Optimal intensity and type of lower limb aerobic training for patients with chronic obstructive pulmonary disease: a systematic review and network meta-analysis of RCTs

Abstract

Background:

Lower limb aerobic exercise is the core component of pulmonary rehabilitation for chronic obstructive pulmonary disease (COPD) patients. The optimal intensity and type (e.g., interval or continuous) of exercise training remains to be determined.

Objectives:

We aimed to evaluate the optimal intensities and types of lower limb aerobic exercise in patients with COPD.

Design:

Systematic review and network meta-analysis of randomized controlled trials.

Data sources and methods:

The PubMed, Web of Science, Embase, and the Cochrane Central Register of Controlled Trials were searched for relevant data. The interventions were classified according to their intensity and type as high-intensity interval training (HIIT), high-intensity continuous training (HICT), moderate-intensity continuous training (MICT), and low-intensity continuous training (LICT). We assessed exercise capacity using peak work rate (Wpeak) and the 6-min walking test (6-MWT). Lung function was evaluated by measuring peak minute ventilation (VE) and the percentage of predicted FEV₁ (FEV₁pred%). Dyspnea was assessed using the Modified Medical Research Council (mMRC) scale. Quality of life was measured with the Chronic Respiratory Questionnaire (CRQ).

Results:

Fifteen studies were identified (979 subjects). HIIT showed the greatest improvement in Wpeak, 6-MWT, VE, and mMRC compared to usual care (MD 18.48 (95% CI 12.35, 24.60), 67.73 (34.89, 100.57), 6.26 (2.81, 9.72), and −0.53 (−0.89, −0.17), respectively) and showed the improvement in CRQ (MD 10.80 (95% CI 1.65, 19.95)). MICT showed improvement in Wpeak and 6-MWT (MD 18.28 (95% CI 11.20, 25.22), 61.92 (28.34, 95.51)) similar to HICT (MD 16.08 (95% CI 8.19, 23.84), 64.64 (28.70, 100.57)) and showed the highest improvement in CRQ compared to usual care (MD 10.83 (95% CI 1.68, 19.98)). LICT significantly improved Wpeak compared to usual care (MD 13.47 (95% CI 4.77, 22.13)). The quality of evidence for outcomes varied from very low to moderate.

Conclusion:

HIIT and MICT might be optimal training approaches for patients with COPD. LICT exhibited limited clinical efficacy. While HICT was as effective as MICT, it caused more dyspnea.

Trial registration:

This systematic review and network meta-analysis was prospectively registered with PROSPERO (No. CRD 42024520134).

Plain language summary

Optimal intensity and type of lower limb aerobic training for COPD patients: a network meta-analysis of RCTs

Pulmonary rehabilitation, which often includes lower limb aerobic exercise, is essential for managing chronic obstructive pulmonary disease (COPD). However, the best intensity and type of exercise for these patients are not yet fully established. This study aimed to identify which exercise intensity and type are most effective for patients with COPD. To achieve this, we reviewed and analyzed data from multiple studies that compared different exercise programs: high-intensity interval training (HIIT), high-intensity continuous training (HICT), moderate-intensity continuous training (MICT), and low-intensity continuous training (LICT). We evaluated their effects on exercise capacity, lung function, breathlessness, and quality of life. Our analysis included 15 studies with 979 participants. The results showed that HIIT led to the most significant improvements in exercise capacity, lung function, and breathlessness, and it also enhanced quality of life. MICT also produced notable benefits, particularly in exercise capacity and quality of life, similar to HICT. LICT was effective in improving exercise capacity but to a lesser extent. Overall, the evidence suggests that HIIT and MICT may be the most effective exercise approaches for patients with COPD. However, the quality of the evidence varied, so further research is needed to confirm these findings.

Graphical abstract

Keywords

chronic obstructive pulmonary disease exercise prescription high-intensity interval training moderate-intensity continuous training pulmonary rehabilitation

Introduction

Chronic obstructive pulmonary disease (COPD) is a pulmonary disease with chronic respiratory symptoms caused by airway and/or alveoli abnormalities, leading to persistent airflow obstruction.¹ As the disease deteriorates, it can lead to complications in various organ systems, such as the cardiovascular and musculoskeletal systems, significantly impacting the patient’s overall quality of life.^2,3 By 2030, COPD will be the third leading cause of death globally, resulting in significant clinical, economic, and social burdens.⁴

The global initiative for COPD guidelines proposed that lower limb aerobic exercise, a cost-effective treatment, could alleviate the clinical symptoms, reduce the risk of mortality and cardiovascular burden among patients with COPD, and prolong their life expectancy.^1,5 However, there is a wide variety of aerobic exercises with significant variations in their content. Exercise prescribing to maximize patient benefit remains unresolved.^6,7

The intensity and type (e.g., interval or continuous) of the exercises play a pivotal role in shaping the exercise prescription.^8
–10 Low-intensity training often produces limited physiological outcomes, whereas high-intensity training invariably results in significant dyspnea, especially in individuals with pre-existing lung conditions.^8,11 Additionally, interval training has been reported to enhance exercise performance more effectively than constant-load training in healthy individuals,¹² as it increases exercise volume and helps to avoid fatigue and breathlessness.¹³ However, it remains unclear whether similar effects are observed in patients with COPD.¹⁴ Some studies have primarily focused on training methods, neglecting the quantitative aspect of intervention intensity, leading to a failure to acknowledge the heterogeneity in different intensity levels.^15,16

This meta-analysis examines the impact of exercise intensity and type on exercise capacity, lung function, dyspnea, and quality of life in patients with COPD. The comprehensive analysis of quantifying the intensity of lower limb aerobic exercise will provide more precise guidance for individuals with COPD.¹⁷

Methods

This systematic review was reported following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) for Network Meta-Analyses (PRISMA-NMA) statement and checklist.¹⁸ The search has been registered in PROSPERO (CRD 42024520134).

Search strategy

Multiple databases, including PubMed, Cochrane Library, Web of Science, and Embase, were searched from their inception to January 25, 2024. The main search strategies were as follows: (“chronic obstructive lung disease” OR “chronic obstructive pulmonary diseases” OR “coad”:ab, ti OR “copd”) AND (“exercise” OR “train” OR “training” OR “lower limb” OR “treadmill” OR “ergometer”) AND (“intensity” OR “endurance” OR “interval” OR “intermittent” OR “continuous”). See Supplemental Part A for a more detailed presentation of the search strategy.

