From December 2019 and December 2020, a total of 53 patients with COPD (≥40 years old) and 23 age-matched healthy controls from the First Hospital of China Medical University (Shenyang, China), the Fourth Hospital of China Medical University (Shenyang, China), and the First Hospital of Dalian Medical University (Dalian, China) were recruited for this prospective observational study. The inclusion criterion was a diagnosis of stable COPD according to the Global Initiative for Chronic Obstructive Lung Disease (GOLD) criteria. The exclusion criteria were as follows: COPD exacerbation within the last 1 month; presence of severe cardiovascular disease or active lung disease; concomitant disease affecting the musculoskeletal system; long-term systemic steroid therapy; and inability to read or understand the informed consent documents. Clinical features, including age, sex, height, and weight, were obtained from medical records. The study was approved by the research ethics committees of the First Hospital of China Medical University (No. 2019-144-2), and written informed consent was obtained from all patients.

Spirometry measurements (pre-bronchodilator) were performed following the American Thoracic Society and the European Respiratory Society guidelines using a Jaeger MasterScreen system (Jaeger, Viasys Healthcare GmbH, Hoechberg, Germany). Dyspnea symptoms were measured using the Chinese version of the modified British medical research council (mMRC) dyspnea scale (Bestall et al., 1999; Cui et al., 2017), and health status was measured using the Chinese version of the COPD assessment test (CAT; Jones et al., 2009; Zhou et al., 2013).

Quadriceps muscle strength (QMS) was measured via a dynamometer (type: microFET2™; Hoggan, Salt Lake City, UT, United States) following the instructions in the manufacturer’s manual and previous studies (O’Shea et al., 2007; Zhang et al., 2018). The patients’ knee was flexed to 90°, and the dynamometer plate was placed. The anterior end of the dynamometer was located 5 cm proximal to the lateral malleolus on the anterior surface of the leg and perpendicular to the long axis of the tibia. The participant was then instructed to generate a maximal knee extension force to hold the line in the same position for a duration of 4 s by pushing against the dynamometer plate to which the investigator applied increasing force with encouragement. The same steps were repeated twice, with an intervening interval of 30–60 s. The average value of the last two assessments for each lower limb was recorded as the maximum unilateral contraction force. Then, the average contraction force on both sides was used to obtain the QMS.

The participant’s body fat rate (BFR) was measured by a bioelectrical impedance meter (InBody770; InBody, Seoul, Korea). The fat-free mass index (FFMI) was calculated as follows: FFMI = weight (kg) × (1-BFR)/height (m)².

The participant was seated on a chair that measured 48 cm high and had no armrests, with their feet on the ground, back supported by the back of the chair, and hands folded in front of their chest. After hearing the test start command, the participant was asked to stand up and sit down five times as quickly as possible; the time needed to complete the five repetitions was recorded. The participant was given verbal encouragement during the test. The test was performed three times, with 1-min rest intervals. The average of three tests was recorded as the result.

According to the 2002 American Thoracic Society (ATS) guidelines (ATS Committee on Proficiency Standards for Clinical Pulmonary Function Laboratories, 2002), a closed, long, and straight 30-m indoor corridor was selected. The test method was explained to the patient before the test, and the patient was told to walk as much as possible. If they felt short of breath or experienced chest pain or dizziness, they were allowed to slow down or to stop to rest. If the above symptoms worsened and were not relieved after rest, the test was stopped immediately, and the patient was supervised by the experimenter and encouraged using standardized language. After 6 min, the patient heard the experimenter say “time is up,” which was their indication to stop. The test personnel recorded the distance travelled in meters.

Measurement of the rectus femoris thickness and cross-sectional area was performed as in previous studies (Shrikrishna et al., 2012; Maynard-Paquette et al., 2020). Grayscale ultrasound was performed with an Aixplorer ultrasound scanning system (SuperSonic Imagine, Aix-en-Provence, France). Our ultrasonographers received formal training, and with >10 years of experience. The patient did not engage in strenuous exercise for 72 h, rested quietly for 15 min and then laid on their back on the operating bed, relaxing all their muscles. The researchers set up a bracket to fix the ultrasound probe in place, thereby reducing muscle deformation due to external forces, and placed the ultrasound probe perpendicular to the patient’s dominant leg. The scanning depth was set such that the femur could be detected for orientation. Gentle contraction-relaxation manoeuvres were employed to delineate the muscle septa prior to image acquisition. RF_thick and RF_csa were calculated after the inner echogenic line of the rectus femoris was outlined by a movable cursor on a frozen image. RF_thick and RF_csa were recorded as the averages of three consecutive measurements within 10%.

