Metabolomics in Traditional Chinese Medicine for Diabetes Mellitus: Mechanistic Insights, Biomarker Discovery, and Clinical Application Prospects Authors: Junhui Song, Gen Lin Li, Suhui Wu, Hanbing Li, Sohini Basu Roy Published: 2026 ABSTRACT AbstractObjectiveThis review provides a comprehensive summary of recent advancements in metabolomic analysis within traditional Chinese medicine (TCM) for the treatment of diabetes mellitus. It focuses on the standardized profiling of syndrome‐specific metabolites, the identification of bioactive compounds using high‐throughput techniques, and the elucidation of mechanisms based on metabolic pathways. The review aims to establish a reproducible analytical framework for TCM metabolomics, which includes identifying active compounds, clarifying mechanisms of action, and assessing therapeutic efficacy. By synthesizing the current body of evidence, this work seeks to provide a scientific foundation for future research, enhance the integration of metabolomics with TCM theory, and support the modernization and global acceptance of TCM in diabetes care. Ultimately, it addresses key challenges, such as the subjective nature of syndrome diagnosis and the complexity of multicomponent interactions.Subjects and MethodsA systematic review of the peer‐reviewed literature was conducted using PubMed, Web of Science, CNKI, and Wanfang Data for studies published up to 2025. The review included original research, reviews, and clinical trials that utilized metabolomic techniques (LC‐MS/MS, GC‐MS, and 600 MHz NMR) and standardized workflows (sample preparation, derivatization, instrument analysis, and data processing) in diabetic models. The qualitative synthesis focused on high‐throughput analytics, multivariate statistics (principal component analysis/partial least squares‐discriminant analysis [PCA/PLS‐DA] with 7‐fold cross‐validation and CV‐ANOVA, p < 0.05), and metabolite identification via the human metabolome database (HMDB)/METLIN metabolite database (mass error < 5 ppm); furthermore, these data were integrated with network pharmacology or multiomics approaches.ResultsMetabolomics has revealed distinct metabolic profiles for TCM syndromes. Specifically, the Qi‐Yin deficiency (QYD) syndrome is associated with altered levels of L‐glutamate (variable importance in projection [VIP] > 1, p < 0.05) and arachidonic acid (limit of detection [LOD] = 0.01–0.1 ng/mL), as well as the accumulation of turbid toxins involving pantothenate and CoA biosynthesis. Key hypoglycemic bioactive compounds, such as epicatechin and berberine, were identified using ultraperformance liquid chromatography‐quadrupole time‐of‐flight mass spectrometry (UPLC‐QTOF‐MS), with a mass error of < 5 ppm. Mechanistic studies have shown that TCM works through multiple pathways, including improving insulin resistance (e.g., mulberry leaves modulating amino acid/lipid metabolism); protecting β‐cells (e.g., timosaponin BII restoring phosphatidylserine levels); regulating glycolipid metabolism (e.g., Huanglian decoction influencing energy‐related metabolites); modulating gut microbiota (e.g., Gegen Qinlian decoction altering bile acids and short‐chain fatty acids); and preventing complications (e.g., Sophora flavescens affecting oxidative stress through glycerophospholipid metabolism). Furthermore, metabolomics enabled efficacy comparisons, highlighting improved outcomes with nanoformulations or exercise‐TCM combinations.ConclusionsMetabolomics provides a powerful approach to objectively assess TCM syndromes, clarify multitarget mechanisms, and comprehensively evaluate efficacy in diabetes treatment, which aligns well with TCM’s holistic principles. Despite existing challenges, such as the lack of standardized TCM syndrome classification and the complexity of metabolomic data (e.g., overlapping metabolite signals and multipathway crosstalk), future studies should focus on rigorous experimental designs, standardized protocols, and multiomics integration to promote biomarker discovery, personalized TCM, and its global integration into diabetes management.Keywords: analytical workflow, biomarker identification, diabetes mellitus, metabolite profiling, metabolomics, traditional Chinese medicine 1. Introduction Diabetes mellitus is a chronic condition characterized by hyperglycemia, which results from either absolute or relative insulin deficiency and impaired insulin utilization. The prevalence of diabetes has been steadily increasing in recent years, becoming a significant global public health issue. According to the 2021 report from the International Diabetes Federation (IDF), approximately 537 million adults (aged 20–79 years) worldwide are affected by diabetes, with China having the highest number of adult diabetes patients. Type 2 diabetes mellitus (T2DM) accounts for over 90% of these cases [1]. Diabetes presents substantial health risks, often accompanied by multiple complications that severely affect patients’ quality of life and overall health. While current therapeutic strategies can effectively control blood glucose levels, they are limited in their ability to fully replicate physiological insulin secretion patterns and cannot completely halt the progression of the disease. The development of metabolomics began in the mid‐1990s as an emerging analytical field focused on the qualitative and quantitative analysis of small‐molecule metabolites (< 1 kDa) in biological systems. In traditional Chinese medicine (TCM) research, metabolomics addresses unique analytical challenges, such as complex matrices with coexisting primary and secondary metabolites, low‐abundance bioactive compounds, and inconsistent sample processing. These challenges represent critical barriers to the modernization of TCM. However, metabolomics provides a holistic and quantitative approach to analyzing metabolite sets in biological systems, offering functional insights into physiological, pathological, and pharmacological processes. After decades of advancement, metabolomics has found broad applications in TCM research. TCM, with its multicomponent, multitarget, and multipathway properties, offers unique advantages in diabetes treatment by regulating systemic metabolic networks. This aligns closely with metabolomics’ ability to capture simultaneous changes across multiple metabolic pathways. This review summarizes recent progress in the application of metabolomics to TCM treatment of diabetes, aiming to provide a comprehensive theoretical foundation for its use in this field. 