Introduction

Heart failure (HF) represents a major global public health challenge, with continuously increasing prevalence and mortality. Among its clinical phenotypes, heart failure with preserved ejection fraction (HFpEF) now accounts for more than 50% of all HF cases and has emerged as one of the most rapidly increasing forms of HF worldwide.1 The pathogenesis of HFpEF is complex, extends beyond hemodynamic abnormalities, and involves multiple mechanisms, such as systemic metabolic disturbances, chronic low-grade inflammation, endothelial dysfunction, and myocardial remodeling.1

Type 2 diabetes mellitus (T2DM) is among the most prevalent and clinically impactful comorbidities in patients with HFpEF. Substantial clinical evidence indicates that, compared with patients with HFpEF who are not diabetic, those with diabetic HFpEF carry a greater burden of comorbidities, including more severe hypertension, pulmonary complications, and renal dysfunction.2,3 In addition, patients with diabetic HFpEF exhibit more pronounced cardiac structural and functional abnormalities, including elevated ventricular filling pressures, more severe diastolic dysfunction, systemic inflammatory activation, endothelial dysfunction, and worse clinical outcomes.2–4 These alterations are closely associated with metabolic stress, oxidative injury, and myocardial remodeling, suggesting that diabetic HFpEF may represent a distinct metabolic and inflammatory phenotype requiring more targeted therapeutic strategies.4

In recent years, the gut microbiota (GMB) has been recognized as a key regulator of host metabolism, immune homeostasis, and cardiovascular function, leading to the concept of the “gut–heart axis.” Gut dysbiosis, generally characterized by reduced microbial diversity, depletion of beneficial bacteria such as short-chain fatty acid (SCFA)-producing taxa, and enrichment of opportunistic pathogens, has been implicated in the development and progression of both T2DM and HFpEF.5,6 GMB-derived metabolites, including SCFAs, trimethylamine N-oxide (TMAO), and bile acids (BAs), have systemic effects on glucose and lipid metabolism, inflammatory signaling, endothelial function, and myocardial remodeling.5,6 These microbiota-related metabolic and inflammatory processes are further influenced by dietary patterns, highlighting the dynamic interaction between environmental factors and the gut–heart axis.7 Despite recent therapeutic advances, treatment options for HFpEF, particularly diabetic HFpEF, remain limited, and direct clinical evidence specifically linking GMB alterations to diabetic HFpEF is still insufficient, underscoring the need for further investigation of the gut microbiota as a potential therapeutic target.

However, despite increasing recognition of the gut–heart axis, the specific characteristics and pathophysiological mechanisms of GMB in patients with diabetic HFpEF have not been systematically summarized. Therefore, this review aims to summarize (1) the epidemiological and pathophysiological links between T2DM and HFpEF; (2) the pivotal role of the GMB in the development and progression of diabetic HFpEF; and (3) the current status of novel antidiabetic agents for the treatment of patients with diabetic HFpEF from a GMB–modulating perspective. We hope that this review will provide a theoretical foundation for future mechanistic studies and the development of targeted therapeutic approaches.

Epidemiology and Prognosis of Diabetic HFpEF

T2DM is a major risk factor and a common comorbidity of HFpEF. In the early stages, patients with T2DM often exhibit preserved systolic function accompanied by progressive deterioration of diastolic function, a pattern that closely aligns with the pathological characteristics of HFpEF. Epidemiological studies have consistently demonstrated that up to 45% of patients with HFpEF have concomitant T2DM, highlighting the existence of diabetic HFpEF as a distinct clinical phenotype.8 A prospective analysis from the Spanish DIABET-IC cohort (n = 1517 T2DM patients) revealed that approximately 30.6% of HF cases at baseline were classified as HFpEF, and during follow-up, HFpEF accounted for 46.6% of newly developed HF cases.9

The coexistence of T2DM and HFpEF significantly worsens patient prognosis. The DIABET-IC study found that patients with HFpEF tend to be older, are more often female, and have higher prevalences of obesity, hypertension, and metabolic syndrome.9 The TOPCAT trial, a randomized controlled study involving 3,385 patients with HFpEF (left ventricular ejection fraction (LVEF) > 45%), demonstrated that T2DM substantially increased the risks of hospitalization, HF re-hospitalization, all-cause mortality, and cardiovascular death in patients with HFpEF. Notably, the risk of adverse cardiovascular events increased further in patients with T2DM with microvascular complications (such as neuropathy, nephropathy, or retinopathy), exhibiting a gradient increase.10 Another analysis from the I-PRESERVE study, which included 4,128 patients with HFpEF (LVEF >50%), unveiled that patients with diabetic HFpEF had significantly higher rates of cardiovascular mortality, hospitalization, and all-cause mortality.11 Overall, a positive correlation exists between T2DM and HFpEF-associated morbidity and mortality.

Despite the high prevalence of HFpEF in patients with T2DM, its clinical recognition remains challenging. The DIABET-IC study revealed that a considerable number of patients with HFpEF did not reach the traditional diagnostic threshold for brain natriuretic peptide levels, potentially leading to under-diagnosis.9 Moreover, factors such as obesity, renal dysfunction, and other metabolic comorbidities, which are common in patients with T2DM, further complicate the diagnosis and management of HFpEF.12 Therefore, greater emphasis should be placed on early screening and identification of HFpEF in patients with T2DM, especially those with multiple metabolic risk factors.

Pathophysiological Mechanisms of Diabetic HFpEF

The development and progression of diabetic HFpEF are driven by distinct and multilayered pathogenic mechanisms, with central processes involving metabolic remodeling, myocardial structural alterations, endothelial injury, amplification of inflammatory responses, and dysregulation of the cardiorenal axis, reflecting the combined impact of multisystem abnormalities3 (Figure 1).

