Short- and Long-Term Responses to Pulmonary Rehabilitation in 922 Patients with COPD: A Real-World Database Study (2002–2019)

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Overall Sentiment: This is a well-conducted retrospective cohort study that addresses a critical gap in pulmonary rehabilitation (PR) research: the identification of "responders" in a real-world setting over both short and long terms. The use of a large, 18-year database (n=922) provides significant statistical power. The findings regarding age, symptom burden, and baseline functional capacity as predictors are clinically relevant and support the move toward personalized PR. However, some methodological clarifications and expanded discussions are needed to strengthen the manuscript.

Specific Comments:

Methods:

• Instrument Transitions: You noted that SGRQ was replaced by CAT in 2012, and ESWT by 6MWT in 2017. While you justified this, please clarify if you performed any sensitivity analysis to ensure that these transitions did not introduce bias into the "responder" classification over the 18-year period.
• Missing Data: A total of 650 patients were excluded due to missing data. This is a substantial number. Please provide a brief comparison of the baseline characteristics (e.g., age, FEV1) between the included cohort and those excluded to address potential selection bias.

Results:

• Figure 1 (Sankey Plot): This is an excellent visualization. However, 23% of patients were "Not assessed" at one year. Please discuss the potential impact of this attrition on the long-term response rates.

Discussion:

• Age and Response: The finding that younger age predicts better response is interesting. Please compare this with existing literature, as some studies suggest older patients can benefit just as much from PR.
• Clinical Implications: You mention the need for an "individualized approach." Based on your data, what specific modifications to PR would you suggest for "Non-responders" (e.g., older patients with lower symptom burden at baseline)?
• Study Limitations: Please explicitly mention the unblinded nature of the data collection and the retrospective design as limitations.

Author Response

Dear Reviewer,

January 14, 2026

We thank the editor for the opportunity to submit a revised version of our manuscript entitled “Short- and Long-Term Response to Pulmonary Rehabilitation in 922 Patients with COPD: A Real-World Database Study (2002–2019)”. We thank the reviewers for thorough and constructive comments that have helped us clarify and improve our manuscript. All changes made in the manuscript are marked in the track changes manuscript. Please find our point-by-point response to specific reviewer comments below (responses in italic bold).

Reviewer 1

Comments and Suggestions for Authors

Overall Sentiment: This is a well-conducted retrospective cohort study that addresses a critical gap in pulmonary rehabilitation (PR) research: the identification of "responders" in a real-world setting over both short and long terms. The use of a large, 18-year database (n=922) provides significant statistical power. The findings regarding age, symptom burden, and baseline functional capacity as predictors are clinically relevant and support the move toward personalized PR. However, some methodological clarifications and expanded discussions are needed to strengthen the manuscript.

Specific Comments:
Methods:

Instrument Transitions: You noted that SGRQ was replaced by CAT in 2012, and ESWT by 6MWT in 2017. While you justified this, please clarify if you performed any sensitivity analysis to ensure that these transitions did not introduce bias into the "responder" classification over the 18-year period.

We thank the reviewer for this important comment. We did not perform formal sensitivity analyses specifically addressing the instrument transitions. However, it is important to clarify that within each patient trajectory, the same instrument was used consistently at baseline, 10-week, and 1-year follow-up. Thus, no individual patient experienced a change in functional test or PROM during follow-up, and MCID-based responder classification was always derived from within-instrument change scores.

Consequently, the transitions from SGRQ to CAT (2012) and from ESWT to 6MWT (2017) do not affect the internal validity of the responder classification at the individual level, but rather reflect changes in routine clinical assessment over time. To further mitigate potential bias, analyses were conducted using instrument-specific models, and we avoided pooling outcomes across different tests or PROMs.
While we acknowledge that instrument transitions may influence the composition of assessed cohorts across calendar periods, this heterogeneity reflects real-world clinical practice and was one of the motivations for using a pragmatic, real-world study design. We have clarified this point in the Methods/Limitations section. (Please read track changes manuscript clarification p. 3-4 line:136-140).

