Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Lung function and peak oxygen uptake in chronic obstructive pulmonary disease phenotypes with and without emphysema

  • Øystein Rasch-Halvorsen ,

    Roles Conceptualization, Formal analysis, Methodology, Writing – original draft, Writing – review & editing

    oystein.rasch-halvorsen@ntnu.no

    Affiliations Department of Circulation and Medical Imaging, Faculty of Medicine and Health Sciences, NTNU, Norwegian University of Science and Technology, Trondheim, Norway, Clinic of Thoracic and Occupational Medicine, St. Olavs Hospital, Trondheim University Hospital, Trondheim, Norway

  • Erlend Hassel ,

    Contributed equally to this work with: Erlend Hassel, Ben M. Brumpton, Martijn A. Spruit, Arnulf Langhammer, Sigurd Steinshamn

    Roles Conceptualization, Methodology, Writing – original draft, Writing – review & editing

    Affiliation Norwegian Armed Forces Occupational Health Service, Trondheim, Norway

  • Ben M. Brumpton ,

    Contributed equally to this work with: Erlend Hassel, Ben M. Brumpton, Martijn A. Spruit, Arnulf Langhammer, Sigurd Steinshamn

    Roles Conceptualization, Methodology, Writing – original draft, Writing – review & editing

    Affiliations Clinic of Thoracic and Occupational Medicine, St. Olavs Hospital, Trondheim University Hospital, Trondheim, Norway, K.G. Jebsen Center for Genetic Epidemiology, Department of Public Health and Nursing, Faculty of Medicine and Health Sciences, NTNU, Norwegian University of Science and Technology, Trondheim, Norway, MRC Integrative Epidemiology Unit, School of Social and Community Medicine, University of Bristol, Bristol, United Kingdom

  • Haldor Jenssen,

    Roles Investigation, Writing – original draft, Writing – review & editing

    Affiliation Telemark Heart Lung and Blood Institute, Skien, Norway

  • Martijn A. Spruit ,

    Contributed equally to this work with: Erlend Hassel, Ben M. Brumpton, Martijn A. Spruit, Arnulf Langhammer, Sigurd Steinshamn

    Roles Conceptualization, Methodology, Writing – original draft, Writing – review & editing

    Affiliations Department of Research and Development, CIRO, Horn, The Netherlands, Department of Respiratory Medicine, Maastricht University Medical Centre, NUTRIM School of Nutrition and Translational Research in Metabolism, Maastricht, The Netherlands, REVAL–Rehabilitation Research Center, BIOMED–Biomedical Research Institute, Faculty of Rehabilitation Sciences, Hasselt University, Diepenbeek, Belgium

  • Arnulf Langhammer ,

    Contributed equally to this work with: Erlend Hassel, Ben M. Brumpton, Martijn A. Spruit, Arnulf Langhammer, Sigurd Steinshamn

    Roles Conceptualization, Methodology, Writing – original draft, Writing – review & editing

    Affiliation Department of Public Health and Nursing, Faculty of Medicine and Health Sciences, NTNU, Norwegian University of Science and Technology, Trondheim, Norway

  • Sigurd Steinshamn

    Contributed equally to this work with: Erlend Hassel, Ben M. Brumpton, Martijn A. Spruit, Arnulf Langhammer, Sigurd Steinshamn

    Roles Conceptualization, Methodology, Writing – original draft, Writing – review & editing

    Affiliations Department of Circulation and Medical Imaging, Faculty of Medicine and Health Sciences, NTNU, Norwegian University of Science and Technology, Trondheim, Norway, Clinic of Thoracic and Occupational Medicine, St. Olavs Hospital, Trondheim University Hospital, Trondheim, Norway

Lung function and peak oxygen uptake in chronic obstructive pulmonary disease phenotypes with and without emphysema

  • Øystein Rasch-Halvorsen, 
  • Erlend Hassel, 
  • Ben M. Brumpton, 
  • Haldor Jenssen, 
  • Martijn A. Spruit, 
  • Arnulf Langhammer, 
  • Sigurd Steinshamn
PLOS
x