2. Overview and Principles of Metabolomics Metabolomics, as a key component of systems biology, aims to systematically analyze all small‐molecule metabolites such as carbohydrates, lipids, amino acids, and organic acids within an organism to reveal its physiological and pathological states and responses to external stimuli. Metabolomics is an emerging discipline that studies small‐molecule metabolites in biological systems, following genomics, transcriptomics, and proteomics [2]. Since its introduction by Nicholson’s team at Imperial College London, it has been widely applied in biomarker discovery, disease mechanisms, drug target identification, and translational research related to drug responses [3, 4]. Metabolomics research relies on standardized high‐throughput analytical techniques: (1) LC‐MS/MS (Thermo Q Exactive): C18 column (2.1 × 100 mm, 1.9 μm); mobile phase (A: 0.1%formic acid; B: acetonitrile); gradient elution (0–5 min: 5%–30%B; 15–20 min: 80%–95%B); and ESI + mode, m/z 50–1000; (2) GC‐MS (Agilent 7890A–5975C): DB‐5MS column and BSTFA + 1%TMCS derivatization (70°C, 30 min); and (3) NMR (Bruker AVANCE III 600 MHz): NOESY‐presat pulse sequence, TSP reference (δ = 0.00), 128 scans. These methods effectively separate and detect metabolites in complex biological samples. For instance, LC‐MS has become a mainstream technique in metabolomics due to its high sensitivity and broad applicability [5, 6]. A summary of the high‐throughput analytical techniques commonly utilized in metabolomics research is presented in Table 1. The abbreviations used in Table 1 are defined as follows: lysophosphatidylcholine (LPC), sphingomyelin (SM), phosphatidylethanolamine (PE), lysophosphatidylethanolamine (LPE), triglyceride (TG), and diglyceride (DG). TABLE 1High‐throughput analytical techniques commonly used in metabolomics research.Analytical techniqueInstrument modelKey experimental parametersRemarksLC‐MS/MSThermo Q ExactiveColumn: C18 (2.1 × 100 mm, 1.9 μm); mobile phase: A (0.1% formic acid), B (acetonitrile); gradient elution: 0–5 min (5%–30% B), 15–20 min (80%–95% B); ionization mode: ESI+; m/z range: 50–1000Mainstream technique with high sensitivity and broad applicabilityGC‐MSAgilent 7890A‐5975CColumn: DB‐5MS; derivatization: BSTFA + 1% TMCS (70°C, 30 min)Suitable for the detection of volatile and semivolatile metabolitesNMRBruker AVANCE III 600 MHzPulse sequence: NOESY‐presat; reference standard: TSP (δ = 0.00); number of scans: 128Nondestructive detection, suitable for qualitative and quantitative analysis of metabolites High‐throughput analytical techniques commonly used in metabolomics research. While these analytical platforms and workflows lay the foundation for metabolomics research, the rigor and reproducibility of results heavily depend on robust quality control (QC) strategies, batch effect correction, and cross‐cohort validation—key aspects that require more critical evaluation in diabetes metabolomics studies. To address this, recent guidelines and studies have delineated clear best practices: the QC omics consortium [7] proposed a comprehensive, implementable QC protocol that systematically addresses batch effects, normalization strategies, internal standard selection, and outlier detection, ensuring data accuracy and reproducibility across experiments and laboratories. This protocol specifically emphasizes the importance of QC sample inclusion, background noise correction, and missing data handling—critical steps for avoiding misleading interpretations in diabetes metabolomics, where subtle metabolic shifts are closely linked to disease progression. Figure 1 depicts the standardized and classical analytical workflow of metabolomics, which consists of five core sequential steps. For data collection, the primary sample types include serum, plasma, urine, feces, and liver tissue. Sample preparation involves targeted preprocessing strategies, such as derivatization with BSTFA containing 1% TMCS (70°C for 30 min) for GC‐MS analysis and protein precipitation for LC‐MS/MS analysis, aiming to improve the detectability of metabolites in complex biological matrices. Data preprocessing is a critical step to ensure data reliability: raw data are processed via XCMS and SIMCA software, combined with rigorous QC strategies including QC sample insertion, batch effect correction, and seven‐fold cross‐validation. Retention time calibration is performed to eliminate the deviations in retention time and m/z values of metabolites across different samples, thus ensuring the accuracy of peak matching. Figure 2 depicts the complete procedure for data preprocessing. The chemical structure identification of metabolites is based on the HMDB and METLIN metabolite database, with a stringent mass error threshold of < 5 ppm. Potential specific biomarkers are screened using the criteria of VIP > 1 combined with p < 0.05. Further metabolic pathway analysis of the identified differential metabolites is conducted by integrating bioinformatic tools. This standardized metabolomic workflow provides a reproducible analytical framework for metabolomic research and serves as a technical basis for elucidating the material basis and therapeutic mechanisms of diabetes treatment. Schematic workflow of metabolomics analysis. Schematic diagram of common types and workflows of metabolomic data pre‐processing. In data analysis, quantitative metabolite profiling combined with multivariate statistics (PCA/PLS‐DA) facilitates the extraction of meaningful information from complex metabolic data, and results interpretation requires integration with biological context to ensure scientific validity [8, 9]. Furthermore, interpreting metabolomics results requires integration with biological context to ensure scientific validity and reliability. Combined with clinical data, metabolomics not only supports early diabetes diagnosis but also informs personalized treatment strategies [1, 10]. 3. Applications of Metabolomics in TCM Research for Diabetes Treatment 3.1. Current Status of TCM in Diabetes Treatment TCM’s understanding of diabetes is rooted in its unique theoretical framework, emphasizing “yin‐yang imbalance” (a core TCM concept referring to the disharmony of opposing yet complementary forces in the body) and “visceral dysfunction” (dysregulation of organs like the spleen, kidney, and lung) as fundamental causes. In TCM, diabetes is classified under “xiao ke” (wasting‐thirst syndrome), with core pathogenesis involving yin deficiency and dryness‐heat, primarily affecting the lungs, stomach, and kidneys, while involving the liver and spleen. Etiological factors include constitutional deficiency, improper diet, emotional disturbances, and irregular lifestyle. Main symptoms include polydipsia, polyuria, and emaciation. TCM theory posits that diabetes arises from dysfunctions in organs such as the spleen, kidneys, and lungs, with spleen deficiency and kidney yin deficiency as primary causes [11]. TCM emphasizes treatments that harmonize yin and yang, tonify the spleen and kidneys, and clear heat and moisten dryness to restore bodily balance. In clinical practice, the efficacy of TCM in diabetes treatment is widely recognized (Table 2). Some classic formulas and traditional herbal compounds have demonstrated potential mechanisms; for example, Gegen Qinlian decoction significantly reduced blood glucose and serum insulin levels in T2DM mice. Network pharmacology and bioinformatics analyses revealed that it regulates 82 corresponding proteins and 59 diabetes‐related biological pathways [12]. Jiangtang Sanhuang tablet, derived from Taohong Chengqi decoction from the Treatise on Cold Damage with modifications, alleviates T2DM by modulating gut microbiota and bile acid metabolism interactions [13]. Danggui Liuhuang decoction has been shown to reduce proinflammatory cytokines and vascular endothelial growth factor release by inhibiting the JAK2/STAT3 signaling pathway, thereby ameliorating endothelial dysfunction in T2DM [14]. TABLE 2Commonly used Chinese medicines for diabetes: A summary.Single herbsTCM herb pairsChinese herbal compound formulas Astragalus membranaceus, Coptis chinensis, Rehmannia glutinosa, Panax ginseng, Panax quinquefolius, Dioscorea opposita, Ophiopogon japonicus, Pueraria lobata, Trichosanthes kirilowii, Anemarrhena asphodeloides, Salvia miltiorrhiza, Lycium barbarum, Scrophularia ningpoensis, Atractylodes lancea, Poria cocos, Alisma orientale, Zea mays Astragalus membranaceus–Coptis chinensis, Astragalus membranaceus–Ophiopogon japonicus, Astragalus membranaceus–Salvia miltiorrhiza, Coptis chinensis–Anemarrhena asphodeloides, Rehmannia glutinosa–Ophiopogon japonicus, Trichosanthes kirilowii–Anemarrhena asphodeloides, Panax ginseng–Schisandra chinensis, Salvia miltiorrhiza–Ligusticum chuanxiong,Liuwei Dihuang pill, Jinkui Shenqi pill, Yuye decoction, Xiaoke formula, Yunü decoction, Shengmai powder, Shenling Baizhu powder, Buzhong Yiqi decoction, Xuefu Zhuyu decoction, Taohong Siwu decoction, Qiwei Baizhu powder, Gegen Qinlian decoction, Commonly used Chinese medicines for diabetes: A summary. 