Pulmonary diffusing capacity for carbon monoxide and nitric oxide after COVID‐19: A prospective cohort study (the SECURe study)

Abstract Many patients exhibit persistently reduced pulmonary diffusing capacity after coronavirus disease 2019 (COVID‐19). In this study, dual test gas diffusing capacity for carbon monoxide and nitric oxide (D L,CO,NO) metrics and their relationship to disease severity and physical performance were examined in patients who previously had COVID‐19. An initial cohort of 148 patients diagnosed with COVID‐19 of all severities between March 2020 and March 2021 had a D L,CO,NO measurement performed using the single‐breath method at 5.7 months follow‐up. All patients with at least one abnormal D L,CO,NO metric (n = 87) were revaluated at 12.5 months follow‐up. The D L,CO,NO was used to provide the pulmonary diffusing capacity for nitric oxide (D L,NO), the pulmonary diffusing capacity for carbon monoxide (D L,CO,5s), the alveolar–capillary membrane diffusing capacity and the pulmonary capillary blood volume. At both 5.7 and 12.5 months, physical performance was assessed using a 30 s sit‐to‐stand test and the 6 min walk test. Approximately 60% of patients exhibited a severity‐dependent decline in at least one D L,CO,NO metric at 5.7 months follow‐up. At 12.5 months, both D L,NO and D L,CO,5s had returned towards normal but still remained abnormal in two‐thirds of the patients. Concurrently, improvements in physical performance were observed, but with no apparent relationship to any D L,CO,NO metric. The severity‐dependent decline in D L,NO and D L,CO observed at 5.7 months after COVID‐19 appears to be reduced consistently at 12.5 months follow‐up in the majority of patients, despite marked improvements in physical performance.

membrane diffusing capacity and the pulmonary capillary blood volume.At both 5.7 and 12.5 months, physical performance was assessed using a 30 s sit-to-stand test and the 6 min walk test.Approximately 60% of patients exhibited a severity-dependent decline in at least one D L,CO,NO metric at 5.7 months follow-up.At 12.5 months, both D L,NO and D L,CO,5s had returned towards normal but still remained abnormal in twothirds of the patients.Concurrently, improvements in physical performance were observed, but with no apparent relationship to any D L,CO,NO metric.The severitydependent decline in D L,NO and D L,CO observed at 5.7 months after COVID-19 appears to be reduced consistently at 12.5 months follow-up in the majority of patients, despite marked improvements in physical performance.

K E Y W O R D S
diffusion, long COVID, SARS-CoV-2

INTRODUCTION
After the initial surge of the coronavirus disease 2019 (COVID-19) pandemic, it became evident that the pulmonary effects of the disease often lasted far beyond the immediate severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection (Crook et al., 2021).
Besides common pulmonary symptoms, such as breathlessness, chest pain and fatigue, grouped under the term 'long COVID' , numerous studies have reported a severity-dependent reduction in pulmonary diffusing capacity (D L ) (Huntley et al., 2022;Katzenstein et al., 2022;Watanabe et al., 2022).
Morphologically, the post-COVID reduction in D L has been shown to be caused by a fibrosis-like restrictive lung disease rather than pulmonary vascular disease (Katzenstein et al., 2022).These changes appear to be reversible, at least in part, although residual parenchymal damage is often observed in severe cases (Lee et al., 2022;Torres-Castro et al., 2021;Zhao et al., 2020).Functionally, dual test gas diffusing capacity assessments using carbon monoxide (CO) and nitric oxide (NO) (i.e., D L,CO,NO ), performed at 2 to 4 months follow-up, suggest that the impairment of pulmonary diffusion is caused mainly by alveolar-capillary membrane diffusing capacity (D M ) changes following mild-to-moderate COVID-19 (Barisione & Brusasco, 2021).In contrast, a combination of D M and capillary blood volume (V C ) changes appear to be involved following severe COVID-19 (Núñez-Fernández et al., 2021;Seccombe et al., 2023).At present, it is unknown whether the recovery of diffusing capacity is caused by an increase in D M and/or V C and whether such changes are related to changes in physical performance.Furthermore, it is unknown whether the post-COVID changes in diffusing capacity are associated with an impairment of alveolar-capillary reserve; that is, the ability to recruit and distend pulmonary capillaries (Behnia et al., 2017;Hsia et al., 1994).
The present paper is based on data from the Danish SECURe (Sequelae of COVID-19 at Copenhagen University Hospital-Rigshospitalet) study, a prospective cohort study designed to track post-COVID complications using comprehensive clinical, physiological and chest imaging evaluations in both hospitalized and non-hospitalized COVID-19 patients (Katzenstein et al., 2022).
Here, we explored: (1) the degree to which the D L,CO,NO metrics were reduced at 5.7 months follow-up in the SECURe cohort; (2) the extent to which the post-COVID decline in D L,CO,NO persisted up to 12.5 months follow-up, including the extent to which any such changes were caused by D M and V C , respectively; and (3) whether any such changes were related to changes in physical performance.

