Diabetes, a chronic metabolic disease prevalent worldwide, is marked by persistently elevated blood glucose levels. This condition results from either inadequate insulin secretion or diminished insulin sensitivity. As per the International Diabetes Federation (IDF), the global adult population affected by diabetes exceeded 536.6 million in 2021, and this figure is expected to rise in the future (1). Diabetes has profound effects on patients’ quality of life and can lead to several complications, such as cardiovascular disease, nephropathy, retinopathy, and neuropathy. These complications present significant challenges to public health (2).

Currently, traditional diabetes treatment strategies primarily include lifestyle interventions (dietary control, exercise therapy) (3) and pharmacological interventions (oral hypoglycemic agents such as metformin, sulfonylureas, thiazolidinediones, SGLT2 inhibitors, GLP-1 RAs, DPP-4 inhibitors, as well as insulin injections, etc.) (4). Although these methods have achieved certain efficacy in glycemic control, they still face numerous challenges and limitations in clinical application. Long-term medication may induce a series of side effects, such as hypoglycemia risk (especially with sulfonylureas and insulin therapy), weight gain, gastrointestinal discomfort (metformin, GLP-1 receptor agonists), genitourinary tract infections (SGLT2 inhibitors), and potential hepatorenal toxicity (5–9). Furthermore, poor long-term treatment adherence is also a widespread issue; complex medication regimens, frequency of drug administration, frequent blood glucose monitoring, and the discomfort associated with injection therapy all impact treatment effectiveness (10–13). More critically, diabetes and its complications impose a substantial economic burden on patients and society, including high drug costs, frequent medical visits, hospitalization expenses, and disability and reduced work capacity due to complications (14). These challenges and limitations in clinical practice underscore the urgent need for effective, safe, economical, and patient-friendly alternative or complementary therapies.

Against this backdrop, traditional Chinese medicine (TCM), with its millennia-long history and unique theoretical framework, offers valuable insights and perspectives for diabetes management. Accumulating clinical and preclinical evidence, particularly in recent years, continues to highlight the potential advantages of TCM interventions in diabetes care. Firstly, unlike single-target Western drugs, TCM formulas—typically comprising multiple herbs—exert synergistic effects on multiple pathways involved in glucose metabolism, insulin resistance, beta-cell function, inflammation, and oxidative stress. This multi-target action aligns with TCM’s holistic philosophy and may address diabetes’ complex pathophysiology more comprehensively (15, 16). Critically, modern phytochemical research has identified a plethora of bioactive metabolites isolated from TCM herbs that underpin these therapeutic effects. Key compounds such as berberine (from Coptis chinensis), astragaloside IV (from Astragalus membranaceus), ginsenosides (from Panax ginseng), and polyphenols (e.g., from Quinoa) have demonstrated significant anti-diabetic properties in mechanistic studies. These include enhancing insulin sensitivity, promoting β-cell regeneration, and suppressing inflammatory cascades (17–19). This scientific validation of active constituents provides a molecular basis for TCM’s efficacy and bridges traditional knowledge with modern pharmacology. Secondly, TCM employs individualized treatment through Syndrome Differentiation and Treatment (A core principle of TCM that involves identifying syndromes based on clinical manifestations and formulating corresponding therapies). Diagnosis classifies diabetic patients into distinct patterns (e.g., Yin Deficiency with Dryness-Heat, Qi and Yin Deficiency, Spleen Deficiency with Dampness), enabling customized herbal prescriptions, potentially yielding better personalized outcomes (20–22). Thirdly, TCM demonstrates potential for reducing complications. Specific herbs and formulas show protective effects against diabetic nephropathy, retinopathy, and neuropathy in preclinical and clinical studies, primarily through anti-inflammatory, antioxidant, and microcirculation-improving mechanisms (23–25).Finally, systematic reviews and meta-analyses indicate TCM interventions, when properly administered, exhibit a relatively favorable safety profile. They are associated with lower incidence of adverse events—particularly severe hypoglycemia and gastrointestinal issues—compared to conventional hypoglycemic agents. This suggests suitability for long-term management or adjunctive therapy (26, 27).

