The protocol for this systematic review and meta-analysis was prospectively registered in the International Prospective Register of Systematic Reviews (PROSPERO No: CRD42024614968) (Ni et al., 2024). This study adheres to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement and its extension for Chinese Herbal Medicines (Page et al., 2021; Zhang et al., 2020).

Studies were eligible for inclusion if they met all of the following criteria:

• Study Designs: RCTs were the primary study design.

Studies were excluded if they focused on complications of metabolic diseases (e.g., diabetic nephropathy, hypertensive heart disease), used multiple CHM treatments together as interventions, were conference abstracts or duplicates, or contained serious methodological errors or flaws.

We systematically searched six medical databases to identify relevant literature, including China National Knowledge Infrastructure, China Biomedical database, Wanfang, PubMed, Embase, and Cochrane Library. Our search strategy incorporated both controlled vocabulary and keywords related to metabolic diseases and CHM interventions targeting dampness. To identify additional records, we also reviewed reference lists of included studies, searched for studies cited in systematic reviews of interest, and examined literature about proprietary Chinese medicines for T2DM, hypertension, dyslipidaemia, and obesity listed in the list of National Basic Medical Insurance Medicine, Employment Injury Insurance Medicine, Maternity Insurance Medicine, and Chinese Pharmacopoeia. The search was conducted from the inception of each database to August 2024. No language restrictions were applied. The detailed search strategy is provided in Supplementary Table S1.

The study was selected by removing duplicates using reference management software, screening titles/abstracts independently by two reviewers, assessing full texts for eligibility, resolving discrepancies via discussion with a senior reviewer, and documenting the process in a PRISMA flow diagram.

Data extraction was performed independently by two reviewers using a standardised electronic form designed in EpiData Software (version 3.1, EpiData Association, Odense, Denmark). A third reviewer conducted validation checks to ensure accuracy and completeness of the extracted data. Data items included bibliographic details, study designs, participant characteristics, Chinese medicine syndrome, therapeutic principle, and intervention details, such as CHM formula composition, dosage, administration route, treatment duration, and preparation method, the control, and outcomes. The therapeutic principles of TCM and the ingredients of proprietary Chinese medicines were verified through consultation of official records from the National Medical Products Administration and Chinese Pharmacopoeia. When necessary, additional clarification was sought directly from study authors or through consultation with TCM experts.

Using the Cochrane Collaboration’s risk of bias tool for randomised trials (Higgins et al., 2011), we assessed potential bias across Seven domains, including random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective reporting, and baseline balance. For blinding assessments, we evaluated impact separately for three outcome types, including objective, clinician-reported, and patient-reported outcomes. Discrepancies between independent reviewers were resolved through structured discussion or consultation with a senior reviewer when consensus was not achieved. Each domain was categorised as low, high, or unclear risk of bias based on predefined criteria.

We also evaluated the evidence body for the primary outcomes using the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) approach (Schünemann et al., 2013). The quality of evidence was graded as high, moderate, low, or very low, considering factors such as the risk of bias, imprecision, inconsistency, indirectness, and publication bias.

All meta-analyses and forest plots were generated using the Review Manager software (version 5.3, The Nordic Cochrane Centre, The Cochrane Collaboration, Copenhagen, Denmark). For dichotomous data, we calculated the relative risks (RRs) with 95% confidence intervals (CIs). For continuous outcomes, we used mean differences (MDs) with 95% CIs. All analyses followed the intention-to-treat principle when possible. Clinical and methodological diversity across included studies—key drivers of statistical heterogeneity—was identified a priori. A random-effects model was employed for all analyses to estimate average effect sizes of CHM interventions targeting dampness in metabolic conditions, accommodating between-study variation and ensuring conservative, robust pooled results by accounting for heterogeneity.

Pre-specified subgroup analyses were also systematically conducted to explore primary clinical sources of heterogeneity, focusing on clinically meaningful effect modifiers—including basic treatment regimens and specific types of controls. Analysing representative CHM intervention (e.g., specific decoctions or proprietary formulations), individually served as an additional strategy to address and explain heterogeneity, where fixed-effect models were used when I² was <50%, and random-effects models were used when I² was ≥50%. Publication bias was assessed using a funnel plot and Egger’s test if more than ten RCTs were included in the meta-analysis.