Ethical approval
The study conformed to the standards set by the most recent version of the Declaration of Helsinki (except for registration in a database) and was approved by the Danish Data Protection Agency and the Regional Ethical Committee of the Capital Region of Denmark (file no.H-20028792).All study participants provided oral and written informed consent.

Study design and setting
The SECURe study is a prospective cohort study of individuals who were diagnosed with SARS-CoV-2 confirmed by PCR testing by an oral or nasopharyngeal swab.The study was conducted as a collaboration between several departments at University Hospital Copenhagen-Rigshospitalet, a specialized University Hospital in Denmark serving mainly as a tertiary health-care service.

Recruitment
All COVID-19 patients who had been admitted.SARS-CoV-2-infected participants without need for hospitalization were also invited to participate.For admitted patients, the initial SECURe study visit was aimed at being scheduled 3-4 months post-discharge, whereas for non-admitted participants, it was scheduled for 3-4 months after testing positive for SARS-CoV-2 (Katzenstein et al., 2022).In the present paper, we report on all participants (n = 148) who had a D L,CO,NO measurement performed at 5.7 months follow-up.All patients who had at least one abnormal D L,CO,NO metric (n = 87) were re-evaluated at ∼12.5 months follow-up.

Clinical scores
Participants were classified as group 1 (asymptomatic), group 2 (mild), group 3 (moderate), group 4 (severe) or group 5 (critical), according to the clinical features during the acute SARS-CoV-2 infection (Gandhi et al., 2020).To assess symptom severity during follow-up, the chronic obstructive pulmonary disease (COPD) assessment test (CAT) was used, which is a questionnaire designed to assess the impact of COPD on everyday life activities and enables the evaluation of changes over time (Jones, Harding, Wiklund et al., 2009).The questionnaire focuses mainly on respiratory symptoms, such as shortness of breath, cough, chest pain and limitations in physical abilities; however, questions also relate to energy levels and sleep patterns (Jones, Harding, Berry et al., 2009).Comorbidity burden was assessed at baseline using the Charlson comorbidity index (Charlson et al., 1987).

Physical performance
Physical performance was assessed using 30 s sit-to-stand (STS) muscle strength test and 6 min walk test (6MWT) (Alcazar et al., 2018;Peeters & Mets, 1996).The STS test was conducted in accordance with standard test procedures.The 6MWT was performed on a 30 m track.Both STS and 6MWT were reported as a percentage of predicted based on the empirical equations provided elsewhere (Enright & Sherrill, 1998;Suetta et al., 2019).

Lung function testing
Lung function testing included dynamic spirometry, body plethysmography and single-breath measurement of D L,CO in accordance with European Respiratory Society (ERS) guidelines (Bhakta et al., 2023;Graham et al., 2019).The tests were conducted at the Department of Clinical Physiology and Nuclear Medicine at University Hospital Copenhagen-Rigshospitalet using an integrated system (MasterScreen; Vyaire Medical, Würzburg, Germany).Before measurements, standing height (to the nearest 1 mm), body mass (to the nearest 100 g) and haemoglobin (Hb) concentration (to the nearest 0.1 mM) were obtained.The [Hb] was measured in capillary blood using