Missing Data: A total of 650 patients were excluded due to missing data. This is a substantial number. Please provide a brief comparison of the baseline characteristics (e.g., age, FEV1) between the included cohort and those excluded to address potential selection bias.

We thank the reviewer for this important comment. We have now compared baseline characteristics between patients included in the analyses and those excluded due to missing data. The groups did not differ statistical significantly with respect to sex, age baseline FEV₁, ESWT, 6MWD CAT, or SGRQ scores. These results have been added to the manuscript to improve transparency and to address potential selection bias.

It now reads: “Patients excluded due to missing data (n = 650) did not differ significantly from the included cohort with respect to sex, age baseline FEV₁, ESWT, CAT, or SGRQ scores (61% female; mean ± SD: age 68.8 ± 10.3 years, FEV₁ 36 ± 12% predicted, ESWT 166 ± 87 s, 6MWD 262 ± 118 m, CAT 19 ± 6 points, SGRQ 55 ± 13 points).” (track changes manuscript clarification p. 4-5 line:182-186).

Results:

Figure 1 (Sankey Plot): This is an excellent visualization. However, 23% of patients were "Not assessed" at one year. Please discuss the potential impact of this attrition on the long-term response rates.

We thank the reviewer for this comment. We acknowledge that loss to follow-up at 1 year may influence estimates of long-term response. Patients classified as “not assessed” were excluded from the 1-year response analyses, and no assumptions were made regarding their outcome status. As such, sustained response rates may be over- or underestimated depending on the clinical trajectories of these patients. We have added a discussion of this limitation to the Discussion section, emphasizing that attrition is inherent to long-term real-world follow-up and that results should be interpreted accordingly. (Please read track changes manuscript clarification p. 14 line:419-421).

Discussion:

Age and Response: The finding that younger age predicts better response is interesting. Please compare this with existing literature, as some studies suggest older patients can benefit just as much from PR.

We thank the reviewer for this important comment and agree that older patients can benefit substantially from pulmonary rehabilitation. In several previous studies, responders were numerically younger than non-responders; however, the observed age differences were small and frequently not statistically significant. Notably, Di Meo et al. (Age and Ageing, 2008) demonstrated that age did not hamper response to pulmonary rehabilitation. Other studies by Spruit et al., Walsh et al., and Souto-Miranda et al., like our study reported modest younger age differences (approximately 1–2 years) that did not reach statistical significance, likely reflecting limited sample sizes and heterogeneous outcomes. We have revised the discussion to place our findings in this context and to emphasize that age should be interpreted as a contextual modifier rather than a standalone determinant of rehabilitation response.
(Please read track changes manuscript clarification p. 13 line:375-381).

Additional references added:
35. Walsh JR, Morris NR, McKeough ZJ, Yerkovich ST, Paratz JD. A simple clinical measure of quadriceps muscle strength identifies responders to pulmonary rehabilitation. Pulm Med. 2014;2014(1): 782702. DOI: 10.1155/2014/782702

Clinical Implications: You mention the need for an "individualized approach." Based on your data, what specific modifications to PR would you suggest for "Non-responders" (e.g., older patients with lower symptom burden at baseline)?

We thank the reviewer for this important question. While causal recommendations cannot be made based on our observational data, our findings suggest that patients classified as non-responders, often older individuals with lower symptom burden, may benefit from an even more individualized pulmonary rehabilitation approach, in line with ERS/ATS recommendations. Potential modifications could include closer monitoring and adjustments of training modality and intensity for optimized targeting of extra-pulmonary manifestations/ limitations, greater emphasis on strength, aerobic or task-specific functional training depending on limitation, and consideration of outcome targets beyond symptom-based MCIDs, acknowledging day-to-day variability in functional walk testing and patient-reported outcomes. In addition, further assessment of frailty, comorbidity burden, or nutritional status may help identify potentially modifiable barriers to response. We have clarified these clinical implications in the Discussion section. (Please read track changes manuscript clarification p. 13-14 line:395-402).