The literature search identified 5,354 relevant records. After removing duplicates and study selection, 122 studies were included to the systematic review including 64 CHM RCTs for T2DM (Chen et al., 2012; Chen et al., 2022; Chen, 2022; Cheng, 2018; Dai, 2022; Deng et al., 2018; Duan et al., 2015; Fan et al., 2017; Fang, 2015; Feng et al., 2016; Fu, 2017; 2021; Ge, 2018; He et al., 2018; He and Li, 2014; Huang and Liu, 2022; Huang and Zhang, 2021; Huang et al., 2017; Jiang, 2023; Ke et al., 2019; Li, 2011; 2018; 2020; Li et al., 2020; Li X. et al., 2021; Liang and Wang, 2016; Luo et al., 2014; Ma, 2016; Ma et al., 2020; Meng et al., 2008; Ni et al., 2021; Pan et al., 2021; Song et al., 2019; Sun, 2018; Tian, 2020; Wang, 2021; Wang et al., 2018; Wang H. Y, 2022; Wang M.K. et al., 2021; Wang Q.Y. et al., 2021; Wang Y, 2022; Wu et al., 2019; Wu, 2021; Xiong, 2019; Xu et al., 2009; 2015; Yan et al., 2019; Yang, 2011; 2016; 2021; Yu et al., 2017; Yu and Chen, 2010; Zeng et al., 2006; Zhang, 2016; Zhang and Cai, 2016; Zhang et al., 2014; Zhang H.F. et al., 2019; Zhang H.J. 2019; Zhang L.N. et al., 2019; Zhang M.Q. 2019; Zheng, 2017; Zhou, 2012; 2020; Zhu, 2018), 28 for hypertension (Cheng et al., 2021; Fang, 2016; Guan and Chen, 2016; Hu, 2015; Huang, 2018; Huang and Li, 2014; Li and Wang, 2021; Lin, 2017; Liu, 2014; 2016; Liu et al., 2007; Lu, 2018; Ma et al., 2019; Mao and Li, 2022; Miao et al., 2017; Pang, 2013; Ren, 2017; 2022; Shi, 2019; Sun et al., 2021; Wang et al., 2023; Wu et al., 2020; Wu, 2019; Xiong, 2010; Yang, 2023; Yuan and Wu, 2016; Zhao et al., 2016; Zhu et al., 2019), 21 for dyslipidaemia (He et al., 2007; Hong, 2007; Hu, 2012; Kong et al., 2008; Li, 2013; Liu and Chen, 2011; Liu et al., 2008; Lu et al., 2017; Lu, 2004; Meng et al., 2004; Pan et al., 2016; Rao et al., 2015; Su et al., 2012; Tan et al., 2006; Wang and Jiao, 2012; Xiao, 2014; Xue, 2015; You, 2015; Yu et al., 2010; Zhao et al., 2021; Zheng et al., 2009) and nine for obesity (Ke et al., 2012; Lenon et al., 2012; Li et al., 2007; Liu et al., 2020; Ruan et al., 2019; Wu, 2016; Yang, 2010; Ye et al., 2016; Yu, 2016). The full process of study screening is shown in Figure 1. The list of excluded references with reasons is attached as Supplementary Table S2.

The included 122 studies were published between 2000 and 2024, of which seven were published in English journals and 103 in Chinese journals and 12 were archived as degree thesis. The total sample size was 11,252, with an average of 92. The participants in the included RCTs were aged between 17 and 86 years, and female accounted for 45.3% of all participants. Regarding study design, all of them were parallel, controlled studies, of which seven were with multiple arms. The most common CM syndrome across four all metabolic conditions was syndrome of phlegm and dampness retained in the internal, and the most common method to assessing a CM syndrome was clinical guiding principles for new TCM drugs. The safety was monitored in 69 studies, of which 32 of them did not identify any adverse events. The details of study characteristics for individual metabolic conditions are shown in Table 1, and the full characteristics of individual studies are shown in Supplementary Tables S3–S4.

