The Relationship Between Exercise Tolerance and Other Outcomes in COPD

Abstract

Exercise training has been shown to be a clinically effective therapeutic intervention for COPD patients resulting in a myriad of beneficial effects. These include improvements in exercise tolerance, health-related quality of life and activity levels. Activity levels can be assessed using health-related quality of life instruments or instruments designed especially for this purpose. Previous studies show that the relationships between activities of daily living, assessed using generic health-related quality of life instruments, and exercise vary considerably with correlations ranging from 0.18–0.72. The relationships between activities of daily living, assessed using disease specific health-related quality of life instruments, and exercise also vary considerably with correlations ranging from 0.14–0.59. The relationships between activities of daily living, assessed using activities of daily living instruments, and exercise are less variable and generally stronger with correlations ranging from 0.34–0.83. Relationships between generic health-related quality of life instruments and exercise vary considerably (0.19–0.65) as do relationships between disease specific health-related quality of life instruments and exercise (0.18–0.61). The correlations between changes in activities of daily living and changes in exercise following pulmonary rehabilitation are generally weak (0.13–0.28). The correlations between changes in health-related quality of life and changes in exercise following pulmonary rehabilitation are also generally weak. The fact that these outcomes are not tightly associated is due, in part, to the variability in responses to the instruments used and the fact that the various instruments were often designed to assess different constructs.

Keywords::

• Exercise
• Activities of Daily Living
• Health-related Quality of Life

INTRODUCTION

Chronic obstructive pulmonary disease (COPD) is a major cause of morbidity in the United States and world and is projected to be the 5th leading disease burden by the year 2020 ([1]). Advances in the treatment of COPD include supplemental oxygen therapy and smoking cessation ([2], [3]). Both have been shown to decrease mortality, and smoking cessation has been shown to improve the rate of decline in the forced expiratory volume in one second (FEV₁) ([3]). While both of these outcomes are important to patients and clinicians, they may not embody the patient's global perspective of his or her health. For the patient, symptom relief and improved physical functioning may be more important outcomes ([4]). Chief complaints from COPD patients are often dyspnea and exercise intolerance. In response to these complaints, pulmonary rehabilitation may be prescribed for these patients.

The assessment of pulmonary rehabilitation as a therapeutic intervention is based on a variety of outcome measurements, and participation in such programs has been shown to be clinically effective and to result in a myriad of beneficial effects. These include improvements in exercise tolerance as measured during cardiopulmonary exercise tests and sub-maximal exercise tests ([5]), improvement in self-reported function ([6]), improvements in health-related quality of life ([5], [7]), improved performance of activities of daily living and instrumental activities of daily living ([6]) and a reduction in patient symptoms ([8]). Preliminary data also suggest that certain types of pulmonary rehabilitation programs may result in greater daily activity levels following such a program ([9]). Limited published data also suggest that pulmonary rehabilitation may lead to a reduction in the use of health care resources ([10], [11]).

Despite improvements in these outcomes, it is also well documented that measures of pulmonary function do not change in response to exercise training. Additionally, there are no definitive data demonstrating decreased mortality resulting from participation in pulmonary rehabilitation programs ([11], [12]). The evaluation of pulmonary rehabilitation as an treatment for COPD presents a problem in that the outcomes that are most clinically relevant to patients and clinicians have not been clearly identified ([13]) nor have the relationships among these various outcomes been well defined. Therefore, the purpose of this review is to examine reported relationships between exercise tolerance and health-related quality of life and between exercise tolerance and activities of daily living.

Defining outcomes

Health outcomes are a broad continuum of end-points often used to test the efficacy of treatments, including pulmonary rehabilitation, and include objective measures such as death and the FEV₁. Increasingly, patient reported outcomes are being utilized to test treatment efficacy ([4], [14]). When examining patient reported outcomes, there can be considerable overlap in their definitions. In order to examine the relationships between exercise and outcomes, it is important that the measures of exercise and outcomes be well defined. Defining and identifying outcomes based on patients' perceptions of symptoms and health status can be confusing. For example, the assessment of activities of daily living may considered in the context of a health status outcome ([15]), a quality-of-life outcome ([16]), a functional performance outcome ([17]) or a functional status outcome ([18]). Additionally, activities of daily living and activity levels during daily life are separate constructs. Because of these inconsistencies and for the purposes of this review, the following definitions of outcomes are provided.

Exercise Tolerance

Patients with COPD present with a reduced exercise tolerance. This reduced exercise tolerance can be quantified by measuring a patient's exercise capacity, sometimes referred to as functional capacity. Exercise capacity is an objective measure of the level of physical functioning that a patient can achieve and can be measured in a variety of ways. These include the use of exercise tests performed on treadmills or cycle ergometers that may be done to a maximal or peak level or done at less than maximal (i.e., submaximal) levels. With these tests, the work rate can either be incremented over a given time period or be unchanged over time (constant work rate test) ([19]). If performed on a treadmill, test duration may serve as the outcome variable. If performed on a cycle ergometer, the external work rate can be measured, and thus may be the variable of interest. Other variables of interest that may be obtained from these tests include measures of ventilation and gas exchange.

