Efficacy and Safety of Traditional Chinese Medicine in Treating Chronic Obstructive Pulmonary Disease with Osteoporosis: A Systematic Review and Meta-Analysis

Abstract

This comprehensive meta-analysis on the efficacy and safety of traditional Chinese medicine (TCM) as an adjunctive therapy to conventional treatment of chronic obstructive pulmonary disease (COPD) with osteoporosis (OP) aimed to provide evidence and guidance for clinical practice. We systematically searched four international databases and four Chinese databases to identify relevant randomized controlled trials (RCTs) published up to November 2025. Meta-analysis and sensitivity analysis were performed in R using the meta and gemtc packages. The revised Cochrane risk of bias tool (ROB2) was employed to assess the quality of included studies to ensure their rigor and clinical applicability. A total of 11 eligible studies involving 944 patients were identified. The results indicated that compared with conventional Western medical therapy (CWMT) alone, CWMT plus TCM demonstrated a certain degree of improvement in the primary outcomes, namely bone mineral density (BMD) (standardized mean difference [SMD] = 0.71, 95% confidence interval [CI]: 0.34–1.07, I² = 83.8%, P_HEW < 0.001) and forced expiratory volume in 1 second (FEV1) (SMD = 0.54, 95% CI: 0.17–0.90, I² = 76.5%, P_HEW = 0.007). Statistically significant improvements were observed in some secondary outcomes. The original studies did not adequately report adverse events. The available data only suggested short-term tolerability. According to network meta-analysis results, among the three treatment approaches, the TLKRBS (tonifying the lung and kidney and resolving blood stasis) approach had the highest probability of being ranked first for increasing BMD (surface under the cumulative ranking curve [SUCRA] = 98.82%) and FEV1 (SUCRA = 99.99%), indicating a substantial effect on improvement. The combination of CWMT and TCM may yield more favorable outcomes compared to CWMT alone for treating COPD with OP. The meta-analysis results indicate that the therapeutic efficacy of TCM combined with CWMT may be improved to a certain extent compared with CWMT alone. However, the certainty of the evidence is limited due to small sample sizes, high heterogeneity, and risks of bias. Future high-quality RCTs with rigorous designs, larger sample sizes, and standardized outcome measures are warranted to strengthen the evidence base for the application of TCM in patients with COPD and comorbid OP and to further confirm its efficacy.

Introduction

Chronic Obstructive Pulmonary Disease (COPD) is a common, preventable, and treatable chronic airway disease primarily manifesting as persistent airflow limitation and associated respiratory symptoms.¹ With a global prevalence of approximately 10.3%,^2–5 COPD affects nearly 391.9 million individuals.⁶ Globally, it ranks third among the top causes of fatality, contributing to approximately 3.2 million deaths annually.⁷ Patients with COPD frequently develop multisystem comorbidities, including cardiovascular and musculoskeletal disorders, severely impairing quality of life and elevating the risk of mortality.⁸ Osteoporosis (OP) is a common extrapulmonary comorbidity of COPD. Epidemiological data have indicated a 38% prevalence of COPD with comorbid OP.⁹ Different prevalence rates are reported in studies from different countries. For example, a South Korean survey showed that the prevalence of OP among COPD patients was approximately 17.7%,¹⁰ while a study of male COPD patients in Japan showed a prevalence of OP of approximately 38.8%.¹¹ However, both studies indicate that COPD patients are significantly more likely to develop OP than the general population without COPD, consistently suggesting that OP is a comorbid condition that cannot be overlooked among COPD patients. Furthermore, COPD and OP share many risk factors, including advanced age, smoking, low body mass index (BMI), reduced physical activity, malnutrition, and vitamin D deficiency. Factors in COPD patients, including systemic inflammation, oxidative stress, reduced physical activity, and glucocorticoid use, contribute to the development or exacerbation of OP. Meanwhile, OP patients typically exhibit metabolic disturbances and dysregulation of immune-related inflammatory cytokines, which in turn aggravate the progression of COPD.^12,13 In addition, some patients undergo repeated courses of systemic glucocorticoids or long-term inhaled corticosteroid therapy, which may further compromise their bone health. Thus, COPD and OP share some common risk factors and pathophysiological mechanisms, forming an intertwined vicious cycle that significantly increases the overall disease burden.

Current management of COPD with comorbid OP primarily follows guidelines for the respective specialty, often requiring long-term, multi-drug combination treatment. In older patients, the risks of drug interactions and adverse reactions rise with diminished physiological reserves and multiple comorbidities. Moreover, the onset and progression of COPD are closely linked to environmental and behavioral risk factors. Smoking is the most significant preventable risk factor, while secondhand smoke, occupational exposure to dust and fumes, and outdoor air pollution can all increase the risk of developing COPD. Therefore, the long-term management of COPD and its associated comorbidities not only relies on medication but must also include comprehensive interventions such as smoking cessation, reduction of environmental and occupational exposure, and pulmonary rehabilitation. Given the potential risks and limitations of conventional long-term polypharmacy, it is of practical significance to explore safe and effective adjunctive interventions that address multiple pathological mechanisms.

Traditional Chinese Medicine (TCM), emphasizing holistic regulation in the management of comorbidities, may offer supplementary therapeutic approaches for COPD with comorbid OP. A network pharmacology study has suggested that TCM formulas, particularly Bu Fei Yi Shen (lung-nourishing and kidney-tonifying) prescriptions, may contribute to the management of COPD with comorbid OP through synergistic effects involving multiple components, targets, and pathways.¹⁴ An animal study has further demonstrated that therapeutic approaches, such as Bu Fei Jian Pi (lung-nourishing and spleen-strengthening) approach, Bu Fei Yi Shen (lung-nourishing and kidney-tonifying) approach, and Yi Qi Zi Shen (qi-tonifying and kidney-nourishing) approach, can improve lung function and alleviate airway inflammation in the COPD rat model while enhancing femoral bone mineral density (BMD) and bone mineral content.¹⁵ A clinical study has also reported that the Bu Shen Yi Fei (kidney-nourishing and lung-tonifying) formula reduces the frequency of acute exacerbation, alleviates symptoms, improves lung function, and optimizes bone metabolism indicators in COPD patients with comorbid OP.¹⁶ However, systematic reviews of multiple TCM interventions for COPD and comorbid OP remain lacking. Existing clinical studies provide insufficient evidence-based support for clinical decision-making due to small sample sizes, inconsistent outcome measures, and the absence of quantitative synthesis of efficacy and safety data.

Network meta-analysis (NMA), as an advanced evidence-based research method, is capable of quantitatively integrating direct and indirect evidence with a common comparator based on Bayesian or frequentist models. It can probabilistically rank the efficacy and safety of multiple interventions within the evidence network of the disease.¹⁷ Therefore, our study sought to systematically review and assess the efficacy and safety of TCM treatment for COPD with comorbid OP through an NMA approach. By integrating multidimensional evidence and identifying better protocols, this study was intended to provide potential evidence-based support for clinicians to develop more effective integrated Chinese and Western medicine treatment strategies.

