Abstract
Background: Chronic obstructive pulmonary disease (COPD) can lead to severe disability as the disease advances. The 6-minute walk test (6MWT) is commonly used to measure functional capacity in COPD patients and has three potential outcomes; walking distance, oxygen desaturation, and self-perceived dyspnea assessed by the Borg scale, all reflecting different aspects of COPD. The aim of this study was to identify predictors of all 3 outcomes of 6MWT in patients with COPD. Methods: 370 COPD patients, aged 40–75 yrs, were included from the first phase of the Bergen COPD cohort study. They were examined with spirometry, bioelectrical impedance measurements, 6MWT, Center for Epidemiologic Studies of Depression (CES-D) Scale, Medical Research Council (MRC) dyspnea scale, Charlson index for co-morbidities, self-reported physical activity questionnaire, plasma levels of C-reactive protein (CRP) and arterial blood gases. Results: Significant predictors in the multivariate analyses were sex, age, FEV1 in% predicted, symptoms of dyspnea (MRC), co-morbidities (Charlson Index) and self-reported physical activity for walking distance, FEV1 in% predicted and PaO2 for oxygen desaturation, and body composition, smoking and co-morbidities for self-perceived dyspnea assessed by the Borg scale. Conclusion: Several COPD characteristics have predictive value for the 6MWT, and some COPD characteristics are more strongly related to specific 6MWT outcomes than others.
| Abbreviations | ||
| ATS | = | American Thoracic Society |
| CES-D | = | Centre for Epidemiologic Studies Depression scale |
| COPD | = | Chronic Obstructive Pulmonary Disease |
| CRP | = | Plasma C - reactive protein |
| ERS | = | European Respiratory Society |
| FEV1 | = | Forced Expiratory Volume in 1 second |
| FFMI | = | Fat Free Mass Index |
| FMI | = | Fat Mass Index |
| FVC | = | Forced Vital Capacity |
| MRC | = | Medical Research Council scale |
| PaCO2 | = | Partial pressure of carbon dioxide in arterial blood |
| PaO2 | = | Partial pressure of oxygen in arterial blood |
| Post-BD | = | Post-bronchodilator |
| RV | = | Residual Volume |
| 6MWT | = | 6-minute walk test |
| DlCO | = | Diffusing capacity of the lung for carbon monoxide. |
Introduction
Chronic obstructive pulmonary disease (COPD) is fast becoming one of the most common diseases in the world, affecting up to 10% of the adult population if defined by spirometry only (Citation1). With increasing loss of lung function, disability increases. However, this correlation is surprisingly weak (Citation2, 3). In fact, one of the important unexplained features of COPD is the large difference between patients in disease progression, symptoms, frequency of exacerbations, and disability.
Arguably, what matters most to the patient is functional capacity. Measures of functional capacity are increasingly used to monitor disease progression (Citation4), to assess capacity improvements following pulmonary rehabilitation programs (Citation5), as outcome measures in clinical trials (Citation6), and as predictors of mortality (Citation7). The most commonly used test of functional capacity is the 6-minute walk test (6MWT).
The 6MWT has 3 potential outcomes; walking distance (6MWD), oxygen desaturation and self-perceived dyspnea assessed by the Borg scale (Citation8).
These three outcomes encompass many pathophysiological and perhaps psychological mechanisms in COPD patients (Citation6, Citation9). Yet, surprisingly little is published on which COPD disease characteristics are related to these 3 outcome measures.
The best-described outcome in the 6MWT is 6MWD. In univariate analyses, 6MWD has been associated with exacerbations (Citation10), disease severity according to Global Initiative for Obstructive Lung Disease (GOLD) criteria (Citation11), transfer factor for carbon monoxide (TlCO) (Citation12), and emphysema and air trapping on CT scans (Citation13). 6MWD and change in Borg score has been associated with fat free mass index (FFMI) (Citation14).
Oxygen desaturation has been associated with gender, resting PaO2, and 6MWD in univariate analyses, but not with body mass index (BMI), forced expiratory volume in one second (FEV1) and forced vital capacity (FVC) (Citation7).
