In the past 20 years, quality of life (QOL), which. Baseline Pulmonary Function and Quality of Life 9 Years Later in a Middle-Aged Chinese Population*

Transcription

1 Baseline Pulmonary Function and Quality of Life 9 Years Later in a Middle-Aged Chinese Population* Gaoqiang Xie, MD, PhD; Ying Li, MD; Ping Shi, MD; Beifan Zhou, MD; Puhong Zhang, MD, PhD; and Yangfeng Wu, MD, PhD Study objective: This research examined the association of baseline pulmonary function with future quality of life (QOL). Methods: We collected baseline pulmonary function data in 1993 and 1994, and assessed QOL using the Chinese 35-Item Quality of Life Instrument in 2002 in a cohort of 1,356 participants. We used Pearson correlation analysis, multivariate analysis of variance, and multivariate linear regression analysis to assess the relationship between pulmonary function and QOL. Results: The baseline percentage of age- and height-predicted FEV 1 (FEV 1 %) was significantly correlated with the resurvey total QOL score (r 0.126, p < 0.001) and with QOL scores for the general (r 0.074, p 0.006), physical (r 0.085, p 0.002), independence (r 0.178, p < 0.001), and psychological (r 0.064, p 0.018) domains but not with the social and environmental domains after adjusting for age and sex. These associations were weaker for the percentage of age- and height-predicted FVC. Multiple linear regression showed that the above associations were independent of baseline and resurvey smoking status. Inclusion of respiratory symptoms in the model reduced the regression coefficients from 0.82 to 0.41 for the total QOL score and from 1.43 to 0.94 for the independence domain score, for a 10% change in FEV 1 %. The age- and sex-adjusted mean total QOL scores were 78, 76, 76, and 69, respectively (p < 0.001), for the groups of normal, symptomatic only, impaired pulmonary function only, and both symptomatic and impaired pulmonary function. This trend was also significant for the general, physical, independence, and psychological domain scores. Conclusion: Impaired baseline pulmonary function has a significant negative impact on QOL in later life that is independent of age, sex, height, and smoking status and is largely mediated through the development of chronic respiratory symptoms. (CHEST 2005; 128: ) Key words: Chinese; prospective study; pulmonary function; quality of life Abbreviations: FEV 1 % percentage of age- and height-predicted FEV 1 ; FVC% percentage of age- and heightpredicted FVC; PRC-USA People s Republic of China/United States; QOL quality of life; QOL-35 Chinese 35-Item Quality of Life Instrument In the past 20 years, quality of life (QOL), which includes general, physical, independence, psychological, social and environmental facets, has been *From the Department of Epidemiology (Drs. Xie, Li, Zhou, Zhang, and Wu), Cardiovascular Institute and Fu Wai Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College; and the Shijingshan Center for Disease Control and Prevention (Dr. Shi), Beijing, People s Republic of China. Supported by the People s Republic of China National Tenth Five-Year Plan Project (grant No. 2001BA703B01). Manuscript received August 6, 2004; revision accepted March 21, Reproduction of this article is prohibited without written permission from the American College of Chest Physicians ( org/misc/reprints.shtml). Correspondence to: Yangfeng Wu, MD, PhD, Department of Epidemiology, Cardiovascular Institute and Fu Wai Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, #167, Beilishi Rd, Xicheng, Beijing, , ROC; Yangfengwu@263.net considered as an important subjective outcome of therapeutic or preventive management and an important predictor for mortality. 1 2 Thus, it is important to explore the predictors of QOL in patients in middle age. Because impaired pulmonary function due to cigarette smoking and other factors is highly prevalent in most countries, 3 it has been extensively studied as For editorial comment see page 1898 a predictor of decreased QOL Such study is of particular significance for China, which is the largest consumer of tobacco in the world. 16 Although many studies, mostly cross-sectional surveys, 4 15 have shown mild-to-moderate positive correlations between pulmonary function and QOL in patients with 2448 Clinical Investigations