However, despite its promising prospects, the application of TCM in diabetes remains challenging. Clinical evidence remains heterogeneous due to differences in study design (e.g., sample size, duration of treatment, and control settings), TCM formulations (standardization, batch variability), and populations (ethnicity, region, and syndrome type). Furthermore, the complex multi-component and multi-target nature of TCM poses significant challenges to elucidating its precise mechanisms of action using traditional single-target approaches, which, to a certain extent, limits its wider understanding and acceptance. In order to more systematically evaluate clinical efficacy, overcome the limitations of individual studies, and deeply explore its complex mechanism of action, this study employs an integrated strategy combining meta-analysis and network pharmacology. This methodological choice is critical: Meta-analysis provides a rigorous quantitative synthesis of existing randomized controlled trial (RCT) data to derive more precise and generalizable estimates of TCM’s overall clinical efficacy and safety profile in DM management (27, 28), overcoming the limitations of individual studies and establishing robust clinical evidence. Meanwhile, network pharmacology offers a powerful systems biology framework to systematically predict and analyze the interactions between bioactive TCM components, their potential targets, and the associated biological pathways and networks involved in DM pathogenesis (29, 30), uniquely suited to decipher the complex, multi-target mechanisms underlying TCM’s therapeutic effects.

Therefore, the objective of this study is to comprehensively assess the effectiveness and safety of TCM for treating diabetes mellitus, and to investigate its mechanism of action through the integration of meta-analysis and network pharmacology. We will analyze the advantages and shortcomings of TCM in diabetes management by reviewing relevant literature and experimental studies, providing scientific basis and new ideas for comprehensive treatment of diabetes, and promoting the application and development of TCM in modern medicine.

Diabetes is a global chronic metabolic condition with a rising incidence, presenting significant health risks and economic challenges for patients. Traditional Chinese medicine (TCM) adheres to a holistic view of health, regarding the human body as an organic whole influenced by both internal and external factors. Its various components are interconnected and mutually influential, forming a complex dynamic system. TCM, with its multi-target, multi-component, and multi-pathway approach, offers considerable promise as a comprehensive treatment for diabetes (94). Research indicates that the potential mechanisms by which TCM regulates blood glucose include: improving insulin resistance (95–97), enhancing insulin sensitivity (98–100), promoting insulin secretion (101–103), and stimulating glucose uptake (104–106). Notably, recent studies have found that TCM can also improve glucose metabolism disorders by remodeling the balance of the gut microbiota (101, 107, 108), working through the perspective of the “Gut-Pancreas” axis. However, although existing retrospective studies have confirmed the clinical value of TCM (26, 28), there remain significant gaps in screening the optimal medication regimens and elucidating the molecular mechanisms of action. Therefore, this study aims to provide new scientific evidence for TCM in treating diabetes, in order to promote more precise individualized treatment strategies.

This meta-analysis provides robust evidence that TCM can significantly improve glycemic control (e.g., HbA1c, fasting blood glucose), lipid profile (e.g., total cholesterol, triglycerides, low-density lipoprotein cholesterol), and renal function indicators (e.g., serum creatinine, urinary albumin excretion rate) in diabetic patients. These improvements are directly linked to the pathophysiological mechanisms of core diabetic complications: optimizing glycemic fluctuations reduces microvascular damage (109); regulating dyslipidemia delays atherosclerosis (110); and protecting nephron function lowers the risk of nephropathy (111). This suggests that TCM intervention holds significant clinical value in the comprehensive management of diabetes and its common complications. This conclusion is supported by the consistently observed effects in the included studies and the results obtained after our rigorous assessment of potential biases and heterogeneity.

The synergistic improvements in blood glucose, lipids, and renal function observed in this study possess dual clinical significance. On one hand, they pertain to the restoration of metabolic homeostasis: the simultaneous optimization of blood glucose and lipids may directly delay vascular endothelial damage by reducing glucolipotoxicity (112, 113). On the other hand, they indicate a forward shift in the window for complication prevention: improvements in Scr and BUN suggest TCM may potentially block the “hyperglycemia-glomerular hypertension-renal fibrosis” pathway, offering a novel strategy for the primary prevention of diabetic nephropathy (114). This provides evidence-based support for an integrated approach combining Chinese and Western medicine, aligning with the “individualized metabolic goal management” proposed in the ADA/EASD guidelines.

