Abstract
Background
Oxygen uptake (V’O2) obtained from expiratory gas analysis is generally calculated using minute ventilation (V’E) and the inspired‒expired mean oxygen (O2) concentration difference (ΔFO2) during cardiopulmonary exercise testing (CPET). We have reported that ΔFO2, which is associated with ventilatory efficiency, is independent of V’E at peak exercise and affects exercise tolerance in respiratory diseases other than idiopathic pulmonary fibrosis (IPF). We hypothesized that similar results are obtained in IPF, and that ΔFO2 is a prognostic factor for survival in IPF.
Methods
Forty-three patients with IPF, who underwent CPET with blood gas analysis were enrolled from our database.
Results
At peak exercise, ΔFO2 was strongly correlated with variables related to ventilatory efficiency, i.e., V’E/carbon dioxide output (V’CO2) ratio at the nadir during exercise (r=‒0.91) and correlated well with peak V’O2 (r = 0.67), but it was independent of V’E (r = 0.24) at peak exercise. Two multivariate Cox proportional hazards models with adjustment for age, including the previously reported prognostic factors, showed that ΔFO2 at peak exercise was a stronger predictor of survival than (1) peak V’O2, V’E at peak exercise in a first analysis (hazard ratio: 0.195, 95% CI 0.070 to 0.500; p = 0.0005) and (2) than tidal volume at peak exercise, body mass index, and arterial oxygen tension (PaO2)-slope, i.e., the decrease in PaO2 per the increase in V’O2 during exercise in a second analysis (hazard ratio: 0.437, 95% CI 0.201 to 0.958; p = 0.0389).
Conclusions
These results show that ∆FO2 at peak exercise, which is correlated with ventilatory efficiency related to carbon dioxide clearance, is independent of ventilatory ability and is a stronger prognostic factor for survival than physiological ventilatory impairments with hypoxemia in IPF. CPET is essential for evaluating exercise alveolar O2 extraction and guiding the optimal management of patients with IPF.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12890-025-04064-3.
Keywords: Cardiopulmonary exercise testing, Exercise tolerance, Hypoxemia, Pulmonary function, Ventilation
Background
In idiopathic pulmonary fibrosis (IPF), progressive fibrosis leads to exertional hypoxemia, exercise intolerance, and premature death. Though median survival is estimated to be about 4.5 years, there is considerable variation in clinical deterioration [1–3]. In light of this clinical heterogeneity, other factors may be affecting exercise intolerance with exertional hypoxemia, which not only reduces an individual’s activity but is also a prognostic factor [2, 4, 5].
Cardiopulmonary exercise testing (CPET) can be used to evaluate whether the patient has adequate exercise tolerance, focusing on the heart, lungs, and skeletal muscles. The therapeutic approach to increase peak oxygen uptake (V’O2) is the key to robust movement, and since peak V’O2 is also a prognostic factor for IPF [4], it might improve the prognosis. Generally, V’O2 is calculated using the following Eqs [6–8].
 |
 |
where FIO2 and FEO2 are inspired and expired mean oxygen (O2) concentrations, respectively, FECO2 is expired mean carbon dioxide (CO2) concentration, V’ E is minute ventilation, and V’CO2 is carbon dioxide output calculated by the product of FECO2 and V’E, all of which are obtained from expiratory gas analysis with a mask covering the mouth and nose [8]. Based on the above equation and defining ΔFO2 = FIO2 − FEO2, V’O2 is considered dependent on V’E and ΔFO2. ΔFO2 reflects alveolar O2 extraction, representing the capacity of peripheral tissues beyond the mouth, that is, throughout the whole body, including the lungs, heart, and skeletal muscle, to consume O2 [7, 8]. Generally, in many pulmonary diseases including IPF, as the disease progresses, exertional ventilation is not adequate. In such cases, we hypothesized that the patients depend on alveolar O2 extraction ability, i.e., ΔFO2, as an alternative means to help increase exercise tolerance. ΔFO2 is approximately the inverse of V’E/V’O2, and it is naturally expected that ΔFO2 relates to the ventilatory efficiency variables regardless of O2 or CO2. However, the relationship between ΔFO2 and V’E remained unclear, even though V’E in the numerator and denominator of the ventilatory efficiency variable cancels out. So far, we have found that ΔFO2 is a factor related to ventilatory efficiency variables independent of V’E in chronic obstructive pulmonary disease (COPD) and nontuberculous mycobacterial pulmonary disease (NTM-PD) [8–10]. In addition, assessing whether improvements in V’E and ΔFO2 contribute to exercise tolerance after pulmonary rehabilitation is possible only with CPET, not with the six-minute walk test or pulmonary function tests (PFTs) [11, 12]. In fact, in severe and very severe COPD patients with ventilatory impairments, the effect of rehabilitation with exercise therapy depended on the change of ΔFO2 rather than V’E at peak exercise [11, 12]. Furthermore, when focusing on prognosis, peak V’O2 has already been reported as a prognostic factor for IPF [4, 5]. In contrast, V’E has rarely been identified as a strong prognostic factor [6, 13], and ΔFO2 has never been studied in this context, despite both being components of V’O2. Assuming the possibility that ΔFO2 at peak exercise is a more significant prognostic factor than V’E at peak exercise, then since V’O2 is an indicator that includes both V’E and ΔFO2, we hypothesized that ΔFO2 at peak is a stronger prognostic factor for IPF than peak V’O2. Separating ΔFO2 from V’E and evaluating which parameter has greater potential for improvement could aid in personalized treatment strategies and prognostic prediction.
The aim of the present study was to investigate whether ΔFO2 is related to ventilatory efficiency variables independent of V’E in patients with IPF and a prognostic factor for IPF.
Materials and methods
Study participants
Of 4749 respiratory disease patients who underwent CPET at the NHO Osaka Toneyama Medical Center between February 1999 and August 2024, 779 patients underwent CPET using a treadmill concurrently with arterial blood gas analysis. Of them, 43 IPF patients in whom it was possible to analyze FIO2 and FEO2 by expiratory gas analysis at the baseline evaluation for IPF were retrospectively included in this study. Data after CPET for each patient were tracked until August 2024. In all cases, the diagnosis of IPF was uniformly made using the clinical practice guideline of 2018 [14] with the baseline high-resolution computed tomography (HRCT) evaluation, and, if necessary, the HRCT evaluation during the disease course. Then, the cases that met the criteria for usual interstitial pneumonia (UIP) not subjected to surgical lung biopsy were selected. Before the actual CPET, comprehensive consent was obtained, focusing on medical conditions, examination details, and handling of obtained data, and all patients gave their written, informed consent. The NHO Osaka Toneyama Medical Center Ethical Review Board approved this study and handling of the data (approval number: TNH-R-2020055-4, final approval date: March 11, 2024), and the protocol complied with the Declaration of Helsinki on research involving human subjects. For details regarding study participants, see the online supplementary Methods.
Pulmonary function tests (PFTs)
PFTs were performed before CPET as previously described [15, 16].
The reference equations used are described in the Online Supplementary Material. The Z-scores calculable in a large Japanese population from the Japanese Respiratory Society were used [17, 18].
