Abstract
BACKGROUND: The incremental shuttle walk test (ISWT) has been widely used to assess exercise capacity of subjects with COPD. However, to date, no studies have assessed its reliability in the asthmatic population. This study aimed to assess the test-retest reliability of ISWT for the distance walked (ISWD) and cardiorespiratory and metabolic responses in adult subjects with asthma.
METHODS: This was a cross-sectional observational study. Thirty-four subjects of both genders, 18–45 y old, with a diagnosis of controlled asthma, were recruited from March 2012 to December 2015. The subjects performed 3 ISWTs on different days, with a minimum interval of 48 h and a maximum of 1 week between the second and third ISWT. For the reliability analysis, the values of the second and third ISWTs were used, eliminating the influence of any learning effect from the first test.
RESULTS: The intraclass correlation coefficients were > 0.75 (P < .001) for ISWD (m), speed (km/h),oxygen uptake (V̇O2), ventilatory equivalent carbon dioxide (V̇E/V̇CO2), and minute ventilation (V̇E) at the ISWT peak. The Bland-Altman plot presented a mean error close to zero, and measurement distribution was within acceptable limits of variation.
CONCLUSION: The ISWT presented excellent reliability for the ISWD and metabolic responses. The cardiorespiratory responses in the ISWTs presented good reliability. We concluded that the ISWT was reliable for young adult subjects with controlled asthma.
Introduction
Asthma is a common respiratory disease characterized by chronic or recurrent airway inflammation associated with variable expiratory flow limitation and bronchial hyper-responsiveness.1 Patients with chronic lung diseases are more likely to present with lower exercise tolerance, explained not only by the air-flow limitation, but also by peripheral muscle dysfunction.2 As a consequence, many patients adopt a sedentary lifestyle, predisposing them to early fatigue and exercise intolerance.1
Regular physical training seems to counteract the consequences of the disease and is a cornerstone in the rehabilitation process. It has been suggested that increased exercise capacity may play an important modulatory role in reducing the degree of systemic inflammation in asthma,3 along with the use of corticosteroids.4 Regular exercise training may also act beneficially by improving asthma-related psychosocial factors5 and reducing the risks of exacerbation in asthmatic patients.6
Assessment of exercise capacity is fundamental to determine the systemic consequences in patients with asthma, as well as to propose personalized intervention strategies to promote improved exercise capacity. The cardiopulmonary exercise test is considered the standard method to determine the presence and etiology of exercise intolerance; however, this test is expensive, requires skilled staff, is not widely available, and is associated with risks inherent to its implementation.7
Field tests, on the other hand, have been widely used to evaluate exercise capacity because they are less expensive, require minimal staff for execution, and are considered safer in that only submaximal stress is induced during the tests.8 The incremental shuttle walk test (ISWT)9 is an externally paced walk test widely used for the evaluation of exercise capacity. This test is considered to be a reliable, valid, and safe instrument for assessing exercise capacity in subjects with COPD.10
Some studies of the ISWT have shown that the peak oxygen uptake (V̇O2) presented moderate to strong correlation with measurements of maximal exercise performance in the cardiopulmonary exercise test.11 Thus, the reliability of physiologic variables during the ISWT is important because it provides information about exercise capacity and allows elaboration of a more specific therapeutic strategy in a physical training program. However, although the ISWT is described as able to evaluate the exercise capacity of patients with cardiorespiratory diseases, there is less clinical evidence regarding the application of ISWT in asthmatic patients. For example, Dyer and colleagues12 studied a small sample of elderly subjects with asthma and verified that the reliability for the distance walked in the ISWT was weak. Another study recently conducted by Costa and colleagues13 showed that the ISWT was reliable to assess the functional capacity of subjects with difficult-to-control asthma.
In addition to the low cost and easy application of the ISWT, it is also safe for assessing exercise capacity in patients with asthma. The test is controlled by symptoms, and the progressive increase in walking speed allows important information about physiological responses during physical exercise, including the possibility of triggering exercise-induced bronchospasm, which is an inherent condition in many asthma patients, to be collected.