Figure 1 Proposed pathophysiological mechanisms underlying diabetic HFpEF. Metabolic dysregulation, chronic inflammation, endothelial dysfunction, oxidative stress, mitochondrial impairment, and myocardial fibrosis interact to promote diastolic dysfunction and the progression of HFpEF in patients with T2DM.

Chronic insulin resistance and hyperglycemia shift myocardial metabolism from glucose oxidation toward a predominance of free fatty acid (FFA) utilization. This increases the mitochondrial burden; reduces adenosine triphosphate (ATP) generation efficiency; and leads to lipid droplet accumulation (lipotoxicity), excessive reactive oxygen species (ROS) production, and mitochondrial damage.13–15 These metabolic disturbances ultimately contribute to cardiomyocyte hypertrophy, increased sarcomere stiffness, and impaired active diastolic relaxation, thus forming the fundamental pathological basis of HFpEF.3

In the context of persistent hyperglycemia, the accumulation of advanced glycation end products (AGEs) promotes collagen cross-linking, thereby resulting in irreversible thickening of the extracellular matrix (ECM) and increased myocardial stiffness. This process is a key contributor to the reduced myocardial compliance characteristic of patients with HFpEF.16 Concurrently, AGEs further impair the nitric oxide-cyclic guanosine monophosphate-protein kinase G (NO–cGMP–PKG) signaling pathway by decreasing NO bioavailability, thereby affecting adjacent cardiomyocytes and cardiac fibroblasts and ultimately leading to changes in cardiac structure and function.17 Impairment of this signaling pathway is recognized as a crucial mechanism underlying diastolic dysfunction in patients with HFpEF and is particularly prominent in patients with T2DM.17

Endothelial dysfunction represents another central pathological component of diabetic HFpEF. Insulin resistance leads to reduced expression of endothelial nitric oxide synthase (eNOS) and decreased NO production, thereby weakening coronary vasodilatory capacity and amplifying inflammatory responses and oxidative stress.18 In HFpEF, myocardial remodeling originates, in part, from coronary microvascular endothelial inflammation associated with T2DM. Patients with HFpEF who exhibit endothelial-dependent microvascular dysfunction experience more severe diastolic impairment and have a poorer prognosis.19,20 Furthermore, elevated AGEs and hyperglycemia may cause vascular injury across different vascular beds, affecting microvascular remodeling and angiogenesis, which further promotes structural remodeling in HFpEF.21 A chronic inflammatory state permeates the entire pathophysiological course of diabetic HFpEF. Persistently elevated systemic inflammatory cytokines (such as tumor necrosis factor-alpha [TNF-α] and interleukin [IL]-6) in patients with T2DM promote cardiomyocyte hypertrophy, interstitial fibrosis, and endothelial damage.3,22 Concurrently, T2DM, often cooccurring with obesity, is characterized by the accumulation of epicardial adipose tissue (EAT), an endocrine organ that directly affects adjacent myocardial tissue by releasing paracrine and autocrine factors, thereby promoting fibrosis and diastolic dysfunction.23,24 Additionally, the mechanical compressive effect of increased EAT can increase pericardial restraint pressure, limit left ventricular diastolic filling and heighten pulmonary venous pressure, providing an additional pathological basis for hemodynamic alterations in HFpEF.24,25

Furthermore, approximately 50% of patients with T2DM develop chronic kidney disease, termed diabetic kidney disease (DKD).26 Chronic renal insufficiency is closely associated with increased hospitalization rates and mortality risk in patients with HFpEF.27 The resulting volume overload, activation of the renin-angiotensin-aldosterone system (RAAS), and sympathetic nervous system overactivation exacerbate cardiac preload and afterload, promoting reduced ventricular compliance and elevated filling pressures.28 This vicious interplay among the heart, kidneys, and metabolic system constitutes a major determinant of poor prognosis in patients with diabetic HFpEF.28

Overall, the pathophysiology of diabetic HFpEF is systemic and progressive in nature and involves multiple interconnected mechanisms related to metabolism, vascular dysfunction, inflammation, and the cardiorenal axis. These features indicate that effective clinical management should target several of these key pathways simultaneously to improve the complex HFpEF phenotype.

Relationships Among GMB, T2DM, and HFpEF

The collective genetic material of the GMB, known as the microbiome, forms a gene pool with a total gene count exceeding that of the human genome by an order of magnitude.29 Consequently, the GMB is regarded as a vital “organ” that plays key roles in enhancing host immunity, promoting food digestion, regulating intestinal endocrine functions and neural signaling, influencing drug actions, and modulating metabolic processes and detoxification. This symbiotic relationship with the host ensures the normal development of the human metabolic system. GMB-derived metabolites absorbed by the host can act on receptors in organs such as the liver, gut, brown and white adipose tissue, and central nervous system and participate in physiological processes, including micronutrient synthesis, intestinal motility, and the absorption of minerals and electrolytes.30,31 A growing body of research has provided compelling evidence that the GMB is closely associated with human diseases and plays a particularly crucial role in the pathogenesis of diabetic HFpEF.5,6

