wore a mask with a turbine for gas exchange analysis (Oxycon Sigma, Jaeger, Germany). Additionally, VCO2(carbon dioxide), minute ventilation (VE), respiratory quotient (RQ) and breathing frequency were measured with readings every 30 seconds. After recording steady-state measurements at rest (approximately 4 min of registration at rest) the patient started pedalling at 20 W. The load was kept constant until the ventilation and oxygen uptake reached a plateau, on average 3–4 min at each level, then the load was increased. To keep testing time within reasonable limits the load was increased until exhaustion by 5, 10, 20 or 30 W depending on the outcome of the first test (ICT).
The reason for discontinuing the cycle tests and the ISWT was stated at the end of each test.
The two different ergometer cycle tests were performed on the same day, but the lung function test and the ISWT were conducted on separate days. The three test days were separated by 1–3 resting days.
W peak was estimated from the measured VO2 peak on CPET by an equation derived from Åstrand : (VO2 peak × 1000-1- 0.1517) × 0.0134-1. The equation from Singh et al for estimating VO2 from ISWT  was multiplied by body weight to express ml/min instead of ml/min/kg: (4.19 + 0.025 × walking distance) × body weight.
Results were expressed as mean and standard deviation (SD) or 95% confidence interval (CI). For simple correlations Pearson's correlation coefficient was calculated. ANOVA and Student's t-test were calculated for paired comparisons, except for RPE and CR-10 scores where Friedman's ANOVA and Wilcoxon's signed rank test were used. Scatterplots as recommended by Bland and Altman were made for comparison of estimated and measured values. The level of significance was set at 5%.
Characteristics of the patients. Values expressed as mean ± SD and (range), n = 93.
Smoking habits, current/ex
VC, % predicted
FEV1, % predicted
MVV, % predicted
DLCO, % predicted
End-exercise values from the three tests: ISWT, ICT and CPET, n = 93.
Heart rate, beats/min
Perceived exertion, RPE
Fifty-two subjects repeated the ISWT within a week. The difference between the two tests was not significant for any variable measured, mean difference in walking distance being 9 ± 38 m or 3% ± 12% (p = 0.09). All calculations were therefore based on the first ISWT. This subgroup did not differ from the larger group in any baseline characteristics.
Correlations and estimated W peak
No significant differences were found between women and men regarding the relationship between the different exercise tests.
Comparisons with estimation based on previous findings
To test whether W peak could be estimated by use of the equation by Singh et al , we estimated VO2 peak from their equation and then converted this estimated VO2 to estimated W peak (as above) . Using the equation by Singh et al, VO2 peak was underestimated compared to our measurements (-171 ± 58 ml/min or -16 ± 15 %, p < 0.0001) which consequently resulted in an underestimation of W peak (-14 ± 13 W, p < 0.0001).
In the present study, ISWT distance × body weight was a good predictor of W peak measured by an incremental cycle test in patients with moderate or severe COPD. The fact that W peak estimated from ISWT was as accurate as W peak estimated from VO2 is of clinical importance, as the ISWT is much simpler and cheaper than a laboratory cycle test.
Our results confirm the findings of others that there is an excellent correlation between performance on ISWT and laboratory testing [5, 6, 24]. Although the correlation in our study between VO2 and distance walked was almost identical with the findings of Singh et al , applying their equation in our material resulted in a significant underestimation. One likely explanation to this could be the difference in the number of subjects, as in their study 19 subjects were included, compared to 93 subjects in the current study. As Singh et al performed two ISWT, and used the second test in their analysis, that might explain some of the inconsistency between their equation and ours, even though no significant improvement on retesting was found in our material. The majority of patients in the study by Singh et al were men while our study was dominated by women. Gender did not significantly affect the relationship between the different tests in our analysis, but this might need further investigation, as only 26% of our subjects were men. The fact that Singh et al compared two walking tests whereas we compared walking and cycling might also explain the difference to some extent, as there is a known difference in metabolic adaptations during walking and cycling [8, 15]. However, as VO2 peak measured on a treadmill test has been found to be higher than  or equal to  VO2 peak measured on a cycle test in patients with COPD, a regression equation derived from a treadmill test could be expected to overestimate VO2 peak on a cycle rather than the opposite.
All patients who were referred to pulmonary rehabilitation and fulfilled the inclusion criteria during the time of the study agreed to participate. Therefore, we conclude that the study is representative of a COPD-population referred to rehabilitation including physical training. The wide range of lung function impairment and age of the participants also enhances the relevance of our sample. In Sweden, women have caught up with men in the prevalence of and mortality in COPD . The majority of women referred to pulmonary rehabilitation during the time of the study is noticeable and we can only speculate that this might indicate that women were either more often offered referral to pulmonary rehabilitation by their physicians or were more likely to accept such an offer than the men.
