American Journal of Respiratory and Critical Care Medicine

This study was undertaken to test the hypothesis that a reduction in midthigh muscle cross-sectional area obtained by CT scan (MTCSACT) is a better predictor of mortality in chronic obstructive pulmonary disease (COPD) than low body mass index (BMI). We also wished to evaluate whether anthropometric measurements could be used to estimate MTCSACT. One hundred forty-two patients with COPD (age = 65 ± 9 years, mean ± SD, 26 F, BMI = 26 ± 6 kg/m2, FEV1 = 42 ± 16% predicted) were recruited from September 1995 to April 2000 with a mean follow-up of 41 ± 18 months. The primary end-point was all-cause mortality during the study period. A Cox proportional hazards regression model was used to predict mortality using the following independent variables: age, sex, daily use of corticosteroid, FEV1, DlCO, BMI, thigh circumference, MTCSACT, peak exercise workrate, PaO2, and PaCO2. Only MTCSACT and FEV1 were found to be significant predictors of mortality (p = 0.0008 and p = 0.01, respectively). A second analysis was also performed with FEV1 and MTCSACT dichotomized. Patients were divided into four subgroups based on FEV1 (< or ⩾ 50% predicted) and MTCSACT (< or ⩾ 70 cm2). Compared with patients with an FEV1 ⩾ 50% predicted and a MTCSACT ⩾ 70 cm2, those with an FEV1 < 50% predicted and a MTCSACT ⩾ 70 cm2 had a mortality odds ratio of 3.37 (95% confidence interval, 0.41–28.00), whereas patients with an FEV1 < 50% predicted and a MTCSACT < 70 cm2 had a mortality odds ratio of 13.16 (95% confidence interval, 1.74–99.20). MTCSACT could not be estimated with sufficient accuracy from anthropometric measurements. In summary, we found in this cohort of patients with COPD that (1) MTCSACT was a better predictor of mortality than BMI, and (2) MTCSA had a strong impact on mortality in patients with an FEV1 < 50% predicted. These findings suggest that the assessment of body composition may be useful in the clinical evaluation of these patients.

Body weight loss adversely affects survival in several chronic diseases such as chronic obstructive pulmonary disease (COPD) (1, 2), chronic heart failure (3), cystic fibrosis (4), and acquired immune deficiency syndrome (AIDS) (5). Although the relationship between emaciation and mortality in chronic diseases is well established, the nature of the link between these two states is not known. In this regard, it is unclear which body compartment loss has more relevance to survival. Given the preferential loss of muscle over other body compartments in wasting related to chronic diseases (6, 7) including COPD (8, 9), we can speculate that the loss of muscle tissue may have more prognostic implication than the loss of other body compartments.

Although body weight is a useful prognosis marker in COPD, it has a number of limitations in estimating muscle mass. For instance, body weight is not sensitive to changes in body composition as it can be normal or even increased despite the presence of muscle wasting (10). Dissociation between body weight and muscle mass can be found in obese subjects or in patients with fluid retention. In these common clinical conditions, a normal body weight may be misleading in suggesting the absence of muscle wasting and falsely reassuring in regards to survival.

Given these potential limitations of body weight measurements, we reasoned that muscle mass could be a better predictor of mortality than body weight, which is insensitive to change in body composition. To address this question, computed tomography (CT) scan of the midthigh was obtained in 142 patients with COPD who were followed for up to 6 years to study the relationship between muscle mass and other clinical parameters and the risk of death. We also wished to determine the usefulness of anthropometric measurements to predict muscle mass in this population.

Subjects

One-hundred forty-two patients with stable COPD were recruited from September 1995 to April 2000 as they entered a 12-week pulmonary rehabilitation program. The diagnosis of COPD was based on current or past smoking history, clinical evaluation, and pulmonary function tests (11). They were followed for a mean of 41 ± 18 months up to the time of data analysis in November 2001. The research protocol was approved by the institutional ethics committee and a written consent form was obtained for each patient.

Protocol
Anthropometric measurements.

Body weight and height were obtained in all subjects. Midthigh circumference and quadriceps skinfold thickness were also measured with standard methods in 118 subjects (12).

CT of the thigh.