Inclusion and exclusion criteria

The eligibility criteria were reported according to the Population, Intervention, Comparison, Outcome, and Study design approach.¹⁹ Studies were included that matched the following criteria:

i. Participants: Patients were diagnosed with COPD defined by best post-bronchodilator forced expiratory volume in 1 s (FEV₁)/forced vital capacity ratio <0.7;

ii. Intervention: The intervention was mainly lower limb aerobic training with quantifiable intensity based on a maximal test (i.e., Wpeak), and the experimental intensities or types were reported; Groups receiving multiple interventions (e.g., resistance training) were not considered.

iii. Comparator: The control group underwent lower limb aerobic training with intensity or type of exercise different from that of the intervention group, conventional health education, and usual care;

iv. Outcomes: We assessed exercise capacity using peak work rate (Wpeak) and the 6-min walking test (6-MWT). Lung function was evaluated by measuring peak minute ventilation (VE) and the percentage of predicted FEV₁ (FEV₁pred%). Dyspnea was assessed using the Modified Medical Research Council (mMRC) scale. Quality of life was measured with the Chronic Respiratory Questionnaire (CRQ). The study should include at least one of the above outcomes.

v. Study design: Randomized controlled trials (RCTs) were included. Articles without an abstract or full text published were excluded.

Those studies that could not meet the above criteria were excluded from this review, for example, COPD with other predominant diagnosis, published abstract only, or unpublished data.

Data extraction

Two researchers independently extracted the primary information from a pre-set standardized form. The following data were extracted: (1) Basic information, including first author and year of publication. (2) Characteristics of participants, including sample size, age, and gender. (3) Intervention details include the treatment, dosage, duration, process, and follow-up period. (4) Outcomes information. If the data were not presented in textual format, the “GetData Graph Digitizer” software was utilized to extract the information from visual representations. If necessary, the corresponding authors were contacted for additional information. According to previous research and guidance, exercise intensity was categorized as high intensity (⩾80% of peak work capacity), medium intensity (>50% and < 80% of peak work capacity), and low intensity (⩽50% of peak work capacity).^11,20,21 Two researchers (ZTQ and ZWK) independently screened the studies obtained and then analyzed the title, abstract, and full text. An independent researcher (KL) will evaluate if conflicting viewpoints arise.

Risk of bias

Cochrane tool for randomized trials (RoB 2.0) was used by two independent researchers (QZT and ZWK) to assess the methodological quality of the RCTs.²² The third researcher (KL) resolved disagreements as required. During the evaluation process, the risk of bias was assessed based on five domains: randomization process, deviations from the expected interventions, missing outcome data, outcome measurement, and selection of reported outcomes.^23,24 Publication bias was assessed via visual inspection of funnel plots. If all areas are classified as “low risk,” the study is considered to have a “low concern.” When an area was categorized as “some concerns,” the study was deemed to possess an “unclear risk of bias” (including inapplicability and lack of information). When one or more areas are rated as having a “high risk,” these studies were regarded as situated in “high concern” areas.

Certainty assessment

The certainty of evidence was evaluated using the Grading of Recommendations, Assessment, Development, and Evaluation (GRADE) methodology.²⁵ In adherence to the GRADE system of rating the certainty of the evidence, we rated the certainty for each comparison and outcome as high, moderate, low, or very low. All comparisons were initially rated as “high” due to the inclusion of RCTs but subsequently downgraded based on risk of bias, inconsistency, intransitivity, heterogeneity, imprecision, and publication bias.²⁶

Statistical analysis

The Stata v16.0 (StateCorp LLC, USA), including the network package and the network diagram package, was utilized for network meta-analysis. The network package employed a random effects model and a frequency-based framework to analyze networks. A visual representation in the form of a network diagram was created, where nodes and lines were used to depict different interventions, with node size indicating population size and line thickness representing the number of studies conducted. In continuous variables, effect sizes were estimated using a mean difference with 95% confidence interval (CI). According to the recommendations of the Cochrane Handbook, if data on the change in standard deviation (SD) of outcomes from baseline were missing, estimates would be based on standard error, 95% CI, p values, or t-statistics. The results obtained from the network meta-analysis encompass all possible pairwise comparisons, including mixed comparisons that combine direct and indirect comparisons. The node-splitting tests assessed local inconsistencies between direct and indirect comparisons. If p < 0.05, it indicates the presence of local inconsistency. In case of observed inconsistency, non-transitivity was suspected, and potential confounding factors that may influence treatment effects were examined. The effectiveness of different interventions is evaluated using the surface under the cumulative ranking curve (SUCRA). SUCRA values range from 0% to 100%, where a SUCRA value of 100% signifies the most effective treatment, while smaller values indicate poorer treatment effects.

Results

Literature screening process and results

The PRISMA flowchart²⁷ was presented in Figure 1. The initial electronic search identified 17,925 potentially relevant publications. After removing 8296 duplicate records, a total of 9629 records were screened based on reading titles and abstracts, resulting in the exclusion of 9592 records. Among the remaining 37 studies eligible for full-text review, 22 were excluded based on inclusion and exclusion criteria, specifically, 4 non-randomized controlled trials,^14,28
–30 1 single-arm experimental study,³¹ 14 studies^32
–45 with interventions lacking quantifiable intensity, 2 studies^46,47 without the inclusion of the purpose outcome, and 2 studies^48,49 with incomplete/abnormal data. Ultimately, 15 RCTs^50
–65 were included.

Figure 1.

Flow diagram of the studies screened and included according to the PRISMA.

Description of included studies

The study was comprised of 15 studies with 979 participants, including two 3-arm studies. A total of 10 countries contributed to the publication of these studies, with Greece being the most prolific (4 studies, 26.7%). The average age of the participants was 67 years. The duration of the interventions in the 15 studies varied from 3 to 16 weeks, with a predominant treatment period of approximately 8 weeks (5 studies, 33.3%). Among these interventions, HIIT was included in 11 studies, HICT in 4 studies, MICT in 9 studies, LICT in 2 studies, self-paced training (SP) in 2 studies, and usual care (UC) in 4 studies. Twelve studies included Wpeak as an outcome measure; eight studies reported 6-MWT outcomes, FEV₁pred% was assessed in four studies, VE was compared across nine studies, mMRC outcomes were reported by three studies, and CRQ results were analyzed by four studies (Figure 2 and Table 1).