Heart rate, RPE- and CR-10-scores were significantly lower in ISWT than in the cycle tests. The cycle tests were mainly limited by breathlessness and/or exertion whereas the ISWT was, in most cases, limited by the incapability to increase the speed of walking. During walking it is difficult to increase walking speed above a certain level. Some treadmill test protocols are therefore constructed to increase inclination rather than speed .
SpO2 decreased more by walking than cycling in the current study, quite in line with previous findings [8, 14, 24]. It has been speculated that the positional differences between walking and cycling could lead to less effective breathing during walking and thus more desaturation . In spite of the above differences between walking and cycling, our results demonstrated that in patients with moderate to severe COPD it was possible to estimate cycle performance from an ISWT quite as accurately as when estimating performance from one laboratory cycle test to the other. As equations work both ways, our findings also make it possible to estimate distance walked (and thereby walking speed) on ISWT from W peak. This could be useful in clinics where laboratory cycle testing is routine practice but exercise training prescription is mainly walking. As the outlier in our study illustrates, the ISWT can be expected to mimic performance in a symptom-limited cycle test only in subjects that reach their exercise limit by brisk walk, i.e. are unable to run, as otherwise the subjects would not be close to their peak capacity during the test. Being able to run is, however, a very rare condition in COPD-patients referred to rehabilitation, implying that this does not undermine the use of ISWT in patients with COPD.
We used the first ISWT (no training test) for our analyses. Due to the patient's poor condition or because of time constraints it is often not feasible to perform a training test in clinical practice. Thus, it was clinically relevant to present our calculations based on the first ISWT. This was also supported by the finding that the walking distance did not increase at the second test performed within a week. Control calculations were done by using results (not shown) from the second ISWT (n = 52) in our material and no differences were found. Significant difference in walking distance has previously been found on repeated testing with ISWT in patients with COPD [4, 28]. It is not clear why our results are inconsistent with previous studies regarding the repeatability of ISWT, but, as noted above, the number of subjects and gender distribution is somewhat different from the other studies, and this might affect the repeatability.
It could be argued that the cycle tests should have been performed in random order. The present design was chosen because it enabled us to adjust the progressive pattern of the CPET to keep exercise time within reasonable limits when W peak from the incremental cycle test was known. Although all subjects rested between the tests, we can not rule out the possibility that the order of the tests could have influenced the results. However, as peak heart rate, CR-10 scores and ratings of perceived exertion (RPE) were almost identical in the two cycle tests, and W peak could be estimated from VO2 peak as expected , we consider this error as relatively small.
Being able to estimate W peak by ISWT for clinical purposes does not make ISWT a perfect substitute for the incremental cycle test or other forms of laboratory tests. There is, of course, some variation between estimated and measured peak values on an individual basis, but most importantly the safety aspect of the laboratory tests is beyond the ISWT. Therefore, for COPD patients considered at increased risk for cardiovascular hazards, patients that have never had proper cardiologic assessment or patients with uncertain diagnosis, laboratory tests must be considered as the first choice. However, for COPD-patients of minor risk of cardiovascular incidents, and where laboratory tests are scarce, the ISWT could be used as an alternative for estimating W peak.
Peak exercise capacity measured by an incremental cycle test could be estimated from an ISWT with similar accuracy as when estimated from peak oxygen uptake.
Special thanks to Ulrike Spetz-Nyström for her valuable help with testing. This paper was supported by grants from the Swedish Heart-Lung Foundation, The County Council of Uppsala and various funds at Uppsala University.
