Glucagon-like Peptide-1-based and Emerging Cardiometabolic Therapies in Heart Failure with Preserved Ejection Fraction: Linking Molecular Mechanisms to Clinical Outcomes

Abstract

Heart failure with preserved ejection fraction (HFpEF) is a heterogeneous clinical syndrome driven by frequent coexistence of cardio-renal-metabolic comorbidities. It has growing epidemiological relevance as global populations age. At its pathophysiological core lies a state of chronic systemic inflammation, coronary microvascular dysfunction and myocardial remodelling, a cascade where epicardial adipose tissue plays a pivotal amplifying role. This complexity is driving a fundamental shift in HFpEF management towards integrated cardiometabolic targeting. Glucagon-like peptide-1 receptor agonists (GLP-1RAs) and the dual GLP1/glucose-dependent insulinotropic polypeptide (GIP)-RA tirzepatide have emerged as compelling multi-system agents, simultaneously addressing several key pathophysiological drivers of the syndrome with consistent improvements in symptoms, functional capacity and clinical outcomes. Nevertheless, questions remain regarding the long-term impact on cardiovascular outcomes, consequences of discontinuing treatment and the potential loss of muscle mass. This review examines the pathophysiological basis of HFpEF within the cardio-renal-metabolic continuum, appraising the clinical evidence, benefits and limitations of GLP-1RAs in HFpEF and discussing future directions including phenotype-directed therapy, mechanism-based combination strategies and next-generation molecules.

Heart Failure with Preserved Ejection Fraction as a Cardio-renal-metabolic Syndrome

Heart failure (HF) with preserved ejection fraction (HFpEF) is increasingly recognised as a systemic, heterogeneous clinical syndrome in which cardiac dysfunction develops within a complex milieu of metabolic disturbances, chronic low-grade systemic inflammation and progressive multi-organ involvement.1

This pathophysiological framework is reflected by the high burden of coexisting comorbidities, most prominently obesity, type 2 diabetes (T2D), arterial hypertension and chronic kidney disease (CKD), which interact and amplify metabolic, inflammatory and neurohormonal pathways in a self-perpetuating cycle described as cardio-renal-metabolic syndrome.2,3

As the global population ages and the prevalence of chronic comorbidities continues to rise, HFpEF now accounts for more than half of all HF hospitalisations and is projected to increase further. Its clinical burden is substantial because patients frequently experience chronic dyspnoea, exercise intolerance and a progressive decline in functional capacity, leading to worsening disability, recurrent hospitalisations and an overall mortality rate of approximately 68% over a median of 5 years following acute decompensation.1

Therapeutic Paradigm Shift in HFpEF

Given its complex clinical presentation, effective management of HFpEF requires a comprehensive approach integrating the optimisation of comorbidities, non-pharmacological interventions and personalised pharmacological strategies. Despite its substantial clinical burden, disease-modifying therapies for HFpEF have remained elusive for decades. Recently, the therapeutic landscape has been reshaped by the introduction of sodium-glucose cotransporter 2 inhibitors (SGLT2i) as first-line therapy.4 However, the significant residual risk for HFpEF patients remains high, largely driven by the underlying cardiometabolic burden.5 This is driving a fundamental shift in the therapeutic paradigm, from a primary focus on haemodynamic management centred on decongestion and the reduction of afterload, to strategies targeting the upstream cardiometabolic and inflammatory drivers of the disease.

To understand this transition and the emergence of cardiometabolic therapies, it is essential to first dissect the complex biological and molecular mechanisms underlying HFpEF pathogenesis.

Pathophysiological Mechanisms of HFpEF

The haemodynamic and structural abnormalities characterising HFpEF, including impaired ventricular relaxation, increased stiffness and elevated filling pressures, are closely intertwined with metabolic and inflammatory processes.

Ventricular-arterial coupling (VAC), expressed as the ratio of effective arterial elastance (Ea) to left ventricular (LV) end-systolic elastance (E_es), has emerged as a central modulator of cardiac function in HFpEF. As highlighted in a recent review by Li et al., age-related arterial stiffening, which can occur especially when obesity, T2D, hypertension and smoking coexist, initially triggers a compensatory increase in E_es.6 However, this adaptive reserve becomes exhausted over time, leading to impaired LV compliance and reduced functional capacity, ultimately resulting in the clinical HFpEF phenotype. Pressure and volume overload increase cardiac workload and promote structural remodelling while coexisting metabolic comorbidities sustain a state of chronic low-grade systemic inflammation, reflected by elevated circulating levels of C-reactive protein (CRP), tumour necrosis factor-α (TNF-α) and interleukin-6 (IL-6), which collectively accelerate disease progression.7 These alterations contribute to VAC impairment by promoting vascular inflammation and arterial stiffening especially in the context of obesity and insulin resistance.6,7 The consequent increase in haemodynamic stress can also lead to left atrial (LA) dysfunction, worsening pulmonary haemodynamics and predisposition to AF, a common precursor to clinically overt HFpEF.