Figure 2.

Network diagram for Wpeak (a), 6-MWT (b), FEV₁pred% (c), VE (d), mMRC (e), and CRQ (f). Each node represents an intervention that has been tested in studies; the size of the nodes is proportional to the number of patients that have received that intervention, and the thickness of the edges is proportional to the number of studies.

Table 1.

Summary of study characteristics.

Author	Country	Intervention	No. of patients	GenderM/F	Age, yearsmean (SD)	Intervention details in Group 1	Intervention details in Group 2	Intervention details in Group 3	Follow-up	Outcomes
Larson et al. (1999)⁵¹	Chicago	HIIT UC	12 14	-	66 ± 6 62 ± 7	20 min period of cycling for 5 min in duration that intensity reaches limits of their dyspnea, separated by rest intervals (2–4 min) of unloaded, 5 d/week	Education with emphasis on how to live a heal their life and how to manage COPD, 1 time/week, 1 h/session	—	4 months	Wpeak, VE, CRQ
Varga et al. (2007)⁵²	Hungary	HIIT HICT SP	17 22 32	11/6 19/3 25/7	67 ± 10 61 ± 12 60 ± 12	30 min period of cycling for 2 min at 90% followed by 1 min at 50% peak work rate, with 7, 5 min warm-up and cool down at 50% of peak work rate, 3 times/week	Exercise intensity was 80% of peak work rate, 45 min/time, 3 times/week	Self-paced training 45 min walking or cycling, or climbing on stairs, 3 times/week	8 weeks	Wpeak, FEV₁pred%, VE
Vogiatzis et al. (2002)⁵³	Greece	HIIT MICT	18 18	14/4 16/2	67 ± 2 69 ± 2	40 min period of cycling for 30 s at 100% followed by 30s at 50% peak work rate, 2 days/week	Exercise at 50% of baseline peak work rate, 40 min/day, 2 days/week	—	12 weeks	Wpeak, VE
Santos et al. ((2015)⁵⁴	Portugal	MICT HICT	17 17	12/5 15/2	66.9 ± 11.4 67.3 ± 10.4	30 min in duration that intensity of 60% Wpeak, 3 times/week	30 min in duration with the intensity of 80% Wmax, 3 times/week	—	7 weeks	6-MWT
Nasis et al. (2009)⁵⁵	Greece	HIIT MICT	21 21	17/4 16/5	65 ± 3 66 ± 3	40 min period of cycling for 30 s at 100% peak W followed by 30s rest, 3 days/week	Exercised at a mean intensity of 76 ± 5% of baseline Wpeak, 30 min/day, 3 days/week	—	10 weeks	Wpeak, 6-MWT, FEV₁pred%, mMRC
Arnardóttir et al. (2006)⁵⁷	Sweden	HIIT MICT	28 32	25/3 26/6	65 ± 7 64 ± 8	39 min training included 6 min of warm-up; 3 min of high-intensity training (80% of the baseline Wpeak); 3 min of low-intensity training (30%–40%); 6 min of cool-down, 2 days/week	27 min each session at their effective training, that intensity was 65% of baseline Wpeak, 2 days/week	—	16 weeks	Wpeak, VE, CRQ
Mador et al. (2009)⁶⁰	USA	HIIT LICT	21 20	—	72.1 ± 6.8 71.5 ± 8.0	3 min blocks with 1 min at 150% and 2 min at 75% of the cycle workload or treadmill speed, 21 min/day, 3 days/week	Exercised at 50% of Wpeak in ergometer, 20 min/day, 3 days/week	—	8 weeks	Wpeak, 6-MWT, VE, CRQ
Vogiatzis et al. (2005)⁶¹	Greece	HIIT HICT	10 9	16/3	64.0 ± 9.5 67.0 ± 6.0	1–3 weeks: 100% of baseline Wpeak; 4–6 weeks: 120% of baseline Wpeak; 7–10 weeks: 140% of Wpeak, work for 30 s alternating with 30-s rest intervals cycle until 45 min, 3 days/week	1–3 weeks: 60% of baseline Wpeak; 4–6 weeks: 70% of baseline Wpeak; 7–10 weeks: 80% of baseline Wpeak, 30 min/day, 3 days/week	—	10 weeks	Wpeak
Coppoolse et al. (1999)⁵⁸	Netherlands	HIIT MICT	10 11	10/0 11/0	63 ± 8 67 ± 3	Nine blocks including 3 min of alternated 90% (1 min) and 45% (2 min) of the Wpeak, 3 days/week; 2 days/week continuous training at 60% of the Wpeak	Exercised continuously at 60% of peak work rate for 30 min each session, 5 days per week	—	8 weeks	Wpeak
Louvaris et al. 2016⁵⁹	Greece	HIIT UC	85 43	68/17 36/7	65 ± 8 67 ± 8	Exercised at a mean intensity of 130 ± 18% of baseline peak work rate for 45 min by alternating 30-s exercise intervals with 30-s rest periods, 3 days/week	Usual care	—	12 weeks	Wpeak, 6-MWT, mMRC, CRQ
Puhan et al. (2006)⁵⁶	USA	HIIT MICT	48 50	29/19 36/14	69.0 ± 9.2 68.9 ± 9.2	2 min of warm-up (10%); 20 cycles of 20 s to high-intensity training (90% to 100% of maximum exercise capacity); 40 s to low-intensity training (50% maximum exercise capacity); 2 min of cool-down; 24 min/day, 4–5 days/week	Cycle ergometers with a target workload of 70% maximum exercise capacity, 24 min/day, 4–5 days/week	—	3 weeks	Wpeak, 6-MWT, CRQ
Jiang et al. (2012)⁶⁵	China	HIIT LICT UC	20 20 20	18/2 16/4 19/1	73.5 ± 7.9 73.6 ± 8.3 75.3 ± 8.5	5 min of warm-up; 4 cycles of 5 min to high-intensity training and 5 min to active recovery, 5 min of cool down, 50 min/day; 3–5 days/week	Training intensity at 50%, with 5 min of warm-up and relaxation activities before and after, 20 min/day, 3–5 days/week	Usual care	8 weeks	Wpeak, VE
Borghi-Silva et al. (2009)⁶³	Brazil	MICT UC	20 20	13/7 12/8	67 ± 10 67 ± 10	The training intensity on the treadmill was set at 70% of the maximal speed achieved during the symptom-limited exercise test, 30 min/day, 3 days/week	Usual care	—	6 weeks	6-MWT, VE
Wang et al. (2017)⁵⁰	China	MICT SP	27 26	—	70.0 ± 6.3 69.8 ± 6.4	Exercise intensity was 70% VO₂max, 30 min/day, 3 days/week	Free walk on the playground of the rehabilitation center, 30 min/day, 3 days/week	—	8 weeks	Wpeak, 6-MWT, VE, mMRC
Maltais et al. (2008)⁶⁴	Canada	HICT MICT	126 126	72/54 68/58	66 ± 9 66 ± 9	Target training intensity was 80% of peak work capacity during incremental exercise, 25–30 min/day, 3 days/weeks	The target intensity was 60% of the maximum work rate, 40 min/day, 3 days/week	—	8 weeks	6-MWT