What is the central question of this study?
To what extent do changes in pulmonary diffusing capacity observed 6 months after coronavirus disease 2019 (COVID-19) persist after a year?
• What is the main finding and its importance?
In patients who have had COVID-19, the changes in pulmonary diffusing capacity for carbon monoxide and nitric oxide, the alveolar-capillary membrane diffusing capacity and pulmonary capillary blood volume observed at approximately 6 months follow-up largely persist 1 year after the acute disease, despite a marked improvement in physical performance.
The variables reported in the present study are forced expiratory volume in 1 s (FEV 1 ), forced vital capacity (FVC), FEV 1 /FVC, total lung capacity (TLC), residual volume (RV) and [Hb]-corrected D L,CO .The D L,CO was based on an 8-12 s breath-hold (denoted as D L,CO,10s ) using a gas mixture of 0.3% CO, 0.3% CH 4 , 20.9% O 2 and balance N 2 .Values were reported as a percentage of predicted, except for FEV 1 /FVC, which was reported as a raw value.Reference values were based on height, sex and age as appropriate (Stanojevic et al., 2022).

Dual test gas diffusing capacity for CO and NO
The D L,CO,NO was assessed in the seated upright position, adhering to the ERS task force recommendations (Zavorsky et al., 2017).The procedure commenced with tidal breathing, transitioning to a maximal expiration to residual volume, followed by a swift maximal inhalation to total lung capacity.During this, subjects inhaled a gas mix mixture of 0.2% CO, 50 p.p.m.NO, 6.3% He, 20.9% O 2 and balance N 2 .This was followed by a stable 4-8 s breath-hold at total lung capacity, then a forceful, uninterrupted maximal expiration.The D L,CO measurements obtained by this method are denoted as D L,CO,5s .
Manoeuvres were deemed acceptable if the inspired volume from TLC was ≥90% of FVC or the plethysmography-derived vital capacity, or if it was ≥85% of the FVC or plethysmography-based vital capacity with a V A within either 200 mL or 5% of the highest alveolar volume (V A ) from previous acceptable manoeuvres (Zavorsky et al., 2017).
Repeatability was confirmed on the basis of two consistent readings that fell within 5.8 mmol min −1 kPa −1 (17.4 mL min −1 mmHg −1 ) for D L,NO , 1.0 mmol min −1 kPa −1 (2.98 mL min −1 mmHg −1 ) for D L,CO,5s , 11.2 mmol min −1 kPa −1 (33.5 mL min −1 mmHg −1 ) for D M and 10 mL for V C (Zavorsky et al., 2017).Readings were repeated post-washout until repeatability was achieved and were limited to 12 attempts.The average of two manoeuvres that satisfied the repeatability criteria was reported for both D L,NO and D L,CO,5s .When these criteria were not met, the data from the best manoeuvre were used, with deviations highlighted for judicious evaluation.
Participants with at least one abnormal D L,CO,NO metric (D L,CO,5s , D L,NO , D M or V C ) at 12.5 months follow-up were asked to perform an additional measurement in the supine position when practically possible (Madsen et al., 2023).For this, participants were placed in the supine position, with slight head elevation using a pillow for comfort; after 15 min of rest in this position, the D L,CO,NO measurements were repeated using the same approach and criteria as described above.

Calculations
The values of D M and V C were calculated from D L,CO,NO measurements using the following equations (Munkholm et al., 2018): where  is the D M ratio (D M,NO /D M,CO ), which is assumed to be 1.97, and k is the ratio between the specific conductance of the two test gases (θ NO /θ CO ).Here, θ NO was assumed to have a finite value of 1.51 mL blood min −1 kPa −1 (mmol NO) −1 [4.5 L blood min −1 mmHg −1 (mL NO) −1 ] ( Carlsen & Comroe, 1958), and θ CO was calculated as: where [Hb] is the concentration of haemoglobin in capillary blood (mM).
Values of D L,NO , D L,CO,5s , D M and V C were reported as raw values, as a percentage of predicted (%pred) and as z-scores using established reference equations to normalize for height, sex and age (Munkholm et al., 2018).

Statistical analyses
All data were entered in REDCap (v.13.7.14For all data, a two-sided P < 0.05 was considered statistically significant.

Participant characteristics
Demographics and clinical characteristics according to disease severity at 5.7 months (median 170 days, interquartile range 46-263 days) follow-up are provided in Table 1.Of note, greater disease severity was associated with lower lung function metrics, including FEV 1 , FVC, TLC, RV and a lower D L,CO,10s .