For the purposes of this review, the variables of peak oxygen consumption (VO_{2 peak}), peak work rate and test duration will be included. Incremental tests performed on a treadmill and cycle ergometer are often considered tests of maximal exercise capacity, although clearly not all such tests are maximal. Another important test of exercise capacity is the timed walk test, with the 6-minute walk test being the most commonly used and standardized test ([20]). With any timed walk test, the variable of interest is the distance walked. The Incremental Shuttle Test has also been used as a test of exercise capacity with the outcome of this test being distance walked ([21]). Timed walk tests are considered a test of sub-maximal exercise capacity, since patients choose their own pace and are allowed to stop and rest during such test ([20]); whereas the incremental shuttle test is considered a test of maximal exercise capacity ([22]). All these above measures of exercise tolerance can be considered as objective measures of physical function, where physical function can be defined as the patient's ability to perform physical activity.

Activities of Daily Living

Because of differences between self-reported and objective measures of physical function, it is important to assess both ([23]). While objective and self-reported measures of physical function are related, research has shown them to be two distinct constructs ([24]). More specifically, patients may self-report high levels of physical function while performing poorly on objective measures of physical function and vice-versa. Differences in performance between these two measures may be attributable to the patient's psychological well-being and self-efficacy beliefs ([25]). Most self-reported measures of physical function assess difficulty in performing activities of daily living and/or instrumental activities of daily living. It is important to measure activities of daily living as they have been shown to be predictors of mortality ([26]). The National Center for Health Statistics defines activities of daily living as being related to personal care and to include bathing or showering, dressing, getting in or out of bed or a chair, using the toilet, and eating ([27]). Instrumental activities of daily living are defined as activities related to independent living and include preparing meals, managing money, shopping for groceries or personal items, performing light or heavy housework, and using a telephone ([28]). Leidy defines the extent to which people are able to perform these activities as functional performance ([17]), whereas Curtis and Patrick define this as functional status ([29]). In addition to assessing the ability to perform activities of daily living and instrumental activities of daily living, self-reported measures of physical function may also provide clinical information on the impact that disease or disease specific symptoms such as dyspnea has on the patient's ability to perform these activities ([16], [30], [31], [32]).

Activity Levels During Daily Living

Recently, a number of techniques have been developed that allow for the monitoring and determination of daily activity levels in patients with COPD. These include radioisotope techniques and the use of motion detectors ([33]). These measures differ from activities of daily living in that they examine the total volume of activity performed by the patient and typically do not capture or provide information with respect to the type or nature of the activity being performed.

Health-Related Quality of Life

Quality of life can be defined as a person's self-determined satisfaction with issues important to them, and it can be influenced by a number of factors including financial status, employment, spirituality, and health ([34]). In assessing medical conditions, health-related quality of life is a more appropriate concept and can be broadly defined as the self-reported assessment of the impact of health status on one's quality of life ([34]). Based on this definition, the terms health-related quality of life and health status are sometimes used interchangeably. Health-related quality of life is thought to include physical, psychological and social dimensions ([35]). More specifically, Shumaker and colleagues define health-related quality of life as including the dimensions of physical function, emotional function, social function, health perceptions and a global assessment of satisfaction with life ([36]). Rejeski and colleagues also include the dimension of cognitive function in their definition of health-related quality of life ([25]).

When determining health-related quality of life, investigators have used both generic and disease specific instruments. Generic instruments are designed to monitor patient outcomes in general medical practice, allowing for comparisons among various diseases and patient groups. Additionally, some generic instruments were developed for use as utility measures to estimate quality-adjusted life years ([37]). Disease-specific instruments have been designed to measure the same dimensions as the generic measures; however, these instruments have increased sensitivity and specificity with respect to a specific disease. For example, most COPD specific instruments measure the severity of symptoms such as dyspnea, wheezing and cough and the effect that these symptoms have on a patient's life.

Measurement instruments

Whereas the assessment and quantification of exercise tolerance relies primarily on objective measures, the same cannot be said of other outcomes such as activities of daily living and health-related quality of life. The majority of instruments used to measure these outcomes are self-reported. Additionally, many of the instruments that examine activities of daily living also contain questions that pertain to quality of life and vice-versa. As a result of this overlap, it can at times be difficult to determine the effect that an intervention, such as exercise training, has on specific measures, such as activities of daily living.

Activities of Daily Living

The measurement of activities of daily living is typically based on patient's responses to questionnaires. As noted before, questionnaires used to measure activities of daily living can be considered measures of functional performance and functional status. Additionally, many instruments designed to measure health-related quality of life will also provide a gauge of activities of daily living. For the purposes of this paper, those instruments that examine dyspnea during the performance of activities of daily living and those instruments that measure changes in activities of daily living as a result of dyspnea will also be included as measures of activities of daily living. Instruments that have been used with COPD patients that were designed to specifically address performance in activities of daily living include: the Additive Activities Profile Test ([38]), the Pulmonary Functional Status Scale ([39]), the Pulmonary Functional Status and Dyspnea Questionnaire ([40]), the Activities of Daily Living List ([41]), the Functional Performance Inventory ([42]), the Physical Functioning Questionnaire ([24]) and the University of California at San Diego's Shortness of Breath Questionnaire ([32]).

Activity Levels During Daily Living

The criterion methods for the determination of physical activity levels during daily life include doubly labeled water, indirect calorimetry and direct observation. These have been shown to be the most reliable and valid measurements although they can be expensive and technically difficult to perform ([43]). The measurement of activity levels during daily life can also be determined using questionnaires. However, these instruments suffer from problems such as patient recall and may have limited validity and reliability ([44]). Because of these problems, the determination of activity levels during daily living is typically performed using activity monitors such as pedometers and accelerometers ([45], [46], [47]).