2 COPD, asthma, cystic fibrosis, and other lung diseases, Engström and colleagues 6 found that QOL was not significantly affected in patients with mildto-moderate impairment of pulmonary function, possibly due to sufficient pulmonary reserve capacity. However, data are scarce on the association between QOL and pulmonary function in normal populations In this article, the relationship between pulmonary function at baseline and QOL 9 years after baseline in a middle-aged Chinese natural population was investigated to understand the longterm effects of impaired pulmonary function. Materials and Methods The People s Republic of China/United States (PRC-USA) collaborative study of cardiovascular and cardiopulmonary epidemiology was initiated in A detailed description of the goals, design, and methods used appears elsewhere Briefly, we selected four population samples in China: industrial (urban) and agricultural (rural) from both Beijing (North) and Guangzhou (South); each had a population sample of 2,000 middle-aged participants (50% men and 50% women). Four cross-sectional surveys on pulmonary function and other risk factor for heart and lung diseases were done from 1983 to 1984, from 1987 to 1988, from 1993 to 1994, and from 1997 to 1998, respectively. In the 1993-to-1994 survey, serum insulin was added to the laboratory tests for the Beijing urban and rural participants to study its association to cardiovascular risk factors. In 2002, only the Beijing rural population from the PRC-USA joint study who participated in the 1993-to-1994 survey were resurveyed to study the longitudinal association of baseline insulin to incidence of hypertension. This sample included all age-eligible (35 to 64 years) men and women in all 11 villages in the Shijingshan District of Beijing. Of the 2,313 participants with available baseline pulmonary data and information about smoking, 39 participants died and 648 participants had missing baseline serum insulin data (nonresponders due to insufficient volume of serum collected at phlebotomy); the remaining 1,626 participants were invited to the resurvey in Of these 1,626 invited participants, 1,356 consented and underwent the resurvey in 2002, at which time QOL assessment was added to study its associates and predictors. The remaining 270 subjects who did not return for follow-up are considered dropouts. In this article, we studied the association of baseline pulmonary function in this cohort of 1,356 responding and consenting adults with their QOL 9 years later. To evaluate the possible bias induced by the dropouts and nonresponders, a detailed comparison of baseline characteristics such as age, sex, height, smoking status, pulmonary function, and pulmonary diseases was made between the responders and nonresponders and the dropouts. Methods of Measurement We based our baseline and resurvey data collection on a standardized protocol developed in the PRC-USA study, except for the QOL questionnaire. The methods are described below for key variables used in this article. Training and certification of interviewers/technicians and equipment calibration were done according to a detailed manual of procedures. Spirometry Spirometry was performed at baseline in 1993 to 1994 in the sitting position using a Collins Stead-Wells water-filled spirometer (including bell and potentiometer) [Collins 10 Liter Survey II; Warren E. Collins; Braintree, MA], which was interfaced to a portable computer using SPIRO software for quality control. (SPIRO software was developed for the PRC-USA collaborative study by Larry Johnson, Peter Boyle, and Paul Enright). Each participant performed at least three acceptable and two reproducible maneuvers in maximal eight forced expirations. Acceptable maneuvers were defined as those with peak expiratory flow within 10% of the maximum observed, a rapid start, absence of major flow fluctuation, and adequate time of expiration. 20 Reproducible maneuvers agreed within 0.1 L or 5% for FVC and the FEV 1. FEV 1 and FVC values in our analyses were calculated from the best volume-time curve, which was corrected according to room temperature and average pressure of air conditions by computer. These values were used to derive the FEV 1 /FVC ratio. In order to adjust for height, age, and sex, FEV 1 and FVC were divided by the predicted values for each individual and multiplied by 100 (percentage of age- and height-predicted FEV 1 [FEV 1 %] and percentage of age- and height-predicted FVC [FVC%]). The predicted values were based on multiple linear regression models using age and height in men and women, respectively. These models were developed using data from Chinese asymptomatic, nonsmoking men and women in the PRC-USA collaborative study In this article, normal pulmonary function was defined as FEV 1 80% predicted, FVC 80% predicted, and FEV 1 / FVC ratio Respiratory Symptoms At both the baseline survey and the resurvey, individual respiratory symptoms were determined using the same standardized questionnaire. Symptomatic individuals were defined as those with one of the three respiratory symptoms: chronic cough, chronic phlegm, or asthma attack. Chronic cough was defined as a reported chronic cough for at least 3 consecutive months in a year. Chronic phlegm