Severity-Stratified Pulmonary Rehabilitation Modulates Diaphragm Function and Oxidative Stress in Hospitalized AECOPD Patients: A Randomized Controlled Trial

Introduction

Chronic obstructive pulmonary disease (COPD) poses a substantial global burden, driven by factors such as smoking and aging, and is a leading cause of mortality worldwide.^1–4 Beyond airflow limitation, extrapulmonary manifestations like respiratory muscle dysfunction—particularly of the diaphragm are central to its pathophysiology.⁵ In COPD, the diaphragm undergoes maladaptive remodeling due to chronic overload, hypoxia, inflammation, and oxidative stress, leading to contractile impairment and increased fatigability.⁶ This dysfunction worsens acutely during exacerbations (AECOPD), exacerbating dyspnea, reducing exercise tolerance, and elevating the risk of respiratory failure.⁷

Oxidative stress is a key driver of diaphragmatic atrophy and functional decline.⁸ During AECOPD, systemic inflammation further amplifies oxidative damage, which impairs contractility and promotes muscle loss via pathways such as protein oxidation and proteasome activation.^9,10 From a clinical perspective, mitigating oxidative stress during AECOPD is potentially important because enhanced antioxidant capacity may protect diaphragmatic myocytes from damage, thereby preserving muscle contractility and potentially improving patient outcomes such as exercise tolerance and recovery from respiratory failure.¹¹ However, direct evidence linking changes in oxidative stress to functional recovery remains limited, and this study aims to explore that relationship.Although pulmonary rehabilitation (PR) improves outcomes in stable COPD,^12,13 its application during acute hospitalization remains challenging due to clinical heterogeneity and the lack of severity-stratified, individualized protocols. Evidence is notably scarce on how early rehabilitation differentially affects diaphragmatic function and oxidative stress across severity grades, and on the interrelationship between these physiological parameters.^14,15 Diaphragmatic ultrasound parameters, including excursion and thickness, were selected as the primary outcomes for this study because they provide a direct, non‑invasive, and real‑time assessment of diaphragmatic contractility and structure, which are critically compromised during AECOPD and are specifically targeted by exercise rehabilitation.¹⁶

To address this gap, the severity‑graded exercise rehabilitation program used in this study was developed based on our previous work. Briefly, the Triangle model was adopted as the theoretical framework to stratify patients by disease severity,¹⁷ and a comprehensive literature synthesis was conducted to summarize available evidence on exercise rehabilitation in AECOPD.¹⁸ The protocol was then refined through a two‑round Delphi expert consultation,¹⁹ resulting in a scientifically robust and clinically applicable graded rehabilitation framework.

Therefore, this study aimed to implement and evaluate a severity-graded exercise rehabilitation program in hospitalized AECOPD patients. We sought to assess: 1) its overall effects on diaphragmatic function and oxidative stress biomarkers; 2) its differential benefits across disease severity strata (Grades I–III); and 3) the correlation between improvements in oxidative stress and diaphragmatic function, thereby exploring potential mechanisms of rehabilitation efficacy. We hypothesized that a graded rehabilitation approach would safely improve physiological outcomes, with efficacy moderated by baseline severity and mediated in part by attenuated oxidative stress.

Methods

Study Design

This study was a single-centre, randomised controlled trial(RCT). Following baseline assessment, all eligible participants were randomly assigned by a research assistant, using a sealed envelope method, to either the study group or the control group. Group allocation was concealed from the investigators. Data were collected on the first day of the intervention and on the day of hospital discharge. Participant recruitment details are presentedin Figure 1.

Figure 1 Consort diagram of the study.

Randomisation and Blinding

The random allocation sequence was generated by computer in a 1:1 ratio. Eligible patients were first classified into one of three severity grades (I, II, and III) based on the predefined clinical and physiological criteria shown in Table 1. Then, within each severity grade, patients were randomly assigned (1:1) to either the study group or the control group. An independent researcher, who was not involved in participant recruitment, implemented allocation concealment using sealed, opaque envelopes. The study employed a single-blind design, whereby the sonographers and personnel responsible for data collection and analysis remained blinded to group assignment throughout the trial.