3.2. Metabolomics Studies on TCM Syndromes in Diabetes Metabolomics shares a close biological association with TCM syndromes, where syndromes represent macroscopic pathological manifestations of the body in specific spatiotemporal contexts, while metabolomics elucidates underlying microscopic metabolic dynamics from a systems biology perspective. This integration is particularly valuable for diabetes, as global metabolomics research has clearly characterized the core metabolic dysregulation of the disease, and TCM syndrome‐specific metabolic signatures can be contextualized within this broader framework to enhance international relevance. Qi‐yin deficiency (QYD) syndrome, a common TCM syndrome characterized by fatigue, dry mouth, and reduced organ function, is the core syndrome in the mid‐stage of diabetes and the most common and prolonged clinical phase, marking the transition from metabolic hyperactivity to organ functional decline. Notably, the metabolic disturbances associated with QYD align with the well‐established metabolic phenotypes of diabetes, including altered amino acid metabolism and oxidative stress‐related lipid metabolism reported in non‐TCM biomedical studies [15]. Huang et al. [16] employed LC‐MS/MS (Thermo Q Exactive) with a C18 column (2.1 × 150 mm, 2.5 μm) and gradient elution (A: 5 mM ammonium acetate; B: methanol) to profile serum metabolites. Data preprocessed by XCMS and SIMCA revealed 7 differential metabolic pathways, with L‐glutamate and arachidonic acid identified as biomarkers (VIP > 1, p < 0.05) via PLS‐DA and t‐test. The high sensitivity of LC‐MS/MS (LOD = 0.01–0.1 ng/mL) ensured detection of low‐abundance arachidonic acid, including D‐glutamine and D‐glutamate metabolism, which is consistent with the core amino acid metabolic dysregulation in diabetes identified by Pandey et al. [17]. Among differential metabolites, L‐glutamate and arachidonic acid were identified as potential biomarkers for T2DM QYD syndrome. Arachidonic acid is a key mediator of inflammatory metabolic disturbances in diabetes and has been extensively characterized by MS‐based metabolomics. Qiu et al. [18] detected 40 differential metabolites in urine from QYD rats, with 18 downregulated and 22 upregulated, enriched in pathways involving carbohydrate, lipid, amino acid, and bioenergy metabolism—consistent with the universal metabolic disturbance patterns of diabetes summarized by Xu et al. [15]. This study further supports the alignment between TCM syndrome‐related metabolic changes and the global cognitive framework of diabetes metabolic pathology, as the identified carbohydrate and energy metabolism disturbances are core features of insulin resistance and β‐cell dysfunction. Turbid toxin accumulation (TTA) syndrome is the core syndrome in the mid‐to‐late stages of diabetes complications, a TCM syndrome associated with metabolic waste accumulation and complications like microangiopathy. Zhao et al. [19] screened 17 potential differential metabolites in plasma from T2DM patients with TTA, with seven upregulated (e.g., D‐4′‐phosphopantothenate and 2,6‐dichloroindophenol) and ten downregulated (e.g., 3‐methoxybenzoic acid and 3‐iodooctadecanoic acid). The key pathway identified was pantothenate and CoA biosynthesis, which is jointly involved by D‐4′‐phosphopantothenate and β‐alanine. Pantothenate and CoA biosynthesis dysfunction has been widely validated in global diabetes metabolomics studies as a key pathway linked to mitochondrial dysfunction, energy metabolism disturbance, and diabetic microvascular complications. It is directly associated with the classical characterization of lipid metabolic remodeling [20]. The syndrome of yin deficiency with excessive heat is one of the common syndromes of diabetes mellitus, which is mostly seen in the early or middle stage of the disease, and a number of metabolomic studies have been carried out on it. Through urinary metabolomics analysis, Huang Hanhui found that this syndrome type is closely associated with disorders in glycerophospholipid metabolism, bile secretion, steroid degradation, and primary bile acid biosynthesis [21]. Glycerophospholipid metabolic disturbance is a core feature of diabetes metabolic dysregulation reported in global metabolomics research, with phosphatidylcholines (PCs) and LPCs being consistently identified as differentially expressed metabolites in diabetic patients across multiethnic cohorts, and their abnormal levels are closely linked to insulin resistance and β‐cell dysfunction [22, 23]. The phlegm‐dampness obstruction syndrome is more common in the early stage of diabetes. Relevant metabolomics studies have mainly focused on diabetic patients with phlegm‐dampness constitution, and it has been found that this syndrome type is closely related to abnormalities in gut microbiota metabolites, with core metabolites involving pathways such as fatty acid oxidation and bile acid metabolism [24]. This confirms that metabolomics technology can reveal the intrinsic correlation between “endogenous phlegm‐dampness” of this syndrome type and metabolic disorders, and the disorder of fatty acid oxidation in this TCM syndrome is highly consistent with the abnormal lipid metabolism of diabetes characterized by elevated free fatty acids (FFAs) and acylcarnitines in mainstream biomedical research. In fact, acylcarnitines, as key metabolites of fatty acid β‐oxidation, have been validated as robust biomarkers for predicting the progression of diabetes and its complications such as diabetic kidney disease (DKD) in large‐scale prospective cohort studies [25]. Figure 3 illustrates the correspondence between TCM syndromes and classic metabolic phenotypes. Schematic of the correlation between traditional Chinese medicine syndromes and classical diabetic metabolic phenotypes. 