A CT scan of the right thigh halfway between the pubic symphisis and the inferior condyle of the femur was performed using a fourth generation Toshiba Scanner 900S (Toshiba, Inc., Tokyo, Japan). The midthigh muscle cross-sectional area (MTCSACT) was obtained as previously described (8).

Pulmonary function tests and arterial blood gases.

Standard pulmonary function tests including spirometry, lung volumes, and DlCO were obtained according to previously described guidelines (11), and related to normal values of Knudson and colleagues (13), Goldman and Becklake (14), and Cotes and Hall (15), respectively. Due to technical limitations, DlCO could only be measured in 119 patients. In 129 patients, arterial PO2, Pco2, and pH were measured with a blood gas machine (AVL 995; AVL Scientific Corp., Roswell, GA).

Exercise test.

All subjects performed a symptom-limited progressive exercise test. After 5 minutes of resting, a progressive and symptom-limited stepwise exercise test was performed starting with a workrate of 10 watts and breathing room air. Because peak VO2 could not be obtained in all subjects due to technical reasons, peak workrate (Wpeak) was used as the index of peak exercise capacity. Wpeak was defined as the highest exercise workload tolerated for at least 30 seconds.

Survival status.

The medical chart of each patient was reviewed. The survival status and the date of mortality, when appropriate, were noted. For patients in whom the survival status and mortality date could not be established from the medical chart, this information was obtained from the Régie de l'Assurance Maladie du Québec (RAMQ), a provincial organization to which all deaths occurring in the province of Québec are reported. For patients who died in our hospital, the exact cause of mortality could be established.

Statistical analysis.

Descriptive statistics were used to describe the study population at baseline. We first conducted univariate analyses based on the Cox proportional hazards model using each of the potential predictors of mortality as independent variables (age, sex, daily use of oral corticosteroids, FEV1, DlCO, body mass index [BMI], thigh circumference, MTCSACT, Wpeak, PaO2, and PaCO2) and the survival status as the dependent variable. Independent variables that were associated with mortality with a p < 0.15 in the univariate analyses were then incorporated into a multivariate analysis also based on the Cox proportional hazards model. A first multivariate analysis was done with FEV1 and MTCSACT considered as continuous variables. A second analysis was performed with FEV1 and MTCSACT dichotomized to facilitate data interpretation. The best cutoff points for FEV1 and MTCSACT were determined using a change-point model for hazard function as proposed by Muller and Wang (16). The maximum likelihood ratio chi-square value was used to estimate the values where there was a shift in the hazard function. The threshold in FEV1 and MTCSACT that provided the strongest association with mortality was 50% predicted and 70 cm2, respectively. Because MTCSACT is not readily available in current clinical practice, we sought to estimate it from simple anthropometric parameters (12, 17). We also developed our own quadratic regression model to estimate MTCSA (MTCSAEST). This was done using regression analyses and using a cross-validation technique to avoid overfitting of the data. To this end, the cohort was divided into 10 subgroups (9 subgroups of 14 and one subgroup of 16). A regression model was obtained from 9 subgroups and was validated in the last group. This procedure was repeated iteratively, 10 times. The equation providing the best average results and its correlation coefficient are reported. The level of agreement between MTCSACT and MTCSAEST was evaluated using the Bland-Altman approach (18). Throughout the analyses, the level of statistical significance was set at the 0.05 level.

Subject Characteristics

The characteristics of the study population are presented in Table 1

TABLE 1. Subject characteristics (n = 142)

Age, yr 65 ± 9
Sex, F/M26/116
BMI, kg/m2 26 ± 6
MTCSACT , cm273 ± 18
FVC, L2.52 ± 0.79
FVC, % predicted64 ± 16
FEV1, L1.11 ± 0.47
FEV1, % predicted42 ± 16
FEV1/FVC, %44 ± 12
TLC, % predicted 119 ± 23
DlCO, % predicted65 ± 28
PaO2, mm Hg84 ± 14
PaCO2, mm Hg 41 ± 6
Peak workrate, watts59 ± 29
Peak workrate, % predicted46 ± 19
Death, no
25

Definition of abbreviations: BMI = body mass index; MTCSACT = midthigh muscle cross-sectional area obtained by CT scan.

Values are mean ± SD.