Cardiopulmonary exercise testing (CPET)
Symptom-limited incremental CPET [19] with blood gas analysis was performed on a treadmill using the Sheffield protocol or two modified Sheffield protocols [16], as shown in the online supplementary Methods. Pre-exercise resting and exertional measurements were performed as previously described [16]. Briefly, the dyspnea intensity level (10-point modified Borg category-ratio scale), percutaneous oxygen saturation (SpO2), arterial blood gas, plasma norepinephrine (NE) level, and plasma lactate level were measured at rest, during the last 15 s, at the first 1 min, then at 3-min intervals, and at the end of exercise. The ventilatory variables V’E, V’O2, V’CO2, respiratory frequency (fR), tidal volume (VT), O2 pulse [V’O2/heart rate (HR)], and ΔFO2 were measured breath-by-breath and collected as 30-s mean values at rest, at 1- and 3-min intervals during exercise, and at the end of exercise (peak). The anaerobic threshold (AT) was identified using the V-slope method, and the ventilatory equivalent method was also checked [7]. V’E-V’CO2-slope was calculated as previously reported [9]. In brief, the V’E-V’co2-slope was calculated by linear regression, excluding the nonlinear part of the data after the onset of the respiratory compensation point. If no respiratory compensation could be identified, the V’E-V’CO2-slope was calculated using the data from the start to the end of the exercise. V’E/V’CO2-nadir was defined as the lowest value during exercise. The predicted maximum voluntary ventilation volume (MVV) was calculated as FEV1 × 35 [7]. Predicted maximum HR was calculated as 220–age in years [7]. The arterial oxygen tension (PaO2)-slope (mmHg·L−¹·min), NE-slope (ng·mL−¹·L−¹·min), pH-slope (L⁻¹·min), and dyspnea-slope (Borg scale·L⁻¹·min) were calculated as the decrease in PaO2, the increase in norepinephrine, the decrease in pH, and the increase in the Borg scale, respectively, per increase in V’O2 from rest to peak exercise.
Statistical analysis
Continuous data are presented as means (standard deviation) if normally distributed (Shapiro-Wilk test) or medians (interquartile range) if not normally distributed. If the data were normally distributed, Pearson’s correlation coefficient was used to quantify correlations; otherwise, Spearman’s correlation coefficient was used.
The receiver-operating characteristic (ROC) curve was used to determine the cut-off point with the most appropriate values of ΔFO2 to divide patients into two groups based on V’E-V’CO2-slope ≥ 34 being a better way of identifying high-risk patients as a commonly recognized indicator [6, 20]. Univariate Cox proportional hazard models were used to examine the relationships of selected variables with mortality. Subsequently, the three variables (V’E, ΔFO2, and V’O2 at peak exercise) as a first analysis, along with the important prognostic factors obtained from univariate analysis, previously reported prognostic factors from CPET in IPF, and general prognostic factors in IPF (PaO2-slope, dyspnea-slope, V’E/V’CO2 at peak exercise, V’E/V’CO2-nadir, VT at peak exercise, FVC, BMI, %DLCO, and ΔFO2 at peak exercise) as a second analysis, were listed. Furthermore, based on the correlation matrix, variables that were strongly correlated (|r| ≥ 0.9) were excluded. In addition, regarding factors affecting survival time, variance inflation factors (VIFs) to evaluate multicollinearity were confirmed, and factors with VIFs < 5 were selected. From the remaining variables falling below both thresholds, a multivariate Cox proportional hazards model was created to assess ΔFO2 as a prognostic factor. In variable selection for the second analysis, as a result of calculating their correlation matrix, V’E/V’CO2 at peak exercise, V’E/V’CO2-nadir, and FVC were excluded, because those variables were strongly correlated with each other, i.e., |r| ≥ 0.9. Furthermore, %DLCO, with VIF > 5, was excluded. Furthermore, given that the principle is to analyze approximately one variable per 10 samples and considering that this study relates to O2 extraction, ultimately, PaO2-slope, VT at peak exercise, BMI, and ΔFO2 at peak exercise were selected. As for DLCO, 10 patients had missing data, but outside of that, there were no missing data. The estimated %DLCO in 10 patients was determining based on the DLCO (ml/min/mmHg) calculated by the following formula using PaO2-slope (mmHg/l/min) and PaO2 at rest (mmHg) [5]:
PaO2-slope = −43.7−0.8×PaO2 at rest + 7.0×DLCO.
A p-value < 0.05 was considered statistically significant (JMP software, version 11, SAS Institute Inc., Cary, NC, USA).
Results
Forty-three patients with IPF were included, of whom 30 (70%) had a history of smoking; the median age when CPET was performed was 72 years (Table 1). The period during which these patients underwent CPET was from February 1999 to February 2011. During follow-up [median follow-up duration, 833 days (2.3 years)], 34 patients (79%) died, with a median survival of 850 days (2.3 years) and a mortality rate of 21.5% from the time of CPET. Causes of death were IPF exacerbation in 21 patients, malignancy in 5, pneumothorax in 4, sudden death in 2, and unknown in 2.
Table 1.
Baseline characteristics of IPF patients (n = 43)
| Age, y | 72 (66; 76) |
| Sex, male/female, n | 37/6 |
| BMI, kg ⋅ m− 2 | 22.9 (3.4) |
| Smoking history, pack year | 25 (0; 45) |
| Smoking history, smoker/ex-smoker/never, n | 4/26/13 |
| Hemoglobin, g⋅dL− 1 | 14.2 (1.36) |
| Cardiopulmonary exercise testing | |
| Peak V’O2, mL⋅min− 1 ⋅ kg− 1 | 18.4 (5.8) |
| Lowest SpO2, % | 83 (8) |
| Patient number with SpO2 < 90%, n (%) | 31 (72) |
| Pulmonary function test | |
| VC, L | 2.17 (0.69) |
| %VC, % | 65.6 (55.9; 85.0) |
| VC, Z-score | ‒3.91 (2.19) |
| FVC, L | 2.04 (0.67) |
| %FVC, % | 62.6 (53.9; 82.2) |
| FVC, Z-score | ‒3.39 (‒3.99; ‒1.94) |
| FEV1, L | 1.74 (0.43) |
| %FEV1, % predicted | 69.3 (15.5) |
| FEV1, Z-score | ‒2.30 (1.22) |