Twenty-two studies assessed the independent effect of CHM for metabolic diseases, whereas 100 examined the add-on effect of CHM for the participants with lifestyle modification (16, 16%), pharmacotherapy (37, 37%), or pharmacotherapy plus lifestyle modification (47, 47%). Fifty-nine distinct formulae of CHM decoction potentially treating dampness were identified in the included RCTs, whereas six different proprietary CHM were evaluated. The therapeutic period of CHM intervention ranged from 2 weeks to 2 years.

In terms of disease diagnostic criteria, all the studies on diabetes adopted the internationally recognised standard promulgated by the World Health Organization in 1999, namely, fasting blood glucose ≥7.0 mmol/L and/or 2hPG level ≥11.0 mmol/L (World Health Organization, 1999). However, the diagnostic criteria of other metabolic diseases varied per different time periods and issuing organisation.Vascular impairments and endpoint events, including mortality, cardiac infarction, stroke, and renal failure were not reported in all the included studies.

The most frequently reported CHM formulae across the included studies were Gegen Qinlian Decoction (GQD) (28 studies, for T2DM), Banxia Baizhu Tianma Decoction (BBTD) (20 studies, for hypertension), and Linggui Zhugan Decoction (LZD) (3 studies, for obesity). Notably, no single CHM formula emerged as predominant in the context of dyslipidaemia, likely attributable to the significant heterogeneity of herbal compositions used in CHM decoctions for this condition. Specific to dyslipidaemia, two proprietary Chinese medicines were most extensively evaluated: Jiangzhi Tongmai Granule (5 RCTs) and berberine hydrochloride (5 RCTs). Among the individual Chinese herbs identified across the included studies, the following exhibited the highest frequency of use: Poria (Chinese pinyin: fuling), derived from the sclerotia of Poria cocos (Schw.) Wolf; Pinellia ternata (Chinese pinyin: banxia), obtained from the tubers of Pinellia ternata (Thunb.) Breit.; and Atractylodis Macrocephalae Rhizoma (Chinese pinyin: baizhu), obtained from the rhizome of Atractylodes macrocephala Koidz. The analysis of herbal usage patterns across different metabolic conditions revealed distinct preferences for specific Chinese herbs in their treatments, whereas Poria (Chinese pinyin: fuling), derived from the sclerotia of P. cocos (Schw.) Wolf., emerged as a consistently prominent component across multiple conditions. The details on the top eight frequently used CHM formulae are presented in the Supplementary Table S5.

Of the 122 RCTs included in this review, only three registered their protocols in clinical trial registries (Chen et al., 2012; Lenon et al., 2012; Zhao et al., 2021). The risk of bias for the remaining 119 studies was assessed based on the published reports. Detailed results of the risk of bias assessment for individual metabolic conditions are presented in Figure 2, with full results for each study provided in Supplementary Table S6.

A meta-analysis of 45 RCTs evaluating CHM for eliminating dampness combined with specialised drugs revealed insights into adverse events. Among these, 22 studies reported no adverse events in either the experimental or control group (Fang, 2015; Feng et al., 2016; Guan and Chen, 2016; He et al., 2018; Huang et al., 2017; Li, 2018; 2013; 2011; Liang and Wang, 2016; Lin, 2017; Liu and Chen, 2011; Ma, 2016; Ma et al., 2019; Mao and Li, 2022; Meng et al., 2008; Shi, 2019; Wu, 2019; Xiao, 2014; Xue, 2015; You, 2015; Zhang H.J., 2019; Zhu et al., 2019), whereas 23 studies documented adverse events (Chen et al., 2022; Dai, 2022; Fan et al., 2017; Fu, 2017; Hu, 2015; Huang and Liu, 2022; Huang, 2018; Huang and Zhang, 2021; Jiang, 2023; Lu, 2018; Miao et al., 2017; Pan et al., 2016; Su et al., 2012; Tian, 2020; Wang et al., 2023; Wang Q.Y. et al., 2021; Wu et al., 2019; Wu, 2021; Yang, 2016; 2023; 2021; Zhang M.Q., 2019; Zhou, 2020). A meta-analysis of these 23 studies indicated that adjunctive CHM therapy resulted in fewer adverse events compared with pharmacotherapy alone (RR = 0.56; 95% CI, 0.39–0.82). In direct comparisons between CHM for eliminating dampness and specialised drugs, 24 RCTs reported adverse events. A meta-analysis of 14 of these studies showed no significant difference in adverse event incidence between the two groups (RR = 0.63; 95% CI, 0.27–1.42). The most common adverse events in the CHM group were mild digestive issues, such as nausea, vomiting, and diarrhoea, which typically resolved with or without intervention. Importantly, no studies reported severe adverse events, indicating a favourable safety profile for CHM in this context. Figure 7 presents the forest plots on the meta-analysis of adverse events, and Supplementary Table S7 details the individual adverse events.