Abstract

Previous studies of associations of forced expiratory lung volume in one second (FEV1) with peak oxygen uptake (VO2peak) in chronic obstructive pulmonary disease (COPD) have not taken sex, age and height related variance of dynamic lung volumes into account. Nor have such demographic spread of spirometric measures been considered in studies comparing VO2peak between COPD phenotypes characterized by degree of emphysema. We aimed to assess the association of FEV1Z-score with VO2peak in COPD (n = 186) and investigate whether this association differs between emphysema (E-COPD) and non-emphysema (NE-COPD) phenotypes. Corresponding assessments using standardized percent predicted FEV1 (ppFEV1) were performed for comparison. Additionally, phenotype related differences in VO2peak were compared using FEV1Z-score and ppFEV1 as alternative expressions of FEV1. E-COPD and NE-COPD were defined by transfer factor of the lung for carbon monoxide below and above lower limits of normal (LLN), respectively. The associations were assessed in linear regression models. One unit reduction in FEV1Z-score was associated with 1.9 (95% CI 1.4, 2.5) ml/kg/min lower VO2peak. In stratified analyses, corresponding estimates were 2.2 (95% CI 1.4, 2.9) and 1.2 (95% CI 0.2, 2.2) ml/kg/min lower VO2peak in E-COPD and NE-COPD, respectively. The association did not differ statistically by COPD phenotype (p-value for interaction = 0.153). Similar estimates were obtained in analyses using standardized ppFEV1. Compared to NE-COPD, VO2peak was 2.2 (95% CI 0.8, 3.6) and 2.1 (95% CI 0.8, 3.5) ml/kg/min lower in E-COPD when adjusted for FEV1Z-score and ppFEV1, respectively. In COPD, FEV1Z-score is positively associated with VO2peak. This association was stronger in E-COPD but did not differ statistically by phenotype. Both the association of FEV1 with VO2peak and the difference in VO2peak comparing COPD phenotypes seems independent of sex, age and height related variance in FEV1. Mechanisms leading to reduction in FEV1 may contribute to lower VO2peak in E-COPD.

Introduction

In current strategic documents provided by the Global Initiative for Chronic Obstructive Lung Disease (GOLD), the post-bronchodilator ratio of forced expiratory lung volume in one second/forced vital capacity (FEV1/FVC) is the key diagnostic variable defining chronic obstructive pulmonary disease (COPD) [1]. Reduced FEV1 denotes the severity of airway obstruction and may reflect the pathophysiological mechanism of expiratory flow limitation (EFL) at rest or during exercise, characteristic of this disease [2].

Dynamic hyperinflation (DH), inflicting restrictive constraints on tidal volume expansion, is a functional consequence of EFL [3] and associated with dyspnea and exercise limitation in COPD [4]. Exercise intolerance from dyspnea on exertion is a major manifestation of COPD leading to physical inactivity, deconditioning and reduced exercise capacity [5]. Assessed by cardiopulmonary exercise testing (CPET), peak oxygen uptake (VO2peak) is a direct measure of exercise capacity and reduced values are associated with increased mortality in COPD [6].

Although integrated in the concept of ventilatory limitation to exercise, variables of airway obstruction and exercise capacity are not strongly associated in COPD [79]. However, previous studies have not adequately taken sex, age and height related variance of dynamic lung volumes into account when reporting associations of FEV1 with VO2peak among those with COPD. Furthermore, the specificity of the diagnostic criterion of airway obstruction, when defined by the fixed ratio of post-bronchodilator FEV1/FVC less than 0.70, is age dependent and may lead to over-diagnosis of COPD in an older population [1012]. In contrast to percent predicted FEV1 (ppFEV1), and the fixed ratio of FEV1/FVC, Z-scores of these measures provide spirometric expressions that are comparable independent of sex, age and height. Reference equations for the calculation of Z-scores of dynamic lung volumes are readily available [13], and both the American Thoracic Society (ATS) and the European Respiratory Society (ERS) have endorsed defining the lower limit of normal (LLN) of lung function measures by the fifth percentile (z-score = -1.645) [14].

COPD is recognized as a heterogeneous disease and different phenotypes have been proposed [15]. In patients with predominantly emphysema (E-COPD), destruction of the pulmonary parenchyma leads to increased compliance from loss of elastic recoil pressure predisposing dynamic compression of the airways. The non-emphysema (NE-COPD) phenotype is characterized by airway inflammation with intrinsic caliber reduction and increased airflow resistance. These two phenotypes represent pathophysiological diversity in COPD highlighting mechanisms contributing to reduced FEV1 and expiratory flow limitation.

The transfer factor of the lung for carbon monoxide (TLCO) has been used to differentiate E-COPD from NE-COPD [14]. Patients with reduced TLCO have been shown to have lower VO2peak than patients with more preserved TLCO when compared at similar ppFEV1 [4]. It is unknown whether this difference in VO2peak is related to normalization of FEV1 by ppFEV1, and/or whether the association of FEV1 with VO2peak is stronger in E-COPD than in NE-COPD. Alternatively, mechanisms unrelated to FEV1 may explain the difference in exercise capacity comparing these COPD phenotypes.

Classification of patients by COPD phenotypes should be related to differences in clinically important outcomes [16]. New knowledge on associations between major diagnostic and prognostic measures in readily identifiable phenotypes of this multifaceted disease is likely to be of scientific and clinical value. We aimed to assess the association of FEV1Z-score with VO2peak in COPD and investigate whether this association differs comparing phenotypes with and without emphysema. Additionally, we explored the potential influence of sex, age and height related variance in FEV1 on both the association of FEV1 with VO2peak and on the difference in VO2peak between E-COPD and NE-COPD.