Study Limitations: Please explicitly mention the unblinded nature of the data collection and the retrospective design as limitations.

We thank the reviewer for this clarification request and have revised the limitations section to explicitly acknowledge the retrospective and unblinded nature of the data collection. (Please read track changes manuscript clarification p. 14; line:426-429).

Reviewer 2 Report

Comments and Suggestions for Authors

The authors have presented an interesting study on the response to rehabilitation in the medium and long term. The work is well conducted, with limitations typical of retrospective studies.

However, the authors must clarify and improve some points.

1. In the abstract, the authors must clarify what the MCDI was measured for.
2. The authors must better clarify the role of patients with CRF and whether chronic respiratory failure was a variable that influenced the results.
3. The authors must emphasize in every part of their work (in the methods and conclusions) that the studied population is an outpatient population.
4. The authors must better clarify the number of sessions proposed to the patients and, in particular, explain in detail what type of training was proposed, how they chose the training level (watts), and how they progressed over time. This is fundamental to understanding whether the patients were actually trained or not, with inevitable repercussions on the results.
5. The authors talk about walking capacity, but it would be better to talk about exercise tolerance.
6. The authors explain that the choice of ESWT versus six minutes and SGRQ versus CAT was made over the years: the authors must explain why.
7. The authors talk about percentages of differences in responses to the MDCI, but it is not clear for which single chosen outcome: these results must be presented in detail.
8. The authors then did not calculate the percentage of cumulative response to all the identified indicators.
9. The studied population shows a superiority in the number of women compared to men, which is different from previous studies: a comment is needed.
10. The tables must be aligned in numbers with respect to pack-years and comorbidities.
11. A comment must be made regarding the medications used by the patients, which do not seem particularly in line with the guidelines of recent years.
12. The authors must absolutely present the percentage of predicted referred to the walking test to better understand the severity level of their patients.
13. From the presented data, it seems that many patients showed desaturation during exercise: the authors must explain how desaturation was measured (at nadir? at the end of the walking test?; How did they behave in the face of exercise desaturation? Was supplemental oxygen prescribed? Were patients who desaturated during exercise trained with supplemental oxygen?
14. The authors talk about pauses during the walking test, and from the analysis of the results, they seem to be quite numerous: this is quite unusual given the meter values presented in the table.
15. Table three is too long: it is recommended to split it into two, one with the data at 10 weeks and another table with the data at one year.
16. The conclusions must be better clarified in terms of MDCI results for all indicators, for partial indicators, or for no indicators.
17. What happened during the patient follow-up? Did they have reinforcements? Did they attend maintenance pathways?
18. 50% of patients improved at least one of the chosen indicators: now this result seems quite modest if we consider that the authors used a value of 30 meters on the walking test and 180 seconds on the ESWT test, and 2 points on the CAT, which are quite achievable cutoffs for patients who are not particularly severe.
19. Furthermore, the authors must better clarify the role of comorbidities in their final results.
20. In their discussion, to comment on the differences between ESWT and six minutes, the authors cite a 2022 paper (Vitacca et al.), claiming that their results contrast with those of the cited work: the cited work does not consider the ESWT value but only the walking test.

Author Response

Reviewer 2

The authors have presented an interesting study on the response to rehabilitation in the medium and long term. The work is well conducted, with limitations typical of retrospective studies.

However, the authors must clarify and improve some points.

1. In the abstract, the authors must clarify what the MCDI was measured for.

We thank the reviewer for this clarification comment. We have revised the abstract to explicitly clarify that MCIDs were defined for walking capacity (6MWT or ESWT) and patient-reported outcomes (SGRQ or CAT). (Please read abstract track changes manuscript clarification p. 1).