So far, only two studies have attempted multivariate analyses to assess the relative contribution of several potential predictors of 6MWD; a report from the National Emphysema Treatment Trial (NETT) (Citation15) and a report from the Evaluation of COPD Longitudinally to Identify Predictive Surrogate Endpoints (ECLIPSE) study (Citation9).
In the NETT report, several clinical variables were associated with 6MWD (Citation15), and in the ECLIPSE report, the outcome was 6MWD below or above 350 meters (Citation9). Neither of these two reports included measures of physical activity or systemic inflammation, and neither examined the outcome measures oxygen desaturation and change in Borg score. In the current report, we expand upon current knowledge by presenting multivariate analyses of several important COPD disease characteristics on all three outcome measures of the 6MWT, in a Norwegian cohort of 370 COPD patients.
Methods
Study population
The selection of the study population and the inclusion and exclusion criteria have been described in detail (Citation16). Briefly, 433 COPD subjects aged 40–75 years were included in the Bergen COPD Cohort Study at baseline in 2006. All patients had a clinical diagnosis of COPD; a smoking history of more than 10 pack-years; a post-bronchodilator (post-BD) FEV1/FVC ratio of less than 0.7, and post-BD FEV1% predicted <80%.
All COPD patients were encouraged to undergo a 6MWT. 389 patients attempted the test, including 13 patients using supplemental oxygen. Five patients were excluded because they were unable to finish the test, 3 due to ‘heavy breathing’, and 2 for ‘other reasons’, and for 1 patient we had missing data on actual completion. Thus, a total of 370 COPD patients that were able to complete the test, without using supplemental oxygen or having to stop before finishing the 6MWT, were included in the current study. Participation in the study was voluntary. Written information was provided and written consent was obtained from each participant prior to inclusion. The Regional Committee of Medical Research Ethics (REK) approved the study (REK Study number 165.08).
Data collection
All subjects were clinically interviewed and examined by a physician, and all other examinations were conducted by trained study engineers. FEV1 and FVC were measured using a Viasys-Jaeger Masterscope (Viasys, Hoechberg, Germany) before and after inhalation of 0.4 mg Salbutamol, and Norwegian pre-BD reference values were used (Citation17). Residual volume (RV) was measured with a Viasys Sensormedics Encore body plethysmograph (Viasys, Hoechberg, Germany) and reference equations recommended by the 1995 ATS/ERS workshop were used (Citation18).
Body composition, hence fat mass (kg) and fat free mass (kg), was determined using bioelectrical impedance measurements with a Bodystat 1500 (Bodystat Ltd, Douglas, Isle of Man, UK). Fat mass index (FMI) and fat free mass index (FFMI) were calculated as the mass (kg) divided by the squared height (m2). Plasma C-reactive protein (CRP) was measured with a high-sensitivity enzyme immunoassay (EIA) (Citation19). Medical Research Council (MRC) breathlessness scale and Center for Epidemiologic Studies Depression Scale (CES-D), were registered using a self-administered questionnaire.
Self-reported physical activity was recorded using the question: “How has your physical activity been in your spare time the last year?”, with the help text “Imagine your weekly average for the year. Way to work is counted as spare time,” and the four categories for response: “Hard physical activity (sweating or experiencing shortness of breath) for: Zero hours per week, less than 1 hour per week, 1–2 hours per week, and 3 or more hours per week.” These questions have been previously validated in a large Norwegian general population study (Citation20). Charlson Index for co-morbidities was determined from the information collected by the physician during the clinical examination (Citation21).
Blood gases were measured in 1.5 mL blood, which were drawn from the radial artery with a radiometer PICO 70 arterial sampler and immediately analyzed on an ABL 520 blood gas analyzer (Radiometer, Copenhagen, Denmark).
The 6-minute walk test
All subjects received a detailed explanation of the test prior to participation. The participants were asked to walk back and forth on a 30-meter long pathway as many times as possible in 6 minutes. Standardized encouragements were given in accordance with the ATS guidelines (Citation6). Oxygen saturation was measured before and after the test using a pulseoxymeter (NONIN Medical Inc., Plymouth, MN), whereas self-experienced dyspnea was measured before and after the test using the Borg CR10 Scale. The scale goes from 0 to 10, where 0 means no dyspnea, and 10 means the most dyspneic the patient has ever felt. Heart rate before and after the test was also recorded, but not blood pressure. All participating study engineers were trained with the same standardized detailed instructions.