Methods

This study was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) Extension for Network Meta-Analyses and the PRISMA Extension for Chinese Herbal Medicines^18,19 (Supplementary Table S1) and was prospectively registered with the International Prospective Register of Systematic Reviews (PROSPERO) (CRD420251250502) to ensure transparency of the study and reliability of results.

Inclusion and Exclusion Criteria

Inclusion criteria were as follows: (i) Study design: Randomized controlled trial (RCT); (ii) Study population: Studies enrolling patients diagnosed with confirmed stable COPD and comorbid OP, with no restrictions on patient age, sex, disease duration, or severity; (iii) Interventions: Studies in which the observation group received TCM combined with conventional Western medical therapy (CWMT), while the control group received CWMT alone; (iv) Outcome measures: Studies reporting BMD and forced expiratory volume in 1 second (FEV1) as primary outcomes, and FEV1/forced vital capacity (FVC) ratio, FEV1% predicted, alkaline phosphatase (ALP), serum calcium (Ca), and COPD assessment test (CAT) scores before and after treatment as secondary outcomes.

Exclusion criteria comprised: (i) Non-RCTs, studies with abstracts only, and review articles; (ii) Studies in which the control group received TCM or integrated Chinese and Western medicine treatment; (iii) Studies with unavailable data, unreliable data sources, or duplicates; (iv) For duplicate or serial studies published by the same author(s) or research team, or for the same study reported repeatedly at different time points, only the study with the longest treatment or follow-up duration and the most complete data was included. (v) Studies with unclear outcome measures.

Search Strategy

Computerized databases were searched from the inception of each database to November 30, 2025, for published RCTs on TCM for the treatment of COPD with comorbid OP. These databases included PubMed, Cochrane Library, Embase, Web of Science, China National Knowledge Infrastructure (CNKI), China Biology Medicine disc (CBM), Wanfang Data (WF), and Chinese Scientific Journals Database (VIP). The search was limited to studies published in English or Chinese. Besides, relevant literature in this field and reference lists from previously assessed studies were manually searched to identify additional potentially relevant articles. The initial search strategy, as presented in Supplementary Table S2, was tailored to specific databases.

Literature Screening

EndNote 2025 was utilized for organizing and managing retrieved articles. Two researchers conducted literature screening and preliminary data extraction independently based on predefined eligibility criteria. First, duplicate articles were removed from the search results. Subsequently, studies clearly ineligible were excluded through title and abstract screening. Full texts of potentially eligible articles were obtained and evaluated to determine their eligibility for final inclusion.

During the literature screening process, two researchers (Xulun Li and Huiqing Zhang) independently recorded the reasons for inclusion or exclusion and performed a cross-check after screening. Any discrepancies were first resolved by reviewing the original article and engaging in thorough discussion. If discrepancies persisted, a third researcher (Bin Wang) with a background in evidence-based medicine made the final decision. The entire screening process adhered to the relevant requirements of the PRISMA guidelines, as illustrated in Figure 1.

Figure 1.

Literature screening flowchart.

Data Extraction

Two researchers (Xulun Li and Huiqing Zhang) independently extracted data from included articles. After data extraction, they compared and cross-checked their data extraction forms. Any disagreements were addressed through thorough discussion and reference to the original article until consensus was reached. If disagreements persisted, they were adjudicated by a third researcher (Bin Wang) with a background in evidence-based medicine. Extracted data included first author, publication year, sample size of included studies, patient age, elements of risk of bias assessment, intervention measures, outcome measures, and treatment duration. After data extraction, researchers verified the drug composition, primary therapeutic approaches, and the basis for pattern differentiation as described in the original studies for all included formulas. The classification of network nodes was not based solely on the names of the formulas, but was primarily based on the therapeutic principles and core efficacies explicitly described in the original studies. Two researchers with backgrounds in clinical research on TCM drugs (Xulun Li and Huiqing Zhang) independently completed the classification. Any disagreement was resolved through consultation with a third researcher (Bin Wang). Formulas were assigned to the same node only if their primary therapeutic approach aligned with the core therapeutic principle of the corresponding category. If a formula possessed multiple therapeutic efficacies, it was classified based on the primary therapeutic approach explicitly stated in the original study and the core therapeutic direction reflected by the monarch-minister drug combination, rather than on the basis of study results or effect sizes. It should be emphasized that this classification reflects similarities in terms of TCM therapeutic approaches and does not imply that different formulas within the same node are completely equivalent in terms of chemical composition, pharmacological actions, or therapeutic effects.

Finally, the researchers categorized the interventions into three distinct treatment groups: tonifying the lung and kidney and diffusing the lung qi (TLKDLQ) group, tonifying the lung, spleen, and kidney (TLSK) group, and tonifying the lung and kidney and resolving blood stasis (TLKRBS) group.

For continuous outcomes, changes from baseline to the end of treatment were used for quantitative pooling. If the original study reported mean changes and their standard deviations (SDs), these data were extracted directly. If the original study did not report the changes before and after treatment, the mean change was calculated by subtracting the baseline mean from the post-treatment mean. If the SD of the change was not reported, it was calculated based on the pre- and post-treatment SDs and the correlation coefficient between the pre- and post-treatment measurements. For studies requiring an estimate of the correlation coefficient, a correlation coefficient of r = 0.5 was prespecified. For the outcome BMD, data were standardized during subsequent processing due to differences in measurement sites and equipment across studies. Given the differences in the measurement scales used for some outcomes across the studies, other continuous outcome measures were also standardized to mitigate potential measurement bias, and standardized mean differences (SMD) and 95% confidence intervals (CIs) were used to pool the effect sizes.

Quality Assessment

The risk of bias was assessed by two independent researchers (Xulun Li and Huiqing Zhang) for each included study using the revised Cochrane risk of bias tool for randomized trials (RoB2).²⁰ RoB2 is an authoritative, reliable, and widely recommended tool for assessing risk of bias in RCTs in medical practice. This tool specifically covered risk of bias in five domains: randomization process, deviations from intended interventions, missing outcome data, measurement of the outcome, and selection of the reported result. The deviations from intended interventions domain was further categorized into two aspects based on the study objective: the effect of intervention assignment and the effect of adherence to intervention. Each domain was assessed and classified as low risk, some concerns, or high risk. Any discrepancies were finally resolved by a third researcher (Bin Wang) with a background in evidence-based medicine.

The certainty of evidence for primary outcomes was evaluated using the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) tool, which covers risk of bias, inconsistency, indirectness, imprecision, and publication bias. The certainty of evidence was categorized as high, moderate, low, or very low. Given that this study involved an NMA, particular emphasis was placed on the indirectness of the evidence, clinical heterogeneity among the treatments being compared, and the rationality of the transitivity assumption. Because no closed loop was present in the network, it was not possible to empirically assess the consistency between direct and indirect evidence.