Dual test gas diffusing capacity for CO and NO
Of the 148 patients initially included, 87 (59%) completed follow-up D L,CO,NO at 12.5 months (median 375 days, interquartile range 328-581 days), of whom only one patient was classified as group 1.At 5.7 months follow-up, group 4 and 5 patients generally had a higher rate of abnormal D L,NO and D L,CO,5s and V C scores than the other severity groups, and the rate of abnormal D M was higher in group 5 than in the other groups (Table 2).A largely similar distribution of abnormal scores was observed in the 87 patients who completed

months follow-up
The D L,CO,NO metrics presented as a percentage of predicted are provided in Table 3, and changes from 5.7 to 12.5 months are shown in Table 4.The absolute values and z-scores are provided in the Online Supplemental File.Overall, there was a small increase in both
At 5.7 months, STS and 6MWT were reduced similarly across the remaining severity groups (Table 3).In the 43 patients who completed 12.5 months follow-up, STS had improved at 5.7 months, primarily driven by an improvement in group 4 (Table 4).For 6MWT, all groups had normal scores at both follow-ups, but it was higher in groups 3-5 at 12.5 than at 5.7 months follow-up (Table 4).At 12.5 months, 30 of those with an available D L,CO,NO measurement had STS performed, and 31 had both a D L,CO,NO measurement and 6MWT performed.No statistically significant correlations were observed between changes in D L,CO,5s %pred and STS (r = −0.10,P = 0.60), D L,CO,5s %pred and 6MWT (r = 0.14, P = 0.47), D L,NO %pred and STS (r = −0.12,P = 0.54) or between D L,NO %pred and 6MWT (r = 0.08, P = 0.69).