3.3. Metabolomics in Studying the Material Basis of TCM for Diabetes Treatment TCM has a long history in diabetes treatment, with its multicomponent and multitarget properties effectively alleviating symptoms and improving metabolic disorders, offering unique advantages for complex diseases. However, insufficient research on the material basis of TCM efficacy has hindered its modernization. Metabolomics, leveraging modern analytical techniques, evaluates metabolic changes holistically and has become a vital tool for elucidating the material basis of TCM compounds, with widespread applications. Khanna et al. [26] used UPLC‐QTOF‐MS to chemically analyze Cupressus torulosa needle extracts, tentatively identifying 50 components. The hypoglycemic active substances were primarily compounds such as epicatechin, amentoflavone, and cupressuflavone, along with natural iridoid glycosides exemplified by harpagoside. Li et al. [27] found that ferulic acid‐4‐β‐glucoside, bufotenine, jatrorrhizine, and berberine (BBR) from Berberis vernae exhibited significant affinity for 30 proteins related to T2DM and 14 differential metabolite‐associated proteins, confirming these as potential active components for T2DM. To further clarify the specific metabolic biomarkers of antidiabetic Chinese medicinal materials (single herbs, herb pairs, and classic formulas) identified by diverse metabolomic approaches, the corresponding detection techniques, biological samples, and characteristic differential biomarkers are summarized in Table 3. TABLE 3Metabolomic detection of specific biomarkers in antidiabetic Chinese herbal medicines.Chinese herbal medicine (single/herb pair/formula)Metabolomic technologyBiological sampleSpecific biomarkers (VIP > 1, p < 0.05)Reference Berberis vernae 1H NMRSerumIsoleucine, trimethylamine N‐oxide, glucose[27]Gegen Qinlian decoction (formula)Targeted LC‐MS/MSFecesDeoxycholic acid, ursodeoxycholic acid, taurodeoxycholic acid, α‐muricholic acid[28] Anemarrhena asphodeloides (single herb)Lipidomics/LC‐MS/MSINS‐1 cell lysatePhosphatidylcholine, phosphatidylserine, phosphatidylethanolamine, phosphatidylglycerol[29]Huanglian decoction (formula)HPLC‐MSUrineCytosine, L‐carnitine, betaine, glucose[30] Morus alba (mulberry leaf, single herb)LC‐MS/MSLiver tissueα‐D‐Glucose, glycerol phosphate, succinate, fumarate[31]Yu‐Ye Tang (formula)LC‐MS/MSSerum1‐Oleoyl‐sn‐glycero‐3‐phosphocholine, 5‐hydroxytryptophan, L‐kynurenine, 1‐acyl‐sn‐glycero‐3‐phosphocholine (18:1)[32] Sophora flavescens (single herb)LC‐MS/MSRetinal tissueTaurine, LPC (18:2 (9Z, 12Z)/0:0)), arginine, vitamin A2[33] Phellodendri chinensis Cortex–Anemarrhenae rhizoma herb pairHPLC‐Q‐TOF/MSHuman serum2‐Hydroxyvaleric acid, hypoxanthine, inosine, arachidonic acid[34] Metabolomic detection of specific biomarkers in antidiabetic Chinese herbal medicines. 3.4. Metabolomics in Elucidating Mechanisms of TCM in Diabetes Treatment TCM, as a traditional therapeutic approach, has developed advantages in regulating glucose and lipid metabolism and combating insulin resistance through sustained practice, experience, and observational treatments, offering mild yet stable long‐term effects [35–37]. Although its efficacy has been clinically validated, the active substances and modes of action have long lacked scientific elucidation. This stems from TCM compounds featuring multiple active components that achieve disease treatment and regulation via multicomponent, multitarget synergies [38]. Thus, interpreting the pathological substances and pathways in complex syndromes to comprehensively evaluate TCM formulas’ effects has become a focus in modern medicine, where metabolomics fulfills this need [39]. Metabolomics has revealed five mechanisms of TCM in diabetes treatment. 3.4.1. Improving Insulin Resistance Insulin resistance reduces physiological sensitivity to insulin, impairing glucose uptake by tissues under normal insulin concentrations, representing a core pathological basis for diabetes onset and progression. Metabolomic investigations have comprehensively delineated the core metabolic dysregulations driving insulin resistance in diabetes, with conserved perturbations in amino acid, lipid, and energy metabolism validated across independent cohorts and populations [40]. Accumulating evidence further identifies dysregulated branched‐chain amino acid (BCAA) and sphingolipid metabolism as two pivotal pathways underlying the pathogenesis of insulin resistance [41, 42]. Metabolomic studies have revealed that TCMs such as mulberry leaf and Pueraria root can reverse the dysregulation of amino acid and lipid metabolic pathways through their regulatory effects, thereby ameliorating insulin resistance. Ma et al. [43] used serum metabolomics to study mulberry leaf effects on T2DM rat serum metabolites, involved in amino acid, carbohydrate, and lipid metabolism, contributing to obesity and insulin resistance amelioration. Cai et al. [44] explored puerarin’s role in diabetes, with metabolomics showing regulation of sphingolipid metabolism and reduced ceramide levels, which positively correlate with insulin resistance index and T2DM risk, as key factors in insulin resistance. Zhou et al. [44] characterized metabolite profiles in high‐fat diet‐induced KK‐Ay mice using LC‐MS‐based untargeted metabolomics, finding ginsenoside Rb1 reduced fecal FFA levels, includingα‐linolenic acid, oleic acid, arachidonic acid, palmitic acid, and stearic acid. FFAs are closely linked to insulin resistance, with high concentrations disrupting insulin signaling via multiple mechanisms, and the downregulation of these FFAs by ginsenoside Rb1 mirrors the therapeutic direction of targeting lipid metabolic dysregulation in diabetes, a core pathway validated by translational metabolomic studies linking lipid signatures to clinical insulin resistance outcomes. 3.4.2. Protecting Pancreaticβ‐Cell Function Pancreaticβ‐cells are central to diabetes pathogenesis, with functional defects and reduced numbers spanning the entire disease process. Global metabolomics and clinical cohort studies have well‐characterized the core metabolic perturbations underlying pancreatic β‐cell dysfunction in diabetes, including disordered glucose metabolism, impaired tricarboxylic acid (TCA) cycle flux, and lipid metabolic remodeling with elevated ceramides and FFAs [17, 20] [1, 2]. Shi et al. [29] extracted timosaponin BII from Anemarrhena asphodeloides and investigated its protective effects onβ‐cells in T2DM using a glycolipid toxicity INS‐1 cell model. Lipidomics identified phosphatidylserine as a lipid marker; supplementation enhanced glucose‐stimulated insulin secretion and reducedβ‐cell apoptosis, confirming timosaponin BII’s role in restoring insulin secretion and cell viability. Zhang et al. [45] studied Coptis deltoidea in T2DM mouse models, with metabolomics revealing inhibition of overactivated glycolysis and regulation of TCA cycle products. Hyperactive glycolysis and impaired TCA cycles lead toβ‐cell dysfunction and apoptosis, elucidating C. deltoidea’s protective mechanism. This regulatory effect on central carbon metabolism is consistent with the established metabolic profile of diabetes, where disrupted glycolysis–TCA cycle coupling is a key driver of β‐cell energy metabolism disorder and functional failure [20]. 