The meta-analysis results further indicate that Astragali Radix (Huangqi), Codonopsis Radix (Dangshen), Poria (Fuling), Corni Fructus (Shanzhuyu), Coptidis Rhizoma (Huanglian), and Alismatis Rhizoma (Zexie) are the high-frequency herbs used in TCM for diabetes treatment. Relevant studies support their efficacy. For instance, Chao et al. found that a traditional Chinese herbal compound (containing Coptidis Rhizoma, Astragali Radix, and Lonicerae Japonicae Flos) ameliorated insulin resistance in T2D patients and also improved glucose metabolism (including FPG, PPG, and HbA1c) and blood pressure to some extent (115). A multicenter study by Chan et al. demonstrated that adding Astragali Radix to standard treatment significantly slowed the decline in renal function in patients with diabetic nephropathy (116). The mechanisms by which these herbs treat diabetes align with those described previously: by optimizing glycemic fluctuations, regulating dyslipidemia, and protecting renal function to reduce the risk of nephropathy. Specifically, Corni Fructus ameliorates diabetic nephropathy through its effects of lowering blood glucose, regulating lipids, and reducing oxidative stress; its renal protective mechanism is closely associated with activating the PPARγ signaling pathway (117). The aqueous extract of Codonopsis Radix ameliorates insulin resistance (IR) by increasing Akt and GSK-3β phosphorylation, reduces hepatic triglyceride content via AMPK phosphorylation, and protects β-cell function by reducing β-cell apoptosis (118). Furthermore, extracts of Codonopsis Radix and Polygonati Rhizoma (Huangjing) can also ameliorate IR, lower blood glucose, and reduce lipid levels by activating the IRS1/PI3K/AKT pathway (119). For Astragali Radix, its polysaccharide component (APS) promotes GLUT4 translocation and upregulates PPAR-γ expression by activating the AMPK/PI3K/AKT pathway, thereby ameliorating IR. It also possesses anti-inflammatory effects, inhibits pancreatic β-cell apoptosis, and promotes insulin secretion (120). Meanwhile, both the saponin (ASS) and flavonoid (ASF) components of Astragali Radix combat hyperglycemia by activating the adiponectin-AMPK pathway and its downstream factors, although their effect intensities vary across different tissues (121).

Notably, the network pharmacology analysis in this study identified quercetin, kaempferol, and stigmasterol as the core active components of the aforementioned high-frequency TCM herbs (e.g., Astragali Radix, Codonopsis Radix, Corni Fructus). A total of 32 potential targets for TCM in treating diabetes were identified. These active components exert their therapeutic effects through synergistic regulation of core pathological pathways in diabetes. Specifically, quercetin activates the AMPK/PI3K/Akt pathway to enhance GLUT4 translocation efficiency in skeletal muscle, promoting glucose uptake and increase insulin sensitivity (122, 123). It also inhibits the activation of the TNF-α/NF-κB pathway by enhancing Akt phosphorylation, thereby alleviating inflammatory insulin resistance (124, 125). Furthermore, quercetin inhibits ferroptosis by activating the Nrf2 pathway (e.g., upregulating GPX4/xCT) and chelating iron ions. In the kidneys, this manifests as protection of renal tubular mitochondria and reduction of proteinuria. In the pancreas, it improves insulin secretion by reducing iron deposition in β-cells, thereby providing protection against diabetic kidney disease (DKD) and for pancreatic β-cells (126, 127). By activating PPARγ, kaempferol enhances downstream PI3K/AKT signaling activity to promote glucose uptake in peripheral tissues and inhibit hepatic gluconeogenesis, thereby effectively lowering blood glucose and improving glucose tolerance, while ameliorating lipid metabolism disorders and reducing lipotoxicity through activation of the PPARγ/LXRα/ABCA1 pathway (128, 129). Kaempferol also inhibits IKKβ/IKKα phosphorylation and activation, blocking the TNF-α/NF-κB pathway, reducing serum pro-inflammatory cytokines, and increasing IRS-1 protein expression, thereby improving insulin resistance (130). It has also been found to modulate the gut microbiota (e.g., reversing the relative abundance of Firmicutes and Bacteroidetes), which may be related to its amelioration of high-fat diet-induced lipid metabolism abnormalities (131). In terms of renal protection, kaempferol inhibits RhoA/ROCK signaling, downregulates pro-fibrotic factors such as TGF-β1 and CTGF, reduces extracellular matrix (ECM) accumulation, protects podocyte structure, and delays renal fibrosis while improving renal function (132). Stigmasterol acts by increasing the expression of SREBP2 and its target gene LDLR while decreasing the expression of the cholesterol efflux transporter ABCA1. This reduces free cholesterol levels induced by glucolipotoxicity. Combined with its antioxidant effects to lower ROS, stigmasterol restores glucose-stimulated insulin secretion (GSIS) capacity, increases total insulin content, and alleviates pancreatic β-cell dysfunction (133). Additionally, stigmasterol increases GLUT4 translocation and expression, enhancing insulin sensitivity to improve insulin resistance (134).