Therefore, the aim of our study was to analyze the test-retest reliability of the distance walked in the ISWT (ISWD) and the cardiorespiratory and metabolic responses in young adults with controlled asthma. We hypothesized that the ISWT would present good reliability for the distance walked and the cardiorespiratory and metabolic responses in a sample of subjects with asthma.
QUICK LOOK
Current knowledge
Air-flow limitation and peripheral muscle dysfunction are responsible for reducing exercise tolerance in patients with chronic respiratory diseases, inducing adoption of a more sedentary lifestyle, predisposing them to early fatigue and exercise intolerance. It is important to evaluate exercise capacity to determine the systemic consequences in patients with asthma and to propose personalized intervention strategies to promote improvement in exercise capacity.
What this paper contributes to our knowledge
The distance walked and cardiorespiratory and metabolic variables measured with the incremental shuttle walk test (ISWT) were shown to be reliable for the assessment of the exercise capacity of adult subjects with controlled asthma. In addition, the ISWT can be considered as a low-cost and safe instrument to assess the exercise capacity in this population.
Methods
Study Design and Participants
This was an observational cross-sectional study (the reporting of this study was guided by the STROBE statements)14 conducted by the Spirometry and Respiratory Physiotherapy Laboratory and Cardiopulmonary Physiotherapy Laboratory at the Federal University of São Carlos. The study included 34 subjects with asthma of both genders, 18–45 y old, recruited from March 2012 to December 2015. Dissemination of the study proposal was performed through posters placed on the university campus, local radio, television, and newspapers advertisements, or after physician evaluation and referral. Subjects were contacted by telephone and email, and they were invited to participate in the study if they met the inclusion criteria.
Subjects with a clinical and functional diagnosis compatible with asthma were eligible to participate in the study. Inclusion criteria involved subjects ≥ 18 y with a previous asthma diagnosis. Asthmatic subjects with different disease severities could be engaged in the study, with the exception of individuals who presented with severe asthma, regardless of whether it was controlled. Subjects were required to have been evaluated by the physician in charge and to have the disease classified as controlled according to the criteria established by the Global Initiative for Asthma.3 Individuals with a medical history of asthma exacerbation < 3 weeks before the scheduled tests were excluded from the study.
Individuals presenting with other respiratory, metabolic, or cardiovascular diseases that could cause or aggravate the sensation of dyspnea during efforts were not eligible for the study. Other exclusion criteria included any other contraindications to perform the cardiopulmonary exercise test7 such as musculoskeletal, neurological, arterial, rheumatological, renal or liver disorders; diabetes mellitus with diabetic neuropathy; difficulty understanding and/or adhering to the study procedures; illegal drug addiction; and pregnancy.
All subjects were informed about all experimental procedures and signed a consent form of a broader study, which included all evaluations and analyzes used in this study. The study was approved by the university human ethics committee (decision number 018/2012).
Experimental Procedures
Subjects with asthma were evaluated on 3 different days with a minimum interval of 48 h between the first and second days. All procedures were carried out during the afternoon. On the first day of the protocol, subjects underwent clinical and physiotherapeutic evaluations as well as pre- and post-bronchodilator pulmonary function tests. Additionally, on the same day, they underwent the first ISWT to become familiar with the test. Subjects received the following recommendations prior to the test execution: avoid consumption of stimulating beverages (caffeine) for 48 h before the test, do not perform physical activity for 24 h before the test, consume light meals on the day of the test, and sleep properly the night before the test (ie, at least 8 h). In this study, short-acting bronchodilators were suspended for at least 6 h and long-acting bronchodilators for 12 h before all assessments because heart rate variability was included in the evaluations. Previous results regarding heart rate variability in asthmatic subjects have already been published by our group.15 However, ß2-agonist bronchodilators could be administered at any time, under a previous medical prescription (investigator and pulmonologist linked to the study), to revert the bronchoconstriction response after exercise if the volunteer presented significant dyspnea or if FEV1 had not returned to within 10% of baseline when the subject was ready to leave the laboratory.16 On the second day of the protocol, the second ISWT was conducted and, after a minimum of 48 h and a maximum of 1 week, subjects performed the third ISWT. All tests were conducted by the same assessor in a level corridor. The portable cardiopulmonary exercise testing system (Oxycon Mobile, Mijnhardt/Jäger, Würzburg, Germany) was used to capture breath-by-breath ventilatory, cardiorespiratory, and metabolic variables. Furthermore, subject responses were assessed at rest, during exercise, and in the recovery period. To eliminate the learning effect that might influence the first test, only the data from the second and third ISWTs were used in the reliability analysis.
