Participants were recruited between October 2022 and May 2025 at the China-Japan Friendship Hospital. Inclusion criteria were as follows: (1) aged 18 years or older; (2) undergone unilateral or bilateral lung transplantation; (3) able to stand unsupported for at least 5 min; (4) able to walk 50 m or more (assistive devices permitted); (5) provided written informed consent. The exclusion criteria referred to post-transplant conditions and were established to ensure patient safety and adherence during PR. Patients were excluded if they met any of the following: (1) presence of severe comorbidities affecting major organ systems (cardiac, cerebral, hepatic, or renal dysfunction); (2) psychiatric, cognitive, or musculoskeletal disorders that would impair adherence to pulmonary rehabilitation protocols; (3) clinically unstable conditions deemed rehabilitation-contraindicated by physicians; (4) inability to provide informed consent; (5) noncompliance with prescribed PR protocols.

All baseline assessments were performed prior to the initiation of the PR program. Demographic data, resting oxygen saturation, transplant type (single/double LTx), lung function, dyspnea severity (via the Borg scale), interval between LTx and PR, duration in intensive care unit. Lung function, specifically forced expiratory volume in the first second (FEV1) and forced vital capacity (FVC), was measured via spirometry. The 6MWT and handgrip assessments were conducted pre and post the PR protocols. The assessment order was randomized, with a minimum 48-h interval.

The outcomes of the calculations for the anchor-based, distribution-based and combined methods were shown in Table 2. We calculated a 6MWT MCIDs of each anchor: left handgrip 58.50 m, right handgrip 56.30 m and average handgrip 52.65 m. With the 0.5*SD_pre distribution-based method, 57.28 m was observed as the MCID. The SEM distribution-based method estimated the 6MWT MCID at 59.40 m. The combined MCID (average of the three anchor-based MCIDs and two distribution-based MCID) was 56.83 m.

Our study confirmed that the 6MWT is highly responsive to PR in LTRs, with a large effect size (Cohen’s d = 1.10). Notably, despite a longer PR intervention duration (12 weeks vs. 8 weeks in a prior study), the mean improvement in 6MWT distance (73.37 m) was smaller than the 174.2 m reported previously (17). Two key factors may explain this discrepancy: first, demographic and clinical differences between cohorts, including older mean age (57.23 years vs. 38.4 years) and a longer interval from LTx to postoperative PR initiation (159.53 days vs. 31.4 days), which have been shown to modulate PR outcomes in LTRs (7); second, a high proportion of our participants (69.14%) had EID at baseline—a condition known to impact functional outcomes in chronic respiratory diseases (19, 20), though its specific effect on PR responsiveness in LTRs warrants further investigation.

Importantly, the interval between lung transplantation and baseline assessment varied across participants, indicating enrollment at different postoperative stages. This reflects real-world clinical practice, in which PR is initiated once clinical stability is achieved rather than at a fixed postoperative time point. Although recovery trajectories and responsiveness to rehabilitation may differ by postoperative stage, the present analysis focused on within-subject pre–post changes in 6MWT distance. This self-controlled design minimizes confounding from baseline functional heterogeneity and transplant timing, thereby facilitating the assessment of PR-related improvement across a heterogeneous post-transplant population.

This study estimated the 6MWT for the MCID to be 56.83 m in postoperative LTRs, which is smaller than the 79 m reported in a prior study using a single distribution-based method (17). Three interrelated factors contribute to this discrepancy: first, MCID estimates are known to be influenced by disease severity and clinical complexity (21, 22), and our cohort included older LTRs (mean age 57.23 years vs. 38.4 years) with higher baseline complexity (69.14% with EID), which likely reduced the magnitude of PR-induced 6MWT improvement and consequently yielded a lower MCID; second, unlike the prior study that relied solely on a distribution-based approach, we adopted a combined anchor-based (handgrip strength) and distribution-based method, which addresses the limitations of single-method estimates and enhances result robustness (16, 23); third, the integration of clinical relevance and statistical rigor avoids the potential overestimation associated with standalone distribution-based calculations (12, 13).

The 56.83 m MCID holds important clinical and research implications: first, it exceeds the MCID for exercise capacity in chronic respiratory diseases (e.g., 26–30 m in COPD) (21), highlighting that LTRs require a greater improvement in 6MWT distance to achieve clinically meaningful benefits from PR—reflecting the unique physiological challenges of post-transplant recovery; second, it provides a practical clinical benchmark: a 6MWT improvement exceeds 56.83 m post-PR can be defined as a “clinically effective response,” enabling clinicians to identify non-responders and adjust rehabilitation protocols in a timely manner; third, it supports rigorous trial design by offering a validated threshold for sample size calculation, reducing the risk of underpowered studies in LTR rehabilitation research.

Several methodological considerations should be acknowledged. First, handgrip was selected as the anchor for MCID estimation, given its close association with skeletal muscle dysfunction in LTRs and its strong correlation with the 6MWD. Although this objective measure strengthens the pathophysiological relevance of the estimates, it does not fully capture patients’ perceived improvement—a core aspect of the MCID concept (10). Considering the heterogeneity in anchor selection and the uncertainty in defining a minimal important difference, future studies should employ multiple anchors (both subjective and objective) to reduce estimation error and enhance robustness. Secondly, we conducted the 6MWT only before and after intervention, and the intraclass correlation coefficient (ICC) for measuring the MCID based on the SEM was derived from a previous study (17). While we did not anticipate that this approach significantly affected the current results, the reliability of the findings would have been enhanced if the ICC had been obtained from the same sample as the standard deviation (24). Thirdly, there is a lack of a universally accepted standard for determining the MCID. Different calculation methods may yield varying results. A MCID that is estimated to be too low might lead to an overestimation of the effects of an intervention. Conversely, a MCID that is estimated to be too high may result in patients who actually benefit from an intervention being incorrectly judged as not responding to it in clinical practice (9). Although, many studies recommend using a combination of different methods to determine the MCID. There is currently no robust research on the appropriate weight allocation. We synthesized methodologies from existing studies and ultimately implemented a mean-weighted approach to calculate the MCID (16, 23). Finally, while anchor-based methods (e.g., linear regression) prioritize clinical interpretability of MCID, their validity remains limited by anchor selection, analytical approaches (absolute/relative change), and population heterogeneity (e.g., disease severity, interventions) (25). Although our study demonstrated significant correlations between grip strength (anchor) and 6MWT improvements, future studies should integrate multiple validated anchors and reduce inter-subject variability to enhance MCID reliability.