Data expressed as mean ± SD.

6-MWT, 6-min walking test; CRQ, Chronic Respiratory Questionnaire; FEV₁pred%, percentage of predicted forced expiratory volume in 1 second; HICT, high-intensity continuous training; HIIT, high-intensity interval training; LICT, low-intensity continuous training; MICT, moderate-intensity continuous training; mMRC, modified Medical Research Council; SP, self-paced training; UC, usual care; VE, peak minute ventilation; VO₂max, maximal oxygen consumption; Wpeak, Peak Work rate (W).

NMA of efficacy outcomes

Peak work rate

Twelve studies involving 595 patients with COPD were analyzed for Wpeak. HIIT showed the highest improvement in Wpeak (85.0%), followed by MICT (80.9%), HICT (61.3%), LICT (50.7%), SP (21.2%), and UC (0.9%; e-Figure 1(A)). HIIT showed a significant improvement compared to SP and UC (MD 11.70 (95% CI 6.93, 16.46) and 18.48 (95% CI 12.66, 24.42)), respectively. Similarly, MICT showed a significant improvement compared to SP and UC (MD 11.43 (95% CI 6.67, 16.20), MD 18.28 (95% CI 11.20, 25.22)), respectively. HICT also showed a significant improvement compared to SP and UC (MD 9.24 (95% CI 4.38, 14.09), MD 16.08 (95% CI 8.19, 23.84)), respectively. However, LICT only significantly improved compared to UC (MD 13.47 (95% CI 4.77, 22.13)). SP was superior to UC but without a significant difference (Figure 3(a) and e-Figure 2).

Figure 3.

Summary effects of interventions on changes in Wpeak (a), 6-MWT (b), FEV₁pred% (c), VE (d), mMRC (e), and CRQ (f). The arrow in the upper right corner indicates the direction of improvement.

Six-minute walking test

Eight studies involving 620 patients with COPD were analyzed for 6-MWT. HIIT (82.9%) was ranked highest in 6-MWT, followed by HICT (74.3%), MICT (63.4%), LICT (39.7%), SP (29.6%), and UC (10.0%; e-Figure 1(B)). HIIT showed a significant improvement compared to SP and UC (MD 40.00 (95% CI 22.60, 57.40), MD 67.73 (95% CI 34.89, 100.57)). HICT was also significantly better than SP and UC (MD 36.90 (95% CI 18.81, 54.99), MD 64.64 (95% CI 28.70, 100.57)). Similarly, MICT showed significantly better than SP and UC (MD 34.19 (95% CI 21.39, 46.99), MD 61.92 (95% CI 28.34, 95.51)). LICT and SP did not significantly improve compared to UC. (Figure 3(b) and e-Figure 3).

Percentage of predicted forced expiratory volume in 1 second

Four studies, including 294 patients with COPD, were analyzed for FEV₁pred%. HICT (59.5%) shows the highest improvement in peak work rate (W), followed by HIIT (58.7%), SP (56.1%), UC (47.4%), and MICT (28.3%) (e-Figure 1(C)). No significant differences were observed between HIIT, HICT, MICT, LICT, SP, and UC (Figure 3(c) and e-Figure 4).

Peak minute ventilation

Nine studies involving 502 patients with COPD were analyzed for VE. HIIT (87.1%) showed the highest efficacy in improving peak VE, followed by MICT (62.4%), HICT (53.4%), SP (44.3%), LICT (41.9%), and UC (10.9%; e-Figure 1(D)). Both HIIT and MICT demonstrated a significant improvement compared to UC, with MD 6.26 (95% CI 2.81, 9.72) and 4.34 (95% CI 0.13, 8.55), respectively. No significant differences were observed among other interventions (Figure 3(d) and e-Figure 5).

Modified Medical Research Council

Three studies involving 223 patients with COPD were analyzed for mMRC. HIIT (83.0%) and MICT (82.8%) demonstrated the highest efficacy in improving mMRC, followed by SP (21.7%) and UC (12.5%; e-Figure 1(E)). No significant differences were observed in the network comparison. MICT and HIIT significantly improved compared to SP (MD −0.44, 95% CI −0.80, −0.08; MD −0.44, 95% CI −0.82 to −0.06). Similarly, both MICT and HIIT showed a significant improvement compared to UC. (MD −0.53, 95% CI −0.91, −0.15; MD −0.53, 95%CI −0.89, −0.17; Figure 3(e) and e-Figure 6).

Chronic Respiratory Questionnaire

Three studies, including 324 patients with COPD, were analyzed for CRQ. MICT (74.1%) and HIIT (69.5%) demonstrated the highest efficacy in improving peak VE, followed by LICT (51.0%). In comparison, UC (5.4%) exhibited the lowest effectiveness (e-Figure 1(F)). Both HIIT and MICT showed a significant improvement compared to UC (MD 10.80 (95% CI 1.65, 19.95), MD 10.83 (95% CI 1.68, 19.98)). Although LICT outperformed UC, the CIs of these differences in CRQ crossed the null value of 0 (Figure 3(f) and e-Figure 7).