| FEV1/FVC, % | 87.4 (9.3) |
| FEV1/FVC, Z-score | 1.64 (1.57) |
| IC, L | 1.19 (1.04; 1.75) |
| %DLCOa, % | 50.6 (24.2) |
| DLCOa, Z-score | ‒26.0 (‒104.3; ‒14.5) |
| Comorbidity | |
| Diabetes mellitus | 14 |
| Cardiovascular disease | 11 |
| Hypertension | 8 |
| Gastroesophageal reflux disease | 3 |
| Chronic obstructive pulmonary disease | 2 |
| Past history | |
| Pneumothorax | 6 |
| Pulmonary tuberculosis | 4 |
| Cerebrovascular disease | 2 |
Data are presented as means (standard deviation) if normally distributed or medians (interquartilerange: 25th percentile;75th percentile) if not normally distributed, unless otherwise stated
BMI Body mass index, DLCO Diffusing capacity for carbon monoxide, FEV1 Forced expiratory volume in one second, FVC Forced vital capacity, IC Inspiratory capacity, SpO2 Percutaneous oxygen saturation, VC Vital capacity, V’ 02 Oxygen uptake
a: analyzed in 33 patients
Baseline CPET results are shown in Table 2. All patients reached their symptomatic threshold at exercise termination, with one patient reaching target heart rate (220-age in years) at the end of exercise. No patients discontinued exercise due to ECG changes. Thirty-seven patients stopped the test due to dyspnea, and five stopped due to leg fatigue. The mean peak V’O2 was 18.4 mL·min⁻¹·kg⁻¹ (73% of predicted; lower limit of normal, 75% [7]), SpO2 fell below 90% in 31 patients (72%), and the anaerobic threshold was determinable in 32 patients (74%). As an overall patient evaluation, exercise ended in metabolic acidosis with a mildly reduced HCO3− level developed at peak exercise. The median V’E-V’CO2-slope was not high, 33.3 (< 34) [6, 20]. Exercise-induced hypoxemia with the mean PaO2 at peak exercise less than 60 mmHg was observed in 36 patients (84%), the mean difference in PaO2 between peak exercise and rest [PaO2 (peak−rest)] was −30.5 (standard deviation: 10.7) mmHg. First, the degrees of dependence of V’E and ΔFO2 on peak V’O2 were determined. The squares of the correlation coefficient (r) values of V’E and ΔFO2 to peak V’O2 were 0.50 and 0.45, respectively. Second, the relationships between exercise-induced hypoxemia and other ventilatory variables were examined. PaO2 (peak−rest) showed a significant negative correlation with peak V’O2 (r = −0.42, p = 0.0054). PaO2 (peak−rest) tended to show a negative correlation with ΔFO2 (r = −0.28, p = 0.0650), but PaO2-slope showed a positive correlation with ΔFO2 (r = 0.45, p = 0.0023) (Table 3). Third, the degree to which ΔFO2 at peak exercise related to the exercise variables obtained from CPET was examined by r2 (Table 3). In the relationship of ΔFO2 with exertional variables, the variables that obtained a strong correlation (r2 > 0.5) were variables related to ventilatory efficiency during exercise (Fig. 1; Table 3). However, ΔFO2 was not correlated with V’E at peak exercise. Fourth, which variables are affected by high and low ΔFO2 at peak exercise was investigated in groups based on the cut-off value (2.46%, area under the ROC curve, 0.85) of ΔFO2 at peak exercise corresponding to a V’E-V’CO2-slope ≥ 34 and < 34 (Table 4). In the ΔFO2 higher group, compared with the lower group: (1) BMI was higher; (2) V’O2 and ΔFO2 were higher at peak exercise, although no significant differences were found in V’E at peak exercise; and (3) the exertional decrease in PaO2 was greater, as indicated by the lower PaO2 (peak-rest), although the dyspnea-slope and the PaO2-slope were more gradual, and the %DLCO was higher (Table 4).
Table 2.
Incremental exercise variables with blood gas analysis in IPF patients (n = 43)
| Reason for discontinuing exercise, dyspnea/leg fatigue, n (%) | 38/5 (88/12) |
| At peak exercise | |
| Dyspnea, Borg scale | 7.0 (2.0) |
| V’O2, mL⋅min− 1 ⋅ kg− 1 | 18.4 (5.8) |
| R | 1.04 (0.99; 1.139) |
| V’E, L ⋅ min− 1 | 38.4 (14.0) |
| VT, mL | 1170 (426) |
| fR, breaths ⋅ min− 1 | 42 (38; 47) |
| ΔFO2, % | 2.91 (0.64) |
| V’E/V’O2 | 43.4 (38.4; 56.3) |
| V’E/V’CO2 | 40.9 (36.6; 50.4) |
| V’E/MVV, % | 79 (69; 91) |
| HR, beats ⋅ min− 1 | 129 (23) |
| HR/predicted maximum HR, % | 86 (15) |
| SpO2, % | 83 (8) |
| O2 pulse, mL ⋅ beats− 1 | 8.5 (2.8) |
| pH | 7.359 (0.038) |
| PaO2, mmHg | 51.0 (9.2) |
| PaCO2, mmHg | 42.1 (6.1) |
| HCO3−, mEq · L− 1 | 23.2 (21.7; 25.1) |
| Plasma lactate, mg ⋅ dL− 1 | 26.6 (18.0; 36.9) |
| Plasma NE, ng ⋅ mL− 1 | 2.47 (1.76; 3.58) |
| During exercise | |
| V’E-V’CO2-slope | 33.3 (30.4; 38.2) |
| V’E/V’CO2-nadir | 42.7 (9.5) |
| PaO2-slope, mmHg⋅ L− 1⋅ min | ‒39.5 (‒52.0; ‒30.7) |
| NE-slope, ng⋅ mL− 1⋅ L− 1⋅ min | 2.24 (1.29; 3.81) |
| pH-slope, L− 1⋅ min | ‒0.062 (0.047) |
| Number of patients reaching AT, n (%) | 32 (74) |
| V’O2 at AT, mL ⋅ min− 1 | 767 (690; 1028) |
Data are presented as means (standard deviation) if normally distributed or medians (interquartile range: 25th percentile;75th percentile) if not normally distributed, unless otherwise stated
Estimated maximal voluntary ventilation (MVV) (L·min-1) was equal to forced expiratory volume in one second (FEV1)×35. The PaO2-slope, NE-slope, and pH-slope were calculated as the decrease in PaO2/the increase in V’02, the increase in norepinephrine/the increase in V’02, and the decrease in pH/the increase in V’02 between at rest and at peak exercise, respectively
AT Anaerobic threshold, ΔF02 Difference between inspired and expired mean oxygen concentration, fR Breathing frequency, HCO3- Bicarbonate ion, HR Heart rate, NE Norepinephrine, O2 pulse V’02/HR Predicted maximum HR: 220−age (y), PaCO2: Arterial carbon dioxide tension, PaO2 Arterial oxygen tension, R Gas exchange ratio, SpO2 Percutaneous oxygen saturation, V’CO2 Carbon dioxide output, V’E Minute ventilation and V’E-V’CO2-slope The slope was determined by linear regression analysis of V’E to V’CO2 obtained during exercise (see the Methods for details), V’E/ V’CO2-nadir The lowest value during exercise (see the Methods for details), V’02 Oxygen uptake, VT Tidal volume
Table 3.