Statistical analyses
Walking distance was normally distributed and bivariate associations were tested using t-tests, Anova and Pearson's correlation tests. Clinically significant oxygen desaturation was defined as a post-test saturation of less than 90% and a change in saturation between pre- and post-test of more than 4% (Citation22). Bivariate associations were tested using chi-square tests for categorical variables and t-tests and Anova for continuous variables. Change in Borg CR10 score before and after the test, was not normally distributed. Bivariate associations were analysed using Kruskal-Wallis and Wilcoxon Mann-Whitney tests for categorical variables, and Spearman's rho test for continuous variables.
In the multivariate analyses, walking distance was analysed using linear regression, oxygen desaturation was analysed using logistic regression, and change in Borg score was analysed using the non-parametric quantile regression. The same set of potentially explanatory variables was included in all 3 models. FFMI and FMI were significantly correlated. To adjust for potential colinearity between FFMI and FMI the method of residualization was used. The residuals from a regression of FMI and FMI squared on FFMI were included in the models predicting the effect of FFMI on the 3 outcomes, and vice versa for FMI. Thus, for each model we examined the effect of either FFMI or FMI, adjusted for the residuals of the other body composition index, as well as all mentioned potential confounders. Significance level was set to 0.05, and we used standard statistical software, Stata 11 for the analyses (StataCorp. LP, College Station, TX).
Results
The baseline characteristics of the study sample are given in . Patients had a mean FEV1 in% predicted of 50%, which implies significant airflow limitation. Two thirds of the patients had dyspnea grade 1 or 2. Slightly more than half of the study sample was ex-smokers, and 20% of the patients were depressed when judged by the CES-D depression score. Significant sex differences were found for age, body composition, RV in% predicted, PaO2 and number of co-morbidities ().
Table 1. Characteristics of the 370 COPD patients
The association between 6MWT outcomes and other variables –bivariate analyses: he bivariate associations between potential explanatory variables and the three outcomes are presented in for the categorical variables and for continuous variables. There were significant sex differences for mean walking distance, 415 metres for women and 444 metres for men, but not for oxygen desaturation and self-experienced dyspnea (). Overall, 86 (24%) of the COPD patients experienced oxygen desaturation. The median change in Borg score for all patients was 2. Being an ex-smoker was associated with oxygen desaturation and change in Borg score, dyspnea was associated with all 3 outcomes, and depression was associated with shorter walking distance and a larger change in Borg score. Self-reported physical activity was associated with walking distance ().
Table 2. Bivariate associations between outcomes and categorical predictor variables
Table 3. Bivariate associations between the three outcomes and continuous predictor variables
Mean walking distance decreased with higher age, higher FMI, higher RV in% predicted and higher plasma CRP, and with lower FFMI, lower FEV1, and lower PaO2 (). Oxygen desaturation was significantly associated with higher FMI, lower FEV1 and higher RV in% predicted, lower PaO2 and higher PaCO2. A larger change in Borg scores before and after the walk test was associated with lower FEV1 and RV in% predicted, and lower PaO2.
The association between 6MWT outcomes and other variables –multivariate analyses: he adjusted multivariable analyses are shown in . After adjustment for all covariates, sex, age, FEV1 in% predicted, number of co-morbidities, and self-reported dyspnea and physical activity remained significant predictors of mean walking distance. Men had on average 46 meter longer walking distance than women, and mean walking distance decreased by 26 meters per 10 years age increase, and by 22 meters per 10% lower FEV1 in% predicted.
Table 4. Multivariate regression analyses with 6 minute walking distance, change in Borg score and oxygen desaturation
Patients with two or more co-morbid disorders had almost 60 meters shorter walking distance, and patients reporting no hard physical activity in an average week had more than 80 meters shorter mean walking distance. Although systemic inflammation assessed by plasma CRP and body composition assessed by FFMI and FMI were correlated with walking distance bivariately, these associations were statistically insignificant when adjusting for the other variables.