In this systematic review, several CHM formulae exhibited notable therapeutic effects across different disease conditions:

• GQD (Chen et al., 2022; Cheng, 2018; Dai, 2022; Fan et al., 2017; Feng et al., 2016; Fu, 2017; Ge, 2018; Huang and Liu, 2022; Huang and Zhang, 2021; Jiang, 2023; Li, 2018; 2020; Ni et al., 2021; Sun, 2018; Tian, 2020; Wang, 2021; Wu, 2021; Xiong, 2019; Xu et al., 2015; Yang, 2021; Zeng et al., 2006; Zhang, 2019; Zhang and Cai, 2016; Zhang H.F. et al., 2019; Zhang L.N. et al., 2019; Zheng, 2017; Zhou, 2012; 2020; Zhu, 2018) for T2DM: Most evaluated in included RCTs, this decoction showed both alternative and add-on effects in reducing FPG levels and exhibited add-on effects on post-treatment HOMA-IR and 2hPG levels in participants using hypoglycaemic agents (Figure 8).

GRADE assessment of evidence quality for CHM in metabolic conditions ranged from very low to moderate. Across metabolic disorders, evidence certainty for adjunctive CHM plus conventional therapy was consistently higher than for CHM monotherapy, with obesity as the sole exception. For individual CHM formulae, evidnce certainty was very low to low, attributable to two trial profiles: small-scale, high-quality RCTs, or large-scale trials hampered by substantial methodological limitations. The full GRADE evidence profile was attached as Supplementary Table S8.

In our study, the meta-analysis suggested that monotherapy with CHM significantly reduced FPG levels in T2DM, although heterogeneity in CHM formulations limits generalisability. For 2hPG, preliminary evidence suggests CHM monotherapy may be effective, but limited placebo-controlled trials hinder conclusive findings. Postprandial regulation is critical in T2DM, as fluctuations increase glycaemic burden and stress, causing β-cell dysfunction and complications, and impaired first-phase insulin response is an early T2DM feature (Shibib et al., 2024). Robust evidence for the role of CHM in 2hPG control requires standardised formulations and large-scale placebo-controlled trials. Overall, CHM shows promise as an adjunct therapy for glycaemic control. When combined with metformin, TCM demonstrates potential to enhance glycaemic control for both FPG and 2hPG. However, outcomes vary when CHM is paired with other antidiabetic agents, underscoring the complexity of TCM interactions in polypharmacy regimens.

The meta-analysis in our study also revealed that CHM might improve FINS levels when used in combination with lifestyle interventions, although evidence remains limited and inconclusive. CHM does not demonstrate superiority over conventional pharmacotherapy in this context. When added to existing pharmacotherapy in patients with diabetes, CHM may further enhance FINS improvement, with the strongest evidence supporting its combination with metformin. However, in patients already receiving insulin therapy, CHM appears to offer no additional benefit for FINS enhancement. The effect of CHM on FINS in the absence of lifestyle interventions remains uncertain owing to limited studies and variability in CHM formulations. Regardless of lifestyle adjustments, CHM appears to enhance the effects of pharmacotherapy on HOMA-IR. However, similar to FINS, CHM provides no additional benefit for HOMA-IR improvement in patients already treated with insulin. Elevated FINS level indicates β-cell compensation or failure, whereas elevated HOMA-IR is strongly associated with metabolic syndrome and early stages of T2DM (Muniyappa et al., 2008). The findings suggest that CHM may enhance glycaemic control by supporting β-cell function and improving insulin sensitivity, although its role in severe diabetes remains limited.