Spirometry
The pulmonary function test was performed with a portable ergospirometry system (Oxycon Mobile), with an oro-nasal mask used as an interface. Technical procedures, acceptability, and reliability criteria followed the American Thoracic Society/European Respiratory Society (ATS/ERS) guidelines.17 Spirometric indices were presented in absolute values and as a percentage of the reference values obtained for the Brazilian population.18
ISWT
The ISWT consists of walking along a 10-m corridor in a space delimited by 2 cones. Walking was performed at a progressive speed, which increased by 0.17 m/s every minute after a sonorous beep. Subjects were instructed, every minute, to increase their walking speed during the 12 stages of the test to reach maximum effort. The test was ended when subjects reported to the examiner an intense sensation of dyspnea, pain, lower limb fatigue,19 or any other symptoms that could prevent them from continuing the tests. In addition, the test was also interrupted if the subject could no longer maintain the required speed for 2 consecutive laps, considered as being > 0.5 m away from the cone when the beep sounded.9 If subjects reached the cone before the beep, they were instructed to remain close to the cone and wait for the signal.
Subjects were monitored at rest, immediately after the test, and in the fourth minute of passive recovery. Heart rate was monitored with a cardiofrequencimeter (RS800CX, Polar Electro Co. Ltd., Kempele, Oulu, Finland), blood pressure (mm Hg) was assessed via auscultation, peripheral oxygen saturation was evaluated with a portable oximeter (Nonin 8500A, Hand Held Pulse Oximeter, Plymouth, Minnesota), and the subjective sensation of dyspnea and lower limb fatigue were monitored according to the Borg dyspnea scale.19 In addition, the following cardiorespiratory and metabolic variables were recorded at rest, during the test, and during the recovery period: expiratory minute volume (V̇E), V̇O2; absolute value and corrected by the body mass, and carbon dioxide production (V̇E/V̇CO2). The highest values recorded in the final 6 s of the peak of all tests were selected for further analysis. The percentage of predicted distance walked for the ISWT was calculated as previously published.20
Statistical Analysis
Statistical analysis was performed with the Statistical Package for Social Sciences (SPSS) software for Windows, version 17.0 (IBM, Armonk, New York). Normality of the data was verified with the Shapiro-Wilk test. Data are organized in tables and graphs. Results are presented as mean± SD or median (interquartile range) for parametric and nonparametric variables, respectively. The significance level was assumed at P < .05.
The sample size was estimated as at least 19 asthmatic subjects with the objective of performing the proposed reliability analysis,21 considering α < 0.05 and β < 0.2, a number of repetitions of 2 tests, the null hypothesis of intraclass correlation coefficient (ICC) < 0.7, and the expected hypothesis of ICC > 0.9. A sample size of 30 subjects was considered moderate and acceptable according to the Consensus-Based Standards for the Selection of Health Status Measurement Instruments checklist.22 The expected ICC was consistent with the mean ICC for the ISWD as verified in 7 previous studies that tested its reliability (range 0.80–0.99).13,23–28
Variance analysis comparing the ISWD and the cardiorespiratory and metabolic responses between the second and third ISWT was performed with a paired t test or its corresponding nonparametric test, the Wilcoxon test.