was defined as a reported expectoration of phlegm for at least 3 consecutive months in a year. Asthma attack was defined as having ever reported an attack of shortness of breath with wheezing or asthma that was not associated with a diagnosed pulmonary infection. Quality of life QOL was self-evaluated only in the resurvey using the Chinese 35-Item Quality of Life Instrument (QOL-35). The 35 items in the QOL-35 are classified into six domains plus one item on the individual s self-evaluation of the changes in his/her QOL in the past year. A brief explanation of these domains and items is shown in Appendix 1. Scores for items, domains, and the whole instrument were transformed to the range from 0 (indicating the worst QOL) to 100 points (indicating the best QOL). The QOL-35 was developed from the 100-Item World Health Organization Quality of Life Instrument and the 36-Item Medical Outcomes Study Short-Form Health Status Survey. The QOL-35 was tailored to include only 35 items adapted to the Chinese culture, and was evaluated formally before use in the study. The index was from 0.86 to 1.00 for items in a test-retest survey in 127 adults selected randomly from a Beijing suburban community neighborhood. The Cronbach coefficients of internal consistency reliability were 0.7 for all the six domains. The total QOL score of the QOL-35 had a Pearson correlation coefficient of with the total QOL score of the 100-item World Health Organization Quality of Life Instrument, and of with that of the 36-item Medical Outcomes Study Short- Form Health Status Survey. The reliability and validity of the QOL-35 was thus considered satisfactory. CHEST / 128 / 4/ OCTOBER,

3 Smoking Smoking status was determined for all participants at both baseline and resurvey. A current smoker was defined as an individual currently smoking an average of one or more cigarettes each day or more than one liang (approximately 50 g) of tobacco leaf each month. An ex-smoker was defined as an individual who previously smoked tobacco leaf or cigarettes but was no longer smoking for at least 1 month. Participants who reported that they were neither current smokers nor ex-smokers were classified as having never smoked. Height Body height was measured to the nearest centimeter using a standard right-angle device in both surveys. Each participant was measured standing without shoes. Statistical Methods Pearson correlation coefficients, partial correlation coefficients, and multivariate linear regression analysis were used to examine the relationship between baseline pulmonary function and the QOL scores 9 years later. In addition, we compared the differences of mean QOL scores among the following four groups: normal, impaired pulmonary function only, chronic respiratory symptoms only, and impaired pulmonary function plus chronic respiratory symptoms, using multivariate analysis of variance, adjusting for potential confounders. All analyses were done by statistical software (SPSS version 10.0; SPSS; Chicago, IL). Results Bias From Dropouts and Nonresponders The results shown in Table 1 show that the nonresponders and dropouts were older, were more frequently smokers, had worse pulmonary function, and had more respiratory symptoms. Descriptive Statistics Table 2 gives the descriptive statistics of the major study variables for 1,356 participants taking part in both surveys. Although FEV 1 and FVC were significant higher for men than for women, FEV 1 % and FVC% were significantly lower. More men were smokers, and men had more respiratory symptoms. However, men had significantly higher total QOL scores and scores for physical, independence, and psychological domains than women. Relationship Between Pulmonary Function and QOL Table 3 shows the Pearson correlation coefficients and partial correlation coefficients between baseline pulmonary function and QOL scores at the resurvey. FEV 1 was significantly correlated with scores for general, independence, physical, psychological, environmental domains, and the total QOL scores, but not for social domain and QOL transition item. The corresponding correlation coefficients for FEV 1 % were decreased but still significant for general, independence, physical, psychological domains, and total QOL scores; however, they were no longer significant for the environmental domain. Similar results were found after adjusting for age and sex. For FVC and FVC%, we found very similar but weaker associations. Table 1 Comparison of the Baseline Characteristics Between Participants Taking Part and Not Taking Part in the Resurvey (Dropouts and Nonresponders)* Participants Not Taking Part in Resurvey Variables Participants at Resurvey Total Nonresponders Dropouts Participants, No. 1, Male gender Age, yr 48.4 (8.4) 49.9 (8.5) 51.0 (8.4) 48.5 (8.8) Height, cm (8.0) (8.2) (8.0) (8.6) FEV 1,L 2.56 (0.69) 2.45 (0.74) 2.41 (0.72) 2.53 (0.78) FVC, L 3.37 (0.83) 3.27 (0.87) 3.23 (0.85) 3.37 (0.92) FEV 1 % (18.6) 94.8 (28.0) 93.2 (30.1) 98.7 (21.9) FVC% (17.0) 99.5 (26.4) 97.7 (28.9) (18.6) Ex-smokers Current smokers Chronic cough Chronic phlegm Asthma attacks# Respiratory symptoms** *Data are presented as mean (SD) or %. p 0.05 in comparison to participants at resurvey using univariate analysis of variance or 2 tests. p 0.01 in comparison to participants at resurvey using univariate analysis of variance or 2 tests. p in comparison to participants at resurvey using univariate analysis of variance or 2 tests. Cough on most days or nights for at least 3 consecutive months in the past year. Expectoration of phlegm as sputum production on most days or nights for at least 3 consecutive months in the past year. #Ever having attacks of shortness of breath or asthma with wheezing not associated with a diagnosed pulmonary infection. **Chronic cough, chronic phlegm, and/or asthma attacks Clinical Investigations

4 Table 2 Descriptive Statistics of the Major Study Variables for 1,356 Participants Taking Part in Both Surveys* Variables Men Women Participants, No Variables at baseline Age, yr 48.0 (9.2) 48.2 (7.7) Height, cm 168 (6) 156 (5) FEV 1,L 3.02 (0.70) 2.29 (0.52) FVC, L 4.05 (0.73) 2.97 (0.60) FEV 1 % (18.4) (18.5) FVC% (15.1) (17.9) Ex-smokers Current smokers Chronic cough** Chronic phlegm Asthma attack Respiratory symptoms Variables at resurvey QOL scores General domain 63 (20) 61 (19) Physical domain 81 (17) 75 (18) Independence domain 88 (16) 85 (15) Psychological domain 73 (17) 67 (17) Social domain 76 (16) 75 (16) Environmental domain 65 (20) 66 (21) QOL transition in past year 56 (24) 59 (29) Total QOL scores for all 79 (12) 76 (12) domains Ex-smokers 15.0# 6.2 Current smokers 60.8# 17.9 Chronic cough Chronic phlegm Wheezing or asthma # Respiratory symptoms 18.0# 13.2# *Data are presented as No. (%) or % unless otherwise indicated. p 0.05 for the difference between men and women using t test or 2 test. p 0.01 for the difference between men and women using t test or 2 test. p for the difference between men and women using t test or 2 test. p 0.05 for the difference between baseline and resurvey using t test or 2 test. p 0.01 for the difference between baseline and resurvey using t test or 2 test. #p for the difference between baseline and resurvey using t test or 2 test. **Cough on most days or nights for at least 3 consecutive months in the past year. Expectoration of phlegm as sputum production on most days or nights for at least 3 consecutive months in the past year. Ever having attacks of shortness of breath or asthma with wheezing not associated with a diagnosed pulmonary infection. Cough, chronic phlegm, and/or asthma attack. Figure 1 displays the relationship of FEV 1 %tothe total QOL scores and scores for the independence domain. Generally, the mean total QOL score decreased more and more rapidly with decline of FEV 1 % after adjusting for age and sex. Above 80% of FEV 1 %, the total QOL score did not significantly change with FEV 1 % (10%) [ 0.28, p 0.25]. However, when FEV 1 % (10%) was 80%, the total score started an accelerated decrease ( 2.76, p 0.001). On average, the total QOL score would decrease 0.82 points (p 0.001) when FEV 1 %declines every 10%. FVC% exhibited a similar trend (data not shown). Similar correlation analyses between pulmonary function and QOL were carried out among asymptomatic never-smokers in both baseline survey and resurvey. A significant association was found between baseline pulmonary function variables and score for the independence domain 9 years later. The Pearson correlation coefficient of FEV 1 % with scores for the independence domain was (p 0.014). For FVC%, the corresponding value was (p 0.001). We did not find significant associations between baseline pulmonary function and total QOL scores and QOL scores for other domains among asymptomatic never-smokers (data not shown). Adjustment for Confounders To further understand the role that smoking might play in these associations, we adjusted for smoking status at both baseline and resurvey in our analyses in addition to the age and sex adjustment. The results showed that ex-smokers at resurvey had a significantly lower QOL (total QOL score and score for the independence domain). Ex-smokers at baseline and current smokers at baseline or resurvey did not exhibit this association (data not shown). On the whole, adding smoking status to the model reduced the regression coefficient of FEV 1 % with QOL independence domain score by 5.6% and reduced the coefficient with total QOL score by 7.3% (Table 4). The corresponding values for the general, physical, and psychological