Bias and quality of evidence ROB

One study was assessed as low risk, nine as of some concern, and five as high risk of bias. Regarding each specific domain, the randomization process was classified as low risk of bias in 14 (93.3%) of studies; the deviations from the intended interventions domain were classified as low risk of bias in 1 (6.7%); the missing outcome data domain was classified as low risk of bias in 10 (66.7%); the measurement of the outcome domain was classified as low risk of bias in 14 (93.3%); and the selection of the reported result was classified as low risk of bias in 13 (86.7%; e-Figures 8–15).

Certainty of evidence

The evidence quality of all 67 comparisons was assessed using the GRADE system. Among these, 42 comparisons involved mixed effects (i.e., a combination of direct and indirect effects), while 26 were indirect. Within the mixed-effect group, 8 were deemed moderate-quality evidence, 21 had low-quality evidence, and 1 had very low-quality evidence. All 26 indirect comparisons were rated as very low-quality evidence (e-Figures 2–7).

Comparing efficacy on pairs of outcomes

We evaluated the efficacy of each intervention pair. HIIT, HICT, and MICT demonstrated higher effectiveness in improving 6-MWT and Wpeak, followed by LICT, SP, and UC (Figure 4(a)). For FEV₁pred% and VE scores, HIIT, HICT, and SP ranked higher, followed by MICT, UC, and LICT (Figure 4(b)). In terms of improving mMRC and CRQ scores, MICT and HIIT ranked higher, followed by LICT, SP, and UC. At the same time, HICT was not included in the rankings due to insufficient data (Figure 4(c)). The comprehensive pair comparison was presented in e-Figure 16.

Figure 4.

Scatterplots of the overall SUCRA score (ranking probability) of improving pairs of efficacy outcomes. (a) 6-MWT and Wpeak. (b) FEV₁pred% and VE. (c) mMRC and CRQ.

Inconsistency test

No inconsistencies were identified in Wpeak, 6-MWT, FEV₁pred%, mMRC, and CRQ. No overall inconsistency was detected in VE, but local inconsistencies in the MICT and LICT of VE were revealed. Only one study directly compared MICT and LICT for VE. After conducting a sensitivity analysis of the literature for indirect comparisons, no sources of heterogeneity were discerned.

Discussion

The network meta-analysis included 15 randomized clinical trials involving data from 979 participants. Results of SUCRA, with very low to moderate evidence, suggested that HIIT emerged as the optimal choice for improving exercise capacity (Wpeak and 6-MWT), lung function (VE), and Dyspnea (mMRC), and MICT was the preferred option for enhancing quality of life (CRQ). Regarding outcomes, no significant differences were observed among HIIT, HICT, MICT, and LICT. As hypothesized, we identified an intensity-dependent response and a lower improvement ceiling, making HIIT and MICT superior and most effective for all outcomes of interest. The intensity-dependent means that higher exercise levels yield better clinical effects. The lower improvement ceiling in patients with COPD is demonstrated by the lack of further improvement with HICT compared to MICT.¹³ This study builds on Zainuldin’s 2011 meta-analysis¹¹ but goes beyond a simple update. While Zainuldin’s analysis compared higher-intensity versus lower-intensity training and continuous versus interval training, it found no significant differences between these methods. The current study expands the scope by incorporating more recent literature, refining the classification methods for the first time, and utilizing an NMA to enable broader and more detailed comparisons. Specifically, this NMA provides an extensive synthesis of quantitative data from RCTs in COPD rehabilitation to examine the different effects of different intensities and types of lower limb aerobic exercise in patients with COPD.

The exercise capacity serves as the most immediate manifestation of the impact of training.⁶⁶ Morris et al. proposed that higher exercise intensity may enhance skeletal muscle oxidative capacity and capillary density.⁶⁷ In exercise capacity, both HIIT and continuous moderate to high-intensity lower limb aerobic exercise significantly improved Wpeak and 6-MWT compared to UC and SP. At the same time, LICT only significantly improved Wpeak compared to UC.⁶⁸ Furthermore, Latimer et al. proposed that whole-body and muscle mitochondrial responses to exercise were robust in healthy individuals, evident in healthy old individuals, but deficient in patients with COPD.⁶⁹ This partly explains the poor performance of the COPD patients to sustain an HICT throughout the whole training program. It supports that improving exercise capacity necessitates a certain level of exercise intensity in patients with COPD.^68,70 However, MICT showed superior efficacy to HICT in enhancing Wpeak, indicating that training intensity may not be the sole determinant of training outcomes in individuals with COPD.⁷¹ Several studies proposed that high-intensity exercise could impair extradiaphragmatic respiratory muscle perfusion in patients with COPD.^71,72 Similarly, HIIT provides appropriate rest during high-intensity training, demonstrating the highest exercise improvement ability. Additionally, respiratory challenges during HICT may impede patient participation, rendering higher-intensity exercises unsuitable for individuals with significant comorbidities.⁷³

In addition, current meta-analysis indicated that HIIT and MICT significantly improve VE, mMRC, and CRQ compared to UC. Butcher et al. proposed that exercise training generally enhances VE by promoting the restoration of elasticity following hyperventilation.⁷⁴ The lack of improvement in VE in the HICT group may be attributed to an excessive burden on the respiratory muscles.⁷⁵ Similarly, this negative effect was observed with high-volume/intensity resistance training. For example, single-limb low-load/high-repetition-resistance training did not further increase mean 6-MWT compared with two-limb low-load/high-repetition-resistance training, but it reduced exertional dyspnea and enabled more people to reach clinically relevant improvements in 6-MWT.⁷⁶ In addition, various lower limb aerobic exercises may not effectively improve FEV₁pred%, suggesting that incorporating pulmonary function-focused training may yield superior effects for patients with COPD.⁷⁷ Furthermore, patients with COPD suffer from respiratory issues that act as a barrier to future planning, substantially impacting the quality of life.⁷⁸ Research indicates that HIIT and MICT, rather than LICT, are feasible for patients with COPD and can comprehensively improve their quality of life.⁷⁹ While the impact of HICT on mMRC and CRQ remains uncertain due to the lack of supporting studies, further comparisons across different intensities are still needed.