Correlations with ΔFO2 at peak exercise in IPF patients (n = 43)
| Peak incremental exercise parameters | |
|---|---|
| V’O2, mL ⋅ min− 1 ⋅ kg | 0.67 (0.45) |
| V’E, L ⋅ min− 1 | 0.24 (0.06) |
| fR, breaths ⋅ min− 1 | ‒0.47 (0.22) |
| V’E/V’CO2 | ‒0.94 (0.88) |
| HR, beats ⋅ min− 1 | 0.25 (0.06) |
| O2 pulse, mL ⋅ beats− 1 | 0.65 (0.42) |
| PaO2, mmHg | 0.37 (0.14) |
| PaCO2, mmHg | 0.15 (0.02) |
| HCO3−, mEq · L− 1 | ‒0.09 (0.01) |
| Plasma lactate, mg ⋅ dL− 1 | 0.37 (0.14) |
| Plasma NE, ng ⋅ mL− 1 | ‒0.05 (0.00) |
| During exercise | |
|---|---|
| V’E-V’CO2-slope | ‒0.82 (0.67) |
| V’E/V’CO2-nadir | ‒0.91 (0.83) |
| V’E/V’CO2 at AT | ‒0.82 (0.67) |
| AT (V’O2), mL ⋅ min− 1 | 0.47 (0.22) |
| PaO2 (peak‒rest), mmHg | ‒0.28 (0.08) |
| PaO2- slope, mmHg ⋅ L−1⋅ min | 0.45 (0.20) |
| NE- slope, ng ⋅ mL− 1 ⋅ L− 1 ⋅ min | ‒0.50 (0.25) |
| pH- slope, L− 1⋅ min | ‒0.07 (0.00) |
| Dyspnea- slope, Borg scale ⋅ L− 1⋅ min | ‒0.68 (0.46) |
| Pulmonary function and others | |
|---|---|
| FEV1, L | 0.32 (0.10) |
| FVC, L | 0.37 (0.14) |
| IC, L | 0.54 (0.29) |
| Age, y | ‒0.24 (0.06) |
| BMI, kg ⋅ m− 2 | 0.34 (0.12) |
Data are presented as the correlation coefficient: r (r2)
AT Anaerobic threshold, BMI Body mass index, FEV1 Forced expiratory volume in one second, FVC Forced vital capacity, ΔF02 Difference between inspired and expired mean oxygen concentration, fR Breathing frequency, HCO3- Bicarbonate ion, HR Heart rate, IC Inspiratory capacity, NE Norepinephrine, O2 pulse V’02/HR, PaCO2 Arterial carbon dioxide tension, PaO2 Arterial oxygen tension, PaO2 (peak−rest) The difference in PaO2 between at peak exercise and at rest, V’CO2 Carbon dioxide output, V’E Minute ventilation, V’E/V’CO2-nadir The lowest value during exercise (see the Methods for details), and V’E-V’CO2-slope The slope was determined by linear regression analysis of V’E to V’CO2 obtained during exercise (see the Methods for details), V’O2 Oxygen uptake
The dyspnea-slope, PaO2- slope, NE-slope, and pH-slope were calculated as the increase in dyspnea/the increase in V’O2 , the decrease in PaO2/the increase in V’O2 , the increase in norepinephrine/the increase in V’O2 , and the decrease in pH/the increase in V’O2 between at rest and at peak exercise, respectively
Bold type indicates significance
Fig. 1.
Correlations of ΔFO2 with (a) V’E/V’CO2 at peak exercise (n = 43), (b) V’E/V’CO2-nadir (n = 43), (c) V’E-V’CO2-slope (n = 43), and (d) V’E/V’CO2 at anaerobic threshold (AT) (n = 32). ΔFO2: difference between inspired and expired mean oxygen concentrations; V’CO2: carbon dioxide output; V’E: minute ventilation; and V’E-V’CO2-slope: the slope was determined by linear regression analysis of V’E to V’CO2 obtained during exercise (see the Material and methods section for details); V’E/V’CO2-nadir: the lowest value during exercise (see the Material and methods section for details)
Table 4.
Comparison of the higher and lower ΔFO2 at peak exercise groups
| Higher ΔFO2 at peak exercise group (> 2.46%, n = 29) |
Lower ΔFO2 at peak exercise group (≤2.46%, n = 14) |
||
|---|---|---|---|
| Age, y | 70.1 (6.5) | 71.9 (6.7) | 0.3878 |
| Sex, male/female | 24/5 | 13/1 | 0.3705 |
| BMI, kg⋅m− 2/ | 23.8 (2.8) | 20.9 (4.0) | 0.0073 |
| %DLCO, % a | 60.2 (21.8) b | 28.1 (23.5; 35.2) c | 0.0005 |
| CPET at peak exercise, n = 43 | |||
| ΔFO2, % | 3.07 (0.37) | 2.08 (0.23) | N.D. |
| V’O2, mL⋅min− 1⋅kg− 1 | 20.8 (16.6; 23.5) | 13.6 (3.4) | 0.0001 |
| V’E, L⋅min− 1 | 48.0 (42.6; 57.3) | 42.7 (13.7) | 0.0825 |
| Dyspnea, Borg scale | 6.9 (2.0) | 7.1 (2.0) | 0.7880 |
| VT, mL | 1146 (933; 1545) | 973 (338) | 0.0677 |
| fR, breaths⋅min− 1 | 42 (37; 47) | 46 (8) | 0.2805 |
| R | 1.04 (1.00; 1.13) | 1.09 (0.13) | 0.6778 |
| PaO2, mmHg | 52.4 (8.6) | 47.9 (10.1) | 0.1337 |
| CPET during exercise | |||
| Dyspnea-slope, Borg scale ⋅ L− 1⋅ min | 7.0 (5.1; 9.2) | 16.7 (6.8) | < 0.0001 |
| PaO2 (peak−rest), mmHg | −34.2 (−38.9; −31.3) | −24 (9.5) | 0.0024 |
| PaO2-slope, mmHg ⋅ L− 1⋅ min | −35.6 (−46.3; −27.2) | −52.7 (18.8) | 0.0427 |
| CPET at rest | |||
| ΔFO2, % | 2.22 (0.41) | 1.75 (0.30) | 0.0005 |
| V’O2, mL⋅min− 1⋅kg− 1 | 4.8 (3.5; 5.5) | 4.6 (1.4) | 0.7955 |
| V’E, L⋅min− 1 | 15.5 (12.8; 18.2) | 16.5 (3.2) | 0.3781 |
| CPET at AT, n = 32 | n = 25 | n = 7 | |
| ΔFO2, % | 3.14 (0.41) | 2.32 (0.26) | < 0.0001 |
| V’O2, mL⋅min− 1⋅kg− 1 | 13.6 (11.5; 16.6)) | 11.3 (2.0) | 0.0255 |
| V’E, L⋅min− 1 | 31.6 (26.8; 36.7) | 34.0 (4.0) | 0.3619 |
Data are presented as means (standard deviation) or medians (25th percentile; 75th percentile)
AT Anaerobic threshold, BMI Body mass index, ΔF02 Difference between inspired and expired mean oxygen concentration, DLCO Diffusing capacity of the lung for carbon monoxide, fR Respiratory frequency, PaO2 Arterial oxygen tension, PaO2 (peak−rest) The difference in PaO2 between at peak exercise and at rest, R Respiratory quotient, V’E Minute ventilation, V’O2 Oxygen uptake, VT Tidal volume
The dyspnea-slope and PaO2-slope were calculated as the increase in dyspnea/the increase in V’O2 , and the decrease in PaO2/the increase in V’O2, respectively
a: analyzed in 33 patients
b: analyzed in 22 patients
c: analyzed in 11 patients. Using the cut-off value (2.46%) of ΔFO2 at peak exercise corresponding to a V’E-V’CO2-slope ≥ 34 and < 34, patients were divided into two groups
The results of the univariate Cox proportional hazards model for mortality of the IPF patients are presented in Table 5. Next, a multivariate Cox proportional analysis of mortality of the IPF patients was performed using V’E and ΔFO2 at peak exercise, which are more directly measured by expiratory gas analysis during CPET, and peak V’O2, with age added as a modifier, as shown in Table 6. The results showed that ΔFO2 at peak exercise was a stronger prognostic factor (hazard ratio: 0.195, 95% CI 0.070 to 0.500; p = 0.0005, C index = 0.77). Furthermore, to confirm ΔFO2 as a prognostic factor, among the important prognostic factors obtained from univariate analysis (Table 5), the previously reported prognostic factors obtained from the CPET results in IPF, and the prognostic factors in IPF in general, variable selection was performed using the method described in the analysis section. As a result, with the four variables, that is, ΔFO2 at peak exercise, VT at peak exercise, PaO2-slope, and BMI with age added as a modifier, a multivariate Cox proportional hazards model analysis of mortality of IPF patients was performed (Table 6). In this analysis, ΔFO2 at peak exercise was also the most significant prognostic factor (hazard ratio: 0.437, 95% CI 0.201 to 0.958; p = 0.0389, C index = 0.79).