To test reliability, we used the ICC, also known as the reproducibility coefficient, which is an estimate of the fraction of the total variability of the measure due to individual variations. Reliability was analyzed using the ICC, classifying the values as low (ICC < 0.4), good (ICC > 0.4 and < 0.75), and excellent (ICC > 0.75).29 Furthermore, we calculated the coefficient of variation, which is the SD expressed as a percentage of the mean (coefficient of variation = ISWD/mean). The error analysis between the second and third test was conducted using the mean error and its 95% limits of agreement presented in a Bland-Altman plot.
Results
Demographic and pulmonary function data were obtained on 34 subjects (Table 1). Ninety-one people were initially contacted, of whom 46 were ineligible for various reasons, leaving 45 subjects in the sample. During the study, 11 subjects were excluded, resulting in the 34 subjects who participated in the study (Fig. 1). Medications being taken included short-acting β2-adrenergics (n = 13), long-acting β2-adrenergics (n = 8), anticholinergic (n = 2), inhaled corticosteroids (n = 16), oral corticosteroids (n = 1), nasal corticosteroids (n = 3), antihistamines (n = 1), and theophylline (n = 1). Nine subjects were not taking any medication.
Demographic, Anthropometric and Spirometric Characteristics of the Subjects With Asthma
Flow chart.
Thus, ISWD (ICC = 0.85), speed, V̇O2 (ICC = 0.86) (absolute value (ICC = 0.95) and corrected for body mass (ICC = 0.88), V̇E/V̇CO2 (ICC = 0.84) and V̇E = 0.88) presented excellent reproducibility, with ICC values > 0.75 and heart rate (ICC = 0.68), being considered a good reproducibility, presenting a value of 0.4 < ICC < 0.75 (Table 2).
Reliability Analysis for ISWD, Cardiorespiratory and Metabolic Responses Between the Second and Third ISWT
Considering other reliability aspects, no statistically significant differences were observed between the ISWDs in second (ISWT-2) and third (ISTW-3) ISWTs (P = .864) and between the percentage of predicted ISWD for ISWT-2 and ISWT-3 (P = .880). A statistically significant difference in the heart rate at rest was verified between ISWT-2 and ISWT-3, although this result was not clinically relevant (Table 3).
Comparative Assessment of the Second and Third ISWT for ISWD, Cardiorespiratory, and Metabolic Responses
Subjects walked an average distance of 541.7 m (95% CI 432.8–650.6) in the ISWT-2 and 540.0 m (95% CI 431.1–648.9) in the ISWT-3. The Bland-Altman plot for the ISWD showed a mean error in the test-retest comparison of 0.69 m, with limits of agreement of −117.9 m to 119.3 m. This indicates that in the retest, 95% of subjects should be able to walk a distance ranging from −117.9 m to 119.3 m of the mean ISWT-2 (Fig. 2).
Reliability by Bland-Altman plot for subjects with asthma. Performance between the second and third incremental shuttle walk test (ISWD). Mean error = 0.69.
The heart rate at ISWT peak was an average of 151.9 beats/min (95% CI 116.6–187.2) in the ISWT-2 and 148.2 beats/min (95% CI 113.0–183.5) in the ISWT-3. In the Bland-Altman analysis, a mean error of −4.4 beats/min with agreement limits of −41.9 to 33.1 beats/min between the ISWT-2 and ISWT-3 (Fig. 3) was found for peak heart rate. This indicates that in the retest, 95% of subjects who underwent ISWT are expected to present a heart rate variation ranging from −41.9 to 33.1 beats/min of the mean ISWT-2 heart rate.
Reliability by Bland-Altman plot for subjects with asthma. A: Peak heart rate (HR), mean error −4.4. B: Peak oxygen uptake (V̇O2), mean error −067.