Health-Related Quality of Life

Generic health-related quality of life instruments that are commonly used with COPD patients include the Medical Outcomes Study 36-item Short-Form Health Survey (SF-36) ([15]), the Sickness Impact Profile ([31]) and the Quality of Well-Being Survey ([48]). All 3 of these instruments were designed to measure health-related quality of life and to assess the patient's ability to perform activities of daily living. Additionally, the Quality of Well Being Survey can be used as a health utility tool for the determination of cost effectiveness ([48]). While information regarding activities of daily living can be gleaned from these instruments, their usefulness for this purpose is limited ([18]). This is primarily due to the narrow scope of questions that query activities of daily living. A review of the literature reveals a number of COPD disease specific instruments for measuring health-related quality of life. Two of the most commonly used instruments with COPD patients are the Chronic Respiratory Disease Questionnaire ([16]) and the St. George's Respiratory Questionnaire ([49]). Other disease specific instruments include the Visual Analogue Scale 8 ([50]) and the Seattle Obstructive Lung Disease Questionnaire ([51]). All of these instruments are pulmonary disease specific and were designed to assess patient's perceptions regarding the effect of both their disease and its treatment on their symptoms and their well being.

Relationships between various outcomes

The correlation coefficient is a common statistical measure used to define the strength and the direction of a relationship between two variables. When interpreting the results of studies that have used this technique, there are a number of points to consider. First, the strength of these correlations needs to be defined. For this review, these will be interpreted as suggested by Guyatt and colleagues where correlations less than 0.20 are very weak, from 0.21 to 0.35 are weak, from 0.36 to 0.50 are moderate and greater than 0.51 are strong ([52]). Second, it should be remembered that the correlation coefficient is influenced by the range of values within a distribution of scores, such that a homogeneous population will have a lower correlation as compared to a heterogeneous population when examining similar variables. Third, some correlations that may be interpreted as weak may be statistically significant whereas, others which are stronger may not be statistically significant. This is due to the effect of the sample size on the power of the statistical test. Fourth, the lack of a strong correlation between measures does not imply that these measures are not valuable outcomes in the assessment of COPD patients. Low correlations between measures underscore the fact that each instrument is providing unique information. Additionally, even instruments designed to assess similar constructs can produce different results due to the fact that specific instruments were designed for different purposes. While the construct of health-related quality of life can be simply defined as the patient's subjective experience of the impact of health status on his/her quality of life ([34]), various instruments have been designed for different purposes.

For example, some instruments were designed for studies of cost-utility analysis, whereas the purpose of others may be to determine the effect of a given intervention on patient symptoms. As such, each instrument provides unique information about the patient. Finally, the fact that certain outcomes are not highly correlated may be due to the signal-to-noise ratio of the instrument being used. If the variability in an individual patient's response to a given test or questionnaire (the noise) is similar to the variability between different patients (the signal), then the instrument will not show similar results in stable patients after repeated administrations. Examining the correlation between instruments with poor signal-to-noise ratios will result in low correlations.

Relationships between exercise and activities of daily living

As previously mentioned, the assessment of activities of daily living is important for a number of reasons. For instruments to be useful as outcome measures, it is important that such instruments demonstrate reliability, validity and responsiveness. Oftentimes, construct validity for questionnaires of activities of daily living is determined by comparing an instrument's results against objective exercise test results such as VO_{2 peak} or 6-minute walk distance. Results from studies that have performed these comparisons are presented next. These studies include results from questionnaires that measure activities of daily living exclusively and from questionnaires that measure health-related quality of life containing an activity of daily living or physical function component.

The SF-36 ([15]) and the Sickness Impact Profile ([53], [54]) are generic measures of health-related quality of life that contain an activity levels component. The SF-36 instrument has undergone factor analysis and can be reduced to 2 summary scores representing a physical and a mental component with the physical component incorporating limitations in performing activities of daily living into its score ([55]). Results from the National Emphysema Treatment Trial show the correlation between the physical component of the SF-36 and either the maximum work performed on cycle ergometer or the 6-minute walk distance to be 0.18 and 0.19, respectively ([56]). Others have reported the correlation between 6-minute walk distance and the physical component of the SF-36 to be higher, r = 0.41 ([57]), r = 0.67 ([45]) and r = 0.38 ([58]). These differences may partially be explained by differences in disease severity among study cohorts. The Sickness Impact Profile scores can be aggregated into a psychosocial domain and a physical domain. Jones et al. ([59]) found the correlation between 6-minute walk distance and the physical domain to be −0.72, whereas Prigatano et al. ([60]) found the correlation between maximum work during a cycle ergometer test and the physical domain to be −0.38. Table 1 summarizes the results of studies which have examined the association between exercise tolerance and activities of daily living as determined from generic health-related, quality-of-life instruments.