2. The authors must better clarify the role of patients with CRF and whether chronic respiratory failure was a variable that influenced the results.

Thank you for this relevant comment. Chronic respiratory failure (CRF) was not systematically recorded in the database and therefore could not be analyzed as an independent predictor of pulmonary rehabilitation response.

Although some patients received long-term oxygen therapy, objective criteria required to define CRF (e.g. arterial blood gases or long-term ventilatory support) were not available. Moreover, oxygen supplementation during pulmonary rehabilitation and exercise in Denmark is prescribed and titrated conservatively by physicians according to routine clinical practice, typically at 1 or 2 L·min⁻¹ above the patient’s prescribed long-term oxygen therapy (LTOT) at home.

1. The authors must emphasize in every part of their work (in the methods and conclusions) that the studied population is an outpatient population.

We have accordingly emphasized the term “outpatient” throughout the revised manuscript, including the Abstract, Methods, Discussion, and Conclusions, to explicitly clarify that the study population consisted exclusively of outpatients undergoing pulmonary rehabilitation.
(Please trach changes throughout the revised manuscript).

1. The authors must better clarify the number of sessions proposed to the patients and, in particular, explain in detail what type of training was proposed, how they chose the training level (watts), and how they progressed over time. This is fundamental to understanding whether the patients were actually trained or not, with inevitable repercussions on the results.
   
   Thank you for this suggestion.
   To clarify training exposure and progression, we have added a dedicated subsection of the PR program to the Methods section. This section provides details the number of supervised sessions (twice weekly for 10 weeks; 20 sessions in total), the type of training provided (endurance and resistance training), and the principles used to prescribe and progress training intensity. Furthermore, we have added a Supplementary with full details and description of exercise content, progression principles, and patient education provided in our out-patient PR program.
   We believe these additions clarify that patients received structured, guideline-based training of sufficient intensity and progression to induce physiological training effects and self-management skill. (Please read track changes manuscript table 1 p. 3, line 103-120 and read added supplement S1 document).
2. The authors talk about walking capacity, but it would be better to talk about exercise tolerance.
   
   We thank the reviewer for this comment and agree that “exercise tolerance” is a broad and clinically relevant concept. However, in the present study we deliberately use the term “walking capacity” because the outcomes assessed (6-minute walk test and endurance shuttle walk test) specifically quantify walking-based functional performance rather than global exercise tolerance. While exercise tolerance may encompass additional physiological domains (e.g. cardiopulmonary responses during walking or cycle ergometry), walking capacity represents a clinically meaningful, task-specific manifestation of exercise limitation that is directly relevant to daily function in patients with COPD.
   
   Notably, the term “walking capacity or exercise capacity” is used consistently in the two ERS/ATS technical standards/guidelines for walking tests, whereas “exercise tolerance” is not used, further supporting our terminology choice.
   We have therefore respectfully retained the term “walking capacity” for accuracy.
3. The authors explain that the choice of ESWT versus six minutes and SGRQ versus CAT was made over the years: the authors must explain why.
   
   The transition from ESWT to the 6-minute walk test and from SGRQ to CAT reflects changes in routine clinical practice over time. These decisions were made by the clinical leadership of the pulmonary rehabilitation programme to reduce patient burden and shorten assessment time in a high-volume outpatient setting, while maintaining clinically validated outcome measures. Both the 6MWT and CAT are widely recommended, time-efficient, and well-established tools for assessing functional walking capacity and symptom burden in COPD, supporting their adoption in routine care. All transitions were implemented independently of the present study and without influence from any member of the research group.
4. The authors talk about percentages of differences in responses to the MDCI, but it is not clear for which single chosen outcome: these results must be presented in detail.
   
   Thank you for this comment. We realise that the description of the MCID-based response percentages may not have been sufficiently clear in the original manuscript. In the present study, response was assessed separately for each predefined outcome indicator using established MCID cut-offs, as these outcomes represent distinct and clinically meaningful domains.