The study has certain limitations. First, the intensity classification is based on previous research. However, this classification may be debatable due to the absence of a universally accepted standard for exercise intensity in patients with COPD based on different research protocols. Additionally, the specificity of the metabolic data collection equipment influenced the decision to focus on lower limb aerobic training. It remains uncertain whether the findings from lower limb aerobic training can be extrapolated to training in other anatomical regions. Meanwhile, caution is warranted when interpreting the comparison results, as they may partially violate the assumptions of consistency and transitivity, and there is an increased risk of bias inherent to the NMA. Finally, the experimental results are limited by the small sample size, and further high-quality studies are necessary to validate the experimental results in the future, which is essential for developing precise exercise programs.

Conclusion

HIIT and MICT might be the optimal training approaches for patients with COPD. HICT did not demonstrate superior performance compared to MICT, potentially leading to more dyspnea. LICT showed only limited improvement. Optimal cost-effectiveness is not achieved with either excessive or insufficient training volumes. Furthermore, the limitations of the current study underscore the necessity for future research to validate these findings.

Supplemental Material

sj-docx-1-tar-10.1177_17534666251323190 – Supplemental material for Optimal intensity and type of lower limb aerobic training for patients with chronic obstructive pulmonary disease: a systematic review and network meta-analysis of RCTs

Supplemental material, sj-docx-1-tar-10.1177_17534666251323190 for Optimal intensity and type of lower limb aerobic training for patients with chronic obstructive pulmonary disease: a systematic review and network meta-analysis of RCTs by Zhengtong Qiao, Ziwei Kou, Jiazhen Zhang, Daozheng Lv, Xuefen Cui, Dongpan Li, Tao Jiang, Xinjuan Yu and Kai Liu in Therapeutic Advances in Respiratory Disease

Supplemental Material

sj-docx-2-tar-10.1177_17534666251323190 – Supplemental material for Optimal intensity and type of lower limb aerobic training for patients with chronic obstructive pulmonary disease: a systematic review and network meta-analysis of RCTs

Supplemental material, sj-docx-2-tar-10.1177_17534666251323190 for Optimal intensity and type of lower limb aerobic training for patients with chronic obstructive pulmonary disease: a systematic review and network meta-analysis of RCTs by Zhengtong Qiao, Ziwei Kou, Jiazhen Zhang, Daozheng Lv, Xuefen Cui, Dongpan Li, Tao Jiang, Xinjuan Yu and Kai Liu in Therapeutic Advances in Respiratory Disease

Footnotes

Acknowledgements

None.

Declarations

ORCID iD

Kai Liu

Supplemental material

Supplemental material for this article is available online.

References

Venkatesan

GOLD COPD report: 2024 update. Lancet Respir Med 2024; 12: 15–16.

Fabbri

Celli

Agustí

, et al. COPD and multimorbidity: recognising and addressing a syndemic occurrence. Lancet Respir Med 2023; 11: 1020–1034.

Maclagan

Croxford

Chu

, et al. Quantifying COPD as a risk factor for cardiac disease in a primary prevention cohort. Eur Respir J 2023; 62(2): 2202364.

Murray

CJL

. The global burden of disease study at 30 years. Nat Med 2022; 28: 2019–2026.

Shu

Lee

Tsai

, et al. The ability of physical activity in reducing mortality risks and cardiovascular loading and in extending life expectancy in patients with COPD. Sci Rep 2021; 11: 21674.

Ward

Plumptre

Fraser-Pye

, et al. Understanding the effectiveness of different exercise training programme designs on V˙O(2peak) in COPD: a component network meta-analysis. Thorax 2023; 78: 1035–1038.

Spruit

Rochester

Pitta

, et al. Pulmonary rehabilitation, physical activity, respiratory failure and palliative respiratory care. Thorax 2019; 74: 693–699.

Priego-Jiménez

Torres-Costoso

Guzmán-Pavón

, et al. Efficacy of different types of physical activity interventions on exercise capacity in patients with chronic obstructive pulmonary disease (COPD): a network meta-analysis. Int J Environ Res Public Health 2022; 19: 14539.

Gloeckl

Pitta

Nyberg

Optimising upper-limb exercise in patients with COPD: another step towards personalised pulmonary rehabilitation?

ERJ Open Res 2024; 10: 01012–2023

10.

Jamnick

Pettitt

Granata

, et al. An examination and critique of current methods to determine exercise intensity. Sports Med (Auckland, NZ) 2020; 50: 1729–756.

11.

Zainuldin

Mackey

Alison

JA.

Optimal intensity and type of leg exercise training for people with chronic obstructive pulmonary disease. Cochr Datab Syst Rev 2011; 2011: Cd008008.

12.

Jacob

Sharma

, et al. Effects of high-intensity interval training protocols on blood lactate levels and cognition in healthy adults: systematic review and meta-regression. Sports Med (Auckland, NZ) 2023; 53: 977–991.

13.

Coates

Joyner

Little

, et al. A perspective on high-intensity interval training for performance and health. Sports Med (Auckland, NZ) 2023; 53: 85–96.

14.

Louvaris

Chynkiamis

Spetsioti

, et al. Greater exercise tolerance in COPD during acute interval, compared to equivalent constant-load, cycle exercise: physiological mechanisms. J Physiol 2020; 598: 3613–3629.

15.

Priego-Jiménez

Cavero-Redondo

Pascual-Morena

, et al. Effect of different exercise programs on lung function in people with chronic obstructive pulmonary disease: a network meta-analysis of RCTs. Ann Phys Rehabil Med 2024; 67: 101792.

16.

Paneroni

Vitacca

Venturelli

, et al. The impact of exercise training on fatigue in patients with chronic obstructive pulmonary disease: a systematic review and meta-analysis. Pulmonology 2020; 26: 304–313.

17.

Wang

Feng

, et al. Effects of exercise-based pulmonary rehabilitation on severe/very severe COPD: a systematic review and meta-analysis. Ther Adv Respir Dis 2023; 17: 17534666231162250.

18.

Hutton

Salanti

Caldwell

, et al. The PRISMA extension statement for reporting of systematic reviews incorporating network meta-analyses of health care interventions: checklist and explanations. Ann Intern Med 2015; 162: 777–784.

19.

Liberati

Altman

Tetzlaff

, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. PLoS Med 2009; 6: e1000100.

20.

