In this study, we analyzed the relationships of exercise capacity and health status to mortality in patients with chronic obstructive pulmonary disease (COPD). We recruited 150 male outpatients with stable COPD with a mean postbronchodilator FEV1 at 47.4% of predicted. Their pulmonary function, progressive cycle ergometry, and health status using the Chronic Respiratory Disease Questionnaire, the St. George's Respiratory Questionnaire (SGRQ), and the Breathing Problems Questionnaire were measured at entry. Among 144 patients who were available for the 5-year follow-up, 31 had died. Univariate Cox proportional hazards analysis revealed that the SGRQ total score and the Breathing Problems Questionnaire were significantly correlated with mortality; however, with the Chronic Respiratory Disease Questionnaire, the total score was not significantly correlated. Multivariate Cox proportional hazards analysis revealed that the peak oxygen uptake and the SGRQ total score were both predictive of mortality, independent of FEV1 and age. Stepwise Cox proportional hazards analysis revealed that the peak oxygen uptake was the most significant predictor of mortality. We found that exercise capacity and health status were significantly correlated with mortality, although different health status measures had different abilities to predict mortality. These results will have a potentially great impact on the multidimensional evaluation of disease severity in COPD.
Mortality has been an important outcome in chronic obstructive pulmonary disease (COPD), as it is currently the fourth leading cause of death in the world (1). The disease severity of COPD has been based on the degree of airflow limitation, as defined by FEV1 because FEV1 has been regarded as the most important predictor of mortality in addition to age (2). However, although many studies investigating factors related to mortality have been performed in COPD, the relationships of exercise capacity and health status to mortality have rarely been evaluated.
There has recently been some debate about the use of FEV1 as the main single evaluative parameter for COPD. Van Schayck commented that when the effects of medication are evaluated, the effects on quality of life and functional status are probably much more important than effects on airflow limitation alone (3). Nishimura and coworkers have recently reported that categorizing patients with COPD on the level of dyspnea was more closely correlated with mortality than classification based on disease severity, as assessed by the percentage of predicted FEV1 (4). Celli proposed a systemic evaluation of COPD patients and stressed that there is a need to seek candidates for multidimensional disease staging (5). Exercise capacity and health status may also be important clinical indices to evaluate disease impairment in addition to FEV1.
Although airflow limitation is the most obvious manifestation of COPD, COPD has other extrapulmonary features and should be regarded as a systemic disorder (6). This multipathophysiologic aspect of COPD can influence exercise capacity (7) and health status (8). Therefore, we hypothesized that they could predict mortality in COPD. In this study, we investigated the relationships between various clinical parameters, including exercise capacity and health status, and mortality in patients with COPD after 5 years. In addition, we examined which clinical parameter was the most important when regarding mortality as the reference outcome.
We recruited 150 consecutive male outpatients with COPD from September 1995 to January 1997 at the Kyoto University Hospital. Entry criteria included (1) smoking history of more than 20 pack-years; (2) maximal FEV1/FVC ratio of less than 0.7 and prebronchodilator FEV1 of less than 80% of the predicted value; (3) no uncontrolled comorbidities likely to affect mortality such as malignant disorders, cardiovascular diseases, or cerebrovascular diseases; (4) regular attendance over 6 months; and (5) no exacerbations in the preceding 6 weeks. Pulmonary function, exercise capacity, health status, smoking status, and body mass index (BMI) were evaluated on the same day. Pulmonary function tests were performed at least 12 hours after the withdrawal of inhaled bronchodilators. According to the recommended method (9), subjects underwent spirometry before and at 15 and 60 minutes after inhaling salbutamol (400 μg) and ipratropium bromide (80 μg) using a metered-dose inhaler with a spacer device (InspirEase; Schering-Plough K.K., Osaka, Japan) (10). The predicted values were those established by the Japan Society of Chest Diseases (11). Self-reported smoking status was confirmed by using a smokerlyzer and measuring plasma cotinine levels as explained elsewhere (12). BMI was calculated by dividing the patient's weight in kg by height squared (m2). Although all patients were advised to exercise during daily living, no established rehabilitation program was administered. Verbal informed consent was obtained from all patients.