This association pattern of “TCM - active component clusters - multi-target pathways - synergistic effects” profoundly elucidates the material basis and mechanisms of action underlying TCM in treating diabetes. The core components (quercetin, kaempferol, stigmasterol) identified by network pharmacology and their 32 regulated targets not only validate the efficacy mechanisms of the high-frequency TCM herbs but also reveal, at the molecular level, how different herbs collectively target core pathological aspects of diabetes. This collective action occurs through shared or complementary active components and action pathways, addressing key pathological processes including insulin resistance, glucose and lipid metabolism disorders, inflammatory responses, oxidative stress, β-cell dysfunction, and renal injury. Ultimately, this leads to the synergistic improvement of blood glucose, lipid profile, and renal function. This provides a critical scientific foundation for understanding the holistic effects of TCM formulas and for developing novel therapeutic strategies based on active component clusters.

Beyond the small-molecule active components mentioned above, macromolecular components in TCM (such as polysaccharides and polypeptides) have been widely confirmed to possess significant anti-diabetic activity. Particularly in the core herbs screened in this study, the effects of APS on improving insulin resistance and modulating gut microbiota (120, 135), the hypoglycemic and renal protective effects of Coptidis Rhizoma polysaccharides (CCPW) (136, 137), and the antioxidant and anti-inflammatory activities of Codonopsis Radix polysaccharides (CPPS) and Corni Fructus polysaccharides (COPs) (138, 139) are all significant contributors to their overall therapeutic efficacy. However, it should be noted that the network pharmacology methodology applied subsequently in this study has an inherent design primarily reliant on small-molecule-oriented databases (e.g., TCMSP, PubChem) and prediction tools (e.g., molecular docking). Consequently, it is challenging to systematically analyze the complex mechanisms of action of these macromolecular components. The primary value of this study lies in its focus on the clusters of small-molecule compounds within these same core herbs (e.g., quercetin/kaempferol in Astragali Radix, berberine/palmatine in Coptidis Rhizoma, alisol derivatives in Alismatis Rhizoma). It reveals the potential molecular mechanisms by which they exert anti-diabetic effects through the regulation of key targets (e.g., AKT1, PPARG) and signaling pathways (e.g., TNF signaling pathway). This provides an important perspective for understanding the small-molecule pharmacodynamic material basis of the core herbs. We emphasize that the overall efficacy of high-frequency herbs like Astragali Radix and Codonopsis Radix stems from the synergistic action of their multiple components. This includes both the small-molecule mechanisms predicted in this study (e.g., targeted regulation of inflammatory factors like TNF or the PI3K-Akt pathway) and the well-documented macromolecular components (e.g., the immune/gut microbiota modulatory functions of polysaccharides). Together, they constitute a modern scientific interpretation of TCM’s characteristic “multi-component, multi-target” integrative regulation. Future studies should integrate multi-omics approaches and macromolecule-specific methodologies to more comprehensively parse the synergistic networks of both large and small molecular components within the core herbs.