Table 1 Correlations between activities of daily living assessed from generic Health-related Quality of Life instruments and measures of exercise tolerance

Activities of daily living	Exercise tolerance	Correlation
SF-36 Physical Component ([56])	6 MW distance	0.19
	Wmax	0.18
SF-36 Physical Component ([57])	6 MW distance	0.41
SF-36 Physical Component ([45])	6 MW distance	0.67
SF-36 Physical Component ([58])	6 MW distance	0.38
Sickness Impact Profile Physical Component ([59])	6 MW distance	−0.72
Sickness Impact Profile Physical Component ([60])	6 MW distance	−0.38

6 MW = 6-minute walk, Wmax = maximal work rate performed on a cycle ergometer.

Two commonly used disease-specific measures of health-related quality of life, which have an activity of daily living component are the Chronic Respiratory Disease Questionnaire and the St. George's Respiratory Questionnaire. The Chronic Respiratory Disease Questionnaire assesses activities of daily living by asking patients to list 5 activities during which they experience dyspnea and to evaluate the degree of dyspnea they experience on a 7-point scale. Guyatt has examined the correlation between dyspnea scores obtained from the Chronic Respiratory Disease Questionnaire and the distance walked in 6 minutes and found correlations ranging from 0.26 to 0.52 ([16], [52], [61]). Hajiro and colleagues examined the relationship between VO_{2 peak} obtained from an incremental cycle ergometer test and the dyspnea score from the Chronic Respiratory Disease Questionnaire and found the correlation to be 0.48 ([62]). Overall results from these studies suggest the correlation between activities of daily living assessed from the dyspnea domain of the Chronic Respiratory Disease Questionnaire and exercise tolerance to range from weak to strong with the majority of correlations being moderate.

The other commonly used disease specific measure of health-related quality of life that has an activities-of-daily-living dimension is the St. George's Respiratory Disease Questionnaire ([49]). In this questionnaire, the domain of Activities is concerned with activities that are either caused by or limited by dyspnea. Jones and colleagues report the correlation between 6-minute walk distance and the activity domain to be −0.59 ([49]). Others examining the correlation between 6-minute walk distance and the activity domain from the St. George's Respiratory Disease Questionnaire have found correlations of −0.45 ([63]), and −0.58 ([50]). Hajiro et al. ([62]) found the correlation between VO_{2 peak} and the activity score to be −0.56 ([62]).

The Seattle Obstructive Lung Disease Questionnaire is a disease specific health-related, quality-of-life instrument that examines 4 domains including physical function. The physical function domain of this instrument has been shown to have a moderate association with 6-minute walk distance (r = 0.38 ([51]) and r = 0.48 ([57])). Table 2 summarizes the results of studies that have examined the association between exercise tolerance and activities of daily living as determined from disease specific health-related, quality-of-life instruments.

Table 2 Correlations between activities of daily living assessed from disease-specific Health-related Quality of Life instruments and exercise tolerance

Activities of Daily Living	Exercise Tolerance	Correlation
CRQ Dyspnea ([16])	6 MW distance	0.46
CRQ Dyspnea ([61])	6 MW distance	0.52
CRQ Dyspnea ([52])	6 MW distance	0.26
CRQ Dyspnea ([62])	Bike VO2 peak	0.48
CRQ Dyspnea ([68])	TM VO2 peak	0.14
SGRQ Activities ([49])	6 MW distance	−0.59
SGRQ Activities ([63])	6 MW distance	−0.45
SGRQ Activities ([50])	6 MW distance	−0.58
SGRQ Activities ([62])	Bike VO2 peak	−0.56
SOLDQ Physical Function ([51])	6 MW distance	0.38
SOLDQ Physical Function ([57])	6 MW distance	0.48

6 MW = 6-minute walk, CRQ = Chronic Respiratory Disease

Questionnaire, SGRQ = St. George's Respiratory Disease

Questionnaire, SOLDQ = Seattle Obstructive Lung Disease

Questionnaire, TM = treadmill, VO_{2 peak} = peak oxygen consumption.

In addition to using specific domains within health-related, quality-of-life questionnaires to ascertain activities of daily living, a number of questionnaires have been developed specifically for this purpose in COPD populations. These have been developed to more clearly define the effect of COPD on activities of daily living and the effect that interventions may have on the performance of these activities. As such, these questionnaires focus primarily on the physical function domain of health-related quality of life. The Additive Activities Profile test has patients score activities, from 105 listed, as “still doing,” “quit doing” or “never did.” Daughton and colleagues report a correlation of 0.83 between scores from this questionnaire and VO_{2 peak} ([38]).

Using 2 different scoring algorithms for the Additive Activities Profile test and comparing these results to 6-minute walk distance, Nield and colleagues reported correlations of 0.45 and 0.61 ([64]). The Pulmonary Functional Status Scale measures activities of daily living related to self-care, transportation, household tasks and meal preparations and has been found to have a strong correlation with 12-minute walk distance (r = 0.67 ([65]) and 0.62 ([39])) and 6-minute walk distance (r = 0.73 ([66])). The Pulmonary Functional Status and Dyspnea Questionnaire measures dyspnea intensity with activities of daily living (dyspnea component) and changes in performance of these activities as a result of COPD (functional component). The correlation between the functional component of this instrument and VO_{2 peak} has been reported as −0.41 ([18], [40]) and 6-minute walk distance as 0.34 ([45]).

The University of California Shortness of Breath Questionnaire is a 24-item instrument that measures dyspnea in patients while performing activities of daily living and queries the patient as to whether or not they have discontinued activities due to dyspnea ([32]). The correlation between this instrument and 6-minute walk distance has been reported to be −0.68 ([32]) and −0.37 ([56]). The correlation between this instrument and maximum work rate performed on a cycle ergometer has been reported to be −0.34 ([56]). These lower correlations are most likely due to the fact that these results are from the National Emphysema Treatment Trial, which had a more homogenous sample of patients, low functioning and severely diseased.