Table 1 Severity-Graded Exercise Rehabilitation Programme for AECOPD Patients

Participants

A total of 150 potential subjects were assessed for eligibility. Of these, 18 were excluded (13 did not meet the inclusion criteria and 5 declined to participate). Consequently, 132 hospitalised patients with AECOPD were recruited from the Department of Respiratory and Critical Care Medicine of a tertiary hospital in Zunyi and were randomly assigned to two groups: the study group (n=66) and the control group (n=66). All participants provided written informed consent. The sample size was calculated a priori using G*Power software. Assuming a medium effect size (Cohen’s d = 0.5), α=0.05, and power=0.80, a minimum of 64 patients per group was required. Therefore, we enrolled 66 patients per group (132 in total).

Inclusion and Exclusion Criteria

The inclusion criteria were as follows: 1) diagnosis of COPD according to theGlobal Initiative for Chronic Obstructive Lung Disease (GOLD) 2024 criteria, with the patient experiencing an acute exacerbation; 2) meeting at least one of the diaphragmatic dysfunction criteria defined by the 2022 EXODUS (EXpert consensus On Diaphragm Ultra Sonography in the critically ill): diaphragmatic excursion < 2 cm, diaphragmatic thickness < 2.2 mm, or diaphragmatic thickening fraction<20%; 3) age between 40 and 80 years; 4) clear consciousness, unimpaired verbal communication ability, and willingness to cooperate with the study procedures.

The exclusion criteria were as follows: 1) presence of any contraindication to exercise, including uncontrolled hypertension (>180/110 mmHg) or hypotension (<90/60 mmHg), severe cardiovascular disease, concurrent acute asthma attack, pulmonary embolism, pneumothorax, or any neuromuscular or osteoarticular disorder affecting exercise capacity; 2) chest deformity, known diaphragmatic paralysis, or chronic neuromuscular disease.

Establishment of the Research Team

A multidisciplinary research team was established to implement the study protocol. The core composition and responsibilities of the team are summarized in Figure 2.

Figure 2 Composition and responsibilities of the multidisciplinary research team.

Study Group Allocation

As most patients with AECOPD experience rapid and significant disease progression, pulmonary function tests could not be completed in the majority of cases and were therefore not used as a criterion for severity stratification. Consequently, the 132 eligible AECOPD patients were classified into three severity grades (I, II, and III) according to the predefined stratification criteria (see Table 1 for eligibility criteria). Patients within each severity grade were then randomly allocated to either the study group or the control group.

Intervention Protocol

The intervention was initiated within 24 hours of hospital admission and continued until the day of discharge, without interruption, and was administered twice daily.The hospitalization-based intervention period was chosen because the acute phase of AECOPD is characterized by rapid physiological changes, and previous studies have demonstrated that early, short-term rehabilitation can induce measurable improvements in diaphragmatic function and oxidative stress biomarkers within this timeframe.^18,20

Standardization, Adherence, and Safety

To ensure standardization, all treating physiotherapists received a detailed manual outlining the procedures for each severity grade and attended training session before the study. Adherence was monitored by a daily log completed by the treating therapist after each session, documenting the exercises performed andtheir duration.

During each exercise session, a trained physiotherapist was present to directly supervise the patient, monitor vital signs and symptoms, and ensure adherence to the safety criteria. Overall safety was overseen by the attending physician. Predefined safety pause criteria included: SpO₂ <85% for >3 minutes; heart rate exceeding the target range for >3 minutes; respiratory rate <5 or >30 breaths/min; systolic blood pressure >180 mmHg or <90 mmHg; mean arterial pressure <65 mmHg; a drop or rise of ≥20% from baseline blood pressure lasting >3 minutes; new need for or increased dose of vasoactive drugs; arrhythmia, acute heart failure, or suspected myocardial infarction; severe dyspnea, chest pain, dizziness, profuse sweating, pallor, intolerance to activity, or patient refusal. Emergency response protocols were established for severe dyspnea, acute myocardial infarction, soft tissue injury, muscle spasm, and falls. If any safety criterion was met, the exercise session was interrupted immediately, and the patient was assessed by the attending physician before considering resumption.