Materials and methods

Study population

We assessed patients referred for evaluation with pulmonary function tests (PFT) and CPET due to exercise intolerance at a private specialist clinic in Norway (Telemark Heart Lung and Blood Institute) between May 1999 and November 2014. The clinical database was accessed in the time period between the fifteenth of July and the thirtieth of August 2015. Patients with COPD (n = 301) were included.

In this study, the diagnostic criterion of airway obstruction in COPD was defined by the post-bronchodilator FEV1/FVC less than LLN. E-COPD and NE-COPD phenotypes were defined by TLCO below and above LLN, respectively [14]. LLN was defined as the fifth percentile (z-score = -1.645). Predicted values and Z-scores of dynamic lung volumes and lung diffusing capacity measures were calculated using the Global Lung Function Initiative (GLI) software [17, 18].

Patients > 80 years of age (n = 6) were excluded due to having an age above the valid range of the GLI 2017 reference values for TLCO. Patients ≤ 45 years of age (n = 28) were excluded due to low prevalence of COPD in this age group. We excluded 33 patients with a clinical diagnosis of asthma. In order to ensure uniform modality of exercise testing, 13 patients were excluded due to having performed the CPET on a treadmill. Furthermore, 19 patients were excluded because of termination of CPET due to chest pain and electrocardiogram (ECG) changes suggestive of myocardial ischemia (n = 2), arrhythmia or high blood pressure (n = 9), leg/knee or hip-pain (n = 4), caution due to known aortic valve disease or aortic aneurysm (n = 2), inadequate cycling technique (n = 1) or lack of motivation (n = 1). We also excluded 11 patients who performed PFT and CPET more than 3 months apart.

Three patients were excluded due to missing values on TLCO and two patients due to missing values on smoking status. After exclusions, 186 patients remained in the statistical analyses.

Pulmonary function tests

Spirometry and lung diffusing capacity (Vmax Legacy/Spectra 229; SensorMedics) were tested on separate visits, but no more than 3 months prior to CPET (exclusion criterion). The procedures adhered to recommendations provided by the American Thoracic Society/European Respiratory Society guidelines [19, 20].

Cardiopulmonary exercise test

An incremental exercise test was performed on a cycle ergometer (ER900; Ergoline). The work rate was increased by 5–25 W/minute on an individualized basis aiming for test termination after 8–12 minutes. The patients reported the dominating symptom limiting further exercise as dyspnea, leg discomfort/fatigue or other. The reasons for test termination in the latter category included pain in lower extremities (n = 9), vertigo (n = 4) and attainment of estimated maximal heart rate (n = 1).

Breath by breath measurements (Vmax Legacy/Spectra 229; SensorMedics) of oxygen uptake (VO2) and carbon dioxide output (VCO2) were averaged over 20 seconds. The highest value of VO2 was normalized by bodyweight and termed VO2peak.

Pulse oximetry (SpO2) was measured continuously from rest to test termination using a finger probe (Model 340; Palco Labs/8600; Nonin). The minimum value was termed SpO2min.

Maximal voluntary ventilation (MVV) was estimated by FEV1 x 40 [21]. Ventilatory reserve (VR) was calculated as VR = 1 –Minute ventilation at peak exercise (VEpeak)/MVV [22]. Maximal heart rate (HRmax) was estimated by 220 –age. Heart rate reserve (HRR) was calculated as HRR = 1 –Heart rate at peak exercise (HRpeak)/HRmax [22].

Statistical analyses

Descriptive statistics, calculated as mean and standard deviation (SD) or number of observations and percentages, are reported for both the total sample and the sample stratified by COPD phenotype. Lung function and exercise variables were compared between patients with E-COPD and NE-COPD using independent samples t-tests.

The association of FEV1Z-score with VO2peak was investigated in linear regression models. In a total sample model, VO2peak was regressed on FEV1Z-score and COPD phenotype. Body mass index (BMI), smoking status (former or current), beta-blocker and bronchodilator treatment (both dichotomous), were considered potential confounders. In separate models adjustments were also made for sex and age due to known strong associations with VO2peak. An interaction term (FEV1Z-score x COPD phenotype) was included in a separate analysis to investigate effect modification of COPD phenotype on the association of FEV1Z-score with VO2peak. Due to potential lack of power in the interaction analysis, the associations of FEV1Z-score with VO2peak were also estimated in analyses stratified by COPD phenotype. Corresponding analyses with standardized ppFEV1, calculated as (observed value–sample mean)/SD, were performed for comparison. Regression coefficients (β) with 95% confidence intervals (CI) are reported.