Each patient had complete data for one functional outcome (either the Endurance Shuttle Walk Test or the 6-minute walk distance) and one health status outcome (either the SGRQ or CAT). These outcomes were subsequently combined into responder categories (full responder: both outcomes; responder: one of two outcomes; non-responder: none of the outcomes), as described in Methods section 2.6 (p. 3, lines 149–156).

To further clarify the reported percentages, we have added an explicit explanation to the Results section: “At 10 weeks, 52% of patients achieved a clinically meaningful response to outpatient PR, of whom 15% were classified as full responders and 37% as responders. Approximately 47% of these patients maintained their gains at 1-year follow-up” (Figure 1; tracked changes p. 7, lines 208–210). The Sankey plot was used to visually display these response proportions over time.

1. The authors then did not calculate the percentage of cumulative response to all the identified indicators.
   
   We thank the reviewer for this comment. The cumulative response across the identified indicators was calculated and is reflected in the reported overall response rate of 52% (p. 7, line 208). We recognize that this cumulative nature was not made sufficiently explicit in the original manuscript and have now clarified this by adding an explanatory sentence, as described in our response to the previous comment (Question 7).

If the reviewer is referring to an alternative composite or cumulative response definition and considers this important for data presentation, we would welcome further clarification.

1. The studied population shows a superiority in the number of women compared to men, which is different from previous studies: a comment is needed.
   
   
   Thank you for highlighting the higher proportion of women in the studied population.

In this clinical database study, 63% of included participants were female, which is higher than the national distribution of individuals diagnosed with COPD in Denmark (approximately 54% female). We believe this difference is plausibly explained by the nature of the data source. The database is derived from an outpatient pulmonary rehabilitation setting, where referral and participation depend on healthcare-seeking behavior, symptom reporting, and engagement with specialist services. It is documented that women, on average, are more likely than men to seek medical care and to engage in preventive and rehabilitative healthcare services in Denmark. This may lead to an overrepresentation of women in real-world outpatient clinical databases compared with population-based registries, without necessarily reflecting differences in disease prevalence, severity or need for PR. We also acknowledge that sex distribution in pulmonary rehabilitation programs varies across countries and cultures.

Importantly, sex distributions in Danish randomized controlled trials and cohort studies on COPD typically range from 50–56% female participants, aligning with national statistics and supporting the interpretation that the higher proportion observed in the present database reflects setting-specific referral and participation patterns rather than selection bias.

Furthermore, as requested by your co-reviewer comment, we examined sex distribution among included versus non-included patients. This analysis did not indicate that male sex was associated with a higher likelihood of incomplete assessment or exclusion from the study.

Overall, we believe the observed female predominance is a plausible consequence of healthcare-seeking behavior and engagement with specialist outpatient rehabilitation services and does not compromise the validity or generalizability of the study findings.

1. The tables must be aligned in numbers with respect to pack-years and comorbidities.

We thank the reviewer for this comment. The tables have been revised to ensure consistent numerical alignment for pack-years and comorbidities. These corrections have been implemented throughout the manuscript in the revised version.

1. A comment must be made regarding the medications used by the patients, which do not seem particularly in line with the guidelines of recent years.

We thank the reviewer for this relevant comment. The prescribed inhaler medication data in this study span an 18-year retrospective period during which international and national COPD treatment guidelines changed multiple times. Therefore, the reported medication patterns should be interpreted in the context of evolving guideline recommendations over time.

In addition, real-world clinical practice does not always strictly follow guideline recommendations. In everyday care, inhaled therapy is often individually tailored based on symptom burden, exacerbation history, treatment response, and patient tolerance. Deviations from guidelines may therefore reflect attempts to optimize symptom control or reduce recurrent exacerbations when guideline-recommended regimens prove insufficient for individual patients.