The London Chest Activity of Daily Living is a 15-item instrument that measures dyspnea in COPD patients while performing activities of daily living and has shown to correlate with the Shuttle Walk Test (r = −0.58) ([67]). Using an instrument designed to assess activity levels in older adults, Berry et al. ([68]) reported the correlation between activities of daily living and VO_{2 peak} to be -0.38. Bendstrup and colleagues ([69]) reported the correlation between scores from the Activities of Daily Living questionnaire and the 6-minute walk distance to be 0.64. Table 3 summarizes the results of studies which have examined the association between exercise tolerance and scores from instruments designed to measure activities of daily living.

Table 3 Correlations between activities of daily living assessed from activity specific instruments and exercise tolerance

Activities of Daily Living	Exercise Tolerance	Correlation
Additive Activities Profile ([38])	VO2 peak	0.83
Additive Activities Profile ([64])	6 MW distance	0.45–0.61
Pulmonary Functional Status Scale ([65])	12 MW distance	0.67
Pulmonary Functional Status Scale ([39])	12 MW distance	0.62
Pulmonary Functional Status Scale ([66])	6 MW distance	0.73
Pulmonary Functional Status Scale and Dyspnea Questionnaire—Function Scale ([40])	VO2 peak	−0.41
Pulmonary Functional Status Scale and Dyspnea Questionnaire—Function Scale ([45])	6 MW distance	0.34
Shortness of Breath Questionnaire ([32])	6 MW distance	−0.68
Shortness of Breath Questionnaire ([56])	6 MW distance	−0.37
	Wmax	−0.34
London Chest Activity of Daily Living ([80])	Shuttle Walk	−0.58
Activities of Daily Living Questionnaire ([69])	6 MW distance	0.64
Physical Function Questionnaire ([68])	VO2 peak	−0.38

6 MW = 6-minute walk, 12 MW = 12-minute walk, VO_{2 peak} = peak oxygen consumption, Wmax = maximal work rate performed on a cycle ergometer.

Relationships between exercise tolerance and activity levels during daily living

Recent advances in the development of motion detectors has led to greater use of these instruments in the assessment of activity levels during daily living. Steele and colleagues recently compared the outputs from triaxial accelerometers (in vector magnitude units) used at home for 3 days to 6-minute walk distance and found a correlation of 0.74 ([46]). Pitta et al. ([47]) recently reported that daily walking time was highly correlated with the 6-minute walk distance (r = 0.76) and more modestly to maximal cycle ergometer work (r = 0.64) and VO_{2 peak} (r = 0.33) ([47]). These results demonstrate that a reduced 6-minute walk distance is a good surrogate marker of reduced physical activity during daily life in COPD patients.

Relationships between exercise tolerance and health-related quality of life

Just as exercise has been used to establish the construct validity of questionnaires assessing activities of daily living, it has also been used to establish construct validity of instruments that determine health-related quality of life. The use of exercise tolerance to help establish construct validity of patient questionnaires is well established, and health-related, quality-of-life instruments and exercise tolerance have both been used as important outcomes in a number of clinical trials with COPD patients. Additionally, exercise training and other forms of physical activity are well-established interventions for the enhancement of both exercise tolerance and health-related quality of life. Based on these facts, the use of exercise tolerance in establishing construct validity seems quite reasonable. For more detailed description of procedures used to establish construct validity see Guyatt et al. ([70]).

Kaplan and colleagues examined the validity of the Quality of Well Being Scale as an outcome measure for COPD patients. They used exercise tolerance, as defined by the time on treadmill during a graded exercise test, to help establish the construct validity of this instrument. The correlation between exercise tolerance and the Quality of Well Being score was 0.41. Following 3 months of exercise training, the correlation was 0.54 ([48]). In a subsequent investigation with the National Emphysema Treatment Trial cohort, Kaplan and colleagues found the correlations between the exercise capacity, as determined from 6-minute walk distance and maximal work capacity on a cycle ergometer, and the Quality of Well Being score to be 0.24 and 0.19, respectively ([56]). Guyatt et al. found the correlation between the Quality of Well Being score and the 6-minute walk distance to be 0.12 ([61]), similar to that reported by Kaplan and colleagues ([56]).

A number of studies have examined the relationship between exercise and health-related quality of life using the Sickness Impact Profile ([49], [52], [59], [60]). Prigatano and colleagues found the correlations between the total and the psychosocial Sickness Impact Profile scores and exercise tolerance as determined from maximal work during a cycle ergometer test to be −0.30 and −0.16, respectively ([60]). Jones and colleagues examined the correlations between 6-minute walk distance and the total and the psychosocial Sickness Impact Profile scores and found them to be −0.64 and −0.50, respectively ([59]). Guyatt and colleagues reported a much weaker correlation between 6-minute walk distance and the total Sickness Impact Profile score (r = 0.16) ([61]). Table 4 summarizes the results of studies that have examined the association between exercise and health-related quality of life as determined from generic instruments.

Table 4 Correlation between generic Health-related Quality of life and exercise tolerance

HRQoL Generic Instruments	Exercise Tolerance	Correlation
Quality of Well-Being ([56])	6 MW distance	0.24
	Wmax	0.19
Quality of Well-Being ([81])	TM time	0.41–0.54
Quality of Well-Being ([52])	6 MW distance	0.12
Sickness Impact Profile ([52])	6 MW distance	0.16
Sickness Impact Profile ([60])	Wmax	−0.30
Sickness Impact Profile ([59])	6 MW distance	−0.65

6 MW = 6-minute walk, HRQoL = Health-related Quality of Life, TM = treadmill, Wmax = maximal work rate performed on a cycle ergometer.