Symptom-limited progressive cycle ergometry was performed 60 minutes after the inhalation of bronchodilators on a calibrated, electrically braked cycle ergometer (13). Patients wore a face mask and began unloaded pedaling for 3 minutes, after which the workload was increased progressively by increments of 1 W every 3 seconds to the limit of tolerance. The exercise data were recorded with an automated exercise testing system. The peak oxygen uptake (V̇o2) that was reached during exercise was calculated.
Health status was measured by three disease-specific measures: the Chronic Respiratory Disease Questionnaire (CRQ) (14), the St. George's Respiratory Questionnaire (SGRQ) (15), and the Breathing Problems Questionnaire (BPQ) (16). The Japanese versions of these questionnaires have been previously validated (12). The CRQ consists of 20 items that were divided into four domains: dyspnea, fatigue, emotional function, and mastery. Each question was scored on a seven-point scale, and each domain and the total score were calculated as the sum. The SGRQ consists of 50 items divided into three components of symptoms, activity, and impacts, and the total score was also calculated, with scores ranging from 0 to 100. The BPQ has 33 items, and its total score was calculated using a scale of 1 to 103. Higher scores indicate less impairment on the CRQ, and the opposite is true of the SGRQ and BPQ.
Results are presented as mean ± SD unless otherwise stated. The survival status of all subjects after 5 years was assessed. The duration from entry to the last attendance or death was recorded. The survival time was calculated with the life table method.
Univariate and multivariate Cox proportional hazards analyses were performed to investigate the relationship between the clinical indices and mortality. Postbronchodilator FEV1 was used as an index of airflow limitation because it is regarded as a better predictor of mortality than prebronchodilator FEV1 (2). Clinical variables were used as continuous variables, except that the categoric variables of smoking status and the use of inhaled corticosteroids were coded as one or zero for the analysis. Results of the regression analysis were presented in terms of the estimated relative risks (RRs) with corresponding 95% confidence intervals; p values of less than 0.05 were considered to be statistically significant.
The baseline characteristics of the 150 male COPD patients are presented in Table 1
Mean ± SD or Number
|Age, yr||68.7 ± 6.9||48–89|
|Cumulative smoking, pack-years||58 ± 31||20–210|
|Body mass index, kg/m2||21.1 ± 2.9||14.0–29.0|
|Prebronchodilator FEV1, L||1.01 ± 0.44||0.34–2.52|
|Prebronchodilator FEV1, % predicted||38.1 ± 15.7||14.7–78.0|
|Postbronchodilator FEV1, % predicted||47.4 ± 17.4||15.7–83.6|
|DLCO, % predicted||64.8 ± 20.2||30.5–131.2|
|DLCO/VA, ml/min/L/mm Hg||3.37 ± 1.17||1.29–7.11|
|Peak V̇O2, ml/min||833 ± 265||202–1,625|
|Health status measures|
|CRQ dyspnea (5–35)||26.0 ± 5.8||11–35|
|Fatigue (4–28)||19.9 ± 5.1||4–28|
|Emotional function (7–49)||39.5 ± 7.0||19–49|
|Mastery (4–28)||22.0 ± 4.3||11–28|
|Total (20–140)||107.5 ± 18.3||65–140|
|SGRQ symptoms (0–100)||52.4 ± 19.9||9.7–100.0|
|Activity (0–100)||42.2 ± 21.1||0.0–94.0|
|Impact (0–100)||24.7 ± 16.9||0.0–75.4|
|Total (0–100)||36.1 ± 16.7||2.5–80.0|
|BPQ Total (1–103)||18.1 ± 12.9||1–62|
When the patients were classified according to COPD severity based on airflow limitation defined by the American Thoracic Society, 63 patients (42%) were in stage I (50% or more predicted). Forty-eight (32%) were in stage II (35–49% predicted), and thirty-nine (26%) were in stage III (less than 35% predicted). Kaplan-Meier survival curves based on the severity of airflow limitation are presented in Figure 1.
Among the 150 patients enrolled, 6 were unavailable for follow-up over the 5-year period (follow-up rate 96%), and 31 had died. The survival rates at 1, 3, and 5 years were 95%, 90%, and 80%. Of the causes of deaths, 20 patients died due to COPD or COPD-related diseases, 4 due to malignant disorders, including 2 lung cancer cases, and 1 each due to myocardial infarction and hepatic failure, respectively. Five deaths were due to unknown reasons.