Linear regression models were also used to investigate whether the difference in VO2peak between E-COPD and NE-COPD is influenced by sex, age and height related variance in FEV1. VO2peak was regressed on COPD phenotype, FEV1Z-score and ppFEV1 in separate models adjusted for sex, age, BMI, smoking status, beta-blocker and bronchodilator treatment.

The potential influence of effort dependent underestimation of maximal exercise capacity from VO2peak was evaluated in a sensitivity analysis including only those with reduced VR and/or HRR. VO2peak was regressed on FEV1Z-score and COPD phenotype with adjustment for age, sex, BMI, smoking status, beta-blocker and bronchodilator treatment.

Residual plots were inspected in all models, and no violations of the assumptions of linear regression were uncovered. Statistical analyses were performed with IBM SPSS Statistics Version 25 (IBM Corp., Armonk, NY, USA).

Ethics approval

The Regional Committee for Medical and Health Research Ethics approved this retrospective study and the use of anonymous data without informed consent (REC-Central 2012/673).

Results

Clinical characteristics and PFT measures

The majority of patients were men, in both the total cohort (65%) and in subgroups stratified by COPD phenotype (63% vs. 67% in E-COPD and NE-COPD, respectively). Mean values of age, BMI and SpO2 at rest were similar in both phenotypes (63.9 vs. 60.4 years, 25.6 vs. 26.7 kg/m2 and 95.1 vs. 96.0% in E-COPD and NE-COPD, respectively). All patients were former or current smokers with similar proportions in both phenotypes (current smokers; 51% vs. 49% in E-COPD and NE-COPD, respectively). The majority of patients did not receive any beta-blocker (67% vs. 76% in E-COPD and NE-COPD, respectively), but had ongoing bronchodilator treatment (73% vs. 53% in E-COPD and NE-COPD, respectively) (Table 1).

thumbnail
Table 1. Clinical characteristics and pulmonary function test measures in total cohort and stratified by emphysema (E) and non-emphysema (NE) COPD phenotypes.

https://doi.org/10.1371/journal.pone.0252386.t001

Compared to patients with NE-COPD, patients with E-COPD had lower FEV1Z-score (difference 0.69, 95% CI 0.36, 1.02, p < 0.001), lower FVCZ-score (difference 0.42, 95% CI 0.05, 0.79, p = 0.028), lower (FEV1/FVC)Z-score (difference 0.62, 95% CI 0.38, 0.86, p < 0.001), lower KCOZ-score (difference 1.74, 95% CI 1.40, 2.07, p < 0.001), lower VAZ-score (difference 0.91, 95% CI 0.56, 1.27, p < 0.001) and, by definition, lower TLCOZ-score (difference 2.41, 95% CI 2.14, 2.68, p < 0.001) (Table 1).

CPET measures

The majority of patients reported test termination due to dyspnea in both phenotypes, but the proportion was higher in E-COPD than in NE-COPD (67 vs. 52%). The proportion reporting test termination due to leg fatigue was lower in E-COPD than in NE-COPD (26 vs. 40%) (Table 2).

thumbnail
Table 2. Cardiopulmonary exercise test measures in total cohort and stratified by emphysema (E) and non-emphysema (NE) COPD phenotypes.

https://doi.org/10.1371/journal.pone.0252386.t002

Compared to patients with NE-COPD, patients with E-COPD had lower VO2peak (difference 4.8 ml/kg/min, 95% CI 3.0, 6.5, p < 0.001), lower RERpeak (difference 0.04 95% CI 0.004, 0.08, p = 0.031), higher HRR (difference 6.1%, 95% CI 2.5, 9.6, p = 0.001), lower VEpeak (difference 14.8 l/min, 95% CI 7.8, 21.9, p < 0.001), lower VR (difference 5.5%, 95% CI -0.5, 11.5, p = 0.073) and lower SpO2min (difference 1.7%, 95% CI 0.9, 2.6, p < 0.001) (Table 2).

Associations of FEV1Z-score with VO2peak and effect modification by COPD phenotype

Adjusted for COPD phenotype, sex, age, BMI, smoking status, beta-blocker and bronchodilator treatment, one unit reduction in FEV1Z-score (i.e. one SD reduction in FEV1) was, on average, associated with 1.9 (95% CI 1.4, 2.5) ml/kg/min lower VO2peak (Table 3).

thumbnail
Table 3. Associations of FEV1Z-score with VO2peak (ml/kg/min) in total cohort and stratified by emphysema (E) and non-emphysema (NE) COPD phenotypes.

https://doi.org/10.1371/journal.pone.0252386.t003

Stratified by COPD phenotype, one unit reduction in FEV1Z-score was, on average, associated with 2.2 (95% CI 1.4, 2.9) and 1.2 (95% CI 0.2, 2.2) ml/kg/min lower VO2peak in E-COPD and NE-COPD, respectively after adjustment for sex, age, BMI, smoking status, beta-blocker and bronchodilator treatment (Table 3). The association of FEV1Z-score with VO2peak did not differ statistically by phenotype (p-value for interaction = 0.153).