GMB and T2DM Alterations in the GMB of Patients with T2DM

Numerous preclinical and clinical studies have suggested a close association between GMB alterations and the development of T2DM. Yu et al revealed significantly altered gut microbiota β diversity in db/db mice compared with that in m/m mice. Furthermore, transplantation of the fecal microbiota from db/db mice into pseudogerm-free mice led to metabolic phenotypes such as increased body weight and elevated fasting blood glucose levels in the recipient mice.32 In a Zucker diabetic fatty rat model, Gu et al revealed that the abundance of Ruminococcus and Allobaculum was positively correlated with random blood glucose levels, while Lactobacillus and Turicibacter were negatively correlated.33 Additionally, in a nonobese T2DM Goto-Kakizaki rat model, Peng et al used multiomics analysis and reported significant enrichment of genera such as Allobaculum, Prevotella, and Roseburia, which were markedly correlated with alterations in glucose metabolism and fecal metabolites.34 Furthermore, in a spontaneous T2DM rhesus macaque model, Jiang et al reported significant gut microbiota dysbiosis characterized by reduced abundance of Phascolarctobacterium and Oribacterium, along with enrichment of opportunistic pathogens, which were closely associated with hyperglycaemia and insulin resistance.35 In human cohort studies (Table 1), Qin et al reported in a metagenome-wide association study involving 345 Chinese patients that the gut microbiome of patients with T2DM exhibited moderate overall dysbiosis. This was characterized by a significant decrease in the abundance of butyrate-producing bacteria (such as Faecalibacterium and Roseburia) and a notable increase in the abundance of various opportunistic pathogens.36 Another study involving patients with newly diagnosed T2DM revealed that the abundance of the genus Lactobacillus was significantly increased and positively correlated with fasting blood glucose and glycated hemoglobin (HbA1c) levels, whereas the abundances of Clostridium coccoides and Clostridium leptum were significantly decreased and negatively correlated with fasting blood glucose, HbA1c, and plasma triglyceride levels.37 Furthermore, Shih et al investigated the gut microbial profile in individuals with refractory T2DM, defined as persistent HbA1c elevation of at least 8% despite ongoing therap.38 Relative to patients with conventional T2DM, those with refractory T2DM exhibited an increased abundance of Bacteroides vulgatus and Veillonella denticariosi, together with a reduced abundance of Akkermansia muciniphila and Fusobacterium. Notably, the proportion of A. muciniphila showed an inverse association with HbA1c levels.38 Collectively, these findings suggest that gut dysbiosis is closely associated with the development and metabolic progression of T2DM.

Table 1 GMB Changes in T2DM and HFpEF

GMB–Derived Metabolites and Their Association with T2DM

In recent years, a growing body of evidence has demonstrated that the GMB contributes to the development and progression of T2DM through the production of multiple metabolites, including SCFAs, endotoxins such as lipopolysaccharide (LPS), BAs, and TMAO, as summarized in Supplementary Table S1.

The GMB ferments dietary fiber to generate SCFAs, primarily acetate, propionate, and butyrate. These fatty acids exert important metabolic effects partly through activation of G protein–coupled receptor 41 (GPR41) and G protein–coupled receptor 43 (GPR43), which are involved in the regulation of insulin sensitivity and metabolic homeostasis.44 Beyond receptor-mediated signaling, propionate suppresses hepatic gluconeogenesis by inhibiting pyruvate carboxylase, whereas butyrate enhances mitochondrial function and fatty acid β-oxidation through the activation of AMP-activated protein kinase (AMPK) and peroxisome proliferator-activated receptor gamma coactivator-1α (PGC-1α), ultimately improving insulin signaling.45 Collectively, these mechanisms may contribute to improved glucose homeostasis through reduced hepatic glucose production and enhanced peripheral glucose utilization. A randomized controlled trial demonstrated that supplementation with inulin-type fructans significantly improved insulin resistance and reduced fasting plasma glucose levels in patients with T2DM.46 In a cross-sectional study conducted in a rural population, individuals in the highest tertile of SCFAs, particularly butyrate and acetate, exhibited a significantly lower prevalence of T2DM than those in the lowest tertile did, and SCFA levels interacted with gut microbial diversity indices, including the Shannon index.47 Moreover, a meta-analysis evaluating the effects of SCFA interventions on fasting insulin and the homeostasis model assessment of insulin resistance (HOMA-IR) revealed that increases in SCFA levels were significantly associated with reductions in fasting insulin concentrations and improvements in HOMA-IR.48 In addition, SCFAs regulate appetite and glucose homeostasis by stimulating intestinal enteroendocrine L cells to secrete anorexigenic hormones, such as glucagon-like peptide-1 (GLP-1) and peptide YY, thereby increasing glucose-stimulated insulin secretion, promoting satiety, and delaying gastric emptying. This process is mediated by FFAR2-dependent intracellular calcium signaling and the activation of proglucagon gene expression in L cells.49 Conversely, gut dysbiosis may impair SCFA-mediated incretin secretion and disrupt satiety signaling, which could contribute to excessive caloric intake and progressive insulin resistance.50

LPS, a major structural component of the cell wall of gram-negative bacteria, has been increasingly implicated in the inflammatory and metabolic disturbances associated with T2DM and is involved in reciprocal regulatory interactions with SCFAs.51,52 Circulating LPS concentrations have been proposed as important biomarkers for predicting the development of multiple inflammation-related diseases linked to innate immunity.53 Experimental and clinical studies have shown that high-fat diet–induced gut dysbiosis upregulates LPS levels, thereby promoting the release of proinflammatory cytokines such as TNF, IL-1, and IL-6, potentially contributing to systemic low-grade inflammation.54 Persistent metabolic endotoxemia may promote activation of host immune responses and maintenance of a chronic proinflammatory state, which may further contribute to the development of metabolic disorders, including T2DM. Toll-like receptor 4 (TLR4), a key transmembrane receptor for LPS, belongs to the Toll-like receptor family. Upon activation, TLR4 enhances the transcription of proinflammatory cytokines, including TNF-α, IL-1, and IL-6, through the nuclear factor-κB (NF-κB) and mitogen-activated protein kinase (MAPK) signaling pathways.55 In patients with T2DM, circulating levels of these inflammatory mediators are markedly elevated, subsequently leading to insulin resistance and pancreatic β-cell dysfunction.55