Shear wave elastography was performed using an Aixplorer ultrasound system (Supersonic Imagine, Aix-en-Provence, France) with ShearWave™ Elastography coupled with a 4–15 MHz linear transducer, as previously described in detail (Bercoff et al., 2004; Tanter et al., 2008; Andonian et al., 2016). Grayscale ultrasound was used to identify anatomical structures and to determine the location of the rectus femoris. Next, the user was positioned such that a fixed-size square region of interest (ROI) delimiting the elastographic field of view (SWE box), i.e., a ROI, where shear-wave propagation was analyzed within the muscle, was visible. Three successive SWE acquisitions were performed for the rectus femoris with the transducer in a fixed position (Figure 1B). Images were transferred to an OsiriX workstation (Pixmeo, Geneva, Switzerland) for analysis using a dedicated analysis plugin (QBox, 1.0, Supersonic Imaging). To avoid artifacts in the circular ROIs, five ROIs (5-mm diameter) were placed within a given square SWE color map (Figure 1B). As three SWE acquisitions were performed in succession for each rectus femoris, 15 stiffness measurements were available at each measurement time point for each subject. Same two observers making all SWE measurements in different site.

The definition of sarcopenia followed the guidelines from the AWGS (Chen et al., 2020): low muscle mass [bioelectrical impedance (M: <7.0 kg/m², F: <5.7 kg/m²)] and low muscle strength [handgrip strength (M: <28 kg, F: <18 kg)] and/or poor physical performance (five-time chair stand test: ≥12 s). Muscle mass was evaluated using bioelectrical impedance analysis (BIA; InBody770; InBody, Seoul, Korea). Muscle strength was assessed by handgrip strength using a JAMAR®Plus⁺ hand dynamometer. The five-time chair stand test was used to assess physical performance.

For analysis of circulating biomarkers, plasma sample from patients with COPD were assayed using ELISA kits. TNF-α were assessed by high-sensitivity ELISA kits (#HSTA00E, R&D, Minnesota, United States) following manufacturer’s instructions. And, the GDF-15 and resistin level was determined via human GDF-15 quantizing ELISA Kit (#DGD150, R&D, Minnesota, United States) and human resistin quantizing ELISA Kit (#DRSN00, R&D, Minnesota, United States) according to the manufacturer’s instructions.

Statistical analyses were performed using SPSS 13.0 software (IBM, Armonk, NY, United States). Interobserver and intraobserver reliability were calculated using a two-way random model based on absolute agreement with an intraclass correlation coefficient (ICC) test. The ICCs were interpreted according to the following published guidelines (Shrout and Fleiss, 1979): less than 0.40, poor reliability; 0.40–0.75, fair-to-good reliability; and greater than 0.75, excellent reliability. Continuous variables are expressed as the median values and IQRs, as well as the minimum and maximum values. The differences in the SWE_mean between healthy controls and patients with COPD were determined using Mann-Whitney U test, and values of p < 0.05 were considered statistically significant. Pearson correlation coefficient analysis was used to compare the correlations of the SWE_mean, RF_thick, and RF_csa with the clinical features of COPD patients. Receiver operating characteristic (ROC) curve analysis and the area under the curve (AUC) were used to determine the ability of SWE_mean, RF_thick, and RF_csa to predict sarcopenia.

A total of 53 COPD patients and 23 age-matched healthy controls were enrolled in the final analysis. The baseline characteristics of the patients are listed in Table 1. There were no statistically significant differences in age, BMI, or sex between COPD patients and healthy controls. Pulmonary function (FEV₁, FEV₁%predicted, FVC, and FEV₁/FVC) was significantly decreased in patients with COPD compared with that in healthy controls.

First, we determined the reliability and repeatability levels for measurements by SWE in the rectus femoris muscle. Rectus femoris muscle measurements were performed by two observers and at three different times and are shown in Table 2. The assessment of the rectus femoris using SWE yielded excellent intraobserver repeatability ICCs (observer 1: ICC: 0.952, p < 0.001; observer 2: ICC: 0.985, p < 0.001). Importantly, it was found that the interobserver reliability for the rectus femoris muscle was excellent (ICC: 0.984, p < 0.001). These results suggest that SWE is a reliable and repeatable technique for measuring rectus femoris muscle stiffness.