Associations of standardized ppFEV1 with VO2peak and effect modification by COPD phenotype

Adjusted for COPD phenotype, sex, age, BMI, smoking status, beta-blocker and bronchodilator treatment, one unit reduction in standardized ppFEV1 (i.e. one SD reduction in ppFEV1) was, on average, associated with 2.5 (95% CI 1.8, 3.2) ml/kg/min lower VO2peak (Table 4).

thumbnail
Table 4. Associations of standardized ppFEV1 with VO2peak (ml/kg/min) in total cohort and stratified by emphysema (E) and non-emphysema (NE) COPD phenotypes.

https://doi.org/10.1371/journal.pone.0252386.t004

Stratified by COPD phenotype, one unit reduction in standardized ppFEV1 was, on average, associated with 2.7 (95% CI 1.8, 3.6) and 1.6 (95% CI 0.3, 2.9) ml/kg/min lower VO2peak in E-COPD and NE-COPD, respectively after adjustment for sex, age, BMI, smoking status, beta-blocker and bronchodilator treatment (Table 4). The association of standardized ppFEV1 with VO2peak did not differ statistically by phenotype (p-value for interaction = 0.220).

Phenotype related differences in VO2peak between FEV1Z-score and ppFEV1

Adjusted for sex, age, BMI, smoking status, beta-blocker and bronchodilator treatment, VO2peak was, on average, 2.2 (95% CI 0.8, 3.6) and 2.1 (95% CI 0.8, 3.5) ml/kg/min lower in E-COPD than in NE-COPD, when comparisons were made at similar FEV1Z-score and ppFEV1, respectively.

Sensitivity analyses

In the sensitivity analyses including only those with reduced VR and/or HRR, one unit reduction in FEV1Z-score was, on average, associated with 2.2 (95% CI 1.5, 2.8) ml/kg/min lower VO2peak after adjustment for COPD phenotype, sex, age, BMI, smoking status, beta-blocker and bronchodilator treatment.

Stratified by COPD phenotype, one unit reduction in FEV1Z-score was, on average, associated with 2.5 (95% CI 1.6, 3.4) and 1.3 (95% CI 0.3, 2.3) ml/kg/min lower VO2peak in E-COPD (n = 68) and NE-COPD (n = 66), respectively after adjustment for sex, age, BMI, smoking status, beta-blocker and bronchodilator treatment. The association of FEV1Z-score with VO2peak did not differ statistically by phenotype (p-value for interaction = 0.051).

Adjusted for sex, age, BMI, smoking status, beta-blocker and bronchodilator treatment, VO2peak was, on average, 2.0 (95% CI 0.3, 3.6) ml/kg/min lower in E-COPD than in NE-COPD, when comparisons were made at similar FEV1Z-score.

Discussion

In this study we found FEV1Z-score to be positively associated with VO2peak. This association was stronger in E-COPD than in NE-COPD but did not differ statistically by phenotype. In corresponding analyses with standardized ppFEV1, similar estimates were observed. Exercise capacity was lower in E-COPD than in NE-COPD, but the phenotype related differences in VO2peak were similar when comparisons were made using FEV1Z-score and ppFEV1.

Previous studies have reported weak positive associations between measures of dynamic lung volumes and exercise capacity in patients with COPD [8, 9]. Weak positive correlations between FEV1 and VO2peak may be expected in a heterogeneous disease where multiple organ systems and pathophysiological mechanisms interact and contribute in a highly variable manner to exercise limitation in individual patients [23]. However, normalization of FEV1 by using percent predicted may also weaken the association with VO2peak, especially in diverse study populations with regards to sex, age and height. The present study is the first to report the association of FEV1 with VO2peak in COPD using Z-scores of FEV1/FVC and FEV1 to define both the presence of as well as the degree of airway obstruction, respectively. Although, we support the use of Z-scores of dynamic lung volumes in the diagnostic process in COPD, the associations of FEV1 with VO2peak were similar using FEV1Z-score and standardized ppFEV1. Thus we have no reason to state that Z-scores are preferable over percent predicted regarding the association of FEV1 with VO2peak in COPD. On average, one unit reduction in FEV1Z-score was associated with 1.9 ml/min/kg (95% CI 1.4, 2.5) lower VO2peak. The minimal clinically important difference in VO2peak has not been established in COPD, but we consider this magnitude, representing 9% of the mean VO2peak (20.9 ml/kg/min) in this sample, likely to represent variability in functional impairment.