Table 5.
Predictors of mortality on univariate analysis in IPF patients (n = 43)
| Univariate analysis | ||
|---|---|---|
| Hazard ratio (95% CI) | p value | |
| Age, y | 1.053 (1.000 to 1.114) | 0.0498 |
| Sex, male/female | 0.658 (0.270 to 1.963) | 0.4189 |
| BMI, kg ⋅ m− 2 | 0.806 (0.708 to 0.914) | 0.0008 |
| Pulmonary function test | ||
| FEV1, L | 0.319 (0.135 to 0.733) | 0.0069 |
| FVC, L | 0.402 (0.214 to 0.713) | 0.0014 |
| %FVC, % | 0.975 (0.956 to 0.994) | 0.0078 |
| %DLCO, % (n = 33) | 0.977 (0.958 to 0.994) | 0.0076 |
| %DLCO, % (including estimated %DLCO, n = 43) | 0.981 (0.965 to 0.996) | 0.0108 |
| CPET at peak exercise | ||
| Dyspnea, Borg scale | 0.998 (0.838 to 1.193) | 0.9829 |
| V’O2, mL⋅min− 1 ⋅ kg− 1 | 0.913 (0.851 to 0.974) | 0.0051 |
| V’E, L ⋅ min− 1 | 0.962 (0.933 to 0.990) | 0.0079 |
| VT, mL | 0.998 (0.996 to 0.999) | < 0.0001 |
| fR, breaths ⋅ min− 1 | 1.029 (1.003 to 1.054) | 0.0308 |
| O2 pulse, mL ⋅ beats− 1 | 0.796 (0.691 to 0.913) | 0.0011 |
| V’E/V’O2 | 1.074 (1.032 to 1.118) | 0.0005 |
| V’E/V’CO2 | 1.075 (1.035 to 1.117) | 0.0003 |
| PaO2, mmHg | 0.962 (0.924 to 0.999) | 0.0464 |
| PaCO2, mmHg | 1.035 (0.967 to 1.107) | 0.3186 |
| HCO3−, mEq · L− 1 | 1.201 (1.063 to 1.344) | 0.0040 |
| Plasma lactate, mg ⋅ dL− 1 | 0.961 (0.931 to 0.988) | 0.0035 |
| ΔFO2, % | 0.300 (0.152 to 0.574) | 0.0003 |
| CPET during exercise | ||
| Dyspnea-slope, Borg scale ⋅ L− 1⋅ min | 1.118 (1.057 to 1.180) | 0.0002 |
| V’E-V’CO2-slope | 1.052 (1.010 to 1.092) | 0.0172 |
| V’E/V’CO2-nadir | 1.080 (1.037 to 1.123) | 0.0002 |
| V’O2 at ATa, mL ⋅ min− 1 | 0.997 (0.994 to 0.999) | 0.0041 |
| PaO2-slope, mmHg ⋅ L− 1⋅ min | 0.975 (0.962 to 0.991) | 0.0029 |
| NE-slope, ng ⋅ mL− 1⋅ L− 1⋅ min | 1.200 (1.000 to 1.410) | 0.0501 |
AT Anaerobic threshold, BMI Body mass index, CPET Cardiopulmonary exercise testing, ΔFO2 Difference between inspired and expired mean oxygen concentration, FEV1 Forced expiratory volume in one second, FVC Forced vital capacity, fR Breathing frequency, HCO3 Bicarbonateion, O2 pulse V’O2/heart rate, PaCO2 Arterial carbon dioxide tension, PaO2 Arterial oxygen tension, V’CO2 Carbon dioxide output, V’E Minute ventilation, and V’E/-V’CO2-slope The slope was determined by linear regression analysis of V’E to V’CO2obtained during exercise (see the Methods for details), V’E/V’CO2-nadir The lowest value during exercise (see the Methods for details), V’O2 Oxygen uptake, VT Tidal volume
The dyspnea-slope, PaO2 slope, and NE slope were calculated as the increase in dyspnea/the increase in V’O2, the decrease in PaO2/the increase in V’O2, and the increase in norepinephrine/the increase in V’O2, respectively. Estimated %DLCO see the Methods and Online Supplementary Material for details
aanalyzed in 32 patients
Table 6.
Multivariate analysis for mortality in IPF patients (n = 43)
| First multivariate analysis | |||
|---|---|---|---|
| Hazard ratio (95% CI) | |||
| Age, y | 1.053 (0.989 to 1.125) | 0.1098 | |
| CPET at peak exercise | |||
| V’E, L ⋅ min− 1 | 0.963 (0.920 to 1.006) | 0.0928 | |
| ΔFO2, % | 0.195 (0.070 to 0.500) | 0.0005 | |
| V’O2, mL⋅min− 1 ⋅ kg− 1 | 1.081 (0.959 to 1.216) | 0.2006 | |
| Second multivariate analysis | |||
|---|---|---|---|
| Hazard ratio (95% CI) | |||
| Age, y | 1.040 (0.979 to 1.107) | 0.2050 | |
| BMI, kg ⋅ m− 2 | 0.899 (0.781 to 1.033) | 0.1326 | |
| CPET at peak exercise | |||
| VT, mL | 0.999 (0.998 to 1.000) | 0.2059 | |
| ΔFO2, % | 0.437 (0.201 to 0.958) | 0.0389 | |
| CPET during exercise | |||
| PaO2-slope, mmHg ⋅ L− 1⋅ min | 0.987 (0.967 to 1.008) | 0.2162 | |
CPET Cardiopulmonary exercise testing, ΔFO2 Difference between inspired and expired mean oxygen concentration, FVC Forced vital capacity, PaO2-slope Slope calculated as the decrease in PaO2/the increase in V’O2, V’E Minute ventilation, V’O2 Oxygen uptake
The first multivariate analysis was performed using components of oxygen uptake. The secondmultivariate analysis was performed including previously reported prognostic factors
Bold type indicates significance
Discussion
There were three key findings in the present study. First, ΔFO2 at peak exercise, which is correlated with the ventilatory efficiency variables using CO2, was independent of V’E at peak exercise. Second, when ΔFO2 at peak exercise was divided into two groups based on the cut-off value of ΔFO2 at peak exercise corresponding to a V’E-V’CO2-slope ≥ 34 and < 34, the group with higher ΔFO2 had higher peak V’O2, despite no significant difference in V’E at peak exercise. Third, of the variables more directly measured and the key variables for exercise tolerance on CPET, ΔFO2 at peak exercise was a stronger predictor of survival in patients with IPF than V’E at peak exercise and peak V’O2, and in the analysis including prognostic factors already reported, ΔFO2 at peak exercise was the stronger predictor of survival in patients with IPF than VT at peak exercise and PaO2-slope, a measure of the degree of exertional hypoxemia. These findings show the possibility that ΔFO2 is not only correlated with exertional ventilatory efficiency, but also may be related to exercise tolerance independently of ventilatory ability, and it may be a prognostic factor for survival of patients with IPF, even if physiological ventilatory impairments with hypoxemia result.