GQD, the most evaluated CHM for resolving dampness in included RCTs for diabetes, shows both alternative and add-on effects in reducing FPG, with add-on effects on post-treatment HOMA-IR and 2hPG levels in hypoglycaemic agent users, which aligned with a previous systematic review (Tan et al., 2023). Its effectiveness is dose-dependent, particularly regarding Coptidis rhizoma (Chinese pinyin: huanglian), as validated by recent RCTs (Kang et al., 2024). GQD lowers blood glucose levels by improving insulin resistance and enhancing insulin sensitivity (Gao et al., 2017; Lu et al., 2021). It modulates gut microbiota, enriching beneficial bacteria, such as Faecalibacterium, to reduce hyperglycaemia and inflammation (Gao et al., 2024). GQD contains active components, such as puerarin, baicalin, and berberine, which exhibit antioxidant, anti-inflammatory, and hypoglycaemic effects by targeting pathways, such as Nrf2 and PI3K/Akt (Lu et al., 2021; Xu et al., 2020). Additionally, GQD inhibits hepatic ferroptosis to reduce oxidative stress and improve iron metabolism and regulates the gut flora–bile acid–TGR5 axis, offering novel anti-diabetic mechanisms (Bao et al., 2022; Liu et al., 2024). These findings highlight GQD’s potential as a multifunctional agent in T2DM management, warranting further exploration of its clinical applications.

Extremely limited evidence from our meta-analysis suggested that CHM did not demonstrate superiority over placebo in reducing SBP when used as an adjunct to lifestyle modifications. Furthermore, CHM as a monotherapy was inferior to RAS inhibitors. However, TCM may augment the SBP-lowering effects of pharmacotherapy, particularly CCBs and RAS inhibitors, with more pronounced reductions observed when combined with lifestyle adjustments, achieving the minimal clinically important difference of 10 mmHg (Ettehad et al., 2016; Thomopoulos et al., 2014). Notably, CHM does not further reduce SBP in patients requiring dual antihypertensive therapy. These findings highlight the potential adjunctive role of CHM in SBP management but underscore its limitations as monotherapy or in combination with dual antihypertensive agents. DBP is closely tied to vascular tone and resistance in smaller arteries and arterioles (Tomiyama, 2023), and elevated DBP is more clinically relevant in younger populations, where vascular elasticity is better preserved (Bello et al., 2020). In our meta-analysis, we found that CHM demonstrated modest superiority over placebo in reducing DBP when combined with lifestyle modifications, and CHM consistently enhanced the DBP-lowering effects of pharmacotherapy, regardless of whether patients were receiving monotherapy or dual antihypertensive regimens. Notably, the additive effects of CHM on DBP highlight its potential to optimise antihypertensive management in patients with preserved vascular elasticity, or with impaired peripheral small arteries.

BBTD, the most evaluated CHM in included RCTs for hypertension, demonstrated a significant add-on effect in lowering SBP and DBP in patients on antihypertensive medications but lacked alternative effects on blood pressure reduction, being consistent with previous evidence (Mohammed et al., 2023). This can be explained by its multi-target and multi-pathway regulatory mechanisms. First, these mechanisms may be supported by its active components, such as flavonoids and triterpenoids, which act on key targets, such as AKT1, NOS3, and ACE, thereby influencing vascular tone and blood pressure regulation (Lin et al., 2022). Additionally, it promotes potassium efflux through potassium channels and inhibits calcium influx via voltage-operated calcium channels and intracellular calcium release from the sarcoplasmic reticulum (Tan et al., 2018). It also helps mitigate oxidative stress and inflammation, which are critical in hypertension pathogenesis. Specifically, BBTD enhances nitric oxide (NO) production via the NO/sGC/cGMP pathway and prostaglandin-I-2 synthesis, leading to vasodilation (Jin et al., 2024).

Previous evidence suggests that a 0.4 mmol/L increase in HDL-C level and a 1 mmol/L reduction in LDL-C and TG levels are associated with a lower risk of cardiovascular events (Emerging et al., 2009; Marston et al., 2019; Patel and Giugliano, 2020). In our meta-analysis, the effects of CHM for resolving dampness on TG, LDL-C, and HDL-C levels were inconclusive owing to the limited number of studies, inconsistent results, and modest effect sizes. Although CHM showed potential to enhance the effects of lipid-lowering drugs, its impact was insufficient to meaningfully alter cardiovascular risk. Notably, Jiangzhi Tongmai Capsule and Berberine, the most extensively studied proprietary CHM agents for dyslipidaemia, align with these observations. These findings highlight the need for further high-quality studies to clarify the role of CHM mainly targeting dampness syndrome in lipid management.