domains were 12.7%, 8.3%, and 8.3%, respectively (data not shown). Using the same strategy of analysis, we tested whether the associations between baseline pulmonary function and future QOL was independent of respiratory symptoms. The results showed that there were still significant associations between baseline FEV1% and later QOL scores after adjustment for respiratory symptoms at both baseline and resurvey (Table 4, Appendix 2). However, adding respiratory symptoms into the model reduced the size of the association of FEV 1 % with total QOL score by 50% and reduced the association with QOL score for independent domains by 34% (Table 4). The corresponding values for the general, physical, psychological domains were 43.0%, 73.8%, and 60%, respectively (data not shown). We further classified our study population into the CHEST / 128 / 4/ OCTOBER,

5 Table 3 Pearson Correlation Coefficients and Partial Correlation Coefficients Between Baseline Pulmonary Function and QOL 9 Years Later in Rural Beijing (n 1,356) Baseline Pulmonary Function Variables Domains in Chinese QOL-35 Instrument General Physical Independence Psychological Social Environmental QOL Transition Total QOL Score FEV FEV 1 * FEV 1 % FEV 1 %* FVC FVC* FVC% FVC%* *Adjusted for age and sex. p 0.05 for correlation coefficients. p 0.01 for correlation coefficients. p for correlation coefficients. following four groups according to pulmonary function and respiratory symptoms at baseline: group 1, asymptomatic with normal pulmonary function; group 2, reported respiratory symptoms but normal pulmonary function; group 3, asymptomatic but with abnormal pulmonary function; and group 4, both respiratory symptoms and abnormal function. Again, we defined respiratory symptomatic as having had one or more of the three respiratory symptoms: chronic cough, chronic phlegm, or asthma attack. Table 5 compares the age- and sex-adjusted means of the QOL indexes among the four groups. From group 1 to group 4, there is a significant decreasing trend of QOL scores for total score as well as for general, physical, independence, and psychological domain scores. Discussion In this Chinese cohort, impaired pulmonary function at middle age was found to be significantly associated with future decreased QOL as measured by total QOL score, as well as scores for general, independence, physical, and psychological domains, but not for social and environmental domains and the QOL transition item. These effects were dependent on the extent of baseline impaired pulmonary function but were independent of the confounding effects of age, sex, and height. In addition, these effects were not fully (totally) explained (replaced) by adjustment of smoking status or respiratory symptoms. In this study, all participants came from a community-based population cohort, while previous reported studies 4 15 have been done in patients with COPD, asthma, or nonspecific lung disease. Since the association between pulmonary function and QOL in patients cannot be directly applied to a general normal population, our results demonstrating that impaired pulmonary function predicts lower future QOL in the general population and even in asymptomatic never-smokers are important. We found that this association is not linear but curvilinear. The changes of pulmonary function in the normal range had little effect on future QOL. But when pulmonary function is lower than a threshold value (FEV 1 80% of predicted value in our analysis), further decreases in pulmonary function would affect significantly QOL scores. This is consistent with previous studies 4 7 in which considerable effects on QOL have been demonstrated in severe pulmonary disease but not in the early milder stages of the disease in Western populations. It is clear that impaired pulmonary function leads to difficulties in performing physical activities, such as items in the independence domain which include running, walking, lifting, shopping, doing homework, bathing, and dressing. Noteworthy were the effects on the physical domain, which include pain, sleep, eating, and fatigue. The effects on the psychological domain, which include self-confidence, living pleasure, nervousness, negative feeling (downhearted, despaired, anxiety, melancholy), memory, and attention span, were relatively small but still significant. These indicate almost every aspect of functional status would be affected at lower levels of pulmonary function at baseline. One striking finding was that men had lower ageand height-adjusted lung function and higher QOL. This is in accordance with the results of studies by Osborne et al 21 and Wijnhoven et al, 4 who found that men with asthma or COPD and lower lung function reported a better QOL than women. However, men and women had similar regression coefficients for 2452 Clinical Investigations