Similar variability in the correlations between exercise capacity and health-related quality of life determined from disease specific instruments have been reported. Jones and colleagues report the correlations between 6-minute walk distance and the symptoms and impact domains and total score from the St. George's Respiratory Disease Questionnaire to be −0.26, −0.59 and −0.61, respectively ([49]). Dourando and colleagues report the correlations between 6-minute walk distance and the impact domain and total score from the St. George's Respiratory Disease Questionnaire to be −0.34 and −0.40, respectively ([63]). Nishiyama and colleagues report the correlation between 6-minute walk distance and the total score from the St. George's Respiratory Disease Questionnaire to be −0.42 ([50]). In 2 separate studies, Hajiro and colleagues examined the relationships between VO_{2 peak} and scores from the St. George's Respiratory Disease Questionnaire ([62], [71]). In both studies a strong correlation was found between the total score and VO_{2 peak}, r = −0.54 ([71]) and r = −0.51 ([62]). When examining the correlation between the impact domain and VO_{2 peak}, the correlation was −0.45 ([62]).

The other commonly used disease specific health-related quality of life instrument is the Chronic Respiratory Disease Questionnaire. Guyatt and colleagues report the correlation between 6-minute walk distance and the domains of fatigue and emotion to be 0.35 and 0.19, respectively ([16]). This same group of investigators later reported the correlation between 6-minute walk distance and the composite Chronic Respiratory Disease Questionnaire score to be 0.52 ([61]). Wijkstra and colleagues report the correlation between 6-minute walk distance and the domains of fatigue, emotion and mastery to be 0.03, 0.02 and 0.25, respectively. Additionally, these same investigators found the correlation between maximal work rate during a cycle ergometer test and the domains of fatigue, emotion and mastery to be 0.01, 0.05 and 0.19, respectively. These results show there to be similar associations between health-related quality of life, assessed using the Chronic Respiratory Disease Questionnaire, and sub-maximal and maximal exercise ([72]).

Belza and colleagues report the correlation between 6-minute walk distance and the composite score from the Chronic Respiratory Disease Questionnaire to be 0.28 ([45]). Bendstrup and colleagues report a similar correlation of 0.18 when comparing 6-minute walk distance to the composite Chronic Respiratory Disease Questionnaire score ([69]).

Several groups of investigators have examined the correlation between Chronic Respiratory Disease Questionnaire scores and VO_{2 peak}. When comparing the domains of fatigue and emotion and the composite score to VO_{2 peak}, Hajiro et al. found correlations of 0.35, 0.25 and 0.40, respectively ([62]). In a subsequent study, these same investigators report a correlation of 0.42 when comparing the composite score with VO_{2 peak} ([71]). Berry et al. report the correlation between VO_{2 peak} and the domains of fatigue, mastery, and emotion to be 0.31, 0.28 and 0.11, respectively ([68]). Table 5 summarizes the results of studies which have examined the association between exercise tolerance and health-related quality of life as determined from the Chronic Respiratory Disease Questionnaire and St George's Respiratory Questionnaire.

Table 5 Correlations between disease specific Health-related Quality of Life and exercise tolerance

HRQoL Disease Specific Instruments	Exercise Tolerance	Correlation
SGRQ ([49])	6 MW distance	−0.61
SGRQ ([63])	6 MW distance	−0.40
SGRQ ([71])	Cycle VO2 peak	−0.54
SGRQ ([62])	Cycle VO2 peak	−0.51
SGRQ ([50])	6 MW distance	−0.42
SGRQ ([56])	6 MW distance	−0.26
	Wmax	−0.23
CRQ ([61])	6 MW distance	0.52
CRQ ([45])	6 MW distance	0.28
CRQ ([69])	6 MW distance	0.18
CRQ ([71])	Cycle VO2 peak	0.42
CRQ ([62])	Cycle VO2 peak	0.40

6 MW = 6-minute walk, HRQoL = Health-related Quality of Life, CRQ = Chronic Respiratory Disease Questionnaire, SGRQ = St. George's Respiratory Disease Questionnaire, VO_{2 peak} = peak oxygen consumption. Wmax = maximal work rate performed on a cycle ergometer.

Other disease specific instruments used to assess health-related quality of life that have been correlated with exercise tolerance include the Seattle Obstructive Lung Disease Questionnaire ([51]), the Airways Questionnaire 20 ([71]), and the Visual Analogue Scale 8 ([50]). Correlations between the emotional function and coping skills domains of the Seattle Obstructive Lung Disease Questionnaire and 6-minute walk distance are 0.26 and 0.12, respectively ([51]). The correlation between the Airways Questionnaire 20 and VO_{2 peak} is −0.49 ([71]). The correlation between the Visual Analogue Scale 8 and 6-minute walk distance, expressed as a percent of predicted, is 0.57 ([50]).

Relationships between changes in exercise tolerance and changes in activities of daily living

The primary goal of most interventions with COPD patients is to improve physical function and health-related quality of life. It has been well established that participation in a pulmonary rehabilitation program with an exercise component will improve physical function and health-related quality of life ([5], [6], [11], [73], [74]). The relationships between changes in exercise capacity and changes in activities of daily living and health-related quality of life have been reported in a number of studies and these are presented next.