Univariate Cox proportional hazards analysis was performed to investigate the relationship between the clinical measures and mortality (Table 2)
95% Confidence Interval
|Cumulative smoking, pack-years||1.010||1.002–1.019||0.022|
|Use of inhaled corticosteroids||1.575||0.764–3.245||0.22|
|Body mass index, kg/m2||0.785||0.685–0.900||0.0005|
|Postbronchodilator FEV1, % predicted||0.940||0.914–0.966||< 0.0001|
|DLCO/VA, ml/min/L/mm Hg||0.413||0.269–0.633||< 0.0001|
|Peak V̇O2, ml/min||0.994||0.992–0.996||< 0.0001|
|Health status measures|
|BPQ total||1.035||1.011–1.059|| 0.0044|
To compare the ability of exercise capacity and health status to predict mortality versus that of airflow limitation and age, which have been considered to be the best predictors of mortality (2), multivariate regression analysis was performed (Tables 3 and 4)
95% Confidence Interval
|Postbronchodilator FEV1, % predicted||0.972||0.943–1.001||0.059|
|Peak V̇O2, ml/min||0.995||0.993–0.998||< 0.0001|
95% Confidence Interval
|Age, years||1.134||1.069–1.204||< 0.0001|
|Postbronchodilator FEV1, % predicted||0.940||0.915–0.966||< 0.0001|
|SGRQ total||1.035||1.008–1.063|| 0.012|
To analyze which index was the most significantly correlated with mortality, stepwise Cox proportional hazards analysis was performed using age, cumulative smoking, BMI, FEV1, diffusing capacity for carbon monoxide/Va, peak V̇o2, and the SGRQ total score. This analysis revealed that peak V̇o2 and age were the most significant factors related to mortality (RR = 0.994, p < 0.0001; RR = 1.077, p = 0.024, respectively) (Table 5)
95% Confidence Interval
|Peak V̇O2, ml/min||0.994||0.992–0.996||< 0.0001|
|Postbronchodilator FEV1, % predicted||0.056|
|DLCO/VA, ml/min/L/mm Hg||0.089|
|Body mass index, kg/m2||0.21|
|Cumulative smoking, pack-years||0.32|
We have evaluated the factors related to mortality in COPD, particularly the relationships of exercise capacity and health status to mortality. We have found that exercise capacity and health status were significantly predictive of mortality in COPD, independent of airflow limitation or age. In addition, this study indicates that exercise capacity can be the best predictor of mortality in patients with COPD. Furthermore, the ability of health status to predict mortality turned out to be different depending on the instruments used.
Regarding exercise capacity, one novel finding in this study was that it could be the most significant criterion related to mortality among age, pulmonary function, BMI, or health status when using peak V̇o2 as an index. Exercise capacity of COPD patients is affected by important, complex factors, including ventilation, gas exchange, circulation, muscular function, nutritional status, and their symptoms (7); hence, COPD can be regarded as a systemic disorder (6). Exercise capacity may thus evaluate the severity of COPD more comprehensively and objectively than airflow limitation defined by FEV1. Furthermore, exercise capacity cannot be accurately predicted from resting physiologic variables (7). Therefore, in addition to resting spirometric testing, exercise capacity can be measured as an index of disease severity from the perspective of mortality in patients with COPD.
A few studies have reported the importance of exercise capacity as a predictor of mortality in COPD (2, 17, 18). However, the indices of exercise capacity in these studies were the maximal work rate (2) and the walking distance (17, 18), and an analysis of the gas expired during exercise was not performed. Peak V̇o2 is the primary measure of exercise capacity; the prognostic significance of peak V̇o2 has not, however, been previously evaluated in COPD. When investigating the relationship with mortality in COPD, peak V̇o2 might be preferable, as it was more significantly correlated to mortality than any of the other measures.