Control Group

In addition to standard medical treatment, the control group received conventional exercise rehabilitation, which comprised the following:

1. Airway Management: Position turning, directed effective coughing.
2. Chest Physiotherapy: Manual percussion and vibration, mechanical vibration.
3. Breathing Exercises: Pursed-lip breathing, diaphragmatic breathing.
4. Limb Training: Upper limb exercises with light dumbbells and seated lower limb exercises.

All sessions were administered twice daily.

Study Group

In addition to standard medical treatment, the study group received theseverity-graded exercise rehabilitation programme as detailed in Table 1.

Outcome Measures and Assessment

Timing of Assessment

Measurements and assessments were performed for all participants in both groups within 24 hours of enrolment and on the day of hospital discharge.

Outcome Measures

Diaphragmatic Function

Diaphragmatic function was assessed on the right hemidiaphragm using a portable colour Doppler ultrasonography system (Mindray UMT-200). DE was measured with a low-frequency convex array probe, while DTei and DTee were measured with a high-frequency linear array probe. All measurements were obtainedby the same sonographer using an identical measurement protocol to ensure consistency. Representative still images illustrating the ultrasonographic measurement of DE, DTei, and DTee are presented in Figures 3–5.

Figure 3 Ultrasonographic measurement of DE. The liver is used as an acoustic window. The M‑mode sampling line is directed toward the diaphragmatic dome. The white undulating wave line indicates the amplitude of diaphragmatic movement.

Figure 4 Ultrasonographic measurement of DTei. The diaphragm is visualized as a hypoechoic layer between two hyperechoic lines: the pleural line (superior) and the peritoneal line (inferior). Thickness is measured as the distance between these two lines at the end of inspiration.

Figure 5 Ultrasonographic measurement of DTee. The same anatomical landmarks as in Figure 4 are used (pleural line and peritoneal line as hyperechoic boundaries). Thickness is measured at the end of expiration, when the diaphragm is at its most relaxed position. Table 1. Severity-Graded Exercise Rehabilitation Programme for AECOPD Patients.

Oxidative Stress Biomarkers

A 5 mL blood sample was collected from the median cubital vein. MDA levels, indicating the degree of cellular damage under oxidative stress, SOD activity,reflecting the body’s capacity to scavenge free radicals and a strong independent predictor of one-year all-cause mortality in AECOPD patients, and TAOC, representing the plasma’s overall antioxidant capability, were measured. All biomarkers were analysed using commercially available ELISA kits purchased from Shanghai Jingkang Bioengineering Co., Ltd.These biomarkers are widely recognized indicators of systemic oxidative stress, and the ELISA method is a standard, validated approach for their quantification. All assays were performed by trained laboratory technicians who were blinded to group allocation, following the manufacturer’s protocols. Standard curves were generated for each plate, and samples were measured in duplicate. Internal quality controls were included in each run to ensure reliability.

Data and Statistical Analysis

Data were analyzed using IBM SPSS 29.0. Continuous and categorical variables are presented as mean±SD and frequencies (percentages), respectively. Baseline characteristics were compared using independent t-tests and Chi-square tests. The primary intervention effects on diaphragm function and oxidative stress biomarkers were assessed with repeated-measures ANOVA (evaluating time, group, and interaction effects). Differences in improvements across severity subgroups (Grades I–III) were compared via one-way ANOVA with LSD post-hoc tests. Additionally, partial correlation analyses examined relationships between changes in diaphragm function and oxidative stress, controlling for severity, age, sex, and disease duration. Statistical significance was set at a two-tailed p<0.05.

Before conducting repeated‑measures ANOVA and partial correlation analyses, the underlying assumptions (normality, sphericity, homogeneity of variances, and linearity) were formally assessed using Shapiro‑Wilk, Mauchly’s, and Levene’s tests, respectively. Violations of sphericity were adjusted using the Greenhouse‑Geisser correction.

Results

Comparison of Baseline Characteristics

A total of 132 patients were enrolled, with 66 allocated to the study group and 66 to the control group. No significant differences were observed in baseline characteristics between the two groups, including age, sex, body mass index (BMI), disease duration, educational level, comorbidities, and history of prior rehabilitation training (all p > 0.05), confirming their comparability.See Table 2 for details.