Obesity measured using BMI was strongly associated with an increased risk of cardiovascular diseases (CVDs), CVD mortality, and all-cause mortality, and the risk was enhanced per 1 unit increase in BMI (Dwivedi et al., 2020). In our study, the meta-analysis suggested that CHM eliminating dampness outperformed placebo in reducing BMI while improving central obesity indicators, such as WC, requiring integration with lifestyle interventions.

LZD most assessed in included RCTs for obesity has demonstrated an additive effect on reducing BMI when combined with lifestyle modifications, such as dietary restriction and physical exercise. This synergistic effect is attributed to LZD’s multifaceted pharmacological mechanisms against obesity. LZD restores glucose homoeostasis and enhances insulin sensitivity while modulating serum lipid profiles and intestinal lipid content, particularly through significant alterations in diacylglycerol and monoacylglycerol levels (Li et al., 2024). Notably, LZD upregulates key thermogenesis-related factors, including uncoupling protein 1, PR domain containing 16, peroxisome proliferator-activated receptor gamma coactivator 1-alpha, and peroxisome proliferator-activated receptors alpha and gamma, in white adipose tissue (Li et al., 2024). This promotes adipose tissue ‘browning’ and increases energy expenditure. Animal studies further support LZD’s efficacy, showing that its combination with dietary restriction and exercise alleviates high-fat diet-induced metabolic complications, including obesity, hyperglycaemia, hyperlipidaemia, and insulin resistance, potentially through downregulation of tumor necrosis factor-alpha, leptin, and protein kinase B (Sun et al., 2022). Although these findings highlight LZD as a promising therapeutic strategy for obesity and metabolic health improvement, high-quality RCTs are required to validate its clinical efficacy and establish it as a viable adjunctive therapy.

In the included RCTs, Poria (Chinese pinyin: fuling), derived from the sclerotia of P. cocos (Schw.) Wolf., emerged as a consistently prominent component across multiple metabolic conditions. Overall, the multifaceted pharmacological profile of Poria cocos positions it as a valuable therapeutic agent for metabolic diseases. Its active components, including triterpenoids and polysaccharides, play pivotal roles in these therapeutic effects. Triterpenoids in Poria cocos restore glucose homeostasis and enhance insulin sensitivity, crucial for diabetes management (Guo et al., 2025). These compounds also modulate serum lipid profiles and intestinal lipid content, particularly influencing diacylglycerols and monoacylglycerols, thus aiding in dyslipidaemia treatment (Li et al., 2023). The polysaccharides in Poria cocos are notable for their role in gut microbiota modulation, which is closely associated with metabolic health benefits, including improved glucose metabolism and reduced inflammation associated with obesity (Lai et al., 2025). Additionally, Poria cocos exhibits anti-inflammatory and antioxidant properties, which help mitigate chronic inflammation, a key factor in obesity and related metabolic disorders (Li et al., 2022). Its diuretic activity contributes to hypertension management by promoting sodium and water excretion, reducing blood volume and pressure (Lee et al., 2012). Furthermore, Poria cocos may protect against obesity-induced renal damage through its antioxidant and anti-inflammatory effects (Nie et al., 2020). Although clinical evidence and pharmacological studies have demonstrated the benefits of Poria cocos for metabolic disorders, further studies are required to clarify its precise molecular mechanisms, particularly involving triterpenoids and polysaccharides. Additionally, investigating how Poria cocos modulates gut microbiota and its subsequent effects on metabolic health may reveal insights into its therapeutic potential.

While most outcome measures in RCTs of metabolic diseases are objective (e.g., lipid and glucose profiles), which are unlikely to be influenced by blinding status, outcome assessors should still remain blinded for clinician-reported outcomes, such as blood pressure and BMI. This will ensures a greater precision in effect estimation. The existing evidence supports the potential benefits of dampness-eliminating CHM for metabolic diseases, but it does not exclude the possibility of supplementary effects when combined with other therapeutic principles in a multi-herbal formula. Vascular impairments and endpoint events were not reported in any included studies, thus the long-term benefits of CHM in individuals with metabolic diseases requires further investigation.