Guyatt et al. examined changes in the dyspnea domain of the Chronic Respiratory Disease Questionnaire with changes in 6-minute walk distance following 12 weeks of pulmonary rehabilitation. With the dyspnea domain of the Chronic Respiratory Disease Questionnaire, patients are asked to list activities of daily living that are important to them and are associated with dyspnea. These authors found statistically significant, although weak, correlations when comparing changes in six minute walk distance with changes in the dyspnea domain (r = 0.26) ([52]). de Torres et al. evaluated changes in dyspnea scores from the Chronic Respiratory Disease Questionnaire and in 6-minute walk distance following 6 to 8 weeks of comprehensive pulmonary rehabilitation ([75]). Both the dyspnea domain and 6-minute walk distance increased following rehabilitation. However, when examining the relationship between changes in exercise capacity and changes in the dyspnea domain, no statistically significant association was found, r = 0.21 ([75]). Similarities in the correlations, yet differences in the significance is likely due to the sample size differences.

Garrod et al. examined changes in scores from the London Chest Activity of Daily Living Scale and shuttle walk test following at least 6 weeks of pulmonary rehabilitation ([76]). A weak (r = −0.28) but statistically significant correlation was found between the change in activities of daily living and the change shuttle walk distance. Kaplan et al examined the changes in activities of daily living, as assessed by the physical domain of the SF-36 and the Shortness of Breath Questionnaire, and changes in 6-minute walk distance in National Emphysema Treatment Trial patients following 6 to 10 weeks of pulmonary rehabilitation prior to randomization into the trial ([56]). The physical domain of the SF-36 and the Shortness of Breath Questionnaire increased significantly following pulmonary rehabilitation. Correlations between changes in the 6-minute walk distance and changes in the SF-36 and the Shortness of Breath Questionnaire were weak (r = 0.10 and −0.13, respectively) but statistically significant ([56]).

Belza and colleagues examined the changes in the physical component of the SF-36 and 6-minute walk distance following 8 weeks of pulmonary rehabilitation, which included exercise training ([57]). Both variables were found to improve significantly following the rehabilitation program. However, when examining the correlation between change in 6-minute walk distance and change in the physical component of the SF-36, the correlation was found to be weak (r = 0.06) and non-significant ([57]). Table 6 summarizes the results of studies that have examined the association between changes in exercise tolerance and changes in activities of daily living following participation in a pulmonary rehabilitation program.

Table 6 Correlations between changes in activities of daily living and changes in exercise tolerance measures following exercise training

Activities of Daily Living	Exercise Tolerance	Correlation
SF-36 Physical ([57])	6 MW Distance	0.06
SF-36 Physical ([56])	6 MW Distance	0.10
CRQ Dyspnea ([52])	6 MW Distance	0.26
CRQ Dyspnea ([75])	6 MW Distance	0.21
Shortness of Breath Questionnaire ([56])	6 MW Distance	−0.13
London Chest Activity of Daily Living ([76])	Shuttle Walk	−0.28

6 MW = 6-minute walk, CRQ = Chronic Respiratory Disease Questionnaire.

Relationships between changes in exercise tolerance and changes in health-related quality of life

Ferrari et al. examined changes in health-related quality of life and in exercise tolerance following 12 weeks of minimally supervised home cycle ergometer exercise ([77]). Health-related quality of life was assessed using the SF-36 and exercise tolerance was quantified as VO_{2 peak} from an incremental cycle ergometer test and maximum work rate achieved during the test. Both health-related quality of life and exercise tolerance were found to increase significantly following the exercise intervention. However, when comparing changes in health-related quality of life and changes in exercise tolerance, no significant associations were found ([77]).

In a similar study, Boueri et al. examined changes in health-related quality of life and in exercise tolerance following 3 weeks of comprehensive pulmonary rehabilitation ([78]). Health-related quality of life was assessed using the SF-36 and exercise tolerance was determined from the 6-minute walk. Both health-related quality of life and exercise tolerance were found to increase significantly following the exercise intervention; however, no significant association was found when comparing the changes (r = 0.16) ([78]).

de Torres et al. evaluated changes in health-related quality of life and in exercise tolerance following 6 to 8 weeks of comprehensive pulmonary rehabilitation ([76]). Health-related quality of life was assessed using the SF-36, the Chronic Respiratory Disease Questionnaire and St. George's Respiratory Disease Questionnaire. Exercise tolerance was determined from the 6-minute walk. Following participation in the pulmonary rehabilitation program, 6-minute walk distance increased as did the mastery domain as determined from the Chronic Respiratory Disease Questionnaire. There were no statistically significant improvements in the domains of fatigue or emotion as determined by the Chronic Respiratory Disease Questionnaire. Nor were there statistically significant improvements in the any of the domains measured by the SF-36 or St. George's Respiratory Disease Questionnaire. When examining the relationship between changes in exercise tolerance and changes in health-related quality of life measures no statistically significant associations were found ([75]).