Peak V̇O2 presented a mean value of 25.9 mL/kg/min (95% CI 19.7–32.1) in the ISWT-2 and 25.3 mL kg/min (95% CI 19.1–31.5) in the ISWT-3. In the Bland-Altman analysis, a mean error of −0.67 mL/kg/min was observed for peak V̇O2, with limits of agreement of −7.2 to 5.9 mL/kg/min between the ISWT-2 and ISWT-3 (Fig. 3). This indicates that in the retest, 95% of subjects who underwent ISWT are expected to present a V̇O2 value in the range of −7.2 to 5.9 mL/kg/min of the mean ISWT-2 V̇O2.
Discussion
The main outcomes of this study were that the ISWT presented excellent reliability for the ISWD, speed, V̇O2, V̇E/V̇CO2, and V̇E; the only exception was heart rate, which presented good reliability. Therefore, the ISWT can be considered a safe and reliable instrument to evaluate the exercise capacity of patients with asthma; however, evidence of the validity of the ISWT for the evaluation of cardiorespiratory and metabolic responses, mainly peak V̇O2, in subjects with asthma is scarce and this topic requires further investigation.
In the study of Dyer and colleagues12 the reliability of the ISWT was evaluated in a small sample of elderly subjects with asthma who performed 2 ISWTs within 2 weeks. However, all of the subjects evaluated in this study completed the tests, contrary to the study by Dyer and colleagues,12 where 12% of the subjects who were evaluated did not complete the tests; nonetheless, the investigators concluded that the ISWT is feasible and reliable to evaluate elderly patients with or without air flow obstruction.
Costa and colleagues13 evaluated the reliability of the ISWT in 45 subjects with difficult-to-control asthma and performed 2 ISWTs on the same day with a 20-min rest period between tests. The investigators concluded that the ISWT is reliable and can be used to evaluate the exercise capacity of patients with difficult-to-control asthma.
The ISWT test-retest reliability performed in our study resulted in ICC > 0.75 for distance walked, indicating that the ISWD in the ISWT is reliable. Seven previous studies tested the test-retest reliability for the ISWD, with ICCs ranging from 0.80 to 0.99.13,23–28 These results are consistent with study findings in other chronically ill populations.
Considering the coefficient of variation results of the ISWD, the observed values were low (0.08, or 8.0%), indicating homogeneous variability between means and a dependence on the application method over time. One study verified the ISWT reliability in subjects with chronic heart failure30 and found coefficient of variation values similar to those found in our study (6.9); however, the sample size was considered small.
Costa and colleagues13 conducted a study with subjects with difficult-to-control asthma and found a mean error of −12.7 m with limits of agreement of 37.8 to −63.2 m, whereas the mean error value of our study was close to zero (0.69 m). Although the limits of agreement of our study were higher, the distributions were within acceptable limits, unlike the findings of the study in question, which presented 5 points below the acceptable lower limit.
Pepera and colleagues,26 when studying subjects with clinically stable cardiovascular disease, found a mean error of −7 m with limits of agreement (Bland-Altman) of −203 to 189 m. Hence, the magnitude of these values (386 m) was higher than that found in our study, which was 241.4 m. Jürgensen and colleagues28 conducted a study with obese women, finding a mean error of −5.9 with limits of agreement of −82.0 to 70.2 m, and the magnitude of these values was 148.2 m, lower than the value obtained in our study (241.4 m).
Another study27 conducted with subjects with bronchiectasis without cystic fibrosis found a mean error of −4.4 m and limits of agreement between −57 and 48 m, which were lower values than those found in our study. The tests were performed on the same day with a 30-min interval between tests. van Bloemendaal and colleagues31 conducted a study with subjects after a stroke and they found a mean error of 27.4 and limits of agreement of −272.3 to 327.0 m; however, the authors did not clearly state the time interval between tests.
For ISWT speed we found an ICC of 0.86, similar to the results observed by Jürgensen and colleagues28 (ICC = 0.84). The coefficient of variation observed in our analysis was low (0.04, or 4.0%), indicating homogeneity between the measurements.