There is a significant bidirectional interaction between GMB and BAs. BAs are organic acids synthesized in the liver and secreted into the small intestine, where GMB converts primary bile acids (PBAs) into secondary bile acids (SBAs), such as deoxycholic acid (DCA) and lithocholic acid (LCA), through dehydroxylation and deconjugation processes. Studies have shown that increased levels of total bile acids (TBA) and unconjugated bile acids (UBA) are associated with increased abundances of specific GMB taxa and that SBA concentrations, including those of DCA, are positively correlated with the relative abundance of Bacteroides.56 Importantly, these molecules also function as endocrine signaling mediators involved in metabolic regulation.57 By activating nuclear receptors, including the farnesoid X receptor (FXR) and the Takeda G protein–coupled receptor 5 (TGR5), SBAs modulate hepatic glucose metabolism, fatty acid oxidation, incretin hormone secretion, and energy expenditure.58 In T2DM, gut dysbiosis shifts the BA pool toward primary conjugated bile acids, which may reduce TGR5 activation, impair GLP-1 secretion, and contribute to postprandial hyperglycemia. Moreover, a reduction in bile salt hydrolase–producing bacteria, such as Lactobacillus and Bifidobacterium, disrupts BA deconjugation and enterohepatic circulation, promoting metabolic inflammation and hepatic insulin resistance.59 In parallel, SCFAs regulate incretin hormone release via GPR43 and interact with BA signaling pathways, highlighting the integrated role of the gut–liver–hormone axis in glucose homeostasis.60,61

Accumulating evidence indicates that circulating levels of branched-chain amino acids (BCAAs) are elevated in patients with T2DM or insulin resistance.62 In high-fat diet-fed mice, BCAAs aggravate hepatic insulin resistance by inhibiting the Akt2 signaling pathway, which is characterized by increased hepatic gluconeogenesis and suppressed lipogenesis.63 In contrast, another study demonstrated that a low-isoleucine diet enhances hepatic insulin sensitivity and energy expenditure in mice through activation of the fibroblast growth factor 21 (FGF21)–uncoupling protein 1 (UCP1) axis, whereas a low-leucine diet does not have similar effects.64 Under physiological conditions, BCAAs are initially transaminated by branched-chain aminotransferases (BCATs) to branched-chain α-keto acids (BCKAs), which are subsequently irreversibly decarboxylated by the branched-chain α-keto acid dehydrogenase (BCKDH) complex.65 However, an animal study revealed that BCKDH complex activity is markedly reduced across multiple organs in db/db mice and is accompanied by significant accumulation of BCKAs.66 Further investigations have shown that BCKAs induce mitochondrial oxidative stress and promote cytokine production by macrophages, thereby exacerbating inflammation and organ damage in the setting of T2DM.66

Another GMB–derived metabolite, TMAO, has been extensively investigated because its circulating levels are significantly elevated in patients with T2DM.67 Dietary nutrients such as choline and L-carnitine are metabolized by the GMB into trimethylamine (TMA) in the intestine, which is subsequently oxidized to TMAO in the liver. Accumulating evidence suggests that elevated TMAO levels are associated with insulin resistance, impaired glucose tolerance, and inflammatory activation in T2DM.68–70 However, conflicting data indicate that TMAO may also exert protective effects under certain conditions. Chronic exogenous TMAO supplementation has been shown to improve glucose tolerance and enhance insulin secretion in some animal models.71 In addition, as a chemical chaperone, TMAO can stabilize protein conformation, alleviate endoplasmic reticulum stress, and preserve β-cell function under glucolipotoxic conditions.72 It also inhibits the misfolding and amyloid deposition of human islet amyloid polypeptide.73 Overall, the metabolic effects of TMAO in T2DM remain incompletely understood and may be context- or dose-dependent.

Furthermore, tryptophan, an essential amino acid, is metabolized by the GMB into indole and its derivatives, including indole-3-lactate, indole-3-propionate, and indole-3-acetaldehyde. Emerging evidence indicates that tryptophan-derived metabolites are closely involved in the pathogenesis of T2DM.74,75 Novel T2DM-associated metabolites continue to be identified, and their underlying mechanisms require deeper exploration, which may ultimately provide new therapeutic targets for T2DM management.

GMB and HFpEF Alterations in the GMB of Patients with HFpEF

Compared with healthy controls, patients with HFpEF exhibit consistent structural and functional alterations in their GMB, suggesting the presence of gut dysbiosis characterized by a reduction in beneficial bacteria and a relative enrichment of harmful or opportunistic pathogens (Table 1). In a 16S rRNA sequencing study involving 26 patients with HFpEF and 67 control participants, significant differences in both α-diversity and β-diversity were observed between the HFpEF and control groups, indicating substantial changes in microbial composition and richness. Notably, the abundance of Ruminococcus, a well-recognized SCFA-producing genus, was markedly depleted and represented a major contributor to the differences in microbes between patients with HFpEF and controls.39 Consistent with these findings, the GUMPTION study (30 patients with HFpEF vs 30 healthy controls) used high-throughput sequencing to demonstrate significantly reduced abundances of anti-inflammatory taxa, including Butyricicoccus, Lachnospira, and Ruminiclostridium, as well as a relative enrichment of proinflammatory genera such as Enterococcus and Lactobacillus. In addition, a declining trend in α diversity was observed in the HFpEF group.40 Furthermore, in a subgroup analysis of patients with HFpEF, Drapkina et al reported a marked increase in the abundance of proinflammatory taxa, including members of the family Erysipelotrichaceae.41 Collectively, these findings support a potential association between gut dysbiosis and altered inflammatory or metabolic signaling in HFpEF. Moreover, gut dysbiosis is associated with established markers of myocardial fibrosis, such as C-terminal propeptide of procollagen type I (PICP), N-terminal propeptide of pro-collagen type II (PIIINP), and left ventricular extracellular volume (ECV), further suggesting a possible link between gut microbial alterations and myocardial remodeling in HFpEF.42,43

GMB–Derived Metabolites and Their Association with HFpEF

The GMB interacts with the host cardiovascular system through its metabolites, among which SCFAs, TMAO, and BAs have attracted substantial research interest and may exert important pathogenic or protective effects in the setting of HFpEF, as summarized in Supplementary Table S1.