Data Analysis

A traditional meta-analysis was performed using the “meta” package in R. Treatment effects of the TCM + CWMT group versus the CWMT alone group were assessed using SMD and 95% CI. Statistical heterogeneity was visually assessed using forest plots and quantified via the chi-square test. The threshold for significance was set at P < 0.10. The I² statistic was used to measure the degree of heterogeneity. A fixed-effects model (FEM) was employed when the heterogeneity was low (I² < 50% and P > 0.1); otherwise, a random-effects model (REM) was adopted. When significant heterogeneity was detected, sensitivity analyses, subgroup analyses, and meta-regression analysis were performed to identify sources of heterogeneity. For outcomes with a sufficient number of studies (generally ≥ 10), funnel plots and Egger’s regression test were used to assess small-study effects. The trim-and-fill method was additionally applied as an exploratory analysis where appropriate.

An NMA was performed within the Bayesian framework of R using the “gemtc” package interfaced with JAGS (version 4.3.1). SMDs and corresponding 95% credible intervals (CrIs) were calculated to compare the efficacy of different Chinese herbal formulas with different therapeutic principles. A fixed-effect Bayesian NMA was performed to compare the relative efficacy of the three TCM treatment categories across the available outcomes. Posterior samples were generated using the mtc.run function, with 10,000 warm-up iterations per chain followed by 40,000 sampling iterations across four concurrently running Markov Chain Monte Carlo chains. Model convergence was evaluated using the Brooks-Gelman-Rubin diagnostics by calculating the potential scale reduction factor (R-hat), supplemented by visual inspection of convergence diagnostic plots. Treatment ranking probabilities were estimated, and the area under the cumulative ranking curve (SUCRA) values were calculated to facilitate comparative assessments of the efficacy of Chinese herbal formulas across different treatment groups. Because no closed loop was present in the network, it was not possible to assess the consistency between direct and indirect evidence using methods such as node splitting. Meanwhile, due to insufficient reporting of key variables, the transitivity component could only be evaluated qualitatively based on a limited number of study-level effect modifiers.

Results

Literature Screening Results

Initially, 484 records were collected. After removing 88 duplicate records, the remaining 396 records underwent title and abstract screening. During the initial screening, 380 records were excluded for the following reasons: disease type not meeting the inclusion criteria (n = 78), animal studies (n = 68), review articles (n = 97), intervention not including TCM treatment (n = 17), non-clinical trials (n = 61), and conference abstracts, experience summaries, case reports, or commentary articles (n = 59). Subsequently, full-text retrieval was conducted for 16 potentially relevant studies, of which 1 was excluded due to the unavailability of full text. The remaining 15 studies underwent full-text assessment. After further review, 1 study reporting duplicate data and 3 studies with subjects not meeting the inclusion criteria were excluded. Ultimately, 11 studies were determined as eligible and were incorporated into this systematic review.^16,21–30

Basic Characteristics of Included Studies

Among the 11 RCTs included in this systematic review, a total of 944 patients with COPD and comorbid OP were randomly assigned. The intervention group comprised 473 patients, and the control group comprised 471 patients. Participant age was reported in 11 studies, with mean ages ranging from 61.3 to 74.2 years in the intervention and control groups. The intervention group received TCM combined with CWMT, while the control group received CWMT alone. Among the included trials, dosage forms comprised decoctions (n = 10) and granules (n = 1), with herbal treatment durations ranging from 4 to 48 weeks. Researchers categorized the herbal formulas based on their composition and therapeutic principles into three groups: TLKDLQ group, TLSK group, and TLKRBS group. An NMA was conducted to identify the potentially most effective and safe TCM therapy (Table 1). The basic characteristics of the patients are shown in Table 2.

Table 1.

Basic Characteristics of Included Studies

Number	First Author and Year	Therapeutic Method	Composition	Conventional Treatment Drugs	Diagnostic Criteria for OP	Were Smoking Cessation, Pulmonary Rehabilitation, or Other Measures Used as Adjuncts to Treatment	Outcome Measures	Safety Evaluation	Time
1	Wang 202025	Tonifying the spleen, lung, and kidney while resolving blood stasis and dissolving phlegm	Huang Qi, Wu Wei Zi, Zi Su Zi, Shu Di Huang, Shan Yu Rou, Hou Pu, Dang Shen, Fu Ling, Long Gu, Huang Jing, Shan Yao, Mu Li, Yin Yang Huo, Di Long, Gu Sui Bu, and Dan Shen	Tiotropium bromide powder for inhalation, Caltrate D, alendronate sodium tablets	DEXA T score (≤ −2.5)	NA	BMD, FEV1/FVC, FEV1%,FEV1, CAT	No adverse events occurred	12 weeks
2	He 201723	Nourishing the kidney, diffusing the lung qi, dispelling wind, and unblocking meridians	Sang Ji Sheng, Chuan Xiong, Qin Jiao, Du Zhong, Du Huo, Bai Hua She, Fang Feng, Mu Xiang, Jie Geng, Bai Zhu, and Gan Cao	Calcium carbonate D3 tablets, salmon calcitonin nasal spray (specific drugs for basic treatment of COPD not listed)	DEXA T score (≤ −2.5)	Yes: a balanced diet, outdoor activities, exposure to sunlight, and smoking and drinking cessation were recommended	BMD, Ca	No adverse events occurred	8 weeks
3	Cai 201526	Clearing the lung, nourishing the kidney, resolving blood stasis, and strengthening the bones	Shan Yao, Zi He Che, Wu Wei Zi, Shan Yu Rou, Mai Dong, Xu Duan, Huang Qi, Bu Gu Zhi, Chuan Xiong, Shu Di Huang, Gu Sui Bu, Dan Shen, and Fu Ling	Calcium carbonate D, calcitriol, alendronate sodium tablets (specific drugs for basic treatment of COPD not listed)	DEXA T score (≤ −2.5)	NA	BMD, FEV1/FVC, FEV1%,ALP, FEV1	No adverse events occurred	12 weeks
4	Zhang 202416	Tonifying the kidney and lung, resolving phlegm and dispersing blood stasis	Sheng Di Huang, Yin Yang Huo, Nv Zhen Zi, Tu Si Zi, Shan Yu Rou, and Huai Niu Xi	Umeclidinium bromide and vilanterol trifenatate powder for inhalation, alendronate sodium tablets	DEXA T score (≤ −2.5)	NA	BMD, FEV1, CAT, FEV1/FVC,FEV1%	No adverse events occurred	12 weeks
5	He 202429	Tonifying the kidney, strengthening the bones, nourishing the lung, and fortifying the spleen	Ren Shen, Xing Ren, Huang Qi, Wu Wei Zi, Fu Ling, Bai Zhu, Ma Huang, Bai Guo, Yin Yang Huo, Gou Qi, Shu Di Huang, Gu Sui Bu, Chuan Xiong, Du Zhong, and Gan Cao	Salbutamol sulfate for inhalation, alendronate sodium tablets	DEXA T score (≤ −2.5)	NA	BMD, FEV1/FVC, ALP, FEV1	NA	12 weeks
6	Zhang 201628	Warming the kidney, strengthening the bones, tonifying qi, and fortifying the spleen	Rou Cong Rong, Shan Yu Rou, Sang Ji Sheng, Du Zhong, Huang Qi, Gan Cao, Bai Zhu, Tao Ren, Dan Shen, and Yin Yang Huo	Calcium carbonate D3 tablets, salmon calcitonin	DEXA T score (≤ −2.5)	NA	BMD, Ca, FEV1	NA	24 weeks
7	Chen 201927	Warming the kidney, nourishing the lung, and resolving blood stasis	Sheng Di Huang, Ze Xie, Shan Yu Rou, Huang Qi, Gui Zhi, Mu Dan Pi, Bai Zhu, Shan Yao, Fang Feng, Fu Ling, and Fu Zi	Budesonide and formoterol fumarate for inhalation, calcium carbonate D, alendronate sodium	DEXA T score (≤ −2.5)	NA	BMD, FEV1/FVC, Ca	NA	48 weeks
8	Fei 200430	Tonifying the kidney, fortifying the spleen, and nourishing the lung	Shu Di Huang, Zi He Che, Shan Yu Rou, Mai Dong, Shan Yao, Wu Wei Zi, Fu Ling, Bai Zhu, Huang Qi, Sang Bai Pi, and Dang Shen	Cefadroxil capsules, aminophylline sustained-release tablets, Gai Er Su	DEXA T score (≤ −2.5)	NA	BMD, ALP	NA	4 weeks
9	Niu 202424	Nourishing the lung and kidney, resolving phlegm, unblocking meridians, and alleviating cough and panting	Jin Fei Cao, Di Long, Shan Yao, Chuan Bei Mu, Shu Di Huang, Ge Jie Fen, Yin Yang Huo, and Tian Jiang Ke	Tiotropium bromide powder for inhalation, calcitriol capsules	DEXA T score (≤ −2.5)	NA	BMD, FEV1/FVC, FEV1%, ALP, CAT	No results reported	12 weeks
10	Xie 201722	Tonifying the kidney and qi while activating blood circulation and strengthening the bones	Yin Yang Huo, Gu Sui Bu, Shu Di Huang, Huang Gua Zi, Huang Qi, Mu Er, and Dan Shen	Calcium carbonate D, alfacalcidol (specific drugs for basic treatment of COPD not listed)	DEXA T score (≤ −2.5)	Yes: smoking cessation was recommended	BMD, FEV1/FVC, FEV1%	No results reported	24 weeks
11	Gao 201621	Tonifying qi, fortifying the spleen, and nourishing the lung and kidney	Huang Qi, Zhi Gan Cao, Ren Shen, Dang Gui, Bai Shao, Chen Pi, Sheng Ma, Chai Hu, Bai Zhu, Shu Di Huang, Da Huang, Tao Ren, and Hong Hua	NA	DEXA T score (≤ −2.5)	NA	BMD, FEV1/FVC, FEV1	No results reported	8 weeks