3.4.3. Regulating Glycolipid Metabolism Interactions between hyperglycemia and dyslipidemia form a vicious cycle of glycolipid toxicity, synergistically driving diabetes progression and the development of macrovascular and microvascular complications, with core perturbations in glucose metabolism, lipid remodeling, and amino acid metabolism being the hallmarks of this pathological process. Numerous studies have demonstrated that the regulatory effects of TCM formulas and bioactive components on glucose and lipid metabolism are highly consistent with the core metabolic dysregulations of diabetes identified in extensive biomedical research: both target key nodes such as glucose catabolism, amino acid transamination, and lipid β‐oxidation. Pan et al. [30] analyzed rat urine post‐Huanglian decoction treatment using HPLC‐MS‐based urinary metabolomics, showing regulation of biomarkers like cytosine, L‐carnitine, phenylalanine, glucose, citric acid, and phenylpyruvic acid, linked to energy metabolism—among which L‐carnitine and phenylalanine are well‐characterized metabolites with altered levels in clinical diabetic populations, and their dysregulation is closely associated with insulin resistance and lipid metabolic disorder. Li et al. [46] focused on alkaloids, flavonoids, and polysaccharides from mulberry leaves, using urinary metabolomics to assess differential metabolites in pathways, demonstrating hypoglycemic activity via glucose, amino acid, and lipid metabolism regulation, confirming mulberry leaves’ efficacy in T2DM; the study’s findings align with global metabolomics evidence that branched‐chain amino acids (BCAAs) and FFAs are key metabolic signatures of T2DM, and the modulation of these metabolic axes is a critical therapeutic target for improving glycolipid homeostasis [20]. Song et al. [47] found Jinqi Jiangtang tablets improve T2DM primarily by regulating glucose and lipid metabolism pathways. Ma et al. [32] used untargeted metabolomics to show Yu‐Ye Tang regulates arachidonic acid metabolism, alanine, aspartate, and glutamate metabolism in T2DM rats, indicating its therapeutic role via glycolipid metabolism modulation. 3.4.4. Regulating Gut Microbiota The association between gut microbiota and diabetes is a frontier in metabolic disease research, forming a complex bidirectional regulatory network via the “microbiota‐gut‐brain/liver/pancreas axis. Environmental factors including diet, microbial composition, and host metabolic interactions are key regulators of the metabolome, and the dysregulated gut microbial profile in obesity and diabetes is closely linked to the onset and progression of metabolic disorders [17].” Liu et al. [28] used targeted metabolomics to show that Gegen Qinlian decoction with dried ginger treats T2DM mice by modulating gut microbiota and bile acid metabolism. Yao et al. [48] demonstrated Cyclocarya paliurus polysaccharides alleviate T2DM symptoms via gut microbiota and short‐chain fatty acid (SCFA) modulation using untargeted metabolomics. Wang et al. [49] detected metabolite changes with metabolomics, showing Danggui Buxue decoction regulates pathways like alanine, aspartate, and glutamate metabolism in gut microbiota, restoring balance in T2DM rats for antidiabetic effects—these amino acid metabolic pathways are core dysregulated pathways in diabetes, with elevated alanine, aspartate, and glutamate being consistent with the plasma metabolomic signatures of insulin resistance and T2DM reported in large‐scale human cohort studies [50]. Ma et al. [51] found PuRenDan modulates gut microbiota metabolites like pantothenic acid, 1‐methylhistamine, and 1‐methylhistidine, involved in pantothenate and CoA biosynthesis, histidine metabolism, and secondary bile acid biosynthesis. Correlation analyses revealed close associations between gut microbiota, metabolites, and T2DM indices. Analyses of these studies on TCM have revealed intimate correlations between the gut microbiota, metabolites, and T2DM‐related indices, which align with the global consensus that host metabolic health can be modulated by reshaping the metabolome. 3.4.5. Improving Microcirculation to Prevent Complications Diabetic microangiopathy, a hallmark chronic complication, involves systemic damage to microvessels under prolonged hyperglycemia, primarily affecting the retina, kidneys, nerves, and skin microcirculation, potentially leading to organ failure. It is important to clarify that diabetic microangiopathy is driven by the interplay of metabolic disorders and chronic low‐grade inflammation. The combined effects of these two factors, induced by long‐term hyperglycemia, lead to endothelial damage and microcirculatory dysfunction, ultimately resulting in diabetic complications. Metabolomic studies have clearly demonstrated that diabetic microangiopathy is closely associated with core metabolic dysregulations, including glycerophospholipid remodeling, amino acid metabolic disorders, mitochondrial energy metabolism dysfunction, and oxidative stress‐induced imbalances in lipid and fatty acid metabolism in diabetes [52]. Additionally, multiple studies have shown that TCM can treat related complications by improving microcirculation. Luo et al. [33] studied Sophora flavescens extract’s material basis for diabetic retinopathy, screening 39 differential metabolites via metabolomics, with 23 influenced by the extract. It acts via pathways like lactose metabolism, bile acid metabolism, and glycerophospholipid metabolism, delaying retinopathy progression through oxidative stress, inflammation, and vascular dysfunction linkages. Xu et al. [53] used metabolomics to study salvianolic acid B and tanshinone IIA from Salvia miltiorrhiza, regulating saturated fatty acid metabolism, glycerophospholipid metabolism, and steroid biosynthesis, modulating glomerular, vascular, and mesangial functions via oxidative stress, inflammation, and fibrosis to improve early diabetic nephropathy in rats. Zhao et al. [54] found Jinmaitong ameliorates diabetic peripheral neuropathy by regulating abnormal energy balance and mitochondrial function, improving peripheral nerve function and symptoms. The mechanisms underlying the therapeutic effects of Sophora flavescens, Salvia miltiorrhiza, and Jinmaitong on diabetes are consistent with the findings of global metabolomic studies [20]. In recent years, metabolomics has revealed mechanisms of various TCMs for T2DM. Details are shown in Table 4. Figure 4 illustrates the mechanisms of TCM in diabetes treatment. The abbreviations used in Figure 4 are defined as follows: nuclear factor‐kappa B (NF‐κB), adenosine 5′‐monophosphate‐activated protein kinase (AMPK), nuclear factor e2‐related factor 2 (Nrf2), antioxidant response element (ARE), and glucagon‐like peptide‐1 (GLP‐1). TABLE 4Examples of metabolomics applications in TCM treatment of diabetes.Drug nameT2DM modelMethodSample detectedMain marker observedSample size (metabolomics analysis)Biological matrix consistencyMechanism of actionAuthorWendan decoctionPhlegm‐dampness obstructive sleep apnea‐hypopnea syndrome with T2DM patientsUntargeted metabolomicsSerumDL‐Arginine, guaiacol sulfate, azelaic acid, phloroglucinol, uracil n = 20 patientsConsistent (serum)Regulates biomarkers like DL‐arginine, guaiacol sulfate, azelaic acid, phloroglucinol, and uracil; significantly improves glycolipid metabolism and insulin resistanceZhou [55] Lomatogonium rotatum High‐fat diet combined with STZ‐induced miceUntargeted metabolomicsSerumVitamin B6, mevalonate‐5P, D‐proline, L‐lysine, taurine10 mice/groupConsistent (serum)Regulates insulin resistance by altering levels of vitamin B6, mevalonate‐5P, D‐proline, L‐lysine, and taurineDai [56]Pi‐Dan‐Jian‐Qing decoctionHigh‐fat diet combined with STZ‐induced miceUntargeted metabolomicsSerum and fecesDL‐Tryptophan, L‐kynurenine, D‐(+)‐malic acid, citric acid6 rats/groupConsistent (serum)Improves gut microbiota and insulin resistance by increasing relative abundance of Lactobacillus, Blautia, Bacteroides, etc., and regulating tryptophan metabolism, histidine metabolism, etc.Xie [57] Cynomorium songaricum polysaccharideHigh‐fat diet‐induced miceTargeted metabolomicsSerumLysophosphatidylcholine (LPC) 14:0, sphingomyelin (SM) d16:1/24:1, lysophosphatidylethanolamine (LPE) 22:0, phosphatidylcholine (PC) 16:1/18:28 