In respiratory diseases with ventilatory impairments, improvement in alveolar O2 extraction ability (ΔFO2) is important to maintain and compensate for exercise tolerance, relatively independent of ventilatory ability (V’E). We have reported that ΔFO2 at peak exercise correlates with peak V’O2 with different dependence on disease or disease stage in 156 patients with chronic obstructive pulmonary disease (COPD) with 16 controls [9], and in 29 patients with nontuberculous mycobacterial pulmonary disease (NTM-PD) [10], almost independently of V’E. Similarly, in the IPF cases examined in the present study, at peak exercise, ΔFO2 was independent of V’E and dependent on peak V’O2. As shown in Table 4, in the higher ΔFO2 group, V’O2 was higher at peak exercise and AT, but no significant differences in V’E were observed between the two groups. Furthermore, it is interesting to see that, unexpectedly, the PaO2 difference between resting and peak exercise was greater in the high ΔFO2 group, despite better PaO2-slope and %DLCO, which are measures of gas exchange during exercise and at rest, respectively. One possible explanation for these findings is that IPF patients with higher ΔFO2 and V’O2 at peak exercise maintain O2 consumption ability throughout the body, even when the peripheral circulation is exposed to hypoxemia due to more strenuous exercise. Given the aforementioned evidence, and that ΔFO2 is only affected by FEO2 since FIO2 is given and fixed, and FEO2 is closely related to mixed venous O2 pressure, because V’O2 is calculated by the product of cardiac output and arterial-mixed venous O2 content difference; one can say that ΔFO2, which is a more directly measured CPET indicator, is related to O2 exchange throughout the body. It may be useful to evaluate the relationship between ΔFO2 and muscle mass evaluated by dual-energy X-ray absorptiometry (DEXA), and the relationship between ΔFO2 and muscle O2 extraction assessing tissue oxygenation in peripheral muscles during exercise, but they will be addressed in the future. These results suggest that evaluation of ΔFO2 separately from V’E in pathophysiological assessment may be useful in respiratory disease. The mean ΔFO2 at peak exercise in patients with respiratory disease including COPD and NTM-PD in our previous studies and in IPF patients in the present study was lower than in controls [9, 10]. However, the dependence of ΔFO2 at peak exercise expressed as r2 on the peak V’O2 in patients with respiratory diseases including COPD, NTM-PD, and IPF was higher than in controls [9, 10]. This may suggest that the arterial-to-venous O2 content difference values may differ between patients and the control group depending on the differences in exercise intensity. Furthermore, this might suggest that respiratory disease patients must rely on increased ΔFO2 to compensate for their ventilatory impairments, despite the disease. Two lines of evidence were identified from our reports as reasons for these. First, in the absence of an increase in V’E as an effect of rehabilitation with exercise therapy, in patients with severe and very severe COPD, the increase in ΔFO2 was highly dependent on the increase in peak V’O2 and AT levels of V’O2 [11, 12]. Second, conversely, when only ventilatory ability was primarily increased as a result of expiratory pressure load training for COPD, no change in ΔFO2 was observed, suggesting that the improvement of ventilatory variables, such as V’E and mean expiratory flow at peak exercise is the primary factor for improving exercise tolerance [21]. From a different perspective, CPET is an appropriate test to determine if there is a potential for improvement in ventilatory ability or O2 extraction ability to improve exercise tolerance in an individualized manner, regardless of disease.
Ventilation efficiency provides clinically important information [13]. In contrast, the physiological meaning of ventilation efficiency may be difficult, especially given that interventions targeting exercise tolerance elicit different responses in ventilation efficiency. For example, bronchodilators and expiratory pressure load training increased ventilation ability (such as V’E), but ventilation efficiency remained unchanged [21, 22]. Heliox provided sufficient ventilation ability with ventilatory inefficiency, despite the improvement in subjective symptoms [23]. Lung volume reduction surgery increased ventilation ability and decreased the ventilatory efficiency value [24]. Combination therapy with medium-chain triglycerides and ghrelin did not affect ventilatory ability but decreased the ventilatory efficiency value [25]. Although ventilatory efficiency is best defined by the relationship of the liters of ventilation required to eliminate a liter of CO2 or O2 [7], these findings do not indicate a strong relationship between ventilation efficiency and ventilation itself, likely because ventilation efficiency may be affected by multiple factors. At the least, the ventilation efficiency obtained from expiratory gas analysis is an indirect indicator calculated from two variables (V’E and V’CO2 or V’O2), whereas ΔFO2 is a directly measured indicator. We believe it is important to recognize that regardless of the disease, ventilation efficiency is strongly associated with ΔFO2, which reflects the O2 extraction ability throughout the body, independent of ventilation (V’E). This recognition might lead to understanding the clinical meaning of ventilatory efficiency.
The O2 extraction ability expressed by ΔFO2 was a better prognostic factor for IPF than ventilatory ability and exercise tolerance. So far, many correlations between survival and CPET variables have been identified, and it has become clear that reduced ventilatory efficiency has a profound impact on IPF patients [6, 13, 26]. In the present study, ΔFO2 at peak exercise showed a strong correlation with the variables related to ventilatory efficiency during exercise, especially, V’E/V’CO2-nadir or V’E/V’CO2 at peak exercise (Table 3), which has been confirmed almost equally in other respiratory diseases in our previous studies [8–12]. Therefore, in the multivariate prognostic analysis, the variables related to ventilatory efficiency were excluded. Interestingly, on the first multivariate analysis, ΔFO2 at peak exercise was a stronger prognostic factor for IPF than V’E at peak exercise and peak V’O2 of the variables that are measured more directly or are key for exercise tolerance in CPET, although as prognostic factors for IPF, peak V’O2 <8.3 mL⋅min− 1 ⋅ kg− 1 as an index of exercise tolerance was first reported by Feel et al. in 2009 [4]. Based on clinical significance, the variables included in the second multivariate analysis were as follows. Although FVC and its decline over time have been recognized as reliable prognostic factors [27, 28], considering the results of univariate analysis, VT at peak exercise was identified in the multivariate analysis of the variables related to each other (r = 0.9) in the present study. In addition, PaO2-slope as an index of exercise-induced hypoxemia reported by us in 2003 was also identified, because, interestingly, PaO2-slope was a more significant prognostic factor than peak V’O2 and V’E/V’CO2 at peak exercise [5]. These above factors have been consistent and clinically important based on well-recognized physiological phenomena in the follow-up of IPF patients. However, on the second multivariate analysis including age and BMI [29], ΔFO2 at peak exercise was a stronger prognostic factor than PaO2-slope, and VT at peak exercise. This means that the prognosis is more affected by the reduced O2 extraction ability of the whole body than by the physiological impairment of the mechanical ventilation with hypoxemia, which has often been seen in clinical practice. We have been unaware of the existence of a variable that underlies physiological phenomena in IPF, is a more directly measured CPET indicator, and is related to prognosis and exercise tolerance, namely ΔFO2, even though it is obtained by CPET.