In the analysis of cardiorespiratory responses obtained during the ISWT, the heart rate in the test retest presented an ICC of 0.68. de Camargo and colleagues27 and Costa and collegues13 found ICCs of 0.92 and 0.97 in the post-exercise, respectively; however their experimental procedures differed from our study. The coefficient of variation value of the heart rate in this study was low (0.12, or 12.0%), indicating homogeneity of our results, which is similar to that found by Green and colleagues.30
We found a heart rate mean error value of −4.4 beats/min, which is close to zero, and limits of agreement ranging from −41.9 to 33.1 beats/min. These values were similar to the results obtained by Jürgensen and colleagues28 and by Booth and Adams.32 The expected heart rate was in the range of 110–190 beats/min, which is similar to that found by Jürgensen and colleagues28 (ie, 120–180 beats/min), but different from the one obtained by Booth and Adams32 (ie, 80–140 beats/min). In this study, the distributions of the heart rate occurred within acceptable limits of variation.
Regarding the metabolic variables, we found an ICC of 0.88 for the test-retest for V̇O2 (mL/kg/min); the coefficient of variation was low (0.09, or 9.0%), and there was no significant difference (P = .28) between the V̇O2 in the ISWT-2 and ISWT-3. The mean error was close to zero (−0.67), and the limits of agreement were between −7.2 and 5.9. The distributions occurred within acceptable limits of variation, thus the 2 measurements had similar results. These finding are in accordance with Jürgensen and colleagues,28 except for the limits of variation of their results.
For V̇O2 (mL/min), we found an ICC of 0.95 in the test-retest comparison, and the coefficient of variation was low (0.08, or 8.0%), indicating homogeneity between the measurements. We found an ICC of 0.84 for V̇E/V̇CO2 and 0.88 for V̇E (L/min) in the test-retest comparison, with the coefficients of variation (0.06, or 6.0%) and (0.12, or 12.0%), respectively, indicating good homogeneity. Thus, our results for the variables above mentioned (except V̇E/V̇CO2) were in accordance with the findings of Jürgensen and colleagues.28
We consider the difficulty of comparing our findings with those of other studies to be a limitation of our study, because we found experimental procedures that varied regarding the intervals (time and days) between tests. In this study, we determined that the ISWT-2 and ISWT-3 should be conducted on different days with a maximum interval of 7 d because we could not predict whether the ISWT would provoke exercise-induced bronchospasm.
Our results are clinically relevant because the distance walked and the cardiorespiratory and metabolic responses were reliable during ISWT in the test-retest analysis. According to Singh and colleagues,11 there is a significant learning effect between the first 2 ISWTs, with a mean difference of 9–25 m, and this magnitude is sufficient to recommend that 2 tests could be performed on the first exposure, attenuating the learning effect over subsequent repetitions of the test. However, Dyer and colleagues33 suggested that after longer periods (> 8 weeks) the learning effect may return to a level similar to that recorded in the first exposure to the test. In our study, we assumed that the tests could be compared if performed after a 1-week interval.
Conclusion
ISWT presented excellent reliability for the ISWD and metabolic responses. The cardiorespiratory responses in ISWT presented good reliability. We conclude that ISWT was reliable for the assessment of the exercise capacity of young adults with controlled asthma.
Acknowledgments
The authors are grateful for the contributions made by Juliano Ferreira Arcuri.
Footnotes
- Correspondence: Ivana Gonçalves Labadessa, Department of Physiotherapy, Spirometry and Respiratory Physiotherapy Laboratory, UFSCar, Rod. Washington Luis, km 235, cx. Postal 696, CEP 13565-905, São Carlos, São Paulo, Brazil. E-mail: ivanalaba{at}yahoo.com.br.
The authors disclose relationships with the Fundação de Amparo a Pesquisa do Estado de São Paulo and the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior.
- Copyright © 2019 by Daedalus Enterprises