Abbreviations: 2/9: TLKDLQ, Tonifying the Lung and Kidney and Diffusing the Lung Qi 1/5/6/8/11; TLSK, Tonifying the lung, spleen, and kidney 3/4/7/10: TLKRBS, Tonifying the Lung and Kidney and Resolving Blood Stasis COPD, chronic obstructive pulmonary disease; DEXA, Dual-energy x-ray absorptiometry; BMD, bone mineral density; FEV1, forced expiratory volume in 1 second; NA, not available; CAT, COPD assessment test; Ca, serum calcium; ALP, alkaline phosphatase.

Table 2.

The Basic Characteristics of the Patients

Number	First Author and Year	COPD Grade	Baseline BMD (g/cm2) (I/C)	Baseline FEV1 (L) (I/C)	Baseline BMI (kg/m2) (I/C)	BMD measurement site	Was a Post-Bronchodilator Pulmonary Function Test Performed to Confirm the Diagnosis of COPD?	ICS Use	Sample Size and Sex (I/C) (M/F)	Age (I/C)	Other Comorbidities
1	Wang 202025	GOLD grade 1/2/3	0.721±0.075/0.721±0.087	1.39±0.37/1.39±0.34	NA	Lumbar spine/femoral neck	Y	NA	33 (17/16)/33 (18/15)	67.74±7.23/67.09±7.10	NA
2	He 201723	NA	0.477±0.064/0.479±0.063	NA	NA	Lumbar spine L2–L4	NA	NA	62(NA)/60(NA)	62.4±4.6/61.3±3.8	NA
3	Cai 201526	NA	0.629±0.031/0.631±0.033	1.28±0.32/1.27±0.33	NA	N/A	NA	Patients on long-term ICS therapy were excluded	45 (25/20)/45 (24/21)	71.00±6.21/73.44±5.32	NA
4	Zhang 202416	NA	0.57±0.07/0.58±0.09	1.11±0.22/1.09±0.17	NA	Lumbar spine L1–L4	NA	NA	38 (20/18)/38 (23/15)	66.3±11.47/64.5±12.6	NA
5	He 202429	GOLD grade 1/2	0.783±0.107/0.774±0.109	1.05±0.22/1.02±0.17	NA	Lumbar spine L1–L4/femoral neck	NA	NA	38 (20/18)/38 (23/15)	62/61	NA
6	Zhang 201628	NA	0.622±0.514/0.627±0.519	1.29±0.67/1.28±0.70	NA	Lumbar spine L1–L4	NA	Patients on long-term ICS therapy were excluded	42 (31/11)/42 (30/12)	65.70±4.81/64.27±5.23	NA
7	Chen 201927	GOLD grade 2/3	2.961±0.352/2.898±0.287	NA	NA	N/A	NA	NA	30 (15/15)/32 (17/15)	62.39±7.39/62.67±6.94	NA
8	Fei 200430	NA	NA	NA	NA	Lumbar spine L1–L2	NA	NA	92 (81/11)/90 (78/12)	73.5/74.2	NA
9	Niu 202424	NA	0.93±0.14/0.89±0.13	NA	NA	Lumbar spine L2–L4/femoral neck	Y	NA	29 (26/3)/29 (25/4)	67.5±9.3/67.6±9.2	NA
10	Xie 201722	GOLD grade 1/2/3	0.574±0.088/0.582±0.078	NA	NA	NA	NA	NA	40 (21/19)/40 (22/18)	65.3±3.5/63.2±4.1	NA
11	Gao 201621	NA	3.31±0.44/3.29±0.68	1.25±0.51/1.25±0.51	NA	NA	NA	Patients on long-term ICS therapy were excluded	32 (19/13)/32 (17/15)	69.4±8.4/69.2±7.1	NA

Abbreviations: COPD, chronic obstructive pulmonary disease; BMD, bone mineral density; FEV1, forced expiratory volume in 1 second; BMI, body mass index; I/C, intervention/control; ICS, inhaled corticosteroid; M/F, male/female; NA, not available.