Casaburi

Patessio

Ioli

, et al. Reductions in exercise lactic acidosis and ventilation as a result of exercise training in patients with obstructive lung disease. Am Rev respiratory disease 1991; 143: 9–18.

21.

Kaminsky

Knyazhitskiy

Sadeghi

, et al. Assessing maximal exercise capacity: peak work or peak oxygen consumption? Respir Care 2014; 59: 90–96.

22.

Brignardello-Petersen

Bonner

Alexander

, et al. Advances in the GRADE approach to rate the certainty in estimates from a network meta-analysis. J Clin Epidemiol 2018; 93: 36–44.

23.

Wei

Guo

Shang

, et al. Effects of different interventions on smoking cessation in chronic obstructive pulmonary disease patients: a systematic review and network meta-analysis. Int J Nurs Stud 2022; 136: 104362.

24.

Cao

Zhang

, et al. Reporting and methodological quality of COVID-19 systematic reviews needs to be improved: an evidence mapping. J Clin Epidemiol 2021; 135: 17–28.

25.

Guyatt

Oxman

Akl

, et al. GRADE guidelines: 1. Introduction-GRADE evidence profiles and summary of findings tables. J Clin Epidemiol 2011; 64: 383–394.

26.

Pitre

Mah

Helmeczi

, et al. Medical treatments for idiopathic pulmonary fibrosis: a systematic review and network meta-analysis. Thorax 2022; 77: 1243–1250.

27.

Page

McKenzie

Bossuyt

, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ (Clinical research ed) 2021; 372: n71.

28.

Sabapathy

Kingsley

Schneider

, et al. Continuous and intermittent exercise responses in individuals with chronic obstructive pulmonary disease. Thorax 2004; 59: 1026–1031.

29.

Nyberg

Saey

Martin

, et al. Acute effects of low-load/high-repetition single-limb resistance training in COPD. Med Sci Sports Exer 2016; 48: 2353–2361.

30.

Hsieh

Lan

Chen

, et al. Effects of high-intensity exercise training in a pulmonary rehabilitation programme for patients with chronic obstructive pulmonary disease. Respirology (Carlton, Vic) 2007; 12: 381–388.

31.

Brunton

Barbour

Gelinas

, et al. Lower-limb resistance training reduces exertional dyspnea and intrinsic neuromuscular fatigability in individuals with chronic obstructive pulmonary disease. J Appl Physiol (Bethesda, Md : 1985) 2023; 134: 1105–1114.

32.

Covey

Collins

Reynertson

, et al. Resistance training as a preconditioning strategy for enhancing aerobic exercise training outcomes in COPD. Respir Med 2014; 108: 1141–1152.

33.

Klijn

van Keimpema

Legemaat

, et al. Nonlinear exercise training in advanced chronic obstructive pulmonary disease is superior to traditional exercise training. A randomized trial. Am J Respir Crit Care Med 2013; 188: 193–200.

34.

Clark

Cochrane

Mackay

Low intensity peripheral muscle conditioning improves exercise tolerance and breathlessness in COPD. Europ Respir J 1996; 9: 2590–2596.

35.

Baltasar-Fernandez

Losa-Reyna

Carretero

, et al. Residual effects of 12 weeks of power-oriented resistance training plus high-intensity interval training on muscle dysfunction, systemic oxidative damage, and antioxidant capacity after 10 months of training cessation in older people with COPD. Scand J Med Sci Sports 2023; 33: 1661–1676.

36.

Baumann

Kluge

Rummel

, et al. Low intensity, long-term outpatient rehabilitation in COPD: a randomised controlled trial. Respir Res 2012; 13: 86.

37.

Zambom-Ferraresi

Cebollero

Gorostiaga

, et al. Effects of combined resistance and endurance training versus resistance training alone on strength, exercise capacity, and quality of life in patients with COPD. J Cardiopulm Rehabil Prev 2015; 35: 446–453.

38.

de Castro

Felcar

de Carvalho

, et al. Effects of land- and water-based exercise programmes on postural balance in individuals with COPD: additional results from a randomised clinical trial. Physiotherapy 2020; 107: 58–65.

39.

Vasilopoulou

Papaioannou

Kaltsakas

, et al. Home-based maintenance tele-rehabilitation reduces the risk for acute exacerbations of COPD, hospitalisations and emergency department visits. Eur Respir J 2017; 49 2017/05/27. DOI: 10.1183/13993003.02129-2016.

40.

Hill

Jenkins

Philippe

, et al. High-intensity inspiratory muscle training in COPD. The Eur Respir J 2006; 27: 1119–1128.

41.

Probst

Kovelis

Hernandes

, et al. Effects of 2 exercise training programs on physical activity in daily life in patients with COPD. Respir Care 2011; 56: 1799–1807.

42.

Vallet

Ahmaïdi

Serres

, et al. Comparison of two training programmes in chronic airway limitation patients: standardized versus individualized protocols. Eur Respir J 1997; 10: 114–122.

43.

Boeselt

Nell

Lütteken

, et al. Benefits of high-intensity exercise training to patients with chronic obstructive pulmonary disease: a controlled study. Respir Inter Rev Thorac Dis 2017; 93: 301–310.

44.

Gimenez

Servera

Vergara

, et al. Endurance training in patients with chronic obstructive pulmonary disease: a comparison of high versus moderate intensity. Arch Phys Med Rehabil 2000; 81: 102–109.

45.

Göhl

Linz

Schönleben

, et al. [Benefits of a multimodular outpatient training program for patients with COPD]. Pneumologie (Stuttgart, Germany) 2006; 60: 529–536.

46.

Puente-Maestu

Sánz

, et al. Effects of two types of training on pulmonary and cardiac responses to moderate exercise in patients with COPD. Eur Respir J 2000; 15: 1026–1032.

47.

Brønstad

Tjonna

Rognmo

, et al. Aerobic exercise training improves right- and left ventricular systolic function in patients with COPD. Copd 2013; 10: 300–306.

48.

Kortianou

Nasis

Spetsioti

, et al. Effectiveness of Interval Exercise Training in Patients with COPD. Cardiopulm Phys Ther J 2010; 21: 12–19.

49.