. Patients had on average moderate to severe airflow obstruction with hyperinflation and reduced DlCO. BMI averaged 26 ± 6 kg/m2 with a mean MTCSACT of 73 ± 18 cm2. For comparison, the mean MTCSACT in 36 healthy individuals of similar age (62 ± 5 years) evaluated in our laboratory was 102 ± 18 cm2. Group mean values for arterial PaO2 and PaCO2 were within normal limit. Moderate to severe decrease in work capacity was observed with Wpeak averaging 59 ± 29 watts or 46 ± 19% of predicted value. Five patients were on long-term oxygen therapy. No patients had α1-antitrypsin deficiency, although 12 patients received chronic oral corticosteroid therapy (5–10 mg of prednisone/day). The mean duration of follow-up was 41 ± 18 months (range = 18 to 73 months). During this period there were 25 deaths, representing a mortality rate of 17.6%. Among the 15 patients for whom the exact cause of death could be ascertained, 10 died of advanced COPD, 2 of coronary artery disease, and 3 of cancer that had not been diagnosed at the time of study entry.

Prediction of Mortality

The univariate analyses indicated that age, sex, FEV1% predicted, BMI, thigh circumference, MTCSACT, Wpeak, and PaCO2 were related to mortality with a p < 0.15 (Table 2)

TABLE 2. Predictors of mortality: univariate analyses




Hazard Ratio

95% CI

p Value
Age, yr1.040.99–1.100.1151
Sex, F/M2.060.86–4.950.1047
FEV1, % predicted0.960.93–0.990.0050
BMI, kg/m20.920.84–1.000.0577
Thigh circumference, cm0.940.90–0.980.0084
MTCSACT, cm20.960.94–0.980.0003
Peak workrate, % predicted0.980.96–1.000.0801
PaCO2, mm Hg
1.09
1.01–1.16
0.0214

Definition of abbreviations: BMI = body mass index; CI = confidence interval; MTCSACT = midthigh muscle cross-sectional area obtained by CT scan.

. These variables were included in the multivariate prediction model. It was found that MTCSACT was the variable with the strongest inverse relationship with mortality (p = 0.0008). FEV1 % predicted was the only other variable with a statistically significant relationship to mortality (p = 0.01). The addition of any other variables into the model did not improve its ability to predict mortality. To facilitate the interpretation of the data, the multivariate analysis was also repeated with dichotomized variables (Table 3)

TABLE 3. Predictors of mortality: multivariate analysis




Hazard Ratio

95% CI

p Value
MTCSACT < 70 cm23.681.52–8.090.0038
FEV1 < 50% predicted
4.78
1.12–20.34
0.0342

Definition of abbreviations: CI = confidence interval; MTCSACT = midthigh muscle cross-sectional area obtained by CT scan.

. A MTCSACT < 70 cm2 was associated with a fourfold increase (95% confidence interval, 1.52–8.09) in mortality rate, independently of any other variables (p = 0.004).

Interaction between MTCSACT and FEV1

Patients were divided into four subgroups based on FEV1 (< or ⩾ 50% predicted) and MTCSACT (< or ⩾ 70 cm2) (Table 4)

TABLE 4. Interaction between fev1 and mtcsaCT




No Deaths/
 N of Patients

Hazard Ratio

95% CI
FEV1 ⩾ 50% and MTCSACT ⩾ 70 cm21/291 (referent)
FEV1 ⩾ 50% and MTCSACT < 70 cm21/162.140.13–34.4
FEV1 < 50% and MTCSACT ⩾ 70 cm26/513.370.41–28.00
FEV1 < 50% and MTCSACT < 70 cm2
17/46
13.16
1.74–99.20

For definition of abbreviations, see Table 3.

. Compared with patients with an FEV1 ⩾ 50% predicted and a MTCSACT ⩾ 70 cm2 (reference group), those with an FEV1 ⩾ 50% predicted and a MTCSACT < 70 cm2 and those with an FEV1 < 50% predicted and a MTCSACT ⩾ 70 cm2 tended to have an increased mortality with a two- and threefold increase in mortality rate, respectively. The combination of low FEV1 and MTCSACT had a profound impact on the mortality odds ratio, which rose to 13.16 (95% confidence interval, 1.74–99.20). When considering only the subgroup of patients with an FEV1 < 50% predicted in the analysis, MTCSACT was also a significant determinant of mortality: compared with patients with a MTCSACT ⩾ 70 cm2, the odds ratio of dying was 3.91 (95% confidence interval, 3.54–4.31) when MCSACT was smaller than 70 cm2. Survival curves for each of the four subgroups of patients are provided in Figure 1 .