Finally, as the database is historical and fully anonymized, it is not possible to re-evaluate or validate individual medication regimens or link changes in treatment to clinical decision-making at the patient level.

Overall, the observed medication patterns reflect real-world clinical practice across a long time span rather than current guideline-adherent treatment and should be interpreted accordingly.

1. The authors must absolutely present the percentage of predicted referred to the walking test to better understand the severity level of their patients.
   
   We thank the reviewer for this suggestion. In response, we have now added percentage of predicted values for the 6-minute walk distance (6MWD) to better characterize the severity of functional limitation in the study population. As reference values for percentage predicted are not available for the Endurance Shuttle Walk Test (EndSWT), we instead report walking speed for the EndSWT, which provides a clinically indicator for the reader. In addition, we have tested and reported differences in these variables across full responders, responders, and non-responders, consistent with the analyses presented for other baseline characteristics in Table 1. We believe these additions improve the interpretability of disease severity and functional impairment in the cohort.
   (Please read numbers in track changes manuscript table 1 p. 6).
   
   
2. From the presented data, it seems that many patients showed desaturation during exercise: the authors must explain how desaturation was measured (at nadir? at the end of the walking test?; How did they behave in the face of exercise desaturation? Was supplemental oxygen prescribed? Were patients who desaturated during exercise trained with supplemental oxygen?
   
   Supplemental oxygen during exercise was provided only to patients receiving long-term oxygen therapy (LTOT) and was conservatively titrated (1–2 L·min⁻¹ above resting flow) during exercising. Patients with exertional desaturation performed endurance training in an interval-based format in accordance with guideline recommendations. As suggested in comment point 4., we have added a dedicated subsection of the PR program to the Methods section.
   We believe these additions clarify that patients received structured, guideline-based training. (Please read numbers in track changes manuscript p. 3, line 103-120 and read added supplement S1 document).
3. The authors talk about pauses during the walking test, and from the analysis of the results, they seem to be quite numerous: this is quite unusual given the meter values presented in the table.
   
   We thank the reviewer for this comment. According to the ERS/ATS guidelines for the 6-minute walk test (6MWT), resting or pausing during the test is permitted while time continues to run, and tests may be stopped before 6 minutes are completed, with the recorded walking distance still considered valid and clinically meaningful. Therefore, pauses during the test are an accepted and relevant part of functional assessment in patients with severe COPD although often not reported in most studies using the 6MWT.

The mean number of pauses observed in our study (1.1 pauses per test) is, from our clinical experience, not unusually high, particularly in a cohort with advanced disease. This interpretation is supported by the very low average walking distance of 299 m, indicating markedly reduced functional walking capacity. Distances below 350 m are well established to be associated with increased risk of adverse outcomes, including hospitalization and mortality within the following 12 months, and the distances observed in our cohort are substantially lower than those reported in many published studies. From daily clinical practice, our empirical impression is that such pauses are typically brief (approximately 10–20 seconds) and mainly reflect short rests to relieve dyspnea rather than prolonged interruptions. Unfortunately, the duration of pauses was not recorded in the database and therefore cannot be analyzed further.

Overall, we believe the frequency of pauses observed is consistent with guideline-based test conduct and reflects the severity of functional walking limitation in this patient population rather than an inconsistency in the data.

1. Table three is too long: it is recommended to split it into two, one with the data at 10 weeks and another table with the data at one year.
   
   This is a fair request and suggestion. Table 3 has now been split into two separate tables, one presenting outcome at 10 weeks and another presenting outcome at 1-year follow-up. We hope this addresses the reviewer’s request and improves the readability and clarity of the results.
   (We have inserted the two new/revised tables 3 and 4 in the revised track changes manuscript. Please read track change revised table 3 and 4)

1. The conclusions must be better clarified in terms of MDCI results for all indicators, for partial indicators, or for no indicators.
   