Body composition was significantly associated with a change in Borg score, together with smoking habits and Charlson's co-morbidity index. Somewhat surprisingly, both higher FFMI and higher FMI were associated with a larger change in Borg score. The only significant predictors of oxygen desaturation were FEV1 in% predicted and PaO2 measured at rest (). The multivariable models explained 46% and 36% of the variation of the walking distance and oxygen desaturation respectively, but only 16% of the variation of change in Borg score.
Discussion
This study shows that in COPD patients, several different COPD disease characteristics are related to the outcomes of the 6MWT. Further, this study shows that the 3 test outcomes walking distance, dyspnea, and oxygen desaturation are related to different disease characteristics.
The 3 outcomes in the 6MWT were related to different factors in the COPD patients. Of the 3 outcomes, walking distance was related to the largest number of factors; sex, age, lung function, self-reported dyspnea, co-morbidities, and physical activity, reflecting that walking distance measures several facets of functional capacity. Two previous studies have examined predictors for walking distance in COPD patients with multivariable analyses (Citation9, Citation15).
In a report from the NETT study comparing the 6MWT and the cardiopulmonary exercise test (CPX) in 1218 severe COPD patients, sex, age, height, weight, FEV1 in% predicted, and quality of life were all related to 6MWD (Citation15). Recently published results from the ECLIPSE study showed that in 1795 COPD patients, GOLD stage, emphysema score on CT scans, inspiratory capacity, self-reported dyspnea, CES-D score and quality of life were related to having a walking distance under 350 meters (Citation9).
Although not directly comparable to our study since the outcome was dichotomous, both in the ECLIPSE study and in our study disease severity assessed by FEV1 and dyspnea assessed by MRC were significant predictors of 6MWD. However, we also found physical activity and Charlson co-morbidity to be significantly associated with 6MWD, whereas CES-D score was not. CES-D score was significant in the bivariate analyses, and in contrast with the ECLIPSE study, we also adjusted for arterial blood gases, body composition indices, CRP and physical activity. A confounding analysis where we excluded one potential confounder from the model and examined the effect on CES-D score revealed that the addition of physical activity had the largest effect on CES-D score, reversing the direction of the coefficients.
Oxygen desaturation was related only to two factors in the multivariable model, FEV1 and resting PaO2. Perhaps the most important point regarding oxygen desaturation are all the factors not related, like self-reported dyspnea and co-morbidities. Thus, in trying to determine which patients should be tested for oxygen desaturation, only a low FEV1 and a low PaO2 are certain factors implying at-risk status. Few studies have examined multivariable predictive models for oxygen saturation during the 6MWT (Citation23–25). However, several studies have established the strong relationship between diffusion capacity (DlCO) and desaturation (Citation26, 27). We did not measure DlCO in our patients, and could not examine this factor.
The third outcome, self-perceived dyspnea during the test, was unrelated to the disease severity variables but instead associated with body composition, smoking status and co-morbidities. The most common co-morbidity among the COPD patients was heart disease, a known cause of exertional dyspnea, and thus a rather unsurprising finding. For smoking status, the ex-smokers had more dyspnea than current-smokers. This finding is likely a ‘healthy’ smoker effect, where patients who had the most symptoms from their disease are more likely to give up smoking. Both higher FMI and higher FFMI were related to more dyspnea during the 6MWT, even after extensive mutual adjustment between the two body composition indices.
This seems counterintuitive, as one would expect a higher FFMI to reflect larger muscle mass and therefore be beneficial. It could be that we were not fully able to adjust for the colinearity between FFMI and FMI. However, the fact that no lung function measurements were related to Borg score implies that it is the perception of dyspnea that determines the score. Patients who perceive themselves overweight may thus experience more dyspnea even though their FFMI are actually high.
One of the strengths of this study is the detailed clinical information available allowing adjusted analyses of the potential relationships between clinical factors and the results from the 6MWT. However, some methodological issues should be considered. First, this is a cross-sectional study, making inferences about causal relationships impossible. For instance, whether low levels of weekly physical activity causes a short walking distance, or vice versa cannot be determined with this study design. Importantly, however, this study confirms the close relationship between physical activity and functional capacity.