Several studies have reported differences in exercise capacity for any given reduction in ventilatory capacity when comparing patients with COPD phenotypically stratified by degree of emphysema. O`Donnell et al. [4] reported lower VO2peak in patients with COPD and reduced TLCO (38 ± 8% predicted) compared to patients with similar ppFEV1 but more preserved TLCO (73 ± 4% predicted). Farkhooy et al. [24] found TLCO to be an independent predictor of exercise capacity, assessed by peak workload, in patients with varying severity of COPD. Consistent with the results of these studies, we found lower VO2peak in patients with E-COPD defined by TLCO below LLN. The estimated difference in VO2peak comparing E-COPD and NE-COPD was 2.2 (95% CI 0.8, 3.6) and 2.1 (95% CI 0.8, 3.5) ml/kg/min when patients were compared at similar FEV1Z-score and ppFEV1, respectively. Thus, the difference in exercise capacity comparing these two COPD phenotypes seems independent of sex, age and height related variance in FEV1. In fact, lower exercise capacity in E-COPD seems to be conditional on pathophysiological mechanisms unrelated to reduction in ventilatory capacity as estimated from FEV1.

In a small study of patients with COPD and moderate to severe emphysema (n = 16), Paoletti et al. [25] compared the ventilatory response to maximal incremental exercise between patients with degree of emphysema above and below 50% as measured by high resolution computed tomography. In the group with over 50% emphysema, there was evidence of more dynamic hyperinflation, higher tidal volume to inspiratory capacity ratio (VT/IC), higher end-tidal partial pressure of CO2 (PETCO2) as well as increasing arterial partial pressure of CO2 (PaCO2) at peak exercise, suggesting greater mechanical constraints and relative hypoventilation in this group. O`Donnell et al. [4] also found more dynamic hyperinflation at lower exercise levels, more rapid attainment of volume constraints and higher degree of exercise induced dyspnea in patients with COPD and lower compared to higher TLCO. We found both lower VR, suggesting higher ventilatory demand at peak exercise relative to estimated maximal ventilatory capacity, and a higher proportion reporting test termination due to dyspnea in E-COPD compared to NE-COPD. Mechanical ventilatory limitation may therefore be a principal pathophysiological mechanism explaining the difference in exercise capacity between patients with E-COPD and NE-COPD found in our study.

Since lower aerobic capacity in E-COPD seems to be unrelated to ventilatory capacity, we may also hypothesize that greater inclination to mechanical ventilatory limitation is caused by increased ventilatory demand due to more pronounced gas exchange abnormalities and ventilatory inefficiency in E-COPD. Unfortunately, due to technical problems in the initial data extracting process, only peak exercise data were available and measures of ventilatory efficiency were therefore not included in this study. However, Rinaldo et al. [26] recently reported differences in ventilatory efficiency when comparing noninvasive measures between patients with emphysema and non-emphysema COPD phenotypes. The ventilatory equivalents for CO2 (VE/VCO2), both nadir and slope below ventilatory compensation point, were higher in emphysema indicating more pronounced ventilatory inefficiency in this phenotype.

Although we did not find strong evidence to support an effect modification of COPD phenotype on the association between FEV1Z-score and VO2peak (p-value for interaction 0.153), the estimated association was stronger in E-COPD (2.2 ml/kg/min, 95% CI 1.4, 2.9) than in NE-COPD (1.2 ml/kg/min, 95% CI 0.2, 2.2). Therefore, we cannot exclude the possibility that, additionally to being a marker of ventilatory capacity, any reduction in FEV1 may either impact or reflect gas exchange, ventilation/perfusion (V/Q) heterogeneity and ventilatory efficiency differently depending on COPD phenotype. Interestingly, although not confirmative of such speculation, we did find lower SpO2min in E-COPD compared to NE-COPD. Future studies are warranted and should aim to address underlying pathophysiological mechanisms explaining differences in exercise capacity comparing E-COPD and NE-COPD. In this regard, we believe invasive measures of gas exchange and calculations of dead space ventilation from arterial blood gasses would be beneficial.

Boutou et al. [27] found percent predicted TLCO (ppTLCO) to be an independent predictor of survival in patients with COPD, and the impact of VO2peak on mortality is well known in this disease [6]. By confirming a positive association between phenotype assessment by TLCOZ-score and VO2peak, the present study provides additional evidence linking two prognostic measures in COPD. In agreement with previous studies [2427], we advocate that assessment of TLCO should be considered in the evaluation of patients with COPD to add clinical information not inferable from simple spirometric measurements alone. Furthermore, the finding of both lower FEV1Z-score and VO2peak in patients with E-COPD, coincide with COPD phenotype being a positive modifier of the association between these key variables. We suggest that differentiation between TLCO below and above LLN should complement measures of dynamic lung volumes in future studies of exercise capacity in COPD.