In contrast to the above studies that have failed to demonstrate a relationship between changes in exercise tolerance and changes in health-related quality of life following pulmonary rehabilitation, there are some studies that have demonstrated such a relationship. Kaplan et al. found a statistically significant moderate correlation (r = 0.40) when examining changes in the treadmill time from an incremental exercise test versus changes in Quality of Well Being Scores following 3 months of exercise training ([48]). Kim examined the relationship between changes in 12-minute walk distance following 6 months of inspiratory muscle training and Sickness Impact Profile scores ([79]). A significant improvement was found in 12-minute walk distance following training. Additionally, the changes in 12-minute walk distance were significantly correlated with the baseline Physical Dimension (r = 0.65) and total score of the Sickness Impact Profile (r = 0.50) ([79]). Guyatt et al. examined changes in the various domains of the Chronic Respiratory Disease Questionnaire, the Quality of Well Being score and the overall Sickness Impact Profile Score with changes in 6-minute walk distance following 12 weeks of pulmonary rehabilitation ([52]). These authors found statistically significant improvements in 6-minute walk distance and all the domains of the Chronic Respiratory Disease Questionnaire. The Quality of Well Being score and the Sickness Impact Profile Scores did not significantly improve following pulmonary rehabilitation. When comparing changes in 6-minute walk distance with changes in the domain of mastery, a statistically significant correlation was found (r = 0.28).

Correlations between changes in 6-minute walk distance and changes in the domains of fatigue (r = 0.19) and emotions (r = 0.07) and changes in the Quality of Well Being (r = 0.12) and the Sickness Impact Profile (r = 0.16) were not significant ([52]). Nishiyama et al. examined changes in 6-minute walk distance and changes in the Visual Analogue Scale 8 and the St George's Respiratory Disease Questionnaire following 12 weeks of pulmonary rehabilitation ([50]). All measures were found to improve significantly following pulmonary rehabilitation. Additionally, the correlation between the change in the Visual Analogue Scale 8 score and the 6-minute walk distance was also significant (r = 0.41).

However, the changes in 6-minute walk distance and the changes in the St. George's Respiratory Disease Questionnaire were not significant (r = −0.28) ([50]). Kaplan et al. examined the changes in the mental domain of the SF-36, the Quality of Well-Being Scale and St. George's Respiratory Disease Questionnaire and changes in 6-minute walk distance in National Emphysema Treatment Trial patients following 6 to 10 weeks of pulmonary rehabilitation prior to randomization into the trial ([56]). All measures of health-related quality of life increased significantly following pulmonary rehabilitation. Correlations between changes in the 6-minute walk distance and changes in the mental component of the SF-36, the Quality of Well-Being Scale and St. George's Respiratory Disease Questionnaire were all weak (r = 0.10, 0.11 and −0.15, respectively) but statistically significant ([56]). Table 7 summarizes the results of studies that have examined the association between changes in exercise performance and changes in health-related quality of life following participation in a pulmonary rehabilitation program.

Table 7 Correlations between changes in health-related quality of life and changes in exercise tolerance measures following exercise training

Health-related Quality of Life	Exercise tolerance	Correlation
SF-36 ([77])	Cycle VO2 peak	NS
SF-36 ([78])	6 MW Distance	0.16
SF-36 ([56])	6 MW Distance	0.10
Quality of Well-Being Scale ([48])	TM Time	0.40
Quality of Well-Being Scale ([56])	6 MW Distance	0.11
Quality of Well-Being Scale ([61])	6 MW Distance	0.12
Sickness Impact Profile ([61])	6 MW Distance	0.16
CRQ ([61])	6 MW Distance	0.07–0.28
SGRQ ([50])	6 MW Distance	0.28
SGRQ ([56])	6 MW Distance	−0.15
Visual Analogue Scale 8 ([50])	6 MW Distance	0.21

6 MW = 6-minute walk, CRQ = Chronic Respiratory Disease

Questionnaire, SGRQ = St. George's Respiratory Disease

Questionnaire, VO_{2 peak} = peak oxygen consumption.

SUMMARY AND CONCLUSIONS

The relationships between activity levels during daily living, assessed using motion detectors, and measures of exercise tolerance are relatively strong. Relationships between activities of daily living, assessed using generic health-related quality of life instruments, and exercise tolerance vary considerably with correlations ranging from 0.18–0.72.

The relationships between activities of daily living, assessed using disease specific health-related quality of life instruments, and exercise tolerance also vary considerably with correlations ranging from 0.14–0.59; whereas the relationships between activities of daily living, assessed using activities of daily living instruments, and exercise tolerance are less variable and generally stronger with correlations ranging from 0.34–0.83. Relationships between generic health-related quality of life instruments and exercise tolerance vary considerably (0.19–0.65) as do relationships between disease specific health-related quality of life instruments and exercise tolerance (0.18–0.61). Relationships between changes in activities of daily living and changes in exercise tolerance following pulmonary rehabilitation are very weak to weak (−0.13–0.28). When examining the relationships between changes in health-related quality of life and changes in exercise tolerance following pulmonary rehabilitation, these correlations vary from very weak to moderate with the majority being very weak or weak.

While the majority of the correlations between measures of exercise and activities of daily living and health-related quality of life vary considerably and some are quite weak, this in no way implies that these measures are not valuable outcomes for COPD patients. The fact that these outcomes are not highly correlated may be due to several factors. First, as mentioned previously, if instruments have poor signal-to-noise ratios, this will result in low correlations between the instruments. This is unlikely to be a reason for the low correlations reported in this review since the majority of instruments reviewed here have undergone extensive psychometric testing. A more likely reason is the fact that the various instruments were designed for different purposes and are in deed measuring different constructs. The finding of low correlations among the various measures underscores the fact that each instrument is providing unique information about the patient. Clinicians and researchers should be cognizant of this fact when selecting instruments to assess outcomes in patients.