Next, we aimed to assess the clinical value of SWE in patients with COPD. The mean elasticity index (SWE_mean) of the rectus femoris was measured via SWE, and the process is shown in Figures 1A,B. Representative pictures of rectus femoris SWE in healthy controls and patients with COPD are shown in Figures 2A,B. The SWE_mean of patients with COPD was significantly lower than that of healthy controls (Figure 2C). In addition, the SWE_mean of patients with GOLD stage A and GOLD stage B was significantly higher than that of patients with GOLD stage C and GOLD stage D (Figure 2D). Overall, these results suggest that the rectus femoris stiffness of patients with COPD was decreased compared with that of healthy controls and associated with the severity of the disease.

Next, we explored the relationship between SWE_mean and clinical features in patients with COPD. In Table 3, the 6-min walk distance (6MWD; r = 0.450, p < 0.001), FEV₁ (r = 0.397, p = 0.003), FEV₁%predicted (r = 0.317, p = 0.021), FEV₁/FVC (r = 0.397, p = 0.002), and QMS (r = 0.533, p < 0.001) were positively correlated with the SWE_mean. The five-repetition sit-to-stand (5STS; r = −0.520, p < 0.001), mMRC scores (r = −0.301, p = 0.029), and CAT scores (r = −0.362, p = 0.012) were negatively correlated with the SWE_mean. These results suggest that a lower SWE_mean of patients with COPD was positively associated with reduced pulmonary function, worse physical function, poor exercise tolerance, decreased muscle function, and more severe clinical symptoms.

We compared the relationship between RF_thick, RF_csa, and clinical features in patients with COPD. In Table 3, height (r = 0.344, p = 0.012), weight (r = 0.299, p = 0.029), 6MWD (r = 0.641, p < 0.001), FEV₁ (r = 0.370, p = 0.006), FEV₁/FVC (r = 0.401, p = 0.003), and QMS (r = 0.398, p < 0.001) were positively correlated with RF_thick. The mMRC scores (r = −0.346, p = 0.004) and CAT scores (r = −0.388, p = 0.006) were negatively correlated with RF_thick. Height (r = 0.281, p = 0.041), weight (r = 0.391, p = 0.004), 6MWD (r = 0.615, p < 0.001), FEV₁ (r = 0.287, p = 0.037), FVC (r = 0.329, p = 0.016), FEV₁/FVC (r = 0.253, p = 0.024), and QMS (r = 0.311, p = 0.028) were positively correlated with RF_csa. The mMRC scores (r = −0.325, p = 0.012) and CAT scores (r = −0.151, p = 0.031) were negatively correlated with RF_csa.

Previous studies (Dalle and Koppo, 2020; Looijaard et al., 2021) indicated that circulating biomarkers, including systemic inflammatory markers (Ferrari et al., 2015) and oxidative stress factors (Sepúlveda-Loyola et al., 2021), were associated with muscle quality in COPD. In this study, we analyzed the correlation between SWE_mean and a series of circulating biomarkers (GDF15, resistin, and TNF-α). As shown in Figure 3, SWE_mean was negatively correlated with serum GDF15 levels (r = −0.472, p < 0.001), serum resistin levels (r = −0.291, p = 0.035), and serum TNF-α levels (r = −0.433, p = 0.001).

Sarcopenia is one of the important systemic symptoms in patients with COPD (Sepúlveda-Loyola et al., 2020). Therefore, we explored the potential value of SWE in the diagnosis of sarcopenia in patients with COPD and compared it with grayscale US. Twenty-one (39.62%) patients with COPD had a diagnosis of sarcopenia following the guidelines from the AWGS. As shown in Table 4, pulmonary function (FEV₁, FEV₁%predicted, FVC, and FEV₁/FVC) was significantly decreased in sarcopenic patients with COPD. Patients with sarcopenia showed poor physical function (5STS) and poor exercise tolerance (6MWD). In terms of body composition, FFMI were markedly decreased in sarcopenic patients. The SWE_mean of patients with sarcopenia were both significantly lower than those of patients without sarcopenia (Figure 4A). ROC curve analysis was used to explore the ability of SWE_mean, RF_thick, and RF_csa to predict sarcopenia. As shown in Figure 4, the AUC values for SWE_mean (Figure 4B), RF_thick (Figure 4C), and RF_csa (Figure 4D) were 0.863, 0.802, and 0.816, respectively. These results suggest that the levels of predictive power of SWE_mean (from SWE) were higher than those of RF_thick and RF_csa (from grayscale US).