A strength of the current study is that we used Z-scores of lung function measures when reporting the association of FEV1 with VO2peak. This approach is novel and E-COPD and NE-COPD may be more correctly differentiated by TLCOZ-score, taking sex, age and height related variance into account, than by ppTLCO. Furthermore, confounding effects of sex and age on the association of FEV1z-score with VO2peak are likely to reflect sampling variation, which was explored in a separate model. Additionally, the patients assessed in this study were referred for evaluation from primary care and may represent patients typically encountered in a specialist clinic for diagnostic or prognostic evaluation of COPD.

This study has limitations. COPD phenotypes were dichotomized into E-COPD and NE-COPD based on TLCO below and above LLN, respectively. Although the pre-test probability for emphysema is assumed high in this study population consisting of patients with COPD, radiological qualitative and quantitative confirmation by thoracic computed tomography was not available. On the other hand, this is a real life study and assessment of emphysematous component by TLCO in obstructive lung disease is recommended in clinical practice [14].

The gold standard to assess exercise capacity is direct measurement of VO2 at maximal exercise (VO2max). VO2peak, used in this study, is susceptible to patient effort and may underestimate true VO2max [28]. Provided that any underestimation of VO2max from VO2peak is unrelated to FEV1Z-score, only the precision of the estimates should potentially be influenced. In contrast, associations between underestimation of VO2max from VO2peak and FEV1Z-score could introduce systematic error and bias the results. We identified 52 patients (E-COPD; n = 33) with both VR above 15% and HRR above 10%, potentially indicative of submaximal effort, i.e. both preserved residual capacity of the ventilatory and cardiac system, respectively [22]. Excluding these patients in the sensitivity analyses did not appreciably influence the estimates.

In this cross sectional study we can only test for associations. The nature of the data does not permit causal inference or definite conclusions on involved pathophysiological mechanisms behind observed results. Therefore, the clinical value of differentiation on COPD phenotype, supported by this study, cannot be extended to suggest specific treatment recommendations in individual patients.

Conclusions

In COPD, FEV1Z-score is positively associated with VO2peak. This association was stronger in E-COPD than in NE-COPD but did not differ statistically by phenotype. Both the association of FEV1 with VO2peak and the difference in VO2peak comparing COPD phenotypes seems independent of sex, age and height related variance in FEV1. Mechanisms leading to reduction in FEV1 may contribute to lower VO2peak in E-COPD. Future studies aiming to address underlying pathophysiological mechanisms of these findings are warranted.

Supporting information

Acknowledgments

The authors would like to acknowledge Marina Vang for her contribution as a research assistant in the initial data extraction process.