mice/groupConsistent (serum)Improves lipid metabolism by regulating phosphatidylcholine, lysophosphatidylcholine, phosphatidylethanolamine, and sphingomyelin metabolitesShi [58] Phellodendri chinensis Cortex–Anemarrhenae rhizoma herb pairT2DM patients, high‐fat diet combined with STZ‐induced miceUntargeted metabolomicsSerumHypoxanthine, inosine, 2‐hydroxyvaleric acid, docosahexaenoic acidPatients: 250; mice: 10/groupConsistent (serum)Improves lipid metabolism and insulin resistance by regulating unsaturated fatty acid biosynthesis and purine metabolismYu [34] Schisandra chinensis polysaccharideHigh‐fat diet combined with STZ‐induced miceUntargeted metabolomicsSerum and fecesVitamin A, all‐trans‐4‐hydroxyretinoic acid, retinoyl b‐glucuronide, dehydroepiandrosterone6 mice/groupConsistent (serum)Improves gut microbiota composition and short‐chain fatty acid levelsGuo [59] Rubus irritans High‐fat diet combined with STZ‐induced miceUntargeted metabolomicsSerumInosine, glycolate, L‐threonate, β‐glycerophosphate6 rats/groupConsistent (serum)Improves glycolipid metabolism and insulin resistance by regulating ascorbate and aldarate metabolism, ABC transporters, glycerophospholipid metabolism, and carbon cycle pathwaysHan [60]Red ginsengHigh‐fat diet combined with STZ‐induced miceUntargeted metabolomicsSerumD‐Ornithine, taurodeoxycholic acid, lysophosphatidylcholine (24:1 (15Z)), 2‐methylbutyroylcarnitine6 rats/groupConsistent (serum)Improves lipid metabolism and insulin resistance by regulating D‐arginine and D‐ornithine metabolism, D‐glutamine and D‐glutamate metabolism, etc.Yang [61]Yuquan pillHigh‐fat diet combined with STZ‐induced miceUntargeted metabolomicsSerumLenticin, 1,2,3‐trihydroxybenzene, sorbitol, methyl (indol‐3‐yl) acetate8 rats/groupConsistent (serum)Improves insulin resistance and gut microbiota by regulating five pathways like ascorbate and aldarate metabolism, nicotinate and nicotinamide metabolism, and abundance of six genera including Firmicutes and BacteroidetesLiu [62]White hyacinth beanHigh‐fat high‐sugar diet with low‐dose STZ‐induced miceTargeted metabolomicsSerumL‐Valine, isoleucine, L‐arginine, LPC (P‐18:1 (9Z))8 rats/groupConsistent (serum)Improves insulin resistance and protects β‐cells by regulating four pathways like phenylalanine metabolism and tryptophan metabolismWang [63] Cornus officinalis High‐fat diet combined with STZ‐induced miceUntargeted metabolomicsSerum and urineTaurine, chenodeoxycholic acid, glycocholic acid, L‐tyrosine9 rats/groupInconsistent (serum + urine)Improves glycolipid metabolism and insulin resistance by regulating 42 metabolites like taurine, glycochenodeoxycholic acid, glucuronic acid, and L‐tyrosineHou [64]Da‐Chai‐Hu decoctionHigh‐fat diet combined with STZ‐induced miceUntargeted metabolomicsPlasmaHydroxyphenylacetylglycine, phenylglyoxylic acid, hippuric acid, p‐cresol sulfate10 rats/groupConsistent (urine)Improves insulin resistance and protects 尾‐cells via AGEs/RAGE/AKT signaling pathwayWang [65]HuidoubaHigh‐fat diet combined with STZ‐induced miceUntargeted metabolomicsPlasmaL‐Valine, betaine, N6‐acetyl‐L‐lysine, creatine13 rats/groupConsistent (plasma)Improves glucose metabolism and microangiopathy by regulating bile acid metabolismLi [66] Salvia miltiorrhiza db/db miceUntargeted metabolomicsUrineNiacinamide, homovanillin, tyramine, dopamine6 mice/groupConsistent (urine)Improves insulin resistance via PTP1B/IRS‐1/AKT signaling pathwayZhang [67]Shouhui Tongbian capsuledb/db mice and high‐fat diet‐fed miceTargeted metabolomicsFeces and plasmaIsobutyric acid, isovaleric acid, 2‐methylbutyric acid, acetic acid8‐10 mice/groupInconsistent (feces + plasma)Improves gut microbiota and insulin resistance by regulating gut BCAAs‐mTORC1/IRS‐1/PI3K/AKT axisWang [68]MulberryHigh‐fat high‐sucrose diet with low‐dose STZ‐induced miceUntargeted metabolomicsFeces and plasmaGlycine, Phe‐Pro, urocanic acid, phylloquinone6 mice/groupConsistent (feces)Improves gut microbiota by regulating abundance of four genera like Prevotella and Parabacteroides Du [69] Cordyceps militaris extractHigh‐fat diet combined with STZ‐induced miceUntargeted metabolomicsFecesCatechin, erucic acid, phenols, pyrimidine nucleosides10 mice/groupConsistent (feces)Regulates gut microbiota by increasing abundance of Firmicutes and BacteroidetesLiu [70]Xiexin decoctionHigh‐fat diet combined with STZ‐induced miceUntargeted metabolomicsFecesValine, aspartic acid, phosphoenolpyruvic acid, succinic acid6 rats/groupConsistent (feces)Improves gut microbiota and insulin resistance and protects 尾‐cells by regulating arachidonic acid metabolism, amino acid metabolism, and bile acid metabolismQian [71] Malus toringoides db/db miceUntargeted metabolomicsFecesAcetic acid, butyric acid, malonic acid, glutaric acid6 mice/groupConsistent (feces)Improves gut microbiota and glycolipid metabolism by regulating abundance of four genera like Alloprevotella and Parabacteroides Zheng [72]Renshenjian decoctionHigh‐fat diet combined with STZ‐induced miceUntargeted metabolomicsLiver tissueCeramide, ornithine, leucine, LPC (16:0)8 mice/groupConsistent (liver tissue)Improves insulin resistance by regulating HIF‐1伪/SPTLC2 pathwayChen [73]Zuogui Jiangtang Qinggan formuladb/db miceTargeted metabolomicsLiver tissueTriglyceride (TG), diglyceride (DG), phosphatidylcholine (PC), phosphatidylethanolamine (PE)6 mice/groupConsistent (liver tissue)Improves lipid metabolism by regulating levels of triglycerides, diglycerides, and phosphatidylcholineZhou [74]Mulberry leafHigh‐fat high‐sugar diet with STZ‐induced miceUntargeted metabolomicsLiver tissueα‐D‐Glucose, glycerol phosphate, succinate, fumarate9 rats/groupConsistent (liver tissue)Improves glycolipid metabolism and insulin resistance by regulating key enzymes like ACSL5, Dlat, Pdhb, G6pc, Mdh2, and CsLv [31]Buyang Huanwu decoctionHigh‐fat diet‐induced miceTargeted metabolomicsLiver tissueTriglyceride (TG) (18:4_18:2_18:3), phosphatidylcholine (PC) (36:2e), diglyceride (DG) (22:6_22:6), sphingomyelin (SM) (d18:1_24:0)8–10 rats/groupConsistent (liver tissue)Improves gut microbiota and lipid metabolism by regulating abundance of Bacteroidetes and FirmicutesLiu [75] Dendrobium officinale db/db miceUntargeted metabolomicsLiver tissueAcetic acid, propanoic acid, butyric acid, PPAR pathway‐related metabolites6 mice/groupInconsistent (liver tissue, faeces)Improves gut microbiota and insulin resistance by regulating short‐chain fatty acids and PPAR pathwaySong [76]Shenqi compoundGK miceUntargeted metabolomicsIntestinal contentsL‐Tyrosine, isoleucyl‐tyrosine, leucyl‐hydroxyproline, cyclohexanecarboxylic acid10 rats/groupConsistent (intestinal contents)Improves gut microbiota and glycolipid metabolism by regulating seven genera like Prevotella and Lactobacillus and five metabolisms like gluconeogenesis/glycolysis, amino acid metabolism, and lipid metabolismZhang [77]Shenshu Tiaopi granuleGK ratsUntargeted metabolomicsIntestinal contents2‐Deoxyglucose, D‐maltose, alpha‐lactose, beta‐muricholic acid8 rats/groupConsistent (ileal contents)Improves gut microbiota and lipid metabolism by regulating abundance of Monoglobus and bile acid biosynthesis and cholesterol metabolismZhao [78] Angelica polysaccharideHigh‐fat high‐sugar diet‐induced miceUntargeted metabolomicsIntestinal contentsL‐Cysteine, glycerophospholipids, N‐methyltryptamine, D‐sedoheptulose 7‐phosphate5 mice/groupConsistent (intestinal contents)Improves gut microbiota by regulating genera like Akkermansia and Lactobacillus Tang [79]Erzhi pillHigh‐fat diet with low‐dose STZ‐induced miceUntargeted metabolomicsHeartGlutamine, taurine, glycine, 3‐hydroxybutyrate (3‐HB)5 rats/groupConsistent (heart tissue)Improves glucose metabolism, insulin resistance, and microangiopathy by regulating AMPK and PI3K/Akt/FoxO3a signaling pathwaysPeng [80] Examples of metabolomics applications in TCM treatment of diabetes. Mechanisms of traditional Chinese medicine in treating diabetes. 