Current evidence largely supports a cardioprotective role of SCFAs in the setting of heart failure. In the failing heart, SCFAs can serve as efficient energy substrates, thereby restoring myocardial contractile function.76 SCFAs also exert their physiological effects through G protein–coupled receptors (GPCRs), including blood pressure-lowering properties and protective effects on endothelial function.77 In a murine study, Marques et al demonstrated that a high-fiber diet or acetate supplementation significantly reduced blood pressure and attenuated cardiac fibrosis and left ventricular hypertrophy by downregulating early growth response protein 1 expression.78 Additionally, SCFAs play a critical role in immune and inflammatory regulation by modulating inflammatory gene expression via GPCR and downstream NF-κB and MAPK signaling pathways or by directly inhibiting histone deacetylase activity after entering cells.79 Moreover, SCFAs have beneficial effects on GMB homeostasis and intestinal barrier integrity. Small intestinal bacterial overgrowth, a common manifestation of gut dysbiosis, has been associated with an increased risk of cardiovascular mortality in patients with HFpEF.80 In a study of 60 patients with acute HF, Mollar et al reported a significant inverse association between butyrate levels and small intestinal bacterial overgrowth.81 SCFAs can also enhance intestinal epithelial barrier function by upregulating the transcription of tight junction proteins, such as claudin-1.82 In patients with HFpEF, however, depletion of SCFA-producing microbial taxa compromises SCFA synthesis,39 thereby weakening the protective effects of SCFAs on both the gut and heart.

Studies have shown that plasma TMAO levels are significantly greater in patients with HFpEF than in healthy individuals, and alterations in GMB composition are considered major drivers of elevated TMAO levels.83 Moreover, increased TMAO levels are independently associated with a higher composite endpoint of rehospitalization and cardiac death in patients with HFpEF, with this effect being more pronounced in malnourished patients.84 Additionally, TMAO can serve as a risk stratification biomarker for HFpEF, particularly providing more sensitive prognostic information when B-type natriuretic peptide levels are not elevated.85 TMAO participates in the development and progression of cardiac dysfunction through multiple mechanisms. First, TMAO may contribute to cardiac diastolic dysfunction. A prospective cohort study involving 112 samples revealed that TMAO was positively correlated with diastolic function indices such as the mitral E-wave velocity to early diastolic mitral annular velocity (E/e’) ratio and left atrial volume index.86 Chen et al also reported that elevation of TMAO levels by a high-glucose and high-fat diet promotes cardiac inflammation and fibrosis, impairing diastolic function.87 TMAO acts by upregulating the expression of transforming growth factor-β (TGF-β)/mothers against decapentaplegic homolog 3 (SMAD3).88 Administration of TMAO synthesis inhibitors can modulate the activity of the TGF-β/SMAD3 signaling pathway, thereby preventing myocardial hypertrophy and fibrosis.89 Other studies have also demonstrated that TMAO inhibits the expression of Sirtuin3 and reduces the activity of superoxide dismutase 2 and mitochondrial aldehyde dehydrogenase 2, leading to the accumulation of mitochondrial reactive oxygen species, activation of the nucleotide-binding domain, leucine-rich-containing family, Nod-like receptor pyro-protein domain-associated protein 3(NLRP3) inflammasome, and generation of N-terminal Gasdermin D, thereby inducing cardiovascular endothelial inflammation.90 Furthermore, TMAO can stimulate hepatocytes to secrete exosomes, which cause vascular endothelial diastolic dysfunction by activating the NF-κB signaling pathway.91 In cardiomyocytes, modulation of the TMAO/Protein kinase C (PKC)/NF-κB pathway can suppress myocardial hypertrophy and fibrosis, thereby improving cardiac functional abnormalities.92 However, controversy remains regarding the specific role of TMAO in HFpEF. Although numerous studies have identified it as a pathogenic factor in cardiovascular diseases, some evidence suggests that TMAO may merely serve as a marker of underlying cardiovascular disease.93 Moreover, certain studies have even indicated that TMAO might exert protective effects in HF.94,95 Therefore, the role of TMAO in HFpEF remains inconclusive, and its precise mechanisms in disease development, as well as its potential as a therapeutic target, warrant further investigation.

Recent studies have indicated that BAs can act as signaling molecules that mediate cardiac function.96 Based on their molecular structure, BAs can be classified into hydrophilic and hydrophobic types, each of which have distinct effects on cardiac health. Hydrophobic BAs are associated with QT interval prolongation, cardiac hypertrophy, cardiomyocyte apoptosis, and cardiac hemodynamic dysfunction.97 In contrast, hydrophilic BAs, as steroid signaling molecules, can regulate cardiac inflammation through multiple pathways, thereby exerting cardioprotective effects. These pathways include activating the nuclear receptors FXR and TGR5 in the heart and inhibiting the NF-κB signaling pathway.98–100 TGR5 activation initiates downstream protein kinase pathways in cardiomyocytes, subsequently alleviating cardiomyocyte inflammation and oxidative stress and improving cardiomyocyte survival.101 Animal studies have shown that treatment with taurodeoxycholic acid can attenuate hypertension-induced cardiac inflammation and myocardial remodeling.102 Cell culture studies have demonstrated that TGR5 activation by LCA can ameliorate high-glucose-induced cardiomyocyte hypertrophy.103 BAs are also involved in regulating the stability of the gut ecosystem. In fact, bile salts possess certain antimicrobial activity and can exert selective pressure on the microbiota, thereby modulating the GMB composition.104 PBAs can promote the recovery of the microbiome after dysbiosis and inhibit the overgrowth of pathogenic bacteria in the small intestine.105 Studies have also shown that mice lacking FXR exhibit increased bacterial numbers in the ileum and impaired intestinal barrier function, indicating that this receptor suppresses bacterial overgrowth and mucosal damage in the ileum.106 TGR5 knockout mice display a disorganized molecular structure of colonic tight junctions and increased intestinal permeability.107