This study demonstrated that health status had a significant correlation with mortality independent of airflow limitation or age. The role of health status in predicting mortality has not been previously evaluated well in COPD, although the association between health status and subsequent mortality has been frequently reported for cancer (19, 20). In addition, a novel finding was the different abilities of health status measures to predict mortality. Predicting future outcomes is one important objective of measuring health status (21), and therefore, it may be insightful to compare different measures from the viewpoint of their predictive properties. In this study, the total scores on the SGRQ and the BPQ were strongly and significantly correlated with mortality; however, the CRQ total score was not significantly correlated. Gerardi and coworkers (17) also failed to show a significant relationship between the CRQ and the 3-year survival rate after pulmonary rehabilitation in patients with advanced pulmonary disease. However, recent studies have reported a significant relationship between health status and mortality using the SGRQ (22, 23).
Why was the CRQ less correlated to mortality than the SGRQ and the BPQ? First, the CRQ does not examine activity restrictions, unlike the SGRQ and BPQ (24). Therefore, the patients may not be well discriminated based on the severity of their functional status, which is a significant factor closely correlated with mortality (18). Wijkstra and coworkers (25) also reported that functional exercise capacity was not adequately evaluated by the CRQ. In addition, Hajiro and coworkers (12) showed that the CRQ had relatively lower correlations with pulmonary function, exercise capacity, and dyspnea but was more influenced by psychologic status than the SGRQ or BPQ. These differences in the physical versus psychologic aspects might have contributed to these results. Second, the CRQ has a seven-point Likert scale but includes only 20 items. In comparison, the SGRQ and BPQ have a lower response range to each item but more items than the CRQ. Therefore, the CRQ would be better at investigating the changes within individuals but weaker at discriminating between patients based on various aspects for assessing health status. This point might also have affected the ability of the CRQ to predict mortality.
In this study, BMI was not a significant prognostic factor in the multivariate analysis as observed by Bowen and coworkers (18), although some studies have reported a significant relationship (26, 27). In the Copenhagen City Heart Study (26), the association between BMI and mortality was especially significant in severe COPD and differed according to the severity of airflow limitation. We made only an overall analysis of COPD due to the smaller sample size in this study. In contrast, we included some detailed potential prognostic factors such as exercise capacity and diffusing capacity, which have been shown to be related to nutritional status (28). These may explain the insignificant relationship to BMI in this study.
This study demonstrated that cumulative smoking was a significant predictor of mortality in the univariate analysis but that smoking status at baseline was not. The effects of smoking cessation on lung function and respiratory symptoms have been demonstrated in patients with mild to moderate COPD in the Lung Health Study (29, 30), and a relationship between smoking cessation and improved mortality was anticipated. However, the effects of smoking cessation in patients with severe COPD may not confirm this observation (31). Symptomatic patients with severe COPD spontaneously tended to quit smoking, and smoking status in some patients may have changed during the 5-year follow-up period in this study. These may be some of the reasons for the insignificant relationship between smoking status and mortality in this study.
The use of inhaled corticosteroids was not related to mortality, although Sin and Tu (32) suggested that it was associated with reduced COPD-related morbidity and mortality in patients with COPD. In this study, although half of the patients were treated with inhaled corticosteroids, they had more severe airflow limitation than patients who were not because our pharmacologic treatment was somewhat based on the degree of airflow limitation. In addition, our sample was much smaller than that of the Sin and Tu study (32). Furthermore, the use of inhaled corticosteroids was not comprehensively evaluated during the 5-year follow-up period, and inhaled corticosteroid therapy was initiated in some patients. Therefore, it was impossible to demonstrate the effects of inhaled corticosteroids on mortality in this study.
In addition to this study, we have recently completed two different prospective studies regarding mortality in COPD patients (4, 33). Although some patients were included in more than one of studies, there was no data overlap between the various studies. Moreover, this study has collected additional baseline information on these patients. Notably, only this study included exercise data at baseline, and this study showed that exercise capacity predicted mortality in COPD most significantly.
Some limitations of this study should be mentioned. First, as the entry criteria excluded the major comorbidities that might affect mortality, we did not investigate the significance of comorbidity factors fully. Comorbidity has been reported to play an important role in the prediction of survival of COPD patients (34). Therefore, we should have searched for a better way to investigate the relationship between comorbidities and survival. Second, it is not well known how accurate a face mask is in the determination of peak V̇o2 in comparison to a mouthpiece. Although breathing pattern may change according to types of breathing assembly (35), it was reported that this did not alter maximal exercise (36). To find the individual maximal V̇o2 is difficult because peak V̇o2 will change depending on exercise apparatus, incremental work rate, and so on (37). However, we believe that the more accurately peak V̇o2 is calculated, the stronger the significant relationship between peak V̇o2 and mortality becomes.