6 Figure 1. The trends of mean total QOL score and score for independence domain scores in 1,356 participants, with every 20% increase of FEV 1 % at baseline, after adjustment for age and sex (p 0.05 for linear association). Vertical bars indicate SE. lung function with QOL after adjustment for potential confounders (data not shown). This suggests that the association between lung function and QOL is similar in men and women. The higher QOL in men might be partly attributed to physical fitness, muscle strength, and emotional well-being. The precise mechanisms by which impaired pulmonary function causes reduced QOL are not clear. 22 However, our data may provide some suggestive evidence for the possible mechanisms linking pulmonary function to QOL. After controlling for the confounding effects of age, sex, and height, adding smoking status into the model could only explain approximately 10% of the association between pulmonary function and QOL, but adding respiratory symptoms could explain about half of the association. Nevertheless, it was ex-smokers but not current smokers who had a significant lower QOL in comparison to nonsmokers (data not shown), suggesting that quitting smoking reflected concomitantly poorer health status. In fact, adding respiratory symptoms into the model removed the significance of smoking status (data not shown). Considering the abundance of evidence, 3,19,23 including our own results for a causal relationship between smoking and poor pulmonary function and CHEST / 128 / 4/ OCTOBER,

7 Table 4 Regression Coefficients of FEV 1 % (10%) to Total QOL Score and QOL Score for Independence Domain in Multiple Regression Models After Adjustment for Different Variables Total QOL Score QOL Score for Independence Domain Model Variables Adjusted for Coefficients ( ) t Test for p Value for Coefficients ( ) t Test for p Value for 1 None Age, sex Age, sex, baseline smoking status Age, sex, baseline smoking status, resurvey smoking status 5 Age, sex, baseline respiratory symptoms* Age, sex, baseline respiratory symptoms*, and resurvey respiratory symptoms* *Chronic cough, chronic phlegm, and/or asthma attack. for a causal relation between impaired pulmonary function and respiratory disease, our results do not imply that smoking is not associated to QOL or is associated less stronger than respiratory symptoms. The findings actually support that impaired pulmonary function is caused by smoking, inflammation, or other factors. Impaired pulmonary function further causes poor QOL either by causing respiratory symptoms (as an intermediate, accounting for approximately half of the effect in our study) or by some other mechanisms that we do not understand (accounting for the other half of the effect in our study). Thus, our findings suggest that smoking cessation and other measures that may prevent impairment of pulmonary function should also be effective in preventing poor QOL later in life. This is of importance from the preventive medicine point of view, because better QOL is becoming an important subjective indicator of therapeutic or preventive management with prolonged human longevity. 24 Physical exercise, quitting smoking, decreasing air pollution, and preventing airway infection should be helpful for improving long-term QOL. In addition, our findings in the relationship between baseline pulmonary function and later QOL were established on group data and hence should not be used for individual prediction, giving the relative small correlation coefficients (Table 3) and the relative large spread of the QOL scores (Table 2, Fig 1). This means that an individual with low pulmonary function has a greater but not definite chance of having worse QOL later in life. Although we are not able to predict an individual s QOL in the future, our findings do have a clear message for clinicians. The comparison of QOL between groups of normal, impaired pulmonary Table 5 Comparison of Adjusted QOL Scores Among Baseline Respiratory Disease Status Groups After Adjustment for Age and Sex* Groups by Baseline Pulmonary Function and Symptoms Participants, No. Domains of Chinese QOL-35 Instrument General Physical Independence Psychological Social Environmental QOL Transition Total Scores Normal pulmonary function, asymptomatic Normal pulmonary function, symptomatic Abnormal pulmonary function, asymptomatic Abnormal pulmonary function, symptomatic F value p Value *Data are presented as mean SE. Abnormal pulmonary function was defined as FEV 1 80% of predicted, FVC 80% of predicted, and/or FEV 1 /FVC Symptomatic was defined as having reported a chronic cough, a chronic phlegm, and/or asthma attack. p 0.05 compared with the group with no lung disease. p 0.01 compared with the group with no lung disease. p compared with the group with no lung disease. Analysis of variance Clinical Investigations