Ultimately, the predictive value of each of the associations found in the present study should be tested against a prospective outcome. For example, the association between a low FEV1 and risk for oxygen desaturation is a likely consequence of a lower gas diffusion capacity during physical activity. However, the clinical consequence is yet controversial, and the consequence of oxygen desaturation during exercise should be tested against later outcomes like mortality, risk for developing cachexia, co-morbidities and exacerbation frequency.
Second, only 1 walk test was performed. Existing studies indicate that walking distance is on average longer in the second test than in the first, but the size of the learning effect varies (Citation28–30). Hernandes et al. (Citation29) found that the 6MWT is reproducible, but the majority of their COPD patients increased the walking distance in the second test with 7%. Limited resources made it impossible to perform two tests in the Bergen COPD cohort study.
We can infer from the previous studies mentioned that the mean walking distance may have been slightly underestimated by performing only one test. However, if this were to impact the relationships with the COPD characteristics examined in our study, it would mean that for instance patients with Charlson Index score 4 learned differently than patients with Charlson Index score 1. We believe this unlikely, although the importance of learning effects of 6MWT in cross-sectional studies have not been clarified.
Third, the use of a self-reported questionnaire on physical activity, may not have given us an accurate evaluation of the subjects’ physical activity level. Fourth, the patient sample was not a random sample of all COPD patients in our hospital district, and thus is vulnerable to selection bias. Patients were selected from several different sources, including a previous COPD study from the Hordaland County conducted at the Haukeland University Hospital (‘GenKOLS’, n = 270), the outpatient clinic at Haukeland University Hospital (n = 107), outpatient clinics from other hospitals in Western Norway (n = 22), and from the clinics of private lung specialists practicing in Hordaland County (n = 34). The patients seen at a hospital clinic or attending a specialist tend to have more complicated disease than patients undiagnosed or seen by general practitioners. Thus, it is possible that our patients have more severe disease than the COPD population as a whole. However, the study population included patients with large differences in symptoms, severity, and co-morbidities; important factors for which we could adjust in the multivariable analyses.
In conclusion, the 6MWT is related to a large range of important clinical COPD disease characteristics, and all the 3 outcomes in the 6MWT contain useful and complementary information about the health status of the patients.
Declaration of Interest Statement:
Ms. Waatevik has received payment for work related to lung function testing of general population from AstraZeneca (unrelated to this study). Dr. Johannessen has no conflicts of interest to disclose. Dr. Bjordal has no conflicts of interest to disclose. Dr. Aukrust has no conflicts of interest to disclose. Dr. Hardie has within the last 5 years received sponsorship for travel and accommodations to the ATS congress from GSK, and received grant money from AstraZeneca, GlaxoSmithKline, Boehringer Ingelheim and Pfizer (unrelated to this study).
Dr. Bakke has received lecture fees from GSK and AstraZeneca. Dr. Bakke is also a principle investigator in a GSK sponsored study. Dr. Eagan has within the last 5 years received sponsorship for travel and accommodations to the ATS congress from GSK, and received grant money from AstraZeneca (unrelated to this study). The authors are responsible for the content and the writing of this paper.
References
- Buist AS, International variation in the prevalence of COPD (the BOLD Study): a population-based prevalence study. Lancet 2007; 370(9589):741–750.
- Brown CD, Wise RA. Field tests of exercise in COPD: The six-minute walk test and the shuttle walk test. COPD 2007; 4(3):217–223.
- Kapella MC, Functional performance in chronic obstructive pulmonary disease declines with time. Med Sci Sports Exerc 2011; 43(2):218–224.
- Celli BR, The body-mass index, airflow obstruction, dyspnea, and exercise capacity index in chronic obstructive pulmonary disease. N Engl J Med 2004; 350(10):1005–1012.
- Lacasse Y, Pulmonary rehabilitation for chronic obstructive pulmonary disease. Cochrane Database System Rev 2006 (4):CD003793.