Risk of Bias Assessment for Included Studies

The RoB2 tool was applied to assess the risk of bias across 11 included RCTs. Regarding the randomization process, 7 studies employed random number tables or random allocation to ensure that patients were randomly assigned to different treatment groups. Among them, 1 study used sealed envelopes to conceal the allocation sequence and was therefore rated “low risk”. The remaining 6 studies did not mention any design to conceal the allocation sequence and were thus categorized as “some concerns”. Additionally, 4 studies failed to explicitly describe their randomization methods and were also rated “some concerns”. For deviations from intended interventions, only 1 study reported the implementation of a double-blind design, while the others did not report the implementation of blinding. However, the primary outcomes were mostly objective indicators measured by an instrument or laboratory; therefore, even if participants or researchers were not blinded, it was difficult to directly alter the measured values. Furthermore, pulmonary function indicators were obtained through standardized pulmonary function testing. The tests relied to some extent on subject cooperation. However, if the same testing procedure, equipment, and quality control standard were employed for all groups, and there was no evidence that the testers were aware of treatment allocation and thus provided varying degrees of guidance or employed different measurement methods, the likelihood of significant measurement bias resulting from the absence of blinding was limited. Consequently, only 1 study was rated “low risk”, while the remaining 10 were rated “some concerns”. With regard to missing outcome data, although participant dropouts were reported in some studies, the dropout rates were minimal and unlikely to compromise the robustness of the results. Thus, all studies were rated “low risk”. For the measurement of the outcome, all studies employed appropriate measurement techniques. Among them, 1 study clearly described the blinding protocol, while 7 studies did not mention blinding of assessors. However, since the absence of blinding alone was insufficient to categorize a study as having a high risk of bias, and because most outcome measures were objective, we considered the impact of the absence of blinding on the results to be limited and therefore rated these studies as “low risk”. The other 3 studies included outcome measures in the form of questionnaires. Although there was no evidence suggesting that the testers were aware of treatment allocation and thus provided different levels of guidance, the possibility of bias still existed; therefore, these studies were classified as “some concerns”. For the selection of the reported result, none of the included studies mentioned the adoption of alternative measurement methods or the selective reporting of results from multiple analyses. Thus, all studies were rated “low risk”. Ultimately, all 11 studies included in the analysis were rated as “some concerns”. Details are depicted in Figures 2 and 3.

Figure 2.

Summary of risk-of-bias judgments across the included studies.

Figure 3.

Risk-of-bias judgments for individual included studies.

The GRADE assessment evaluated the certainty of evidence for primary outcomes, taking into account network indirectness and transitivity. The overall certainty was predominantly low or very low. Downgrades were primarily associated with methodological limitations, substantial heterogeneity, concerns regarding indirectness or transitivity, imprecision, or potential publication bias. Detailed results are presented in Supplementary Table S3.

Meta-Analysis Results

Traditional Meta-Analysis Results (TABLE 3)

Table 3.

Results of the Traditional Meta-Analysis

Outcome Measure	Subgroup	Number of Studies Included	Sample Size	Effect Size Value	95% CI	I2	P	P for Between-Group Comparison
BMD	Total	11	944	0.71	[0.34, 1.07]	83.80%	p < 0.001	p = 0.1660
BMD	TLSK	5	472	0.4	[0.10, 0.70]	59.70%	p = 0.042
BMD	TLKDLQ	2	118	0.72	[−0.58, 2.02]	91.30%	p = 0.007
BMD	TLKRBS	4	354	1.1	[0.43, 1.78]	86.60%	p < 0.001
FEV1/FVC	Total	8	618	0.74	[0.22, 1.26]	89.90%	p < 0.001	p = 0.1474
FEV1/FVC	TLSK	3	206	0.75	[0.01, 1.49]	85.10%	p = 0.0012
FEV1/FVC	TLKDLQ	1	58	0.05	[−0.46, 0.56]	NA	NA
FEV1/FVC	TLKRBS	4	354	0.9	[0.01, 1.78]	93.60%	p < 0.001
FEV1	Total	6	502	0.54	[0.17, 0.90]	76.50%	p = 0.007	p = 0.0442
FEV1	TLSK	4	290	0.31	[0.01, 0.62]	41.20%	p = 0.1644
FEV1	TLKRBS	2	212	0.96	[0.41, 1.51]	72.60%	p = 0.0559
FEV1%	Total	5	416	1.35	[0.41, 2.29]	94.4%	p < 0.001	p < 0.001
FEV1%	TLSK	1	66	0.04	[−0.44, 0.52]	NA	NA
FEV1%	TLKDLQ	1	58	0.45	[−0.07, 0.98]	NA	NA
FEV1%	TLKRBS	3	292	2.08	[1.54, 2.63]	71.80%	p = 0.029
Ca	Total	3	206	0.71	[0.42, 0.99]	0%	p = 0.5423	p = 0.5423
Ca	TLSK	1	84	0.67	[0.23, 1.11]	NA	NA
Ca	TLKDLQ	1	60	0.94	[0.41, 1.48]	NA	NA
Ca	TLKRBS	1	62	0.54	[0.03, 1.04]	NA	NA
CAT	Total	3	246	−2.12	[−4.34, 0.09]	97.40%	p < 0.001	p < 0.001
CAT	TLSK	1	66	−1.25	[−1.78, −0.72]	NA	NA
CAT	TLKDLQ	1	58	−0.75	[−1.29, −0.22]	NA	NA
CAT	TLKRBS	1	122	−4.38	[−5.04, −3.72]	NA	NA
ALP	Total	4	406	−0.04	[−0.74, 0.66]	89.60%	p < 0.001	0.633
ALP	TLSK	2	258	0.32	[−1.10, 1.74]	96.1%	p < 0.001
ALP	TLKDLQ	1	58	−0.40	[−0.92, 0.12]	NA	NA
ALP	TLKRBS	1	90	−0.38	[−0.80, 0.04]	NA	NA

Abbreviations: BMD, bone mineral density; FEV1, forced expiratory volume in 1 second; FVC, forced vital capacity; Ca, serum calcium; CAT, COPD assessment test; ALP, alkaline phosphatase; TLKDLQ, Tonify the Lung and Kidney and Diffuse the Lung Qi; TLSK, Tonify the Lung, Spleen, and Kidney; TLKRBS, Tonify the Lung and Kidney and Resolve Blood Stasis.

Changes in BMD levels were reported in 11 studies, with 6 studies showing positive results. The meta-analysis suggested that, compared to the CWMT alone group, the TCM + CWMT group showed a significant improvement (increase) in BMD levels (SMD = 0.71, 95% CI: 0.34–1.07, I² = 83.8%, P_HEW < 0.001). Subgroup analysis showed that the TLKRBS group and the TLSK group yielded positive results, while only the TLKDLQ group yielded negative results (Table 3).