DISCUSSION
The main findings of the present prospective cohort study are that ∼60% of COVID-19 patients exhibited at least one abnormal D L,CO,NO metric at 5.7 months follow-up.In the subgroup of patients with an abnormal D L,CO,NO metric at 5.7 months follow-up, both D L,NO %pred and D L,CO,5s %pred had increased at 12.5 months follow-up, but still remained abnormal in two-thirds of the patients.Concurrently, an improvement in physical performance according to both STS and 6MWT was observed, but with no apparent relationship to any D L,CO,NO metric according to correlational analyses.
Both D L,NO %pred and D L,CO,5s %pred differed significantly at 5.7 months in an apparently severity-dependent fashion.Hence, they were mostly reduced in groups 3-5 (Table 3) and accompanied by a classic restrictive lung function pattern with reduced TLC (Table 1).A comparable severity-dependent reduction in D L,CO,10s at 5.7 months follow-up has likewise been linked directly to a restrictive lung disease pattern with a decreased TLC, ground glass opacities and fibrosis-like changes in the lungs on high-resolution chest CT imaging, indicating ongoing interstitial inflammation and remodelling (Nowers et al., 2002;Verleden et al., 2023).Although it is evident that a pathologically reduced D L,CO,10s is exceedingly common post-COVID-19, the present findings indicate that D L,NO , D L,CO,5s , D M and V C are affected in a similar manner.This is in contrast to findings from two previous studies on 94 and 200 patients from Italy and Spain, respectively (Barisione & Brusasco, 2021;Núñez-Fernández et al., 2021).In those two studies, D L,CO,NO metrics were abnormal in 30%-60% of cases up to 9 months follow-up, and reduction in D M was found to be the dominant mechanism.However, in accordance with our findings, a study from Australia on 49 patients found that D M and V C were equally affected at 2 and 4 months follow-up (Seccombe et al., 2023), indicating that neither pathological changes in the alveolar-capillary membrane nor in the blood available for gas exchange dominates as the underlying mechanism.These variances in study outcomes are not attributable to disparities in COVID-19 severity or follow-up durations, because these elements are comparable across the studies.
Nonetheless, directly comparing the prevalence of abnormal D L,CO,NO measurements between countries is challenging, owing to variations in the magnitude of the COVID-19 outbreak, health-care capabilities and a range of preventive, diagnostic and therapeutic approaches, including criteria for hospital and intensive care unit admissions.Despite these differences, it can be deduced that a significant reduction in D L,NO and D L,CO,5s at both 5.7 and 12.5 months post-COVID-19 is remarkably frequent, and its prevalence escalates in tandem with the severity of the acute phase of the disease.However, definitive conclusions about the specific alterations in D M and V C remain elusive at this time.
As for the gradual improvement in D L,NO %pred and D L,CO,5s %pred from 5.7 to 12.5 months follow-up, this is in agreement with previous studies on D L,CO,10s , which generally show improvements over time in most patients with mild-to-moderate disease, while a concomitant remission of ground glass opacities and fibrosis-like changes are observed on chest imaging (Huntley et al., 2022;Lee et al., 2022).
However, it must be noted that the observed changes in the present study showed no apparent relationship to disease severity, and in any event, the absolute improvement was miniscule.Hence, D L,NO showed an increase of 0.5-1.7 mmol min −1 kPa −1 and D L,CO,5s of 0.1-0.3mmol min −1 kPa −1 (see Online Supplemental File, Table C), both of which are well below the smallest real difference of 5.4 mmol min −1 kPa −1 for D L,NO and 1.0 D L,CO,5s as established in a recent test-retest reliability study on healthy volunteers (Madsen et al., 2023).Neither D M or V C showed any consistent changes from 5.7 to 12.5 months follow-up, thus indicating that the changes in diffusing capacity with time were not dominated by either one in any of the severity groups.
Although D L,CO,10s abnormalities have been linked mechanistically to persistent symptoms of dyspnoea and functional limitations as part of the long-COVID syndrome (Schwendinger et al., 2022), we did not find any association between changes in D L,NO or D L,CO,5s with either STS or 6MWT.However, it must be noted that the main improvement was observed in STS, probably because most patients already had a normal 6MWT at 5.7 months.In any event, the main mechanisms of impaired exercise capacity after COVID-19 typically do not reside in the lungs, but rather in the circulation (Baratto et al., 2021), and are likely to involve a combination of deconditioning and disease-specific mechanisms (Jahn et al., 2022;Rinaldo et al., 2021).Accordingly, it has recently been documented that although targeted rehabilitation with high-intensity interval training might reduce symptom severity and improve physical performance in many cases after COVID-19, this is not associated with any changes in D L,CO,10s , but rather an increase in left ventricular mass (Rasmussen et al., 2023).In any event, relatively few patients were included in the regression analysis reported here, and there is therefore a high risk of type II error (Berg et al., 2024), thus precluding any firm conclusions regarding changes in D L,NO or D L,CO,5s and physical performance after COVID-19.
The study has several limitations impacting its generalizability.
Initially, although an invitation was extended to all patients discharged from University Hospital Copenhagen-Rigshospitalet, the analysis excluded certain patient demographics, notably those with dementia and residents of elderly care facilities.These groups are typically more susceptible to severe COVID-19 and, potentially, more pronounced long-term effects.In contrast, individuals experiencing symptoms of COVID-19 might have a higher likelihood of participation.
Additionally, a considerable number of patients opted out of the study.Notably, among patients who did not require hospitalization, there was a disproportionately high number of health-care workers.
Methodologically, there are also several issues to consider.Firstly, systematic errors are introduced to D M and V C , because their measurement and derivation involves several assumptions and empirical constants (Borland & Hughes, 2020).For example, the prevailing scientific consensus acknowledges the diffusivity ratio, α, as 1.97, representing the ratio of physical solubilities of NO and CO in tissue (Wilhelm et al., 1977).Several studies have challenged this value, with some proposing higher α values to reconcile discrepancies between different measurement methods.However, these propositions are predominantly dismissed because they deviate from the physical diffusivity ratio, leading to inconsistent α values (Zavorsky et al., 2017).Furthermore, θ NO is assumed to have a finite value, but was historically presumed infinite owing to its rapid reaction rate with free haemoglobin.However, comprehensive debates and recent studies have contested this assumption, establishing θ NO as finite, with 1.51 mL blood min −1 kPa −1 (mmol NO) −1 providing the best current estimate, because it aligns well with theoretical predictions and with extensive in vitro and in vivo experimentation (Zavorsky et al., 2017).Likewise, the equations for θ CO are based on empirical constants obtained at pH 7.4, rejecting other proposed values based on less accurate and non-physiological pH measurements (Forster, 1987).
However, for physiological interpretation, it must be kept in mind that for the calculation of θ CO , pulmonary capillary oxygen tension is assumed to be 13.33 kPa, although alveolar oxygen tension might cause it to fluctuate through between-day and between-manoeuvre variations in ambient pressure and alveolar ventilation.Nevertheless, none of the relationships embedded in these empirical constants and derivations is markedly affected by changes in oxygen tension.Until more substantial evidence is presented, it has been recommended that these empirical constants and derivations for α, θ NO and θ CO are retained (Zavorsky et al., 2017).Secondly, when it comes to assessing alveolar-capillary reserve by placing patients in the supine posture, this approach has recently been shown to exhibit high test-retest reliability.However, because the underlying changes in cardiac output are modest, it might provide a more representative measure of the different D L,CO,NO metrics during submaximal exercise (Madsen et al., 2023), which was not attempted in the present study.
In conclusion, among the ∼60% of COVID-19 patients who exhibited at least one abnormal D L,CO,NO metric at ∼6 months follow-up, there was no consistent pattern in terms of underlying contributions from D M or V C .Furthermore, although D L,NO %pred and D L,CO,5s %pred did improve at ∼12.5 months, the changes were miniscule and neither physiologically nor clinically significant, and they appeared to be unrelated to any changes in physical performance.
It remains indeterminate whether post-COVID changes (or lack thereof) in D L,CO,NO are mechanistically linked to changes in physical performance.
consecutive COVID-19 patients admitted to the Department of Infectious Diseases at University Hospital Copenhagen-Rigshospitalet from March 2020 to March 2021 were invited to participate.At discharge and/or during the first post-discharge telephone consultation, patients were informed about the study.The telephone consultation was provided as a standard procedure to all