3.5. Metabolomics in Evaluating TCM Efficacy for Diabetes Treatment Traditional TCM efficacy assessments primarily rely on subjective symptom scores and routine biochemical indicators (e.g., blood glucose and serum insulin). While valuable, these are limited by subjectivity, individual variability in symptom perception, and lagged, simplistic evaluation criteria. Metabolomics, through objective dynamic metabolite changes, reveals holistic effects of herbal interventions, providing modern scientific support for optimizing classic formulas and personalized medication. Global metabolomics studies have well‐established that T2DM is characterized by core dysregulations in amino acid metabolism (especially BCAAs), lipid remodeling, energy metabolism, and inflammatory metabolic disturbances [81]. Currently, a large‐scale population cohort study has confirmed the associations between amino acid metabolic dysregulation, lipid disorders, and the risk of T2DM onset and clearly identified metabolites as robust biomarkers for disease progression and therapeutic response [82]. These universal metabolic signatures provide an essential reference framework for objectively evaluating the efficacy of TCM interventions. Based on the above elucidated antidiabetic mechanisms of TCM, the therapeutic efficacy of potent TCM interventions (including bioactive extracts, classic formulas, and optimized formulations) and their core metabolic modulation effects identified by metabolomics are summarized in Table 5. TABLE 5Therapeutic application of potent antidiabetic Chinese herbal medicines elucidated by metabolomics.Potent TCM interventionsDiabetic model/subjectsMetabolomic approachKey therapeutic efficacyCore metabolic modulationClinical application prospectReferenceBerberine‐baicalin (10∶1)T2DM patients, damp‐heat internal accumulation type T2DM fecal microbiota transplantation pseudo‐sterile miceUntargeted UPLC‐QTOF‐MSImprove insulin resistance, regulate gut microbiota structureMainly modulate arachidonic acid metabolism, α‐linolenic acid metabolismPersonalized hypoglycemic therapy[83] Pueraria lobata polysaccharidesC57BL/KsJ‐db/db mice + high‐fat‐diet + STZ‐induced T2DM miceUPLC‐QTOF‐MSImprove insulin resistance, regulate lipid metabolismPPAR pathway, sphingolipid metabolismNovel strategy for diabetic hepatopathy[44, 84]Jinqi Jiangtang tabletsHFD + STZ miceUPLC‐QTOF‐MSRegulate glycolipid metabolism, reduce hyperglycemiaTarget glucose and lipid metabolic pathwaysClinical oral preparation for T2DM[47] Posidonia oceanica extract (nano‐formulation)STZ miceUntargeted GC‐MSEnhance hypoglycemic effect, improve metabolic homeostasisDownregulate glucose/leucine levels, modulate lipid metabolismFormulation optimization for TCM bioactive compounds[85] Salvia miltiorrhiza extractdb/db miceUPLC‐TOF‐MSImprove glucose/lipid metabolism; alleviate liver/kidney damageRegulate glycerophospholipid metabolism, inhibit oxidative stressPotential therapy for early diabetic nephropathy[53] Therapeutic application of potent antidiabetic Chinese herbal medicines elucidated by metabolomics. Chen et al. [83] used UPLC‐Q‐TOF‐MS to study BBR and baicalin (BAI) combinations in improving T2DM, identifying 28 differential metabolites and 14 pathways consistent with gut microbiota findings, multidimensionally indicating that Coptis–Scutellaria compatibility protects against T2DM by modulating gut microbiota and fecal metabolites. This aligns with global metabolomic evidence that gut microbiota–host co‐metabolism is a key node in T2DM pathogenesis, and restoring microbial–metabolic homeostasis is a validated strategy for improving insulin resistance and metabolic disorders [17, 85]. Ammar et al. compared antidiabetic potential of Posidonia oceanica extract and its gelatin nanoparticle formulation using untargeted GC‐MS metabolomics, showing the nanoparticle formulation as more effective. Such comparative metabolomic evaluation echoes translational metabolomics principles that metabolic profile shifts can sensitively reflect therapeutic potency and formulation optimization, which is increasingly adopted in international antidiabetic drug development [20]. Zhao et al. [86] evaluated lifestyle and TCM interventions on prediabetes indicators, identifying N‐neuraminylmethionine and 5‐hydroxy‐L‐tryptophan as core potential biomarkers for Xiaokeqing granule and lifestyle efficacy. This practice of defining efficacy‐related metabolites is consistent with international standards for diabetes metabolomics, where metabolite panels are validated across cohorts to predict therapeutic responses and stratify high‐risk individuals [87]. Zhang et al. [88] assessed exercise therapy’s potential impact on T2DM via untargeted metabolomics of serum and urine from healthy, Tai Chi, and brisk walking groups, revealing impaired carbohydrate, lipid, and amino acid metabolism in T2DM, improved by exercise. Tai Chi ameliorates disrupted BCAA metabolism, while brisk walking regulates steroid hormone biosynthesis and arachidonic acid metabolism. BCAA elevation is a well‐documented metabolic hallmark of insulin resistance and T2DM risk in large‐scale prospective cohorts worldwide [20], so the regulation of BCAAs by Tai Chi directly targets a core pathogenic metabolic abnormality, providing high‐level mechanistic evidence for its antidiabetic efficacy. According to the Chinese Pharmacopoeia, Pueraria lobata and Pueraria thomsonii treat xiao ke to lower blood glucose; Wang et al. [84] compared Pueraria lobata and Pueraria thomsonii polysaccharides via gut microbiota and metabolomics, showing no significant efficacy differences, both treating T2DM by modulating peroxisome proliferator‐activated receptors (PPAR) signaling to regulate insulin resistance. PPAR signaling is a central pathway governing lipid and glucose homeostasis, and its metabolic consequences (e.g., fatty acid oxidation and glucose uptake) are consistently mapped in global diabetes metabolomics as key pathways for therapeutic intervention [89]. The summary of metabolomics studies on the treatment of diabetes mellitus with TCM is presented in Table 6. TABLE 6Summary table of metabolomics studies on metabolic regulation and efficacy evaluation of traditional Chinese medicine and related interventions in diabetes mellitus.Researchers (year)Research method (metabolomics technology)Intervention/study subjectsObserved biomarkers/differential metabolitesAssociated metabolic pathways/mechanismsChen et al. [83]UPLC‐Q‐TOF‐MSCombination of berberine (BBR) and baicalin (BAI) for intervention in Type 2 diabetes mellitus (T2DM)Alpha‐linolenic acid, 15‐keto‐prostaglandin E2, coprocholic acid, PE [16:0/20:4 (8Z, 11Z, 14Z, 17Z)]14 differential metabolic pathways including arachidonic acid metabolism and alpha‐linolenic acid metabolism; gut microbiota–host co‐metabolismAmmar et al. [85]Untargeted GC‐MS metabolomicsComparative