Egger’s test did not detect any statistically significant funnel plot asymmetry (P = 0.6927). The trim-and-fill analysis did not identify any potential missing studies that needed to be imputed, and the adjusted pooled effect size was consistent with that of the primary analysis (SMD = 0.71, 95% CI 0.34–1.07). However, given the high heterogeneity among studies, the possibility of publication bias could not be completely ruled out (Supplementary Figure S1).

Changes in FEV1 levels were reported in 6 studies, with 3 (n = 3) showing positive results. The meta-analysis suggested that, compared to the CWMT alone group, the TCM + CWMT group showed a significant improvement (increase) in FEV1 levels (SMD = 0.54, 95% CI: 0.17–0.90, I² = 76.5%, P_HEW = 0.007). Subgroup analysis showed that both TCM treatment groups yielded positive results (Table 3).

Changes in FEV1/FVC ratio were reported in 8studies, with 5 (n = 5) showing positive results. The meta-analysis suggested that, compared to the CWMT alone group, the TCM + CWMT group showed a significant improvement (increase) in FEV1/FVC ratio. SMD = 0.74, 95% CI: 0.22–1.26, I² = 89.9%, P_HEW < 0.001). Subgroup analysis showed that the TLKRBS group and the TLSK group yielded positive results, while only the TLKDLQ group yielded negative results (Table 3).

Changes in FEV1% were reported in 5 studies, with 3 (n = 3) showing positive results. The meta-analysis suggested that, compared to the CWMT alone group, the TCM + CWMT group showed a significant improvement (increase) in FEV1%. SMD = 1.35, 95% CI: 0.41–2.29, I² = 94.4%, P_HEW = p < 0.001). Subgroup analysis showed that the TLKRBS group and the TLSK group yielded positive results, while only the TLKDLQ group yielded negative results (Table 3).

Changes in Ca levels were reported in 3 studies, all showing positive results. The meta-analysis suggested that, compared to the CWMT alone group, the TCM + CWMT group showed a significant improvement (increase) in Ca levels (SMD = 0.71, 95% CI: 0.42–0.99, I² = 0%, P_HEW = 0.5423). Subgroup analysis revealed positive outcomes for TCM therapies across all three subgroups (Table 3).

Changes in ALP levels were reported in 4 studies, with 3 (n = 3) showing positive results. The meta-analysis suggested no statistically significant difference in ALP levels between the TCM + CWMT group and the CWMT alone group (SMD = −0.04, 95% CI: −0.74–0.66, I² = 89.6%, P _HEW< 0.0001) (Table 3).

Changes in CAT levels were reported in 3 studies, with 3 (n = 3) showing positive results. The meta-analysis indicated no statistically significant difference in CAT levels between the TCM + CWMT group and the CWMT alone group (SMD = −2.12, 95% CI: −4.34–0.09, I² = 97.4%, P _HEW< 0.0001) (Table 3).

Exclusion of He et al 2024 materially altered the pooled ALP estimate, changing the result from non-significant to statistically significant and reducing heterogeneity to I² = 0% (Supplementary Figure S2). This indicated that the pooled ALP result was sensitive to this individual study and should therefore be interpreted cautiously. The results for the remaining 6 outcome measures remained stable (Supplementary Figures S3–S8).

Meta-Regression Analysis

Given the high heterogeneity observed for multiple outcomes, we further explored potential sources of heterogeneity. Univariate meta-regression analysis was used to evaluate clinically significant potential effect modifiers, including mean patient age, sex, and treatment duration. Due to the limited number of studies included for baseline FEV₁, baseline BMI, and glucocorticoid exposure, regression analysis could not be performed for these variables. The meta-regression yielded regression coefficients and P-values.

The meta-regression analysis revealed that the mean age of participants was a factor influencing the heterogeneity in the effect size of ALP across the included studies (β = −0.1117, P = 0.0077). As the mean age of participants increased, the pooled effect size of ALP showed a downward trend. The mean age of participants was a potential source of heterogeneity in the effect size of CAT (β = 1.6266, P < 0.0001). As the mean age of participants increased, the pooled effect size of CAT showed an upward trend. However, these findings represented observational associations at the study level and could not be interpreted as indicating that the increase in patient age directly led to enhanced treatment efficacy.

No clear statistical associations were observed for the remaining outcomes, as detailed in Supplementary Table S4. However, due to the small number of studies included in some analyses and missing data for certain variables, the statistical power of the meta-regression was limited. The above negative results could not rule out the possibility that these clinical factors may have contributed to the heterogeneity across studies.

Network Meta-Analysis Results

A Bayesian NMA was further employed to compare the relative efficacy of the three TCM therapies across different outcome measures. CWMT served as the common comparator to form an indirect comparison network of all TCM therapies (Figure 4). The network structure for FEV1 is shown in Figure 5. A fixed-effect Bayesian NMA model was used for the network analyses. All the models converged well (R-hat = 1); We conducted a qualitative evaluation of the transitivity assumption by comparing clinically significant potential effect modifiers across different treatment nodes, including mean patient age, COPD subtype, definition of OP, baseline pulmonary function, and treatment duration. While patient age, the definition of OP, and COPD subtype overlapped to some extent across most treatment comparisons, the studies were not entirely consistent regarding baseline pulmonary function, treatment duration, and background therapy. Furthermore, some studies inadequately reported glucocorticoid exposure. Therefore, although the available data did not indicate clear systematic imbalances across all available characteristics, the transitivity assumption could only be considered reasonable to a certain extent and could not yet be fully validated.

Figure 4.

Network meta-analysis diagram.

Figure 5.

Network meta-analysis diagram for FEV1.

The NMA results revealed that the TLKRBS group exhibited a greater effect on the improvement of BMD than the TLSK group (SMD = 0.63, 95% CrI: 0.34–0.92). Among the markers of pulmonary function, the TLKRBS group exhibited a greater effect on the improvement of FEV1/FVC (SMD = 0.85, 95% CrI: 0.28–1.41), and FEV1% (SMD = 1.64, 95% CrI: 1.04–2.23) than the TLKDLQ group, and a greater effect on the improvement of FEV1% (SMD = 2.05, 95% CrI: 1.49–2.62), and FEV1 (SMD = 0.73, 95% CrI: 0.36–1.09) than the TLSK group, whereas no statistically significant difference was identified between the TLKDLQ group and the TLSK group. For the symptom score (CAT), results pooled with a fixed-effects model similarly indicated a greater effect on improvement in the TLKRBS group. No significant differences in the outcomes ALP and Ca were observed among the three TCM therapies. The NMA comparisons are presented in Supplementary Tables S5–S11.