Estimation of MTCSACT

We wished to evaluate the ability of three potentially useful clinical indicators of muscle mass (BMI, midthigh circumference, and MTCSAEST) to predict MTCSACT. The correlation between BMI, midthigh circumference, and MTCSACT was weak, with an r2 of 0.16 and 0.18, respectively. The ability of two previously published equations (12, 17) to estimate MTCSACT was also modest, with an r2 of 0.27 and 0.25, respectively. The following equation was found to be the best predictor of MTCSACT in our patients: MTCSAEST = 905 + 16 × sex − 63 × x1 − 2.2 × x2 + 79 × (x1/x2) + 1.48 × x12 – 0.01 × x13 (where x1 = mid thigh circumference and x2 = quadriceps skinfold; sex: male = 1, female = 0).

The correlation between MTCSACT and MTCSAEST was moderate but statistically significant (r2 = 0.40, p < 0.001). The level of agreement between MTCSACT and MTCSAEST was evaluated using the Bland-Altman approach (Figure 2)

. The difference between MTCSAEST and MTCSACT was plotted against the mean value of MTCSA ([MTCSACT + MTCSAEST]/2). The mean difference between MTCSAEST and MTCSACT was 0 ± 13 cm2. There was a systematic variation in the difference between MTCSAEST and MTCSACT; in patients with a small thigh, MTCSAEST overestimated MTCSACT, whereas the opposite was true in patients with larger muscle mass.

Body weight is a prognostic factor in patients with COPD (1, 2). The present study extents this notion by showing that midthigh muscle cross-sectional area, an index of muscle mass, is more closely related to survival than body weight. This suggests that the loss of muscle has more implication for prognosis than the loss of other body compartments and that assessment of body composition may be useful in the clinical evaluation of patients with COPD. A low MTCSACT had a strong impact on mortality in patients with severe COPD, emphasizing the multisystemic nature of COPD. The second important message of the present study is that the prediction of MTCSACT from anthropometric parameters was not sufficiently accurate for clinical purposes in patients with COPD.

On average, our patients had slightly elevated BMI, whereas their MTCSACT represented 72% of normal values in our laboratory. The preferential loss of muscle mass over body weight found in this and other studies (8, 9) may have important implications as to the origin of wasting in COPD. For instance, a preferential loss of muscle tissue over fat with enhanced protein degradation presumably related to systemic inflammation are characteristic features of cachexia (6, 7). This is in contrast to reduced nutritional intake or starvation, in which adaptive mechanisms tend to preserve muscle tissue (6, 7). Another interesting issue pertains to the clinical relevance of a reduced muscle mass when total body weight is preserved. Low fat-free mass has been associated with functional impairment and poor quality of life independently of total body weight (19, 20). Compared with patients with preserved muscle mass, evidence of systemic inflammation with increased blood level of cytokines was found in patients with COPD and reduced muscle mass, irrespective of body weight (21). Our results extend these observations by showing that low muscle mass may have more important prognostic implications than body weight. Although correlation analyses never prove causality, a mechanistic link between muscle wasting and survival is biologically plausible because muscle tissue is an important pool of essential amino acids. A decrease in this amino acid reservoir may alter several key physiologic functions such as tissue regeneration and immune defense (7). Altogether, these data confirm that muscle wasting may occur independently from body weight loss, and suggest that the physiopathology and consequences of muscle wasting seem to be similar whether or not body weight is normal. Accordingly, body composition assessment should be more informative than simple body weight measurements in the clinical evaluation of COPD.