Table 2 Comparison of Baseline Characteristics Between the Two Groups

Changes and Effects on Diaphragmatic Function

Statistical analysis revealed that the graded exercise rehabilitation significantly improved diaphragmatic function in COPD patients (see Table 3). Repeated-measures analysis of variance confirmed significant time-by-group interaction effects for all three diaphragmatic parameters-DE, DTei, and DTee (all p<0.001), with medium to large effect sizes. Post-intervention measures were significantly improved compared to baseline in the study group (all p<0.001), whereas no significant changes were observed in the control group. Furthermore, the study group demonstrated significantly better outcomes than the control group across all measures at post-intervention (all p<0.001). These findings indicate that the improvements in diaphragmatic function are attributable to the graded exercise rehabilitation intervention rather than natural recovery.

Table 3 Changes in Diaphragmatic Function Before and After Intervention (n=66/Group, )

Changes and Effects in Oxidative Stress Biomarkers

Repeated-measures analysis of variance was used to evaluate the impact of the graded exercise rehabilitation on oxidative stress markersAs presented in Table 4, significant time-by-group interaction effects were observed for both SOD (p=0.001) and TAOC (p=0.024), but not for MDA (p=0.965).Simple effects analysis indicated that, compared to the control group, the study group demonstrated significant post-intervention increases in SOD activity and TAOC levels (all p<0.05). However, no significant changes were observed in MDA levels, a marker of lipid peroxidation, in either group. These results suggest that the rehabilitation programme effectively enhanced the body’s antioxidant defence capacity, but its effect on mitigating oxidative damage was not significant.

Table 4 Changes in Oxidative Stress Biomarkers Before and After Intervention (n=66/Group, )

Heterogeneity in Improvements of Diaphragmatic Function and Oxidative Stress Biomarkers Across Severity Subgroups Within the Study Group

To further evaluate patient heterogeneity in rehabilitation response, we analysed the differential benefits among patients stratified by disease severity. The results demonstrated that the intervention effects varied depending on the specific outcome and baseline severity: Regarding diaphragmatic function, the improvement in DE was significantly greater in patients with mild (Grade I) disease (0.295±0.218) compared to those with moderate (Grade II) disease (p=0.018), suggesting they derived the greatest benefit in diaphragmatic contractile function. In contrast, for oxidative stress, although a significant difference in MDA change was observed across severity groups (p=0.001), thetrendwas severity-specific-MDA levels decreased in patients with mild and moderatedisease (improvement values: −1.506 and −0.421, respectively), whereas they increased in patients with severe (Grade III) disease (0.853±1.298), a change significantly worse than in the mild and moderate groups (all p< 0.05). This indicates that while the graded protocol ameliorated oxidative damage in patientswith mild-to-moderate disease, it failed to reverse the pre-existing oxidative stress burden in severe patients, demonstrating that severity-graded rehabilitation achieves differential physiological effects.Detailed results are presented in Table 5.

Table 5 Comparison of Improvements in Diaphragmatic Function and Oxidative Stress Biomarkers by Disease Severity

Specific Effects of Graded Exercise Rehabilitation on the Diaphragm-Oxidative Stress Relationship

Partial correlation analysis revealed a significant specific effect of the graded exercise rehabilitation (see Table 6). Only in the study group were improvements in DE significantly correlated with improvements in all oxidative stress biomarkers (MDA, SOD, TAOC; r = −0.314 to −0.331, all p< 0.05). In contrast, none of thesecorrelations were statistically significant in the control group. This indicates that the rehabilitation intervention did not uniformly improve all aspects of diaphragmatic function, but rather specifically and significantly enhanced the negativeassociation between diaphragmatic excursion (DE) and systemic oxidative stressstatus. This implies that, post-intervention, a greater amelioration of oxidative stress was associated with better recovery of diaphragmatic mobility. Furthermore,no significant associations were observed between the diaphragmatic thickness measures (DTei, DTee) and oxidative stress biomarkers in either group, suggesting that the rehabilitation intervention likely exerts its effects primarily by improving diaphragmatic contractile function and efficiency, rather than by altering static structural thickness.