8 function only, respiratory symptoms only, and impaired pulmonary function plus respiratory symptoms showed that only participants with both impaired pulmonary function and chronic respiratory symptoms were at high risk for having poor QOL in the future. However, the participants with either impaired pulmonary function only or respiratory symptoms only had a very mild risk of poor QOL in the future. Thus, the clinical use of spirometric testing and questions about chronic respiratory symptoms will help in identifying those who need early intervention. In terms of effective interventions, findings from the recent large clinical trails 23,25,26 were quite disappointing: smoking cessation was the only effective measure found to result in a deceleration of pulmonary function decline. 3,23 However, interventional treatments that are not effective in preventing impaired pulmonary function may be helpful in maintaining higher QOL by relieving chronic respiratory symptoms. Our findings are in agreement with those of Boom et al 25 and Grunsven et al, 26 who found that early treatment with fluticasone propionate was not effective in treatment of pulmonary function decline but was effective in increasing QOL (by reducing dyspnea). 25 Our study has some limitations. First, there were dropouts and nonresponders. The results showed that the nonresponders and dropouts were older, were more frequently smokers, had worse pulmonary function, and had more respiratory symptoms. Thus, the responders were healthier than the original study population. Using only the responders would be expected to dilute the association between baseline pulmonary function and future QOL because the subgroup with lower baseline pulmonary function (dropouts and nonresponders) would have had an even lower QOL than the responders. Thus, a stronger association would be expected if there had been full participation. We did not have QOL measured at baseline and pulmonary function measured at resurvey. This prevented us from being able to better separate the independent effect of baseline pulmonary function and changes of pulmonary function from confounders and inter-mediates. Despite these limitations, our data provide strong support for the conclusion that baseline impaired pulmonary function is associated with decreased future QOL. These effects are dependent on the extent of impaired pulmonary function but are independent of other potential confounders. These results help us to understand the long-term effects of impaired pulmonary function and its modifiable risk factor (tobacco consumption) on QOL. More interventional programs on smoking and other risk factors should be carried out in general population, as well as in those with impaired pulmonary function and chronic respiratory symptoms. ACKNOWLEDGMENT: The authors thank Dr. Robert Detrano for revision of the manuscript and consultation. CHEST / 128 / 4/ OCTOBER,

9 Appendix 1 Description of the Chinese QOL-35 Domains General domain Physical domain Independence domain Psychological domain Social domain Environmental domain QOL transition Description 1. General health status during the past 1 month 2. General quality of life during the past 1 month 1. Bodily pain during the past 1 month 2. Pain interfered with normal life during the past 1 month 3. Appetite during the past month 4. Difficulty sleeping during the past 1 month 5. Fatigue tired during the past 1 month 1. Difficulties in the following activities during the past one month: (A) heavy physical activities; (B) moderate physical activities; (C) lifting daily necessities; (D) climbing several flights of stairs; (E) climbing one flight of stairs; (F) bending, kneeling, or stooping; (G) walking three miles; (H) walking one or two miles; (I) walking around the house; (J) bathing or dressing 2. Need for medicines or treatment in your daily life during the past 1 month 3. Satisfaction in independent living ability during the past 1 month 1. Self-confidence 2. Living pleasure 3. Nervousness 4. Negative feeling (downhearted, despair, anxiety, melancholy) 5. Memory 6. Attention span 1. Connections within the following: (A) family; (B) relatives and friends; (C) colleagues 2. Help or support from your family members or friends in your life during the past 1 month 3. Help or support for your family members or friends during the past 1 month 4. Satisfaction with sex life during the past 1 month 5. Loneliness during the past 1 month 1. Financial condition 2. Condition of residence Compared to 1 year ago, how would you rate your QOL now? Appendix 2 Factors Associated With the QOL Score (Total QOL Score and Score for Independence Domain) in Multiple Variable Regression Analysis Total QOL Score QOL Score for Independence Independent Variables Coefficients ( ) t Test for p Value for Coefficients ( ) t Test for p Value for (Constant) FEV 1 % (10%) Age Sex (1 men, 2 women) Respiratory symptoms at baseline (0 no, 1 yes)* Respiratory symptoms at resurvey (0 no, 1 yes)* *Chronic cough, chronic phlegm, and/or asthma attack. References 1 Fayers PM, Machin D. Quality of life: assessment, analysis and interpretation. 