Several methods can be used to assess body composition, but it is unclear which one represents the best approach. Apart from selected groups of patients such as lung transplantation candidates, we recognize that a CT scan of the lower limb would be difficult to incorporate routinely in the evaluation of all patients with COPD. This issue was addressed by evaluating whether anthropometric measurements could accurately predict MTCSACT. According to the above discussion, the weak correlation between BMI and muscle mass was somewhat expected, implying that the magnitude of muscle wasting could not be adequately predicted by body weight measurements. Similarly, anthropometric measurements did not predict MTCSACT with sufficient accuracy for clinical purposes. Even using our own equation to estimate MTCSACT, the difference between MTCSAEST and MTCSACT could reach as much as 26 cm2, representing approximately a 30% error. Anthropometric equations to predict thigh muscle mass are based on several assumptions that are probably invalid in the majority of patients. For instance, it can be easily appreciated from CT scan (8) that the midthigh cross-section is not circular and that fat is not evenly distributed around thigh muscles. Lastly, the amount of fat infiltration within the muscle cannot be taken into account in anthropometric estimates of muscle mass.

Among other available methods of assessing body composition, dual-energy X-ray absorptiometry (DEXA) is appealing because it may provide, in a single study, precise quantification of body compartments including bone mineral and soft tissue, which can be further divided into fat and fat-free tissue (22). Another interesting feature of DEXA scanning is that tissue composition can be obtained for the whole body and also specifically for each limb. As DEXA scanning is becoming more readily available, this technique may become a useful clinical tool to assess body composition and quantify the degree of osteoporosis, another common problem in COPD (23). Because of its simplicity, biolectrical impedance is another promising method to study body composition. However, this technique may be confounded by the presence of fluid shift and retention (24). Further work should be done to evaluate and compare the potential advantages and disadvantages of the currently available methods to study body composition.

The additive effects of FEV1 and MTCSACT on mortality (Table 4) suggest that estimation of prognosis in COPD could be improved by incorporating parameters reflecting the impairment in lung function with others accounting for systemic consequences of the disease into a patient's evaluation. An important finding was that MTCSACT could be used to discriminate for mortality patients with severe airflow obstruction. In patients with an FEV1 lower than 50% predicted, an MTCSACT lower than 70 cm2 was a powerful predictor of mortality. In the subgroup of patients that could be considered as potential lung transplant candidate based on their FEV1 (FEV1 ⩽ 25% predicted), the mortality rate was 50% (5/10) in patients with MTCSACT lower than 70 cm2, and 12% (2/17) in those with a MTCSACT greater than 70 cm2. Although patients of the present study were on average older than those usually considered for transplantation, our results suggest that that the presence of low muscle mass could help determine the adequate timing for lung transplantation in advanced emphysema. This is often a difficult task as exemplified by a recent study showing that, although lung transplantation should be considered when expected to confer a survival advantage over conventional treatment (25), it may, in fact, be difficult to identify patients with emphysema whose survival will be improved by transplantation (26). Survival is difficult to predict in patients with advanced COPD, and it would be extremely useful to enable clinicians to do so with better accuracy.

Wasting associated with COPD has been recognized for years by clinicians (27), but its relevance to patients' outcome and management has been overlooked. The poor attention to body weight and muscle loss in chronic diseases can be attributed, in part, to the lack of effective treatment. However, the clinical importance of the systemic manifestations related to this disease is becoming more widely recognized (28). Now that the negative impact of these manifestations for patient survival has been confirmed (1, 2), there is a renewed interest in this problem in respiratory and other chronic diseases. Our understanding of how muscle mass is maintained in chronic diseases is also evolving (29, 30). New and more effective therapies aimed at improving muscle mass should become available in the near future (30). This will likely impact our approach to patients with COPD because, as implied by our study and those of others (2), improvement in muscle mass should result in better survival despite the absence of beneficial effects on lung function.

Methodologic Considerations

We acknowledge the retrospective nature of this study and that a prospective evaluation of the prognostic value of muscle wasting will have to be undertaken to confirm the present findings. In this study, CT scan of the thigh was used to assess muscle mass because of its simplicity and accuracy (8, 31). This methodology was also selected because previous studies indicate that, in COPD, muscle dysfunction is unevenly distributed among muscle groups and is more evident in the lower limb than in the upper limb muscles (8, 3234). Accordingly, the thigh appears as an area of predilection to evaluate muscle wasting in COPD. However, one potential limitation of using MTCSACT in the clinical evaluation of COPD is that there are no reference values for this variable. Because muscle mass is influenced by age and sex (35) among other things, the ability to predict mortality with MTCSACT was probably influenced by using absolute values of MTCSA. However, it is important to note that potential influence of age, sex, and body stature on survival was taken into account because these variables were included in the multivariate analysis.