Borghi-Silva

Mendes

Trimer

, et al. Potential effect of 6 versus 12-weeks of physical training on cardiac autonomic function and exercise capacity in chronic obstructive pulmonary disease. Eur J Phys Rehab Med 2015; 51: 211–221.

50.

Wang

Zeng

, et al. Cycle ergometer and inspiratory muscle training offer modest benefit compared with cycle ergometer alone: a comprehensive assessment in stable COPD patients. Int J Chronic Obstruct Pulmon Dis 2017; 12: 2655–2668.

51.

Larson

Covey

Wirtz

, et al. Cycle ergometer and inspiratory muscle training in chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1999; 160: 500–507.

52.

Varga

Porszasz

Boda

, et al. Supervised high intensity continuous and interval training vs. self-paced training in COPD. Respir Med 2007; 101: 2297–2304.

53.

Vogiatzis

Nanas

Roussos

Interval training as an alternative modality to continuous exercise in patients with COPD. Eur Respir J 2002; 20: 12–19.

54.

Santos

Rodrigues

Santos

, et al. Pulmonary rehabilitation in COPD: effect of 2 aerobic exercise intensities on subject-centered outcomes: a randomized controlled trial. Respir Care 2015; 60: 1603–1609.

55.

Nasis

Vogiatzis

Stratakos

, et al. Effects of interval-load versus constant-load training on the BODE index in COPD patients. Respir Med 2009; 103: 1392–1398.

56.

Puhan

Büsching

Schünemann

, et al. Interval versus continuous high-intensity exercise in chronic obstructive pulmonary disease: a randomized trial. Ann Intern Med 2006; 145: 816–825.

57.

Arnardóttir

Boman

Larsson

, et al. Interval training compared with continuous training in patients with COPD. Respir Med 2007; 101: 1196–1204.

58.

Coppoolse

Schols

Baarends

, et al. Interval versus continuous training in patients with severe COPD: a randomized clinical trial. Eur Respir J 1999; 14: 258–263.

59.

Louvaris

Spetsioti

Kortianou

, et al. Interval training induces clinically meaningful effects in daily activity levels in COPD. Eur Respir J 2016; 48: 567–570.

60.

Mador

Krawza

Alhajhusian

, et al. Interval training versus continuous training in patients with chronic obstructive pulmonary disease. J Cardiopulm Rehabil Prevent 2009; 29: 126–132.

61.

Vogiatzis

Terzis

Nanas

, et al. Skeletal muscle adaptations to interval training in patients with advanced COPD. Chest 2005; 128: 3838–3845.

62.

Alcazar

Losa-Reyna

Rodriguez-Lopez

, et al. Effects of concurrent exercise training on muscle dysfunction and systemic oxidative stress in older people with COPD. Scand J Med Sci Sports 2019; 29: 1591–1603.

63.

Borghi-Silva

Arena

Castello

, et al. Aerobic exercise training improves autonomic nervous control in patients with COPD. Respir Med 2009; 103: 1503–1510.

64.

Maltais

Bourbeau

Shapiro

, et al. Effects of home-based pulmonary rehabilitation in patients with chronic obstructive pulmonary disease: a randomized trial. Ann Intern Med 2008; 149: 869–878.

65.

Jiang

Bao

Wang

, et al. Effect of aerobic exercises in different intensities on patients with mild to moderate chronic obstructive pulmonary disease. Chin J Rehabil Med 2012; 27: 120–124.

66.

Moncion

Rodrigues

Wiley

, et al. Aerobic exercise interventions for promoting cardiovascular health and mobility after stroke: a systematic review with Bayesian network meta-analysis. Br J Sports Med 2024; 58: 392–400.

67.

Morris

Walsh

Adams

, et al. Exercise training in COPD: What is it about intensity? Respirology 2016; 21: 1185–1192.

68.

Jenkins

Groenen

MTJ

Vaes

, et al. Baseline dependent minimally important differences for clinical outcomes of pulmonary rehabilitation in people with COPD. Pulmonology 2024; 30: 24–33.

69.

Latimer

Constantin-Teodosiu

Popat

, et al. Whole-body and muscle responses to aerobic exercise training and withdrawal in ageing and COPD. Eur Respir J 2022; 59:

70.

Gao

Huang

Wang

, et al. Effects of high-intensity interval training on pulmonary function and exercise capacity in individuals with chronic obstructive pulmonary disease: a meta-analysis and systematic review. Adv Ther 2022; 39: 94–116.

71.

Louvaris

Rodrigues

Dacha

, et al. High-intensity exercise impairs extradiaphragmatic respiratory muscle perfusion in patients with COPD. J Appl Physiol (Bethesda, Md : 1985) 2021; 130: 325–341.

72.

Rodrigues

Louvaris

Dacha

, et al. Differences in respiratory muscle responses to hyperpnea or loaded breathing in COPD. Med Sci Sports Exerc 2020; 52: 1126–1134.

73.

Morris

Hill

Walsh

, et al. Exercise & sports science australia (ESSA) position statement on exercise and chronic obstructive pulmonary disease. J Sci Med Sport 2021; 24: 52–59.

74.

Butcher

Jones

RL.

The impact of exercise training intensity on change in physiological function in patients with chronic obstructive pulmonary disease. Sports Med (Auckland, NZ) 2006; 36: 307–325.

75.

Troosters

Janssens

Demeyer

, et al. Pulmonary rehabilitation and physical interventions. Eur Respir Rev 2023; 32: 220222.

76.

Nyberg

Martin

Saey

, et al. Effects of Low-load/high-repetition resistance training on exercise capacity, health status, and limb muscle adaptation in patients with severe COPD: a randomized controlled trial. Chest 2021; 159: 1821–1832.

77.

Nymand

Hartmann

Ryrsø

, et al. Exercise adaptations in COPD: the pulmonary perspective. Am J Physiol Lung Cell Mol Physiol 2022; 323: L659–l666.

78.

Hurst

Skolnik

Hansen

, et al. Understanding the impact of chronic obstructive pulmonary disease exacerbations on patient health and quality of life. Eur J Intern Med 2020; 73: 1–6.

79.

Neunhäuserer

Reich

Mayr

, et al. Impact of exercise training and supplemental oxygen on submaximal exercise performance in patients with COPD. Scand J Med Sci Sports 2021; 31: 710–719.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

1.10 MB

0.03 MB