This multi-target and multi-component characteristic is the core of the unique advantages of TCM in treating diabetes. Unlike Western medicine with a single target, TCM for the treatment of diabetes usually does not rely on a single ingredient or a single medicine, but adopts a compound form. TCM can intervene in the complex pathological mechanism of diabetes in an all-round way through the synergistic effect of multiple active ingredients. The multiple drugs in the combination can synergize or inhibit each other, enhancing the efficacy and reducing the side effects. This not only helps to lower blood sugar, but also may have a protective effect on diabetes-related complications such as neuropathy, nephropathy and cardiovascular disease. While this study highlights the considerable potential of TCM in treating diabetes, its clinical application continues to encounter several challenges. First of all, the efficacy and safety of TCM are affected by many factors, including the origin of the medicinal materials, processing methods and individual differences. Secondly, most current clinical trials have problems such as small sample sizes and loose designs, which limit the wide application of the results. Therefore, future research should include more large-scale, high-quality randomized controlled trials to further validate the effectiveness and safety of traditional Chinese medicine for treating diabetes. In addition, diabetes is a highly heterogeneous disease with significant differences in pathophysiology between patients. The personalized treatment approach of TCM, which involves modifying drug formulas based on the patient’s unique condition, constitution, and other factors, may lead to more effective outcomes. Additionally, since diabetes is a chronic condition necessitating ongoing management, it is important to assess the long-term efficacy and safety of TCM treatments.

It is important to specifically note that the meta-analysis in this study only included adult patients aged 18 and above. While this aligns with the population scope of most current randomized controlled trials, it may limit the applicability of the study conclusions to adolescent patients with youth-onset type 2 diabetes mellitus (T2DM). Notably, the disease burden of this special population has been rapidly increasing worldwide in recent years, with overweight/obesity further elevating the risk (140, 141). Existing evidence indicates that, compared to youth-onset type 1 diabetes mellitus (T1DM), youth-onset T2DM not only has a worse clinical prognosis but also exhibits more pronounced metabolic abnormalities and potentially distinct pathogenesis compared to adult-onset T2DM. These characteristics collectively lead to an elevated risk of vascular complications in young patients (142). More concerningly, the current treatment options approved by the U.S. FDA, the European Medicines Agency (EMA), and Health Canada for youth-onset T2DM are very limited. This scarcity of treatment choices may lead to more adverse clinical outcomes for this population. Based on this, our future research will expand the cohort of young patients to systematically evaluate the synergistic effects of TCM-modern drug combination therapy, analyze differential treatment responses among different age groups, and elucidate the mechanisms of action of TCM interventions for youth-onset T2DM. These studies will provide crucial evidence for developing targeted intervention strategies, holding significant clinical value for preventing disease progression and reducing the risk of complications in adolescent patients.

Of course, this study also has some limitations. First, our meta-analysis and network pharmacology analysis are based on existing literature and databases, which may have problems with publication bias and incomplete data. Second, the prediction results of network pharmacology analysis need further experimental verification. Moreover, the complexity of TCM involves various components that may undergo intricate metabolic processes and interactions within the body. These processes not only affect the efficacy of the drug, but may also cause adverse reactions. Our understanding of its multi-component and multi-target mechanisms is still limited, and further research is needed in combination with more systems biology approaches. Future research should focus on the following aspects: First, conducting large-scale, high-quality clinical trials to provide more reliable evidence support; Second, conducting an in-depth analysis of the multi-component and multi-target mechanisms of TCM using multi-omics technologies and systems biology approaches.; Third, exploring the combined application of TCM and modern drugs to give play to their synergistic and enhancing effects; Fourth, strengthening the standardization and quality control of TCM to ensure the safety and effectiveness of clinical application.

This study systematically evaluated the efficacy and potential mechanisms of TCM in the treatment of diabetes through meta-analysis and network pharmacology. The results indicated that TCM can significantly improve blood glucose control in diabetic patients, reduce glycated hemoglobin levels, and alleviate insulin resistance. Additionally, network pharmacology analysis revealed that TCM exerts its effects through multiple targets and pathways, including the regulation of insulin signaling pathways, inflammatory responses, and oxidative stress. These findings provide scientific evidence for the use of TCM in diabetes treatment, supporting its clinical application.