antidiabetic potential of Posidonia oceanica extract and its gelatin nanoparticle formulationGlucose, leucine, oleic acid, butane‐2,3‐diolAmino acid metabolism pathway and lipid metabolism pathway are highly correlated with its therapeutic effectsZhao et al. [86]Not specified; focusing on efficacy‐related biomarker screeningLifestyle intervention + Xiaokeqing granules for intervention in prediabetes indicatorsN‐Nervonoyl methionine, 5‐hydroxy‐L‐tryptophan (core potential efficacy biomarkers)Tryptophan metabolism pathway and pantothenate and CoA biosynthesis pathway provide a basis for efficacy prediction and high‐risk population stratificationZhang et al. [88]Untargeted metabolomics (serum + urine samples)Comparison of healthy population, Tai Chi group, and brisk walking group to evaluate the effects of exercise therapy on T2DMLeucine, isoleucine, taurine, 9‐oxo‐octadecenoic acidBranched‐chain amino acid metabolism pathway, steroid hormone biosynthesis pathway, and arachidonic acid metabolism pathway; Tai Chi directly targets the core metabolic abnormality associated with insulin resistance (elevated BCAAs)Wang et al. [84]Metabolomics + gut microbiota analysisComparison of Pueraria lobata and Pueraria thomsonii polysaccharides for intervention in T2DM (both recorded in Chinese Pharmacopoeia for “alleviating xiao ke and lowering blood glucose”)Alpha‐linolenic acid, tauroursodeoxycholic acid, uric acid, lysophosphatidylcholinePPAR signaling pathway (regulates insulin resistance and participates in lipid and glucose homeostasis) Summary table of metabolomics studies on metabolic regulation and efficacy evaluation of traditional Chinese medicine and related interventions in diabetes mellitus. In summary, metabolomics development offers a novel perspective for TCM diabetes research. It precisely identifies differential metabolites and change patterns across syndromes, enhancing the scientificity and objectivity of TCM syndrome diagnosis and reducing the subjective bias of traditional diagnostics [90, 91]. Metabolomics’ vast data systematically depict dynamic metabolic network changes pre‐ and post‐TCM intervention, enabling broad applications in clarifying material basis, mechanisms, and efficacy evaluation of TCM compounds. 4. Summary and Prospects Diabetes has become a major global health issue, yet Chinese herbal medicine boasts millennia of clinical experience. Due to the complexity of components and mechanisms, deciphering the scientific basis and systemic features of TCM in diabetes treatment remains challenging. Current research indicates metabolomics can elucidate TCM’s material basis and mechanisms for diabetes efficacy, aiding in identifying diagnostic TCM syndromes, advancing compound improvements, and developing novel antidiabetic TCM therapies. Metabolomics identifies metabolic pathways and biomarkers in forms understandable and acceptable to the international scientific community, greatly facilitating publication and exchange of TCM research in high‐impact journals and enhancing its global recognition. The application of metabolomics in TCM for diabetes treatment faces several challenges. TCM’s reliance on subjective syndrome differentiation, without objective quantitative indicators, complicates the standardization of diagnoses and confounds metabolic markers. The multicomponent nature of TCM further complicates attributing metabolic changes to specific compounds. Additionally, metabolomics faces limitations such as the masking of low‐abundance metabolites by high‐abundance signals and the need for high‐sensitivity platforms. Small sample sizes and complex data analysis also hinder progress. To address these issues, there is a need for a unified research workflow, standardized criteria, and multicenter validation. Interdisciplinary collaboration among TCM, metabolomics, clinical medicine, and bioinformatics should be strengthened to integrate metabolic mechanisms with TCM’s holistic and syndrome differentiation principles. Focusing on internationally recognized metabolic pathways and biomarkers, a system of syndrome‐ and efficacy‐specific metabolic biomarkers should be developed. Moreover, metabolomics‐guided biomarkers could significantly enhance personalized diabetes management, not just within the TCM framework but also in broader, globally recognized clinical settings. By identifying individual metabolic profiles and responses to therapy, these biomarkers can support the customization of treatment plans, ensuring more effective and targeted interventions. High‐sensitivity and high‐specificity detection technologies should be developed, and a comprehensive database of TCM bioactive components and their metabolic regulatory effects should be established to ensure accurate traceability of metabolic changes induced by TCM. In conclusion, metabolomics offers a novel approach to understanding TCM’s antidiabetic mechanisms and facilitates personalized treatment. With continued technological advancement and interdisciplinary collaboration, metabolomics will play a central role in advancing both TCM‐based and global diabetes therapy. Author Contributions All the authors contributed equally to each subtopic, reviewing, and final editing of the manuscript. Funding This study was funded by the National Key Research and Development Program of China, 2020YFE0201800, and the Henan Province Science and Technology Research and Development Program Joint Fund (Advantageous Discipline Cultivation Category) Project, 242301420107. Disclosure All contents were carefully reviewed and verified by the authors. Conflicts of Interest The authors declare no conflicts of interest. Acknowledgments This work was supported by the National Key Research and Development Program (2020YFE0201800) and the Henan Province Science and Technology Research and Development Program Joint Fund (Advantageous Discipline Cultivation Category) Project (242301420107). We used AI tools for language polishing and grammar checking. Declaration of Artificial Intelligence (AI) Software Usage. The authors used ChatGPT 5.2 solely for grammatical polishing and language refinement, with all AI‐generated revisions reviewed and adjusted for consistency with the original data, scientific accuracy, and academic ethics. The core aspects of the study, including research design, experimentation, data analysis, and conclusion formulation, were completed independently by the authors without AI intervention. The authors confirm that no AI tools were used for data fabrication, result manipulation, literature misrepresentation, or generating unsubstantiated academic viewpoints. Song, Junhui , Li, Gen Lin , Wu, Suhui , Li, Hanbing , Metabolomics in Traditional Chinese Medicine for Diabetes Mellitus: Mechanistic Insights, Biomarker Discovery, and Clinical Application Prospects, International Journal of Analytical Chemistry, 2026, 8261773, 19 pages, 2026. 10.1155/ianc/8261773 Academic Editor: Sohini Basu Roy Contributor Information Suhui Wu, Email: wusuhui@hactcm.edu.cn. Hanbing Li, Email: lhb8899@163.com. Sohini Basu Roy, Email: sbasuroy@wiley.com. Data Availability Statement The data that support the findings of this study are openly available in PubMed at https://pubmed.ncbi.nlm.nih.gov/. References Associated Data Data Availability Statement The data that support the findings of this study are openly available in PubMed at https://pubmed.ncbi.nlm.nih.gov/.