Table 6 Specific Effects of Graded Exercise Rehabilitation on the Relationship Between Diaphragmatic Function and Oxidative Stress

Interrelationship Analysis of Oxidative Stress Biomarkers

Internal correlation analysis revealed that the improvement values of MDA, SOD, and TAOC exhibited significant positive correlations in both the study and control groups (r = 0.332–0.486, p< 0.01). This indicates an intrinsic coordinated relationship between the oxidative damage and antioxidant defense systems, irrespective of rehabilitation intervention. This consistent pattern across groups reflects an inherent physiological linkage within the oxidative stress system, independent of the intervention, thereby providing a stable biological reference context for the study.Detailed results are presented in Table 7.

Table 7 Specificity of Interrelationships Among Oxidative Stress Biomarkers

Discussion

Graded Exercise Rehabilitation Significantly Improves Diaphragmatic Function in AECOPD Patients

Our study robustly demonstrates that a severity-graded rehabilitation program during hospitalization yields significant, clinically meaningful improvements in diaphragmatic function among AECOPD patients. Compared to usual care, the intervention group exhibited greater increases in diaphragmatic excursion (DE: 1.556±0.260 cm to 1.800±0.303 cm), end-inspiratory thickness (DTei), and end-expiratory thickness (DTee), with medium to large effect sizes and significant time-by-group interactions (all p<0.001). The most pronounced improvement was observed in DE, a dynamic marker of contractile efficiency. This finding confirms that early, personalized exercise is safe and effective even in the acute phase, extending benefits previously observed in stable COPD to a more vulnerable inpatient population.^21–23 The rapid gain in DE, as opposed to thickness parameters which may reflect slower structural adaptation, likely stems from enhanced neuromuscular drive, fatigue resistance, and local perfusion, which are crucial for alleviating dyspnea and maintaining respiratory stability.

Differential Modulation of the Oxidative Stress Response

Our results reveal a dual effect on oxidative stress. The intervention significantly enhanced the endogenous antioxidant defense system, evidenced by a greater increase in serum superoxide dismutase (SOD: p=0.005 vs. control) and total antioxidant capacity (TAOC: p=0.025). This aligns with the exercise hormesis concept, whereby moderate physiological stress upregulates cytoprotective pathways.^24,25 Conversely, no significant between-group difference was found in malondialdehyde (MDA), a marker of lipid peroxidation (p=0.965). This indicates that while short-term rehabilitation effectively bolsters antioxidant capacity, it may be insufficient to rapidly clear pre-existing oxidative damage during an acute exacerbation. The primary therapeutic effect on oxidative stress thus appears to be the augmentation of defense mechanisms rather than the reversal of accumulated damage.

Heterogeneity in Intervention Effects: The Role of Disease Severity

Subgroup analysis underscores that intervention benefits are severity-dependent. Patients with mild (Grade I) disease derived the greatest functional benefit, showing a superior improvement in DE (0.295±0.218 cm) compared to those with moderate disease (0.171±0.150 cm; p=0.018). At the molecular level, a concerning divergence was observed: MDA levels decreased in mild and moderate patients (improvement values: −1.506 and −0.421) but increased in severe (Grade III) patients (0.853±1.298; p<0.05 vs. mild). This indicates that while the graded protocol alleviates oxidative damage and improves function in mild-to-moderate disease, patients with severe AECOPD may have an overwhelming oxidative stress burden that is not mitigated by short-term exercise alone. This dissociation validates the necessity of severity-stratified rehabilitation protocols.

Correlation Between Diaphragmatic Function and Oxidative Stress

A pivotal finding is the specific correlation, observed only in the intervention group, between improved diaphragmatic function and attenuated oxidative stress. Partial correlation analysis revealed that the increase in DE was significantly associated with favorable changes in all oxidative biomarkers (MDA: r = −0.314, p=0.013; SOD: r = −0.331, p=0.009; TAOC: r = −0.304, p=0.016). No such correlations were found in the control group or for diaphragmatic thickness indices. This suggests that exercise-induced enhancement of antioxidant defenses may protect diaphragmatic myocytes, potentially improving calcium handling and mitochondrial function,^8,26 thereby translating into better contractile performance. This evidence chain links molecular adaptation to organ-level recovery, deepening our understanding of pulmonary rehabilitation. However, it is important to note that these findings are correlational and do not prove causality. The proposed mechanistic interpretation remains hypothesis-generating and requires further experimental validation.