In conclusion, we demonstrated significant relationships of exercise capacity and health status to mortality in COPD patients, independent of FEV1 or age. Laboratory exercise capacity using the cycle test could be the most significant predictor of mortality in COPD. With respect to health status, the ability of the CRQ to predict mortality was weaker than the SGRQ or BPQ. Although airflow limitation has been traditionally used as the index of disease severity in COPD, as it is regarded as the most significant predictor of mortality, the findings of this study will have a potentially great impact on the multidimensional evaluation of the disease severity in COPD from the perspective of mortality.
|1.||World Health Report. Geneva: World Health Organization; 2000.|
|2.||Anthonisen NR, Wright EC, Hodgkin JE. Prognosis in chronic obstructive pulmonary disease. Am Rev Respir Dis 1986;133:14–20.|
|3.||Van Schayck CP. Is lung function really a good parameter in evaluating the long-term effects of inhaled corticosteroids in COPD? Eur Respir J 2000;15:238–239.|
|4.||Nishimura K, Izumi T, Tsukino M, Oga T. Dyspnea is a better predictor of 5-year survival than airway obstruction in patients with COPD. Chest 2002;121:1434–1440.|
|5.||Celli BR. The importance of spirometry in COPD and asthma: effect on approach to management. Chest 2000;117:S15–S19.|
|6.||Gross NJ. Extrapulmonary effects of chronic obstructive pulmonary disease. Curr Opin Pulm Med 2001;7:84–92.|
|7.||Nici L. Mechanisms and measures of exercise intolerance in chronic obstructive pulmonary disease. Clin Chest Med 2000;21:693–704.|
|8.||Jones PW. Health status measurement in chronic obstructive pulmonary disease. Thorax 2001;56:880–887.|
|9.||Standardization of spirometry: 1994 update: American Thoracic Society. Am J Respir Crit Care Med 1994;152:1107–1136.|
|10.||Tobin MJ, Jenouri G, Danta I, Kim C, Watson H, Sackner MA. Response to bronchodilator drug administration by a new reservoir aerosol delivery system and a review of other auxiliary delivery systems. Am Rev Respir Dis 1982;126:670–675.|
|11.||Japan Society of Chest Diseases. The predicted values of pulmonary function testing in Japanese [appendix]. Jpn J Thoracic Dis 1993;31.|
|12.||Hajiro T, Nishimura K, Tsukino M, Ikeda A, Koyama H, Izumi T. Comparison of discriminative properties among disease-specific questionnaires for measuring health-related quality of life in patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1998;157:785–790.|
|13.||Oga T, Nishimura K, Tsukino M, Hajiro T, Ikeda A, Izumi T. The effects of oxitropium bromide on exercise performance in patients with stable chronic obstructive pulmonary disease: a comparison of three different exercise tests. Am J Respir Crit Care Med 2000;161:1897–1901.|
|14.||Guyatt GH, Berman LB, Townsend M, Pugsley SO, Chambers LW. A measure of quality of life for clinical trials in chronic lung disease. Thorax 1987;42:773–778.|
|15.||Jones PW, Quirk FH, Baveystock CM, Littlejohns P. A self-complete measure of health status for chronic airflow limitation: the St. George's Respiratory Questionnaire. Am Rev Respir Dis 1992;145:1321–1327.|
|16.||Hyland ME, Bott J, Singh S, Kenyon CA. Domains, constructs and the development of the breathing problems questionnaire. Qual Life Res 1994;3:245–256.|
|17.||Gerardi DA, Lovett L, Benoit-Connors ML, Reardon JZ, ZuWallack RL. Variables related to increased mortality following out-patient pulmonary rehabilitation. Eur Respir J 1996;9:431–435.|
|18.||Bowen JB, Votto JJ, Thrall RS, Haggerty MC, Stockdale-Woolley R, Bandyopadhyay T, ZuWallack RL. Functional status and survival following pulmonary rehabilitation. Chest 2000;118:697–703.|