The GMB is known to be closely associated with amino acid metabolism. Dysregulation of amino acid metabolism is linked to the pathophysiological mechanisms of HFpEF. For instance, patients with HF who have higher plasma phenylalanine levels exhibit elevated levels of C-reactive protein and inflammatory cytokines (IL-8, IL-10), as well as higher mortality,108 whereas glycine demonstrates anti-inflammatory effects and exerts protective effects on cells and the heart.109 Amino acid metabolites derived from the gut microbiota also influence HFpEF. In animal experiments, indoxyl sulfate, a metabolite of tryptophan, may increase the expression of pro-inflammatory cytokines (such as TNF-α and IL-1β) in cardiomyocytes. This upregulation is mediated by the activation of signaling pathways, including p38 MAPK, p42/44 MAPK, and NF-κB, thereby inducing myocardial fibrosis and hypertrophy, ultimately leading to diastolic dysfunction.110 Wang et al reported that exogenous supplementation with another tryptophan metabolite, indole-3-propionic acid, alleviated inflammation, oxidative stress, diastolic dysfunction, and myocardial remodeling in a HFpEF mouse model.111 Furthermore, phenylacetylglutamine, a fermentation product that promotes thrombosis, may also be associated with HFpEF.112 Additionally, the accumulation of BCAAs and their intermediate metabolites in the myocardium is related to HF development and progression.113,114

Additional metabolites with potential associations with HFpEF continue to be discovered. For example, polyphenols are plant-derived antioxidants that can be converted into bioactive metabolites by specific enzymes of gut microbial origin.115 Polyphenols can improve HF-induced gut dysbiosis and exert multiple effects, such as antioxidant, anti-inflammatory, endothelial-protective, and anti-myocardial fibrosis actions, which may be beneficial for HFpEF.116–119 Moreover, other GMB-derived metabolites, such as trimethyl-5-aminovaleric acid, are associated with cardiac energy metabolism.120 Current research on these newly identified metabolites remains insufficient, and their association with HFpEF requires further experimental evidence for validation.

Role of the GMB in Diabetic HFpEF

The GMB is increasingly recognized as being closely linked to the pathophysiology of diabetic HFpEF (Figure 2). However, direct evidence linking gut microbiota alterations to diabetic HFpEF remains limited, and many proposed mechanisms are extrapolated from studies of T2DM and HFpEF separately. The GMB maintains intestinal barrier integrity by regulating the expression of epithelial tight junction proteins, such as occludin and claudins, a process that is influenced by diet, microbial diversity, and metabolite availability. In individuals with T2DM, the GMB is typically characterized by reduced microbial diversity, depletion of SCFA-producing bacteria (such as Faecalibacterium prausnitzii and Roseburia), and a relative enrichment of opportunistic pathogens and gram-negative bacteria (such as Escherichia/Shigella and certain Proteobacteria).121 These alterations are increasingly recognized as being closely associated with insulin resistance, chronic low-grade inflammation, and metabolic dysregulation.121 Reduced SCFA production compromises intestinal barrier function and anti-inflammatory signaling, whereas an increased abundance of gram-negative bacteria increases the LPS burden, leading to increased intestinal permeability.122 Consequently, luminal contents, including LPS, translocate across the epithelium into the portal and systemic circulation, potentially contributing to metabolic endotoxemia, systemic inflammation, and vascular injury.123

Figure 2 The role of GMB in the pathogenesis of diabetic HFpEF. Gut dysbiosis promotes metabolic endotoxemia, chronic inflammation, endothelial dysfunction, and myocardial fibrosis through altered microbiota-derived metabolites (eg, reduced SCFAs and increased TMAO), ultimately contributing to HFpEF progression.

Systemic low-grade inflammation is considered a central pathological feature of HFpEF, and gut dysbiosis may represent one of its important upstream contributors. Following disruption of the intestinal barrier, LPS and other microbe-associated molecular patterns enter the circulation and activate TLR4 via the CD14 coreceptor, leading to activation of the NF-κB signaling pathway and sustained release of proinflammatory cytokines, including IL-6, TNF-α, and IL-1β.124 In parallel, macrophages and monocytes are activated, inducing NLRP3 inflammasome activation and further promoting the release of IL-1β and IL-18.124 This inflammatory response is not acute but rather chronic, low-grade, and systemic, closely mirroring the inflammatory phenotype observed in patients with HFpEF. They progressively drive structural and functional remodeling in HFpEF by impairing endothelial function, disrupting metabolic homeostasis, and activating cardiac fibroblasts.125

The GMB–inflammation axis further regulates a key pathological hallmark of HFpEF: microvascular endothelial dysfunction. Proinflammatory cytokines markedly suppress eNOS activity through the activation of protein kinase C/NF-κB signaling pathways, thereby reducing NO bioavailability and leading to impaired vasodilation, decreased vascular compliance, and increased oxidative stress.126 In addition, inflammatory mediators induce endothelial expression of adhesion molecules, such as vascular cell adhesion molecule-1 and intercellular adhesion molecule-1, facilitating leukocyte adhesion and infiltration within the microvascular bed and further exacerbating microvascular rarefaction and focal ischemia.126 In contrast, SCFAs help preserve vascular tone and hemodynamic balance by activating GPR41 and GPR43 and enhancing NO production. When gut dysbiosis leads to reduced SCFA generation, this protective effect is lost, rendering the endothelium more susceptible to inflammation-induced injury.127 These microvascular and endothelial alterations may contribute to impaired myocardial relaxation and increased ventricular stiffness in HFpEF and may be further exacerbated by chronic inflammatory signaling associated with gut dysbiosis.