Clinical and Mechanistic Implications

Clinical relevance of short‑term improvements – Although the intervention period was limited to the hospital stay, the observed improvements in diaphragmatic excursion and antioxidant markers are clinically relevant. Enhanced diaphragmatic excursion is a direct indicator of improved contractility and has been associated with reduced dyspnea and better exercise capacity in AECOPD patients. Furthermore, serum SOD is a strong independent predictor of one‑year all‑cause mortality in AECOPD, and increased TAOC has been linked to better survival. Therefore, the short‑term improvements observed in our study may translate into meaningful clinical benefits, including better symptom control and potentially improved prognosis, though longer‑term follow‑up studies are needed to confirm this.

Robustness of the correlation and influence of baseline heterogeneity – The intervention‑specific partial correlations between improvements in diaphragmatic excursion and oxidative stress biomarkers remained significant after controlling for disease severity, age, sex, and disease duration, supporting their robustness. However, the magnitude of these correlations varied across severity subgroups, likely due to baseline heterogeneity. Patients with mild (Grade I) disease showed the greatest improvement in DE and a reduction in MDA, whereas patients with severe (Grade III) disease did not exhibit MDA reduction despite some improvement in DE. This suggests that the association between enhanced diaphragmatic mobility and attenuated oxidative stress is most evident in mild‑to‑moderate AECOPD, where the endogenous antioxidant system retains some capacity to respond to exercise. In severe patients, the overwhelming oxidative burden may overwhelm the protective effects of short‑term rehabilitation, highlighting the need for adjunctive therapies in this population.

Clinical guidance for severity‑stratified rehabilitation – Our findings offer concrete guidance for clinical decision‑making based on disease severity. For patients with mild (Grade I) AECOPD, early initiation of progressive resistance and aerobic exercise is recommended. For moderate (Grade II) AECOPD, moderate‑intensity exercise focusing on respiratory muscle training and resistance exercises should be prioritized. For severe (Grade III) AECOPD, exercise should be started at low intensity with close safety monitoring, and adjunctive strategies such as nutritional support or antioxidant supplementation should be considered.

Limitations

This study has several limitations: (1) Single-center design may affect generalizability. (2) Lack of post-discharge follow-up limits assessment of long-term outcomes. (3) While antioxidant capacity improved, no significant reduction in oxidative damage (MDA) was observed. (4) The study did not investigate downstream cellular mechanisms linking oxidative stress to diaphragmatic dysfunction. (5) Due to the inherent differences in exercise intensity and content between the study group and the control group, it was not possible to blind the treating physiotherapists or the patients to group allocation, which may introduce performance bias. However, outcome assessors (sonographers, laboratory technicians, and statisticians) were kept blinded to group assignment throughout the trial to minimize detection bias, and the primary outcomes were objective measurements.Future research should focus on multi-center validation, long-term follow-up, optimized intervention strategies (eg., combining exercise with antioxidants), and mechanistic studies to elucidate underlying pathways.

Conclusion

This study demonstrates that a severity-graded exercise rehabilitation program is a safe and effective strategy for improving diaphragmatic function and endogenous antioxidant capacity in hospitalized patients with AECOPD, with treatment responses varying according to baseline disease severity. These findings provide a practical, evidence-based framework for implementing early, individualized pulmonary rehabilitation in the inpatient setting, supporting a severity‑stratified approach to clinical decision‑making. It is important to note that these conclusions are based on short‑term, in‑hospital physiological outcomes and do not directly address long‑term prognosis. Future multicenter studies with extended follow‑up are warranted to confirm these physiological benefits and evaluate their impact on long‑term clinical outcomes, as well as to optimize intervention strategies for patients with severe AECOPD, in whom exercise alone may be insufficient.