D
L,NO %pred and D L,CO,5s %pred, of which the former did not show any distinct pattern with disease severity, while the latter was mostly observed in groups 4 and 5.However, no distinct change in either D M or V C was observed, either in the overall group of 87 patients or in any of the disease severity groups.TA B L E 1Participant demographics and clinical characteristics.
distributed variables based on visual inspection of histograms and probability plots.Differences between severity groups were assessed using Fisher's exact test for sex, one-way ANOVA for normally distributed data and the Kruskal-Wallis test for non-normally distributed data.An abnormal diffusion score was defined as a z- Participant characteristics, CAT score, Charlson comorbidity index score, physical performance, lung function and D L,CO,NO metrics (including the number of abnormal scores and change from abnormal to normal) were summarized as percentage (n), mean (SD) for normally distributed variables or median [interquartile range] for non-normally FVC, ratio of forced expiratory volume in the first second to forced vital capacity; FVC 1 %pred, forced vital capacity as a percentage of that predicted according to age, height and sex; RV%pred, residual volume as a percentage of that predicted according to height and sex; TLC%pred, total lung capacity as a percentage of that predicted according to height and sex.Differences between severity groups were assessed using Fisher's exact test for sex, one-way ANOVA for normally distributed data and the Kruskal-Wallis test for non-normally distributed data.Number of abnormal pulmonary diffusing capacity scores.Pulmonary diffusing capacity and physical performance at 5.7 and 12.5 months follow-up.Changes are the percentage of predicted (%pred) according to age, height and sex and presented as mean [95% confidence interval (CI)].Group 1, asymptomatic COVID-19; group 2, mild COVID-19; group 3, moderate COVID-19; group 4, severe COVID-19; group 5, critical COVID-19.Abbreviations: D Note: Assessments were made at 5.7 months follow-up after COVID-19.Values are presented as percentage (n), mean (SD) for normally distributed data or median [interquartile range] for non-normally distributed data.Values of CCI higher than three were recorded as four for calculation of the median.Group 1, asymptomatic COVID-19; group 2, mild COVID-19; group 3, moderate COVID-19; group 4, severe COVID-19; group 5, critical COVID-19.Abbreviations: CAT score, chronic obstructive pulmonary disease assessment test; CCI, Charlson comorbidity index; D Note: Values are presented as % (n).Group 1, asymptomatic COVID-19; group 2, mild COVID-19; group 3, moderate COVID-19; group 4, severe COVID-19; group 5, critical COVID-19.Abbreviations: D L,CO,5s , pulmonary diffusing capacity for carbon monoxide (during a 5 s breath-hold); D L,NO , pulmonary diffusing capacity for nitric oxide; D M , alveolar-capillary membrane diffusing capacity; V C , pulmonary capillary blood volume.TA B L E 3 M , alveolar-capillary membrane diffusing capacity; STS, 30 s sit-to-stand test; V C , pulmonary capillary blood volume; 6MWT, 6 min walk test.