Concurrently, multiple GMB-derived metabolites act as signaling molecules that directly mediate the pathophysiology of HFpEF. Among them, TMAO is one of the most robustly implicated microbiota-derived metabolites conferring cardiovascular risk. TMAO contributes to HFpEF progression by activating the NLRP3 inflammasome, aggravating the inflammatory burden, impairing endothelial function and promoting vascular stiffness, and directly activating cardiac fibroblasts via the TGF-β/Smad signaling pathway, thereby accelerating collagen deposition.125,126 In patients with T2DM, dietary patterns and GMB composition often favor increased TMA production, resulting in elevated circulating TMAO levels. These findings support a potential mechanistic link between elevated TMAO levels and metabolic-cardiovascular dysfunction in diabetic HFpEF.128 Under physiological conditions, SCFAs exert multiple protective effects by activating GPR41/GPR43 and inhibiting histone deacetylases, thereby suppressing inflammation, improving insulin sensitivity, and maintaining intestinal barrier integrity. SCFAs also directly protect the cardiovascular system by preserving endothelial function, reducing oxidative stress, and inhibiting inflammasome activation.129,130 In patients with diabetic HFpEF, however, reduced SCFA production weakens these protective mechanisms. In addition, the GMB modulates bile acid metabolism and signaling through FXR and TGR5, which play critical roles in glucose and lipid metabolism as well as inflammatory control.131 When gut dysbiosis disrupts bile acid signaling, host metabolic and immune regulation are impaired, further accelerating the progression of diabetic HFpEF.

Under the combined influence of chronic inflammation, endothelial injury, and metabolite imbalance, the myocardium undergoes metabolic remodeling, characterized by reduced efficiency of fatty acid and glucose utilization, mitochondrial dysfunction, and insufficient energy supply.132 Mitochondrial impairment and lipid toxicity further induce oxidative stress and aberrant intracellular signaling, creating a permissive metabolic and oxidative environment for fibroblast activation and myocardial interstitial remodeling.132 Moreover, the combined effects of TMAO, chronic inflammation, and intestinal barrier disruption activate cardiac TGF-β/Smad signaling, driving excessive collagen production by fibroblasts and leading to diffuse interstitial fibrosis, histopathologically recognized as a key substrate underlying diastolic dysfunction in the setting of HFpEF.132 Notably, HFpEF-related microvascular dysfunction, reduced exercise tolerance, and persistent inflammation may, in turn, exacerbate insulin resistance and impair metabolic control, thereby establishing a vicious cycle.133

In summary, accumulating evidence suggests that the GMB is closely associated with the metabolic, inflammatory, and cardiovascular alterations observed in diabetic HFpEF. Nevertheless, most currently available evidence is derived from experimental studies or relatively small clinical cohorts, and substantial interindividual variability in microbial composition, dietary patterns, and analytical methodologies remains a major challenge in this field.39

Role of Novel Antidiabetic Drugs in Diabetic HFpEF

In recent years, multiple clinical trials have demonstrated that novel antidiabetic agents, including sodium–glucose cotransporter 2 inhibitor (SGLT2i), glucagon-like peptide-1 receptor agonist (GLP-1RA), and dual glucose-dependent insulinotropic polypeptide (GIP)/GLP-1 receptor agonists, not only exert robust glucose-lowering effects but also improve clinical outcomes in patients with HFpEF.134–138 The cardiovascular benefits conferred by these agents may be mediated, at least in part, through modulation of the GMB composition and function. Specifically, alterations in certain microbial taxa and microbiota-derived metabolites have been associated with improvements in metabolic status, systemic inflammatory burden, and vascular biomarkers,139 highlighting the gut–heart axis as a potential mechanism underlying the therapeutic effects of these novel antidiabetic drugs.

SGLT2i

SGLT2i are a class of antidiabetic agents that primarily lower blood glucose levels by inhibiting glucose reabsorption in the proximal renal tubules and promoting urinary glucose excretion. Robust evidence has demonstrated clear cardiovascular benefits of SGLT2i in patients with diabetic HFpEF.134,135 Notably, the therapeutic benefits of SGLT2i in patients with HFpEF are not entirely dependent on their glucose-lowering effects. Meta-analyses have shown that, in patients with diabetic HFpEF, SGLT2i significantly improve LVEF and exercise tolerance while reducing the levels of markers of myocardial fibrosis and the risk of heart failure-related hospitalization.140 Mechanistically, SGLT2i induce natriuresis and osmotic diuresis, leading to a reduction in plasma volume, a decrease in blood pressure, and an improvement in both preload and afterload.141 In addition, SGLT2i have been shown to increase myocardial utilization of ketone bodies and fatty acids, improve mitochondrial energy metabolism, increase cardiac energy efficiency, and reduce lactate accumulation and oxidative stress, thereby alleviating myocardial stress and providing systemic support for HFpEF improvement.142