Conclusion

In summary, midthigh muscle area was a strong predictor of mortality in this cohort of patients with COPD. The assessment of body composition in the clinical evaluation of patients with COPD may prove to be useful in predicting outcome in these individuals. Although the present results need prospective confirmation in a larger and multicenter trial, they nevertheless support the concept that systemic manifestations of COPD such as muscle wasting should be taken into account in the clinical evaluation of patients with this disease.

The authors thank Marthe Bélanger, Marie-Josée Breton, Gisèle Deshaies, Anita LeBlanc, Jacqueline Lepage, and Jacynthe Rondeau for their assistance, Gaëtan Daigle for the statistical analysis, and Drs. Simon Martel and Bruno Raby for helpful comments on the manuscript. The authors would also like to thank the anonymous reviewers for their excellent suggestions.

This work has been supported by CIHR grant no 36331.

1. Wilson DO, Rogers RM, Wright EC, Anthonisen NR. Body weight in chronic obstructive pulmonary disease: The National Institutes of Health Intermittent Positive-Pressure Breathing Trial. Am Rev Respir Dis 1989;139:1435–1438.
2. 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.
3. Anker SD, Ponikowski P, Varney S, Peng Chua T, Clark AL, Webb-Peploe KM, Harrington D, Kox WJ, Poole-Wilson PA, Coats AJS. Wasting as independent risk factor for mortality in chronic heart failure. Lancet 1997;349:1050–1053.
4. Ionescu AA, Nixon LS, Evans WD, Stone MD, Lewis-Jenkins V, Chatham K, Shale DJ. Bone density, body composition, and inflammatory status in cystic fibrosis. Am J Respir Crit Care Med 2000;162: 789–794.
5. Moldawer LL, Sattler FR. Human immunodeficiency virus-associated wasting and mechanism of cachexia associated with inflammation. Semin Oncol 1998;25:73–81.
6. Tisdale MJ. Biology of cachexia. J Natl Cancer Inst 1997;89:1763–1773.
7. Kotler DP. Cachexia. Ann Intern Med 2000;133:622–634.
8. Bernard S, Leblanc P, Whittom F, Carrier G, Jobin J, Belleau R, Maltais F. Peripheral muscle weakness in patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1998;158:629–634.
9. Pouw EM, Schols AMWJ, Deutz NEP, Wouters EFM. Plasma and muscle amino acid levels in relation to resting energy expenditure and inflammation in stable chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1998;158:797–801.
10. Schols AMWJ, Soeters PB, Dingemans MC, Mostert R, Frantzen PJ, Wouters EFM. Prevalence and characteristics of nutritional depletion in patients with stable COPD eligible for pulmonary rehabilitation. Am Rev Respir Dis 1993;147:1151–1156.
11. American Thoracic Society. Standards for the diagnosis and care of patients with chronic obstructive pulmonary disease (COPD) and asthma. Am Rev Respir Dis 1987;136:225–244.
12. Heymsfield SB, McManus C, Smith J, Stevens V, Nixon DW. Anthropometric measurements of muscle mass: revised equations for calculating bone-free muscle area. Am J Clin Nutr 1982;36:680–690.
13. Knudson RJ, Slatin RC, Lebowitz MD, Burrows B. The maximal expiratory flow-volume curve: normal standards, variability and effects of age. Am Rev Respir Dis 1976;113:587–600.
14. Goldman HI, Becklake MR. Respiratory function tests: normal values at median altitudes and the prediction of normal results. Am Rev Tuberc 1959;79:454–467.
15. Cotes JE, Hall AM. The transfer factors for the lung; normal values in adults. In: Arcangeli P, editor. Normal values for respiratory function in man. Torino: Panminerva Medica; 1970. p. 327–343.
16. Muller HG, Wang JL. Estimation in change-point models for hazard functions. In: Carlstein E, Muller HG, Siegmund D, editors. Change-point problems. Hayward, CA: Institute of Mathematical Statistics; 1994. p. 224–241.