- ATS. ATS Statement: Guidelines for the six-minute walk test. Am J Respir Crit Care Med 2002; 166(1):111–117.
- Casanova C, Distance and oxygen desaturation during the 6-min walk test as predictors of long-term mortality in patients with COPD. Chest 2008; 134(4):746–752.
- Enright PL. The six-minute walk test. Respir Care 2003; 48(8):783–785.
- Spruit MA, Determinants of poor 6-min walking distance in patients with COPD: The ECLIPSE cohort. Respir Med 2010; 104(6):849–857.
- Carr SJ, Goldstein RS, Brooks D. Acute exacerbations of COPD in subjects completing pulmonary rehabilitation. Chest 2007; 132(1):127–134.
- Huijsmans RJ, The clinical utility of the GOLD classification of COPD disease severity in pulmonary rehabilitation. Respir Med 2008; 102(1):162–171.
- Wijkstra PJ, Relation of lung function, maximal inspiratory pressure, dyspnoea, and quality of life with exercise capacity in patients with chronic obstructive pulmonary disease. Thorax 1994; 49(5):468–472.
- Lee YK, Quantitative assessment of emphysema, air trapping, and airway thickening on computed tomography. Lung 2008; 186(3):157–165.
- Ischaki E, Body mass and fat-free mass indices in COPD: relation with variables expressing disease severity. Chest 2007; 132(1):164–169.
- Brown CD, Exercise testing in severe emphysema: association with quality of life and lung function. COPD 2008; 5(2):117–124.
- Eagan TM, Systemic inflammatory markers in COPD: results from the Bergen COPD Cohort Study. The Euro Respir J: Off J Euro Soc Clin Respirat Physiol 2010; 35(3):540–548.
- Johannessen A, Post-bronchodilator spirometry reference values in adults and implications for disease management. Amer J Respir Crit Care Med 2006; 173(12): 1316–1325.
- Stocks J, Quanjer PH. Reference values for residual volume, functional residual capacity and total lung capacity. ATS Workshop on Lung Volume Measurements. Official Statement of The European Respiratory Society. Eur Respir J 1995; 8(3):492–506.
- Wu TL, Development of ELISA on microplate for serum C-reactive protein and establishment of age-dependent normal reference range. Clin Chim Acta 2002; 322(1–2):163–168.
- Holmen J, Midthjell K, Kruger Ø, Langhammer A. The Nord-Trøndelag Health Study 1995–1999 (HUNT2). Norsk Epidemiol 2003; 13:19–32.
- Charlson M, Validation of a combined comorbidity index. J Clin Epidemiol 1994; 47(11):1245–1251.
- Wedzicha JA. Domiciliary oxygen therapy services: clinical guidelines and advice for prescribers. Summary of a report of the Royal College of Physicians. J R Coll Phys Lond 1999; 33(5):445–447.
- Mak VH, Effect of arterial oxygen desaturation on six minute walk distance, perceived effort, and perceived breathlessness in patients with airflow limitation. Thorax 1993;48(1):33–38.
- Panos RJ, Eschenbacher W. Exertional desaturation in patients with chronic obstructive pulmonary disease. COPD 2009; 6(6):478–487.
- Knower MT, Baseline oxygen saturation predicts exercise desaturation below prescription threshold in patients with chronic obstructive pulmonary disease. Arch Inter Med 2001; 161(5):732–736.
- Owens GR, The diffusing capacity as a predictor of arterial oxygen desaturation during exercise in patients with chronic obstructive pulmonary disease. New Engl J Med 1984; 310(19):1218–1221.
- Hadeli KO, Predictors of oxygen desaturation during submaximal exercise in 8,000 patients. Chest 2001; 120(1):88–92.
- Leach RM. Portable liquid oxygen and exercise ability in severe respiratory disability. Thorax 1992; 47(10):781–789.
- Hernandes NA, Reproducibility of 6-minute walking test in patients with COPD. Eur Respir J 2011; 38(2): 261–267.
- Knox AJ, Morrison JF, Muers MF. Reproducibility of walking test results in chronic obstructive airways disease. Thorax 1988; 43(5):388–392.