Emerging evidence suggests that some of the cardioprotective effects of SGLT2i may be closely related to the modulation of the GMB and its metabolites. SGLT2i therapy has been shown to increase the abundance of SCFA-producing bacteria, such as Faecalibacterium, Roseburia, and Bifidobacterium, while reducing the Firmicutes-to-Bacteroidetes ratio and the abundance of potentially pathogenic taxa, including members of the Enterobacteriaceae family, thereby promoting a shift toward a more balanced gut microbial ecosystem.143–145 Following SGLT2i treatment, the levels of colonic SCFAs, particularly butyrate and acetate, significantly increase. These metabolites contribute to improved insulin sensitivity, glucose homeostasis regulation, and appetite control and exert anti-inflammatory and cardioprotective effects.146,147 Concurrent reductions in systemic LPS levels further support these metabolic and anti-inflammatory benefits.148 Moreover, SGLT2i may modulate the gut microbial environment by altering intestinal pH, bile acid composition, and local nutrient availability, thereby further influencing microbial growth conditions and metabolic activity. Preliminary clinical evidence indicates that the cardiovascular and metabolic benefits of SGLT2i in patients with T2DM may be partially mediated through GMB modulation and systemic metabolic remodeling. For example, in a randomized clinical trial involving 76 treatment-naïve adults with T2DM and at least one cardiovascular risk factor, compared with metformin, empagliflozin uniquely improved cardiovascular risk profiles, increased GMB diversity and the abundance of SCFA-producing taxa, and shifted circulating metabolites toward a more favorable cardiometabolic profile.149 Similarly, an 8-week single-arm clinical study including 12 patients with T2DM who had been treated with metformin reported that dapagliflozin treatment reduced the abundance of Fusobacterium, which was associated with improvements in glycemic control, lipid metabolism, inflammatory markers, and indices of endothelial function.150 In addition, an ongoing clinical trial (the EMPAGUM study) is designed to directly evaluate the effects of SGLT2i on GMB composition and microbiota-derived metabolites in patients with HFpEF, representing a critical step toward validating the SGLT2i–gut–heart mechanistic axis.151

Although direct evidence remains limited, accumulating data suggest that SGLT2i may attenuate key pathological drivers of diabetic HFpEF by improving GMB composition and modulating microbiota-derived metabolic pathways.

GLP-1RA

GLP-1RA are another class of glucose-lowering agents widely used to treat T2DM. Their physiological basis lies in the incretin hormone GLP-1, which is secreted by intestinal L cells. In addition to stimulating glucose-dependent insulin secretion, GLP-1 delays gastric emptying, suppresses appetite, and improves lipid metabolism. Cardiovascular outcome trials have demonstrated multiple benefits of GLP-1RA, including body weight reduction, blood pressure lowering, attenuation of inflammation, and improvement in cardiac remodeling.140 Notably, the STEP-HFpEF trial program has provided direct evidence that GLP-1RA significantly improves heart failure-related symptoms, physical limitations, and exercise capacity in patients with HFpEF.136,137 Mechanistically, the beneficial effects of GLP-1RA on HFpEF may involve several pathways, such as a reduction in epicardial adipose tissue thickness, inhibition of renin–angiotensin–aldosterone system activation, optimization of myocardial energy metabolism, and alleviation of systemic inflammation and cardiac oxidative stress.152

An increasing body of evidence indicates that GLP-1RA may exert systemic beneficial effects through modulation of the GMB. Several mechanisms have been proposed. First, GLP-1RA treatment can improve the GMB composition by restoring microbial balance and increasing bacterial diversity. For example, liraglutide therapy has been associated with increased abundance of beneficial taxa such as Lactobacillus, Bacteroides, and Akkermansia muciniphila, along with a reduction in the Firmicutes-to-Bacteroidetes ratio.148,153 Second, GLP-1RA can modify the intestinal milieu by regulating luminal pH, BA profiles, and gut motility, thereby creating favorable ecological niches for specific microbial populations and influencing overall GMB richness.154,155 In both human and animal studies, GLP-1RA treatment increased the abundance of SCFA-producing bacteria, such as Roseburia and Faecalibacterium, which are associated with improved glycemic control, enhanced intestinal barrier integrity, and reduced inflammatory burden.156,157 At the level of microbial metabolites, GLP-1RA induce marked shifts in the GMB-derived metabolic profile, including increased concentrations of SCFAs such as butyrate and propionate and reduced levels of gut-derived endotoxins, particularly LPS.153,155 Importantly, a bidirectional interaction exists between GLP-1, bile acids, and the gut microbiota. Bacteria-derived SBAs (including DCA) can promote GLP-1 secretion by modulating intestinal BA receptors such as TGR5 and FXR,158 while liraglutide treatment has been shown to increase circulating deoxycholic acid levels,159 suggesting reciprocal regulation through BA signaling pathways. These GMB-related alterations and metabolite changes may actively contribute to the systemic effects of GLP-1RA and partially explain their cardiovascular benefits beyond glycemic control. For instance, improvements in endothelial function and attenuation of atherosclerotic progression have been linked to increased abundance of Akkermansia muciniphila,153,160 whereas enrichment of SCFAs has been associated with reduced oxidative stress and more stable glycemic variability.161

However, similar to SGLT2i, direct clinical evidence demonstrating that GLP-1RA improves diabetic HFpEF through GMB-mediated mechanisms remains limited. Future studies incorporating high-resolution microbiome profiling and mechanistic analyses in well-characterized patient populations are needed to substantiate this proposed pathway.

GIP/GLP-1 Dual-Receptor Agonists

The GIP/GLP-1 dual receptor agonist tirzepatide has not only been approved for the treatment of T2DM but has also demonstrated cardiovascular benefits in recent studies. The SUMMIT trial reported that tirzepatide reduced the risk of worsening HF or cardiovascular death in obese patients with HFpEF while alleviating symptom severity and improving exercise tolerance.138 These benefits are likely mediated not only through improvements in glucose and lipid metabolism but also via modulation of host GMB composition and function.

Preclinical studies have highlighted the ability of tirzepatide to modulate the gut microbial ecosystem. In high-fat diet-induced diabetic mouse models, tirzepatide ameliorated gut dysbiosis by promoting the restoration of ben