17. Housh DJ, Housh TJ, Weir JP, Weir LL, Johnson GO, Stout JR. Anthropometric estimation of thigh muscle cross-sectional area. Med Sci Sports Exerc 1995;27:784–791.
18. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;8476:307–310.
19. Schols AMWJ, Mostert R, Soeters PB, Wouters EFM. Body composition and exercise performance in patients with chronic obstructive pulmonary disease. Thorax 1991;46:695–699.
20. Mostert R, Goris A, Weling-Scheepers C, Wouters EFM, Schols AMW. Tissue depletion and health related quality of life in patients with chronic obstructive pulmonary disease. Respir Med 2000;94:859–867.
21. Eid AA, Ionescu AA, Nixon LS, Lewis-Jenkins V, Mathews SB, Griffiths TL, Shale DJ. The inflammatory response and body composition in chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2001;164:1414–1418.
22. Visser M, Fuerst T, Lang T, Salamone L, Harris TB. Validity of fan-beam dual-energy X-ray absorptiometry for measuring fat-free mass and leg muscle mass. J Appl Physiol 1999;87:1513–1520.
23. McEvoy CE, Ensrud KE, Bender E, Genant HK, Yu W, Griffith JM, Niewoehner DE. Association between corticosteroid use and vertebral fractures in older men with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1998;157:704–709.
24. Schols AMWJ, Wouters EFM, Soeters PB, Westerterp KR. Body composition by bioelectrical-impedance analysis compared with deuterium dilution and skinfold anthropometry in patients with chronic obstructive pulmonary disease. Am J Clin Nutr 1991;53:421–424.
25. Trulock EP. Lung transplantation. Am J Respir Crit Care Med 1997;155: 789–818.
26. Hosenpud JD, Bennett LE, Keck BM, Edwards EB, Novick RJ. Effect of diagnosis on survival benefit of lung transplantation for end-stage lung disease. Lancet 1998;351:24–27.
27. Sadoul P. Les cachexies respiratoires. Bull Physiopathol Resp 1969;5:3–11.
28. Agusti AG. Systemic effects of chronic obstructive pulmonary disease. Novartis Found Symp 2001;234:242–249.
29. Buck M, Chojkier M. Muscle wasting and dedifferentiation induced by oxidative stress in a murine model of cachexia is prevented by inhibitors of nitric oxide synthesis and antioxidants. EMBO J 1996;15:1753–1765.
30. Musaro A, McCullagh K, Paul A, Houghton L, Dobrowolny G, Molinaro M, Barton ER, Sweeney HL, Rosenthal N. Localized Igf-1 transgene expression sustains hypertrophy and regeneration in senescent skeletal muscle. Nat Genet 2001;27:195–200.
31. Sipilä S, Suominen H. Muscle ultrasonography and computed tomography in elderly trained and untrained women. Muscle Nerve 1993;16: 294–300.
32. Gosselink R, Troosters T, Decramer M. Peripheral muscle weakness contributes to exercise limitation in COPD. Am J Respir Crit Care Med 1996;153:976–980.
33. Newell SZ, McKenzie DK, Gandevia SC. Inspiratory and skeletal muscle strength and endurance and diaphragmatic activation in patients with chronic airflow limitation. Thorax 1989;44:903–912.
34. Gea J, Pasto M, Blanco L, Barreiro E, Hernandez N, Palacio J, Broquetas J. Biomolecular metabolism in a peripheral skeletal muscle of the upper limbs in COPD patients. Am J Respir Crit Care Med 1999;159:A579.
35. Rogers MA, Evans WJ. Changes in skeletal muscle with aging: effects of exercise training. In: Holloszy JO, editor. Exercise and Sports Sciences Reviews, 21st ed. Philadelphia: Williams and Wilkins; 1993. p. 65–102.
Correspondence and requests for reprints should be addressed to Dr. François Maltais, Centre de Pneumologie, Hôpital Laval, 2725 Chemin Ste-Foy, Ste-Foy, PQ, G1V 4G5 Canada. E-mail:

Related

No related items
American Journal of Respiratory and Critical Care Medicine
166
6

Click to see any corrections or updates and to confirm this is the authentic version of record