- Clinical exercise testing with reference to lung diseases: indications, standardization and interpretation strategies. ERS Task Force on Standardization of Clinical Exercise Testing. European Respiratory SocietyEur Respir J 1997,10(11):2662–2689.Google Scholar
- ATS statement: guidelines for the six-minute walk testAm J Respir Crit Care Med 2002,166(1):111–117.Google Scholar
- McGavin CR, Gupta SP, McHardy GJ: Twelve-minute walking test for assessing disability in chronic bronchitis. Br Med J 1976,1(6013):822–823.View ArticlePubMedPubMed CentralGoogle Scholar
- Singh SJ, Morgan MD, Scott S, Walters D, Hardman AE: Development of a shuttle walking test of disability in patients with chronic airways obstruction. Thorax 1992,47(12):1019–1024.View ArticlePubMedPubMed CentralGoogle Scholar
- Singh SJ, Morgan MD, Hardman AE, Rowe C, Bardsley PA: Comparison of oxygen uptake during a conventional treadmill test and the shuttle walking test in chronic airflow limitation. Eur Respir J 1994,7(11):2016–2020.PubMedGoogle Scholar
- Onorati P, Antonucci R, Valli G, Berton E, De Marco F, Serra P, Palange P: Non-invasive evaluation of gas exchange during a shuttle walking test vs. a 6-min walking test to assess exercise tolerance in COPD patients. Eur J Appl Physiol 2003,89(3–4):331–336.View ArticlePubMedGoogle Scholar
- Page E, Cohen-Solal A, Jondeau G, Douard H, Roul G, Kantelip JP, Bussiere JL: Comparison of treadmill and bicycle exercise in patients with chronic heart failure. Chest 1994,106(4):1002–1006.View ArticlePubMedGoogle Scholar
- Christensen CC, Ryg MS, Edvardsen A, Skjonsberg OH: Effect of exercise mode on oxygen uptake and blood gases in COPD patients. Respir Med 2004,98(7):656–660.View ArticlePubMedGoogle Scholar
- Myers J, Buchanan N, Walsh D, Kraemer M, McAuley P, Hamilton-Wessler M, Froelicher VF: Comparison of the ramp versus standard exercise protocols. J Am Coll Cardiol 1991,17(6):1334–1342.View ArticlePubMedGoogle Scholar
- Zhang YY, Johnson MC 2nd, Chow N, Wasserman K: Effect of exercise testing protocol on parameters of aerobic function. Med Sci Sports Exerc 1991,23(5):625–630.View ArticlePubMedGoogle Scholar
- Revill SM, Beck KE, Morgan MD: Comparison of the peak exercise response measured by the ramp and 1-min step cycle exercise protocols in patients with exertional dyspnea. Chest 2002,121(4):1099–1105.View ArticlePubMedGoogle Scholar
- Chuang ML, Lin IF, Wasserman K: The body weight-walking distance product as related to lung function, anaerobic threshold and peak VO2 in COPD patients. Respir Med 2001,95(7):618–626.View ArticlePubMedGoogle Scholar
- Carter R, Holiday DB, Stocks J, Grothues C, Tiep B: Predicting oxygen uptake for men and women with moderate to severe chronic obstructive pulmonary disease. Arch Phys Med Rehabil 2003,84(8):1158–1164.View ArticlePubMedGoogle Scholar
- Man WD, Soliman MG, Gearing J, Radford SG, Rafferty GF, Gray BJ, Polkey MI, Moxham J: Symptoms and quadriceps fatigability after walking and cycling in chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2003,168(5):562–567.View ArticlePubMedGoogle Scholar
- Palange P, Forte S, Onorati P, Manfredi F, Serra P, Carlone S: Ventilatory and metabolic adaptations to walking and cycling in patients with COPD. J Appl Physiol 2000,88(5):1715–1720.PubMedGoogle Scholar
- Åstrand PO, Rodahl K: Textbook of work physiology: Physiological Bases of Exercise. 3rd edition. New York , McGraw-Hill; 1986.Google Scholar
- Wallin L, Brudin LH: Physical working capacity determined by different types of bicycle exercise tests. Clin Physiol 1988,8(5):529–537.View ArticlePubMedGoogle Scholar
- BTS guidelines for the management of chronic obstructive pulmonary disease. The COPD Guidelines Group of the Standards of Care Committee of the BTSThorax 1997, 52 Suppl 5:S1–28.Google Scholar
- Standardization of spirometry--1987 update. Official statement of American Thoracic SocietyRespir Care 1987,32(11):1039–1060.Google Scholar
- Hedenstrom H, Malmberg P, Agarwal K: Reference values for lung function tests in females. Regression equations with smoking variables. Bull Eur Physiopathol Respir 1985,21(6):551–557.PubMedGoogle Scholar
- Hedenstrom H, Malmberg P, Fridriksson HV: Reference values for lung function tests in men: regression equations with smoking variables. Ups J Med Sci 1986,91(3):299–310.View ArticlePubMedGoogle Scholar
- Borg G: Perceived exertion as an indicator of somatic stress. Scand J Rehabil Med 1970,2(2):92–98.PubMedGoogle Scholar
- Borg GA: Psychophysical bases of perceived exertion. Med Sci Sports Exerc 1982,14(5):377–381.View ArticlePubMedGoogle Scholar
- Turner SE, Eastwood PR, Cecins NM, Hillman DR, Jenkins SC: Physiologic responses to incremental and self-paced exercise in COPD: a comparison of three tests. Chest 2004,126(3):766–773.View ArticlePubMedGoogle Scholar
- Mathur RS, Revill SM, Vara DD, Walton R, Morgan MD: Comparison of peak oxygen consumption during cycle and treadmill exercise in severe chronic obstructive pulmonary disease. Thorax 1995,50(8):829–833.View ArticlePubMedPubMed CentralGoogle Scholar
- Nagle FJ, Balke B, Naughton JP: Gradational step tests for assessing work capacity. J Appl Physiol 1965,20(4):745–748.PubMedGoogle Scholar
- Eiser N, Willsher D, Dore CJ: Reliability, repeatability and sensitivity to change of externally and self-paced walking tests in COPD patients. Respir Med 2003,97(4):407–414.View ArticlePubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.