Based on the above NMA results, among the three treatment approaches, the TLKRBS approach showed the greatest potential for alleviating disease symptoms. However, given that no closed loop was present in the NMA and that indirect comparisons between different categories of TCM treatments were primarily derived from common controls, the relative effects and SUCRA rankings described above only reflected the model-estimated relative ranking probabilities and should be considered exploratory findings regarding the therapeutic efficacy of the different therapies.

Safety Reporting

Regarding safety, the available data were insufficient to support a reliable quantitative meta-analysis. Therefore, the evidence on safety was summarized using a descriptive approach only. None of the 11 included studies explicitly reported any adverse reactions. Some studies, such as the study by He et al (2017), reported that “no significant adverse reactions were observed”. This suggested that TCM combined with CWMT may be tolerated during short-term use.

However, safety reporting in the included studies was generally limited and inconsistent. None of the studies specified predefined definitions of adverse events, and only a few studies vaguely reported that “no significant adverse reactions were observed” or “no serious adverse events occurred”. Moreover, the existing studies varied significantly in the definition of adverse events, observation periods, and reporting completeness. Therefore, the current evidence only suggested that TCM combined with CWMT exhibited safety signals in the short term but was insufficient to confirm overall long-term safety.

Discussion

This study systematically integrated data from existing RCTs, employing traditional pairwise meta-analysis and Bayesian NMA to explore the efficacy and safety of TCM combined with CWMT in treating COPD with comorbid OP. Overall data indicated that compared with CWMT alone, TCM + CWMT appeared to show a certain degree of superiority in increasing BMD, improving pulmonary function (FEV1, etc), and reducing symptom burden. These findings not only provide preliminary evidence-based support for the comprehensive management of these complex comorbidities but also offer a crucial opportunity to delve into the multidimensional pharmacologic mechanisms of TCM in the concurrent treatment of a pulmonary disease and a bone disease.

However, while TCM combined with CWMT showed statistically significant improvements in BMD and some pulmonary function indicators, no significant differences were observed in CAT scores or ALP levels, suggesting that the potential effects may be outcome-specific. The statistically significant improvements in FEV1 and FEV1/FVC observed in this study should be interpreted with caution regarding their clinical significance. The pooled effect size for FEV1 was SMD = 0.54, which fell only within the moderate effect range. SMD reflects the standardized difference relative to within-study variability and cannot be directly converted into an absolute increase in FEV1, nor is it equivalent to the minimum clinically important difference (MCID) perceivable by patients. Similarly, because different pulmonary function indicators have different standard deviations, their SMDs should not be directly compared. A higher SMD for FEV1/FVC does not necessarily imply a greater actual improvement. Given that COPD is characterized by persistent, partially reversible airflow limitation, and that the included trials primarily evaluated the combination of TCM and CWMT, the observed effects may also be influenced by baseline medications, smoking cessation, pulmonary rehabilitation, and other measures, as well as differences in disease severity, baseline pulmonary function, sample size, treatment duration, and the composition and dosage of TCM formulas. Some studies did not adequately report post-bronchodilator measurements, quality control of testing, or the handling of endpoint values or changes, further increasing heterogeneity across measurements and studies. Therefore, the current results are better suited as a statistical signal of potential pulmonary function benefits rather than confirmatory evidence of reversing fixed airflow limitation or achieving clear clinical benefits.

From the perspective of potential mechanisms, the pathogenesis of COPD with comorbid OP is not yet fully understood, but chronic systemic inflammation is considered one of the proposed mechanisms contributing to this comorbid state.^31,32 COPD is a chronic airway inflammatory disease associated with the persistent presence of inflammatory cells and cytokines within the body. Long-term airway inflammation can induce systemic inflammatory responses through the spillover of inflammatory mediators.³¹ Among these, core inflammatory mediators such as interleukin-6 (IL-6) and tumor necrosis factor alpha (TNF-α) not only drive airway remodeling but also directly disrupt the dynamic equilibrium of bone metabolism. By excessively activating osteoclasts and inhibiting the functions of osteoblasts, they ultimately lead to progressive bone loss.^12,33 Furthermore, shared risk factors, including oxidative stress, long-term glucocorticoid use, and reduced physical activity, further contribute to this vicious cycle. In COPD-OP comorbidity, the osteoprotegerin (OPG)/receptor activator of nuclear factor kappa-B (RANK)/receptor activator of nuclear factor kappa-B ligand (RANKL) axis is recognized as a pivotal link between inflammation and bone resorption. Bai et al have discussed the association between COPD-related OP and the disruption of this pathway.³⁴ The clinical study by Zhang et al has also reported abnormal levels of RANKL or OPG in COPD patients with comorbid OP, suggesting a more prominent driving role of bone resorption.³⁵ However, the RCTs included in this analysis did not directly investigate these mechanisms. Therefore, it is not possible to draw definitive conclusions regarding causal mechanisms from the current meta-analysis.

Given the findings of our study, TCM formulas may serve as adjunctive therapies for COPD and comorbid OP. The main pharmacological components in these TCM formulas potentially represent the active ingredients for COPD and comorbid OP. This meta-analysis included 11 TCM herbal formulas, with herbs and their functions listed in Table 1. Among the formulas, 8 were self-devised formulas, while the remaining 3 were classical formulas based on the Chinese Pharmacopoeia. The frequency of each herb used in the TCM formulas is shown in Figure 6. Huang Qi (Radix Astragali), Shu Di Huang (Radix Rehmanniae Preparata), and Bai Zhu (Rhizoma Atractylodis Macrocephalae) are the most commonly used herbs in our meta-analysis. Modern pharmacological research has indicated that astragaloside IV, the core component of Huang Qi, directly stimulates the differentiation and maturation of osteoblasts through the Wnt/β-catenin signaling pathway, accelerating bone formation.³⁶ Besides, it can also protect lung tissue from COPD-induced injury by targeting the guanosine triphosphate (GTP)-guanosine diphosphate (GDP) domain of RAS proteins, thereby downregulating the RAS/RAF/FoxO signaling pathway.³⁷ Meanwhile, total flavonoids of Astragalus significantly suppress inflammatory responses and prevent damage to bone joints by modulating the OPG/RANKL/nuclear factor kappaB (NF-κB) signaling pathway.³⁸ The water extract of Shu Di Huang significantly elevates serum OPG levels, reduces RANKL levels, and interferes with the Wnt/β-catenin signaling pathway, thereby reversing bone loss and increasing bone density.³⁹ The core component of Bai Zhu, atractylenolide III, can potently inhibit oxidative stress and excessive inflammatory responses in the lungs by activating the Nrf2-related antioxidant pathway, thereby alleviating lung tissue injury and fibrotic remodeling. This mechanism aligns closely with the core pathological processes of COPD alleviation.⁴⁰ Meanwhile, atractylenolide III also alleviates inflammation-driven bone and joint degeneration by re-polarizing macrophages and inactivating the Toll-like receptor 4 (TLR4)/NF-κB signaling pathway.⁴¹

Figure 6.

Frequency distribution of traditional Chinese medicine therapies.