Some limitations should be considered in the present study. First, this study was performed in a single center with a small number of cases. In addition, this study involved patients in whom inspired and expired O2 concentrations analysis and arterial blood gas analysis were feasible. However, arterial blood gas analysis is not routinely performed during CPET, so there is a possibility of bias in case enrollment. Second, DLCO [30] could not be measured in 10 patients (23%) due to their inability to perform, although the estimated %DLCO [5] values in these 10 patients were used for reference purposes. Third, clarifying the relationship between ΔFO2 and muscle mass and the relationship between ΔFO2 and exertional muscle O2 extraction in peripheral muscles not only in respiratory diseases, but also in other conditions such as cardiovascular diseases, holds the potential to lead to new therapeutic strategies for improving prognosis, as well as exercise tolerance. Admittedly, we have observed that ghrelin i.e., an endogenous ligand for growth hormone secretagogue receptor, which has a variety of cardiovascular activities and muscle mitochondrial oxidative capacity [31, 32], long-term acupuncture, which affects muscle tissue oxygenation [33], and exercise therapy improved ΔFO2, but we have yet to see sufficient improvement [11, 12, 34, 35]. Fifth, although it was possible to examine CPET results only for IPF patients who were not treated, the number of confirmed cases in which antifibrotic agents were started during the clinical course in the present study was only 1; thus, failure to include the effect of antifibrotic drugs may have affected prognosis in terms of ventilation. However, given that alveolar O2 extraction ability is independent of ventilation, as shown in the present study, there is a growing expectation that developing therapeutic interventions to enhance O2 extraction ability throughout the body, including peripheral muscles, independent of ventilation-related treatments such as antifibrotic treatments that suppress ventilation decline, could further improve exercise tolerance and prognosis.
Conclusion
The present study demonstrates that O2 extraction ability, which is more directly measured by CPET, is independent of ventilatory ability, affects exercise tolerance, and is also a stronger prognostic factor than the commonly recognized hypoxemia-associated exercise physiology indices. Further investigation of O2 extraction ability in different clinical scenarios and the development of therapies to improve it is, of course, desirable, but without CPET, it may be impossible to establish individualized therapies based on ventilation and O2 extraction ability to improve exercise tolerance and prognosis in IPF.
Supplementary Information
Acknowledgements
The authors would like to thank Y. Matsumoto (Department of Medical Biochemistry, Graduate School of Medicine, Osaka Metropolitan University) for the acquisition and analysis of data; and H. Yanagi, R. Yonezawa, S. Sakaguchi, and K. Koyama for help with the measurements of cardiopulmonary exercise testing. The authors also wish to thank the patients who participated in the present study.
Abbreviations
- AT
Anaerobic threshold
- COPD
Chronic obstructive pulmonary disease
- CPET
Cardiopulmonary exercise testing
- ΔFO2
Inspired‒expired mean oxygenation concentration difference
- Dyspnea-slope
Increase in Borg scale/increase in oxygen uptake between at rest and at peak exercise
- FEO2
Expired mean oxygen concentration
- FIO2
Inspired mean oxygen concentration
- IPF
Idiopathic pulmonary fibrosis
- NE
Norepinephrine
- NE-slope
Increase in norepinephrine /increase in oxygen uptake between at rest and at peak exercise
- NTM-PD
Nontuberculous mycobacterial pulmonary disease
- PaO2-slope
Decrease in PaO2/increase in oxygen uptake between at rest and at peak exercise
- pH-slope
Decrease in pH/increase in oxygen uptake between at rest and at peak exercise
Authors’ contributions
K.M. developed the concept and design of the study, wrote the manuscript, and is the author responsible for the content of the manuscript. K.M., R.N., K. S., Y.N., Y.M., T.N., T.M., K.T., and H.K. acquired and analyzed the data, and were involved in the interpretation of data. All authors participated in the development of the manuscript, gave final approval of the manuscript for submission, and agree to be accountable for the integrity of the work.
Funding
This work was supported by a Grant-in-Aid for Clinical Research from the National Hospital Organization (number, not applicable).
Data availability
The data that support the findings of this study are available from the corresponding author upon reasonable request.
Declarations
Ethics approval and consent to participate
The study was approved by the ethics committees at NHO Osaka Toneyama Medical Center (TNH-R-2020055-4) and was conducted in accordance with the Declaration of Helsinki. All patients gave written, informed consent.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Kaunisto J, Salomaa ER, Hodgson U, Kaarteenaho R, Kankaanranta H, Koli K, et al. Demographics and survival of patients with idiopathic pulmonary fibrosis in the FinnishIPF registry. ERJ Open Res. 2019. 10.1183/23120541.00170-2018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Khor YH, Ng Y, Barnes H, Goh NSL, McDonald CF, Holland AE. Prognosis of idiopathic pulmonary fibrosis without anti-fibrotic therapy: a systematic review. Eur Respir Rev. 2020. 10.1183/16000617.0158-2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Tran T, Šterclová M, Mogulkoc N, Lewandowska K, Müller V, Hájková M, et al. The European multipartner IPF registry (EMPIRE): validating long-term prognostic factors in idiopathic pulmonary fibrosis. Respir Res. 2020;21(1):11. 10.1186/s12931-019-1271-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Fell CD, Liu LX, Motika C, Kazerooni EA, Gross BH, Travis WD, et al. The prognostic value of cardiopulmonary exercise testing in idiopathic pulmonary fibrosis. Am J Respir Crit Care Med. 2009;179(5):402–7. 10.1164/rccm.200802-241OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Miki K, Maekura R, Hiraga T, Okuda Y, Okamoto T, Hirotani A, et al. Impairments and prognostic factors for survival in patients with idiopathic pulmonary fibrosis. Respir Med. 2003;97(5):482–90. 10.1053/rmed.2002.1469. [DOI] [PubMed] [Google Scholar]
- 6.Laviolette L, Laveneziana P. Exercise Testing in the prognostic evaluation of patients with lung and heart diseases. In: Palange P, Laveneziana P, Neder JA, Ward SA, editors. Clinical Exercise Testing (ERS Monograph). edn. Sheffield: European Respiratory Society; 2018. p. 222–34.