References

  1. 1. Vogelmeier CF, Criner GJ, Martinez FJ, Anzueto A, Barnes PJ, Bourbeau J, et al. Global Strategy for the Diagnosis, Management, and Prevention of Chronic Obstructive Lung Disease 2017 Report. GOLD Executive Summary. Am J Respir Crit Care Med. 2017;195(5):557–82. pmid:28128970
  2. 2. Calverley PM, Koulouris NG. Flow limitation and dynamic hyperinflation: key concepts in modern respiratory physiology. Eur Respir J. 2005;25(1):186–99. pmid:15640341
  3. 3. Pellegrino R, Brusasco V, Rodarte JR, Babb TG. Expiratory flow limitation and regulation of end-expiratory lung volume during exercise. J Appl Physiol (1985). 1993;74(5):2552–8. pmid:8335591
  4. 4. O’Donnell DE, Revill SM, Webb KA. Dynamic hyperinflation and exercise intolerance in chronic obstructive pulmonary disease. Am J Respir Crit Care Med. 2001;164(5):770–7. pmid:11549531
  5. 5. Pitta F, Troosters T, Spruit MA, Probst VS, Decramer M, Gosselink R. Characteristics of physical activities in daily life in chronic obstructive pulmonary disease. Am J Respir Crit Care Med. 2005;171(9):972–7. pmid:15665324
  6. 6. Oga T, Nishimura K, Tsukino M, Sato S, Hajiro T. Analysis of the factors related to mortality in chronic obstructive pulmonary disease: role of exercise capacity and health status. Am J Respir Crit Care Med. 2003;167(4):544–9. pmid:12446268
  7. 7. Cooper CB. The connection between chronic obstructive pulmonary disease symptoms and hyperinflation and its impact on exercise and function. Am J Med. 2006;119(10 Suppl 1):21–31. pmid:16996896
  8. 8. Bauerle O, Younes M. Role of ventilatory response to exercise in determining exercise capacity in COPD. J Appl Physiol (1985). 1995;79(6):1870–7. pmid:8847246
  9. 9. Foglio K, Carone M, Pagani M, Bianchi L, Jones PW, Ambrosino N. Physiological and symptom determinants of exercise performance in patients with chronic airway obstruction. Respir Med. 2000;94(3):256–63. pmid:10783937
  10. 10. Hardie JA, Buist AS, Vollmer WM, Ellingsen I, Bakke PS, Morkve O. Risk of over-diagnosis of COPD in asymptomatic elderly never-smokers. Eur Respir J. 2002;20(5):1117–22. pmid:12449163
  11. 11. Miller MR, Quanjer PH, Swanney MP, Ruppel G, Enright PL. Interpreting lung function data using 80% predicted and fixed thresholds misclassifies more than 20% of patients. Chest. 2011;139(1):52–9. pmid:20522571
  12. 12. Vollmer WM, Gislason T, Burney P, Enright PL, Gulsvik A, Kocabas A, et al. Comparison of spirometry criteria for the diagnosis of COPD: results from the BOLD study. Eur Respir J. 2009;34(3):588–97. pmid:19460786
  13. 13. Quanjer PH, Ruppel G, Brusasco V, Perez-Padilla R, Fragoso CA, Culver BH, et al. COPD (confusion over proper diagnosis) in the zone of maximum uncertainty. Eur Respir J. 2015;46(5):1523–4. pmid:26521284
  14. 14. Pellegrino R, Viegi G, Brusasco V, Crapo RO, Burgos F, Casaburi R, et al. Interpretative strategies for lung function tests. Eur Respir J. 2005;26(5):948–68. pmid:16264058
  15. 15. Augustin IML, Spruit MA, Houben-Wilke S, Franssen FME, Vanfleteren L, Gaffron S, et al. The respiratory physiome: Clustering based on a comprehensive lung function assessment in patients with COPD. PLoS One. 2018;13(9):e0201593. pmid:30208035
  16. 16. Han MK, Agusti A, Calverley PM, Celli BR, Criner G, Curtis JL, et al. Chronic obstructive pulmonary disease phenotypes: the future of COPD. Am J Respir Crit Care Med. 2010;182(5):598–604. pmid:20522794
  17. 17. Quanjer PH, Stanojevic S, Cole TJ, Baur X, Hall GL, Culver BH, et al. Multi-ethnic reference values for spirometry for the 3-95-yr age range: the global lung function 2012 equations. Eur Respir J. 2012;40(6):1324–43. pmid:22743675
  18. 18. Stanojevic S, Graham BL, Cooper BG, Thompson BR, Carter KW, Francis RW, et al. Official ERS technical standards: Global Lung Function Initiative reference values for the carbon monoxide transfer factor for Caucasians. Eur Respir J. 2017;50(3). pmid:28893868
  19. 19. Quanjer PH, Tammeling GJ, Cotes JE, Pedersen OF, Peslin R, Yernault JC. Lung volumes and forced ventilatory flows. Report Working Party Standardization of Lung Function Tests, European Community for Steel and Coal. Official Statement of the European Respiratory Society. Eur Respir J Suppl. 1993;165–40. pmid:8499054
  20. 20. Miller MR, Hankinson J, Brusasco V, Burgos F, Casaburi R, Coates A, et al. Standardisation of spirometry. Eur Respir J. 2005;26(2):319–38. pmid:16055882
  21. 21. Campbell SC. A comparison of the maximum voluntary ventilation with the forced expiratory volume in one second: an assessment of subject cooperation. J Occup Med. 1982;24(7):531–3. pmid:7119913
  22. 22. American Thoracic S, American College of Chest P. ATS/ACCP Statement on cardiopulmonary exercise testing. Am J Respir Crit Care Med. 2003;167(2):211–77. pmid:12524257
  23. 23. Pepin V, Saey D, Laviolette L, Maltais F. Exercise capacity in chronic obstructive pulmonary disease: mechanisms of limitation. COPD. 2007;4(3):195–204. pmid:17729063
  24. 24. Farkhooy A, Janson C, Arnardottir RH, Malinovschi A, Emtner M, Hedenstrom H. Impaired carbon monoxide diffusing capacity is the strongest predictor of exercise intolerance in COPD. COPD. 2013;10(2):180–5. pmid:23547629
  25. 25. Paoletti P, De Filippis F, Fraioli F, Cinquanta A, Valli G, Laveneziana P, et al. Cardiopulmonary exercise testing (CPET) in pulmonary emphysema. Respir Physiol Neurobiol. 2011;179(2–3):167–73. pmid:21840426
  26. 26. Rinaldo RF, Mondoni M, Comandini S, Lombardo P, Vigo B, Terraneo S, et al. The role of phenotype on ventilation and exercise capacity in patients affected by COPD: a retrospective study. Multidiscip Respir Med. 2020;15(1):476. pmid:32153779
  27. 27. Boutou AK, Shrikrishna D, Tanner RJ, Smith C, Kelly JL, Ward SP, et al. Lung function indices for predicting mortality in COPD. Eur Respir J. 2013;42(3):616–25. pmid:23349449
  28. 28. Poole DC, Jones AM. Measurement of the maximum oxygen uptake Vo2max: Vo2peak is no longer acceptable. J Appl Physiol (1985). 2017;122(4):997–1002.