Furthermore, among the three TCM treatment regimens, the TLKRBS regimen exhibited greater effects on improving primary outcomes, namely BMD and FEV1. This statistical advantage is grounded in both a foundation of TCM theory and increasingly clarified pharmacological mechanisms. In TCM, the pathogeneses of COPD and OP both fall under the category of Ben Xu Biao Shi (deficiency in the root and excess in the branch). The progression of both diseases is attributed to Zang Fu Xu Sun (deficiency and damage of visceral organs) in the early stage, which subsequently leads to the accumulation of pathological products and ultimately triggers the diseases. Therefore, treatment should not be limited to merely Bu Xu (tonifying deficiency). The clearance of pathological products within the body should also be included. In TCM, the lung governs qi and thus promotes blood circulation. Lung deficiency and qi deficiency weaken the propulsion of blood circulation, leading to a pathological state of blood stasis. Consequently, bone marrow cannot be adequately nourished, leading to bone atrophy. This represents the pathophysiological evolution of COPD and comorbid OP from the perspective of TCM. Therefore, the therapeutic approach of Hua Yu (resolving stasis) is of critical significance for the management of this comorbidity. Modern pharmacological studies have confirmed that active ingredients in blood-invigorating and qi-tonifying herbs commonly used in formulas, such as Dan Shen (Salvia miltiorrhiza Bunge) and Huang Qi, exhibit potent anti-inflammatory, antioxidant, and immunomodulatory activities.^42,43 These ingredients effectively downregulate the expression of pro-inflammatory factors, including TNF-α and IL-6, and interrupt systemic inflammatory cascades, thereby alleviating lung tissue injury at its source and reversing inflammation-induced imbalance of bone metabolism.^44,45 This “tonifying deficiency while resolving stasis” intervention strategy closely aligns with the pathological characteristics of late-stage COPD, namely, inevitable stasis after prolonged illness and qi deficiency with blood stasis, thereby providing a high possibility of improvement as demonstrated by clinical data.

From the perspective of clinical application prospects, the combination of TCM and CWMT may offer distinct advantages in the holistic management of the complex comorbidity of COPD and OP. Compared to traditional management based on guidelines for a single disease separately, TCM treatment emphasizes a holistic perspective and multi-target intervention, which may help improve impaired lung function and abnormal bone metabolism simultaneously.^46,47 It offers a feasible supplementary strategy for comorbidity management by playing an adjunctive role in reducing bone resorption and alleviating symptom burden.

Furthermore, from an epidemiological perspective, it is worth noting that the disease burden of COPD may be underestimated due to underdiagnosis. In a recent nationwide population-based study in Kazakhstan, pulmonary function tests were performed before and after bronchodilator use in 4626 participants with complete data. The results showed that the prevalence of COPD among individuals aged 40 and older was 10.7%, while only about 30% of those diagnosed with COPD had previously been informed that they had the disease.⁴⁸ This suggests that relying solely on previous clinical diagnoses or patient self-reports may significantly underestimate the true burden of COPD. It also highlights that early screening and standardized quality control of pulmonary function testing are crucial for both epidemiological studies and clinical diagnosis of COPD. That study not only used post-bronchodilator pulmonary function testing to confirm COPD but also explicitly evaluated the quality of the pulmonary function tests, with most FEV₁ and FVC measurements rated as high quality. This approach reduced misclassification resulting from poor procedural quality, inadequate patient cooperation, or the absence of post-bronchodilator validation. In contrast, some of the original trials included in this study merely cited COPD diagnostic guidelines without adequately describing post-bronchodilator pulmonary function testing or other quality control procedures, which may have led to heterogeneity in patient classification and the measurement of pulmonary function outcomes.

Limitations

This study has several limitations. First, included studies were limited to two languages only (English and Chinese), which may have introduced language bias. In addition, because comorbidity-related variables were inadequately reported in some original studies, the potential confounding effects of factors such as comorbidities on the meta-analysis results could not be entirely ruled out. The small overall sample size and significant variations in TCM formula composition, dosage, and treatment duration led to high heterogeneity among studies, potentially affecting the stability of the pooled effect sizes. Moreover, inadequate reporting of random sequence generation, allocation concealment, and implementation of blinding in some included studies may have introduced selection bias and performance bias. Safety reporting in the included studies was generally limited and inconsistent. Therefore, the findings on safety should be interpreted with caution, and the current evidence is insufficient to confirm the overall safety of the interventions. Furthermore, the important clinical outcomes of the patients were inadequately evaluated. Existing studies have primarily assessed treatment efficacy using surrogate indicators such as BMD, bone metabolism markers, and pulmonary function indicators, while systematic evaluation of clinical outcomes such as fracture, Fracture Risk Assessment Tool (FRAX) score, fall risk, hospitalization, and mortality remains lacking. Although a few studies reported acute exacerbations of COPD, the relevant data were limited in number, reported inconsistently, and insufficient to provide reliable pooled evidence. Meanwhile, the relatively short durations of treatment and follow-up in some studies limited the evaluation of long-term changes in BMD, fracture risk, and long-term COPD prognosis. Finally, according to the GRADE assessment, the certainty of evidence for most outcomes was low or very low. Even though the point estimates suggested potential benefits, the results remain subject to considerable uncertainty and cannot be regarded as confirmatory evidence of treatment superiority.

Conclusion

In summary, existing evidence suggests that CWMT combined with TCM may play a favorable role in the comprehensive management of COPD with comorbid OP. Particularly, this combination demonstrates potential advantages in improving pulmonary function and bone metabolism-related outcomes. However, the benefits observed in this study are primarily considered improvements in short-term surrogate outcomes, and it is not yet possible to directly infer long-term clinical benefits. To validate these conclusions, future high-quality, long-term RCTs with clear blinding protocols, larger sample sizes, and standardized outcome measures should be conducted. Furthermore, clinical outcomes such as fractures and acute exacerbations should be prioritized to assess the clinical benefits of the interventions. In addition, pharmacological studies should be standardized, and different TCM interventions should be directly compared to clarify the relative efficacy of different treatment approaches and provide more reliable evidence-based support for clinical practice.

Funding Statement

This work was supported by CACMS Innovation Fund, Project Number: CI2021A00601; CACMS Innovation Fund, Project Number: CI2026A02028;Innovative Team for Traditional Chinese Medicine Prevention and Treatment of Complications in Chronic Airway Diseases (NO.WJCXTD-202408).

Declaration of Artificial Intelligence

The author declare that no artificial intelligence was used at the time of writing the paper.

Data Sharing Statements

The datasets used and analysed during the current study are available from the corresponding author Bin Wang on reasonable request.

Ethics Approval and Consent to Participate

Ethical approval and consent to participate were deemed not applicable.

Consent for Publication

Consent for publication was deemed not applicable.

Author Contributions

Xulun Li and Haiqiang Wang contributed equally to this work and should be considered as co-first authors. All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Disclosure

The authors declare that they have no competing interests.