- 7.Wasserman K, Hansen J, Sue D, Stringer W, Sietsema K, Sun X-G. Principles of exercise testing and interpretation: including pathophysiology and clinical applications. Philadelphia: Lippincott Williams and Wilkins; 2012. [Google Scholar]
- 8.Yanagi H, Miki K, Koyama K, Miyamoto S, Mihashi Y, Nagata Y, et al. Gas exchange efficiency slopes to assess exercise tolerance in chronic obstructive pulmonary disease. BMC Pulm Med. 2024;24(1):550. 10.1186/s12890-024-03312-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Miki K, Tsujino K, Maekura R, Matsuki T, Miki M, Hashimoto H, et al. Oxygen extraction based on inspiratory and expiratory gas analysis identifies ventilatory inefficiency in chronic obstructive pulmonary disease. Front Physiol. 2021;12:703977. 10.3389/fphys.2021.703977. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Miyamoto S, Miki K, Kitada S, Nagata Y, Kijima R, Mihashi Y, et al. Exercise intolerance and oxygen dynamics in nontuberculous mycobacteria with bronchiectasis. Respir Investig. 2024;62(6):1058–63. 10.1016/j.resinv.2024.08.017. [DOI] [PubMed] [Google Scholar]
- 11.Miyazaki A, Miki K, Maekura R, Tsujino K, Hashimoto H, Miki M, et al. Increased oxygen extraction by pulmonary rehabilitation improves exercise tolerance and ventilatory efficiency in advanced chronic obstructive pulmonary disease. J Clin Med. 2022. 10.3390/jcm11040963. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Sumitani H, Miki K, Yamamoto Y, Mihashi Y, Nagata Y, Miyamoto S et al. Increased oxygen consumption ability with pulmonary rehabilitation improves submaximal exercise capacity in advanced COPD. Respir Care.2025 Jul;70(7):887–895. 10.1089/respcare.12303. [DOI] [PubMed]
- 13.Gille T, Laveneziana P. Cardiopulmonary exercise testing in interstitial lung diseases and the value of ventilatory efficiency. Eur Respir Rev. 2021. 10.1183/16000617.0355-2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Raghu G, Remy-Jardin M, Myers JL, Richeldi L, Ryerson CJ, Lederer DJ, et al. Diagnosis of idiopathic pulmonary Fibrosis. An official ATS/ERS/JRS/ALAT clinical practice guideline. Am J Respir Crit Care Med. 2018;198(5):e44–68. 10.1164/rccm.201807-1255ST. [DOI] [PubMed] [Google Scholar]
- 15.American Thoracic Society. Standardization of Spirometry, 1994 update. Am J Respir Crit Care Med. 1995;152(3):1107–36. 10.1164/ajrccm.152.3.7663792. [DOI] [PubMed] [Google Scholar]
- 16.Miki K, Maekura R, Hiraga T, Hashimoto H, Kitada S, Miki M, et al. Acidosis and raised norepinephrine levels are associated with exercise dyspnoea in idiopathic pulmonary fibrosis. Respirology. 2009;14(7):1020–6. 10.1111/j.1440-1843.2009.01607.x. [DOI] [PubMed] [Google Scholar]
- 17.Hanaoka M, Wada Y, Goto N, Kitaguchi Y, Koarai A, Kubota M, et al. Referential equations for pulmonary diffusing capacity generated from the Japanese population using the Lambda, Mu, or Sigma method and their comparisons with prior referential equations. Respir Investig. 2023;61(6):687–97. 10.1016/j.resinv.2023.07.009. [DOI] [PubMed] [Google Scholar]
- 18.Kubota M, Kobayashi H, Quanjer PH, Omori H, Tatsumi K, Kanazawa M. Reference values for spirometry, including vital capacity, in Japanese adults calculated with the LMS method and compared with previous values. Respir Investig. 2014;52(4):242–50. 10.1016/j.resinv.2014.03.003. [DOI] [PubMed] [Google Scholar]
- 19.Radtke T, Crook S, Kaltsakas G, Louvaris Z, Berton D, Urquhart DS, et al. ERS statement on standardisation of cardiopulmonary exercise testing in chronic lung diseases. Eur Respir Rev. 2019. 10.1183/16000617.0101-2018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Arena R, Myers J, Aslam SS, Varughese EB, Peberdy MA. Peak VO2 and VE/VCO2 slope in patients with heart failure: a prognostic comparison. Am Heart J. 2004;147(2):354–60. 10.1016/j.ahj.2003.07.014. [DOI] [PubMed] [Google Scholar]
- 21.Miki K, Tsujino K, Fukui M, Miki M, Kitajima T, Sumitani H, et al. Laryngeal widening and adequate ventilation by expiratory pressure load training improve aerobic capacity in COPD: a randomised controlled trial. Thorax. 2023;79(1):23–34. 10.1136/thorax-2022-219755. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.O’Donnell DE, Voduc N, Fitzpatrick M, Webb KA. Effect of salmeterol on the ventilatory response to exercise in chronic obstructive pulmonary disease. Eur Respir J. 2004;24(1):86–94. 10.1183/09031936.04.00072703. [DOI] [PubMed] [Google Scholar]
- 23.Palange P, Valli G, Onorati P, Antonucci R, Paoletti P, Rosato A, et al. Effect of heliox on lung dynamic hyperinflation, dyspnea, and exercise endurance capacity in COPD patients. J Appl Physiol (1985). 2004;97(5):1637–42. 10.1152/japplphysiol.01207.2003. [DOI] [PubMed] [Google Scholar]
- 24.Armstrong HF, Dussault NE, Thirapatarapong W, Lemieux RS, Thomashow BM, Bartels MN. Ventilatory efficiency before and after lung volume reduction surgery. Respir Care. 2015;60(1):63–71. 10.4187/respcare.03233. [DOI] [PubMed] [Google Scholar]
- 25.Miki K, Kitada S, Miki M, Hui SP, Shrestha R, Yoshimura K, et al. A phase II, open-label clinical trial on the combination therapy with medium-chain triglycerides and Ghrelin in patients with chronic obstructive pulmonary disease. J Physiol Sci. 2019;69(6):969–79. 10.1007/s12576-019-00716-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Barratt SL, Davis R, Sharp C, Pauling JD. The prognostic value of cardiopulmonary exercise testing in interstitial lung disease: a systematic review. ERJ Open Res. 2020. 10.1183/23120541.00027-2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.du Bois RM, Weycker D, Albera C, Bradford WZ, Costabel U, Kartashov A, et al. Ascertainment of individual risk of mortality for patients with idiopathic pulmonary fibrosis. Am J Respir Crit Care Med. 2011;184(4):459–66. 10.1164/rccm.201011-1790OC. [DOI] [PubMed] [Google Scholar]
- 28.Martinez FJ, Flaherty K. Pulmonary function testing in idiopathic interstitial pneumonias. Proc Am Thorac Soc. 2006;3(4):315–21. 10.1513/pats.200602-022TK. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Zinellu A, Carru C, Pirina P, Fois AG, Mangoni AA. A systematic review of the prognostic significance of the body mass index in idiopathic pulmonary fibrosis. J Clin Med. 2023. 10.3390/jcm12020498. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Collard HR, King TE Jr., Bartelson BB, Vourlekis JS, Schwarz MI, Brown KK. Changes in clinical and physiologic variables predict survival in idiopathic pulmonary fibrosis. Am J Respir Crit Care Med. 2003;168(5):538–42. 10.1164/rccm.200211-1311OC. [DOI] [PubMed] [Google Scholar]
- 31.Akalu Y, Molla MD, Dessie G, Ayelign B. Physiological effect of Ghrelin on body systems. Int J Endocrinol. 2020;2020:1385138. 10.1155/2020/1385138. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Müller TD, Nogueiras R, Andermann ML, Andrews ZB, Anker SD, Argente J, et al. Ghrelin. Mol Metab. 2015;4(6):437–60. 10.1016/j.molmet.2015.03.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Kaneko Y, Kime R, Furuya E, Sakamoto A, Katsumura T. Effects of acupuncture stimulation on muscle tissue oxygenation at different points. Adv Exp Med Biol. 2016;923:327–33. 10.1007/978-3-319-38810-6_43. [DOI] [PubMed] [Google Scholar]
- 34.Maekura T, Miki K, Miki M, Kitada S, Maekura R. Clinical effects of acupuncture on the pathophysiological mechanism of chronic obstructive pulmonary disease during exercise. Int J Chron Obstruct Pulmon Dis. 2019;14:2787–98. 10.2147/copd.S225694. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Miki K, Maekura R, Nagaya N, Kitada S, Miki M, Yoshimura K, et al. Effects of ghrelin treatment on exercise capacity in underweight COPD patients: a substudy of a multicenter, randomized, double-blind, placebo-controlled trial of ghrelin treatment. BMC Pulm Med. 2013;13:37. 10.1186/1471-2466-13-37. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.