   Thank you for this relevant comment. We changed the first statement in the conclusion according to your comment to clarify the results of response: “Following outpatient PR, 52% of patients with COPD demonstrated clinically meaningful improvements, 15% as full responders and 37% as responders. Approximately 47% of patients maintained these gains at one year.” (track changes p.14 line 432-434).
2. What happened during the patient follow-up? Did they have reinforcements? Did they attend maintenance pathways?
   
   Thank you for this relevant question. Following completion of the 10-week pulmonary rehabilitation program, patients did not receive any structured reinforcement, booster sessions, or formal maintenance pathways. However, all patients were encouraged to continue regular weekly exercise independently after program completion. In Denmark, no reimbursement or organized funding is provided for continued exercise in fitness centers or community-based exercise programs following pulmonary rehabilitation, and participation in such activities is therefore voluntary and self-initiated. For clarity we have specified this in the revised manuscript (Please read abstract track changes manuscript clarification p. 3, line 117-120).
3. 50% of patients improved at least one of the chosen indicators: now this result seems quite modest if we consider that the authors used a value of 30 meters on the walking test and 180 seconds on the ESWT test, and 2 points on the CAT, which are quite achievable cutoffs for patients who are not particularly severe.
   
   We thank the reviewer for this thoughtful comment. The MCID thresholds applied in the present study were based on well-established and widely accepted cut-off values for the respective outcome measures (30 m for 6MWD, 180 s for ESWT, and 2 points for CAT), which are commonly regarded as reference standards in the pulmonary rehabilitation literature. We deliberately chose these evidence-based thresholds to ensure an objective and comparable definition of clinically meaningful improvement.

Although these cut-offs may appear achievable, it is important to note that the majority of patients in our cohort had advanced disease, with most classified as GOLD stage III or IV, reflecting a clinically severe population. In this context, achieving clinically meaningful improvements across functional and health status domains remains challenging.

As a real-world study, our aim was to report response rates using established MCID criteria rather than more stringent or study-specific thresholds, thereby providing an objective and clinically interpretable representation of the actual effectiveness of outpatient pulmonary rehabilitation in routine practice.

1. Furthermore, the authors must better clarify the role of comorbidities in their final results.
   
   We agree that comorbidities may influence response to pulmonary rehabilitation and potentially affect the final results. Unfortunately, standardized comorbidity indices such as the Charlson Comorbidity Index or the COPD-specific COTE index were not recorded or calculated at the time the database was established. As a result, we are only able to report comorbidity information available in the database, limited to the presence of cardiovascular disease (unspecified) and musculoskeletal disease (unspecified). This represents a limitation of the study, as more detailed and standardized comorbidity data could have provided clinically relevant insights into the influence of comorbidities on rehabilitation outcomes. We have therefore explicitly acknowledged this limitation in the manuscript, as we are unable to further elaborate on the impact of comorbidities on the final results.

The added limitation reads:
“In addition, detailed and standardised comorbidity indices (e.g. Charlson or COTE) were not available in the database, limiting our ability to assess the impact of comorbidities on rehabilitation outcomes.” (track change abstract p.14, line: 423-426.)

1. In their discussion, to comment on the differences between ESWT and six minutes, the authors cite a 2022 paper (Vitacca et al.), claiming that their results contrast with those of the cited work: the cited work does not consider the ESWT value but only the walking test.
   
   We thank the reviewer for pointing this out and agree that the originally cited study did not assess the Endurance Shuttle Walk Test (ESWT).

To address this, we have revised the Discussion to clearly distinguish between the outcome measures used in the two studies. Specifically, we now state that our findings relate to baseline ESWT, whereas the study by Vitacca et al. assessed baseline functional status using the six-minute walk test (6MWT). The revised wording avoids implying that ESWT was
analyzed in the cited study and frames the findings as contrasting but methodologically distinct. (track change abstract p.12, line: 319-321.)

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

Authors have answered to all my observations improving their manuscript