1st ed. Chichester, NY: Wiley & Sons, 2000; Domingo-Salvany A, Lamarca R, Ferrer M, et al. Healthrelated quality of life and mortality in male patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2002; 166: Pauwels RA, Buist AS, Calverley PM, et al. Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease: NHLBI/WHO global initiative for chronic obstructive pulmonary disease (GOLD) workshop summary. Am J Respir Crit Care Med 2001; 163: Wijnhoven HAH, Kriegsman DMW, Hesselink AE, et al. Determinants of different dimensions of disease severity in asthma and COPD: pulmonary function and health-related quality of life. Chest 2001; 119: De JW, Kaptein AA, Van DSCP, et al. Quality of life in patients with cystic fibrosis. Pediatr Pulmonol 1997; 23: Engström CP, Persson LO, Larsson S, et al. Functional status and well being in chronic obstructive pulmonary disease with regard to clinical parameters and smoking: a descriptive and comparative study. Thorax 1996; 51: Kaptein AA, Brand PLP, Dekker FW, et al. Quality of life in a long-term multicentre trial in chronic nonspecific lung disease: assessment at baseline. Eur Respir J 1993; 6: Clinical Investigations

10 8 Cox CE, Donohue JF, Brown CD, et al. Health-related quality of life of persons with sarcoidosis. Chest 2004; 125: Monso E, Fiz JM, Izquierdo J, et al. Quality of life in severe chronic obstructive pulmonary disease: correlation with lung and muscle function. Respir Med 1998; 92: Prigatano GP, Wright EC, Levin D. Quality of life and its predictors in patients with mild hypoxemia and chronic obstructive pulmonary disease. Arch Intern Med 1984; 144: Schrier AC, Dekker FW, Kaptein AA, et al. Quality of life in elderly patients with chronic nonspecific lung disease seen in family practice. Chest 1990; 98: Hajiro T, Nishimura K, Tsukino M, et al. 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: Boueri FM, Bucher-Bartelson BL, Glenn KA, et al. Quality of life measured with a generic instrument (Short Form-36) improves following pulmonary rehabilitation in patients with COPD. Chest 2001; 119: Spencer S, Calverley PMA, Burge PS, et al. Health status deterioration in patients with chronic with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2001; 163: Daughton D, Fix AJ, Kass I, et al. Relationship between a pulmonary function test (FEV 1 ) and the ADAPT quality-oflife scale. Percept Mot Skills 1983; 57: Yang G, Fan L, Tan J, et al. Smoking in China: finding of the 1996 National Prevalence Survey. JAMA 1999; 282: People s Republic of China-United States Cardiovascular and Cadiopulmonary Epidemiology Research Group. An epidemiological study of cardiovascular and cardiopulmonary disease risk factors in four populations in the People s Republic of China: baseline report from the P.R.C.-U.S.A. Collaborative Study. Circulation 1992; 85: Vollmer WM, Tsai R, Wu Y, et al. Pattern of pulmonary function in asymptomatic nonsmoking men and women in the People s Republic of China. Ann Epidemiol 2002; 12: Rao X, Cai R, Huang Z, et al. Effects of smoking on pulmonary function in populations of Beijing and Guangzhou. Zhonghua Jie He He Hu Xi Za Zhi 1996; 19: American Thoracic Society. Standardization of spirometry, 1987 update. Am Rev Respir Dis 1987; 136: Osborne ML, Vollmer WM, Linton KL, et al. Characteristics of patients with asthma within a large HMO: a comparison by age and gender. Am J Respir Crit Care Med 1998; 157: Rahman I, MacNee W. Oxidant/antioxidant imbalance in smokers and chronic obstructive pulmonary disease. Thorax 1996; 51: Anthonisen NR, Connett JE, Kiley JP, et al. Effects of smoking intervention and the use of an inhaled anticholinergic bronchodilator on the rate of decline of FEV 1 : the Lung Health Study. JAMA 1994; 272: Robert P. The future of global ageing. Int J Epidemiol 2002; 31: Boom GVD, Mölken MPMHR, Molema J, et al. The cost effectiveness of early treatment with fluticasone propionate 250 g twice a day in subjects with obstructive airway disease, results of the DIMCA program. Am J Respir Crit Care Med 2001; 164: Grunsven PV, Schermer T, Akkermans R, et al. Short- and long-term efficacy of fluticasone propionate in subjects with early signs and symptoms of chronic obstructive pulmonary disease: results of the DIMCA study. Respir Med 2003; 97: CHEST / 128 / 4/ OCTOBER,