|19.||Ganz PA, Lee JJ, Siau J. Quality of life assessment: an independent prognostic variable for survival in lung cancer. Cancer 1991;67:3131–3135.|
|20.||Coates A, Gebski V, Bishop JF, Jeal PN, Woods RL, Snyder R, Tattersall MHN, Byrne M, Harvey V, Gill G. Improving the quality of life during chemotherapy for advanced breast cancer: a comparison of intermittent and continuous treatment strategies. N Engl J Med 1987;137:1490–1495.|
|21.||Mahler DA, Jones PW. Measurement of dyspnea and quality of life in advanced lung disease. Clin Chest Med 1997;18:457–469.|
|22.||Carone M, Bertolotti G, Donner CF. Mortality in chronic respiratory failure is detected better by health status (QoL) than by functional parameters. Am J Respir Crit Care Med 2001;163:A13.|
|23.||Domingo-Salvany A, Lamarca R, Ferrer M, Garcia-Aymerich J, Alonso J, Félez M, Khalaf A, Marrades RM, Monsó E, Serra-Batlles J, et al. Health-related quality of life and mortality in male patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2002;166:680–685.|
|24.||Singh SJ, Sodergren SC, Hyland ME, Williams J, Morgan MDL. A comparison of three disease-specific and two generic health-status measures to evaluate the outcome of pulmonary rehabilitation in COPD. Respir Med 2001;95:71–77.|
|25.||Wijkstra PJ, TenVergert EM, van der Mark ThW, Postma DS, Van Altena R, Kraan J, Koëter DH. Relation of lung function, maximal inspiratory pressure, dyspnoea, and quality of life with exercise capacity in patients with chronic obstructive pulmonary disease. Thorax 1994;49:468–472.|
|26.||Landbo C, Prescott E, Lange P, Vestbo J, Almdal TP. Prognostic value of nutritional status in chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1999;160:1856–1861.|
|27.||Schols AMWJ, Slangen J, Volovics L, Wouters EFM. Weight loss is a reversible factor in the prognosis of chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1998;157:1791–1797.|
|28.||Gray-Donald K, Gibbons L, Shapiro SH, Martin JG. Effect of nutritional status on exercise performance in patients with chronic obstructive pulmonary disease. Am Rev Respir Dis 1989;140:1544–1548.|
|29.||Scanlon PD, Connett JE, Waller LA, Altose MD, Bailey WC, Buist AS. Smoking cessation and lung function in mild-to-moderate chronic obstructive pulmonary disease: the Lung Health Study. Am J Respir Crit Care Med 2000;161:381–390.|
|30.||Kanner RE, Connett JE, Williams DE, Buist AS. Effects of randomized assignment to a smoking cessation intervention and changes in smoking habits on respiratory symptoms in smokers with early chronic obstructive pulmonary disease: the Lung Health Study. Am J Med 1999;106:410–416.|
|31.||Pride NB. Smoking cessation: effects on symptoms, spirometry and future trends in COPD. Thorax 2001;56:ii7–10.|
|32.||Sin DD, Tu JV. Inhaled corticosteroids and the risk of mortality and readmission in elderly patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2001;164:580–584.|
|33.||Oga T, Nishimura K, Tsukino M, Sato S, Hajiro T, Ikeda A, Mishima M. Health status measured with the CRQ does not predict mortality in COPD. Eur Respir J 2002;20:1147–1151.|
|34.||Antonelli Incalzi R, Fuso L, De Rosa M, Forastiere F, Rapiti E, Nardecchia B, Pistelli R. Co-morbidity contributes to predict mortality of patients with chronic obstructive pulmonary disease. Eur Respir J 1997;10:2794–2800.|
|35.||Hirsch JA, Bishop B. Human breathing patterns on mouthpiece or face mask during air, CO2, or low O2. J Appl Physiol 1982;53:1281–1290.|
|36.||Evans BW, Potteiger JA. Metabolic and ventilatory responses to submaximal and maximal exercise using different breathing assemblies. J Sports Med Phys Fitness 1995;35:93–98.|
|37.||Wasserman K, Hansen JE, Sue DY, Whipp BJ, Casaburi R. Principles of exercise testing and interpretation. Philadelphia: Lea & Febiger; 1994.|