At the myocardial level, this inflammatory environment promotes coronary microvascular endothelial dysfunction characterised by reduced bioavailability of nitric oxide, impaired cyclic guanosine monophosphate-protein kinase G signalling and increased oxidative stress. These alterations compromise cardiomyocyte energetics and favour interstitial fibrosis by infiltration of immune cells and activation of profibrotic pathways, including macrophage-mediated transforming growth factor-β (TGF-β) signalling and maladaptive differentiation of resident perivascular progenitor populations into matrix-producing cells.7–11

Within this framework, adipose tissue, particularly visceral and epicardial fat, plays a pivotal and increasingly recognised role, especially in the obese-related phenotype. Under physiological conditions, epicardial adipose tissue (EAT) functions as a biologically active organ that contributes to thermoregulation and exerts anti-inflammatory and cardioprotective effects. However, in obesity, EAT undergoes maladaptive expansion and acquires a proinflammatory phenotype. As EAT accumulates, its physiological benefits and mechanical protection of the heart are progressively outweighed by its harmful biological activity. Dysfunctional EAT can directly influence the adjacent myocardium through paracrine signalling, which promotes endothelial dysfunction, oxidative stress and myocardial fibrosis. Obesity is associated with increased EAT mass and cardiac and renal fibrosis, supporting its role as a major contributor to the development of incident HFpEF.12,13 Through these mechanisms, Li et al. identified EAT as a pathophysiological amplifier of VAC dysfunction in HFpEF, where EAT-derived cytokines promote profibrotic signalling and oxidative stress accelerating myocardial fibrosis and microvascular rarefaction. EAT burden is consistently increased in patients with HFpEF compared to matched controls, and is associated with a worse haemodynamic and inflammatory profile.6

Visceral adiposity can exert pathogenic effects that extend beyond the heart, promoting metabolic dysfunction-associated steatotic liver disease (MASLD). Accordingly, MASLD and HFpEF are increasingly recognised as organ-specific manifestations of the same systemic metabolic disorder, interconnected through a bidirectional heart-liver axis. Hepatic release of hepatokines and extracellular vesicles amplifies systemic inflammation and shapes myocardial structure, energetics and diastolic function. In turn, cardiac congestion and altered haemodynamics further aggravate hepatic stress, reflecting maladaptive inter-organ communication. Improving hepatic inflammation may translate into clinical benefits in HFpEF, while treating MASLD may improve HF outcomes.14

Finally, the role of visceral adiposity is also recognised in the development of CKD. Obesity induces a state of glomerular hyperfiltration and the activation of the sympathetic nervous and renin-angiotensin systems. The expansion of perirenal fat contributes to intrarenal haemodynamic derangements, promoting progressive decline in renal function and reinforcing the cardio-renal interactions that characterise HFpEF.15

Incretin-based Therapies: ‘Multi-system’ Agents

Glucagon-like peptide-1 receptor agonists (GLP-1RAs) and dual GLP-1/glucose-dependent insulinotropic polypeptide (GIP) receptor agonists are gaining recognition as compelling therapeutic classes in HFpEF. By simultaneously targeting multiple molecular pathways, they induce a broad improvement in systemic metabolic homeostasis.16

Mechanistically, these agents are incretin mimetics that replicate and enhance the physiological actions of gut-derived incretin hormones released in response to nutrient intake. Activation of the GLP-1 receptor in the gastrointestinal tract, pancreatic islets and central nervous system slows gastric emptying, enhances glucose-dependent insulin secretion, suppresses glucagon release and promotes satiety via hypothalamic pathways. Collectively, these effects result in reduced appetite, improved glycaemic control and substantial weight loss.17

Incretin receptor signalling also extends to the cardiovascular (CV) system, vascular smooth muscle cells, adipose tissue, including EAT, and kidney, liver and immune cells. This distribution provides a mechanistic basis for pleiotropic effects extending beyond weight loss and glycaemic control, including improvements in endothelial function, systemic vascular resistance, atherosclerotic burden and lipid metabolism.18–23 More broadly, GLP-1RAs have emerged as ‘multi-system’ agents with potential effects across a wide range of conditions, including CV diseases, MASLD, musculoskeletal inflammation and neurodegenerative disorders.24

Dual GLP-1/GIP receptor agonism with tirzepatide may further amplify these benefits through pharmacologically distinct and complementary receptor signalling. At the GLP-1 receptor, tirzepatide acts as a biased agonist, preferentially activating cyclic adenosine 3,5-monophosphate (cAMP)-mediated signalling pathways, while showing limited recruitment of β-arrestin compared with native GLP-1. This is pharmacologically relevant since β-arrestin engagement promotes receptor internalisation and desensitisation and its relative attenuation may sustain receptor responsiveness and prolong insulinotropic activity. At the GIP receptor, tirzepatide acts as a full agonist, promoting glucose-dependent insulin secretion and glucagon regulation while also exerting direct effects on adipose tissue and skeletal muscle that improve insulin sensitivity and lipid metabolism.25 This synergistic receptor activity is likely to explain the marked metabolic efficacy of tirzepatide and its potential superiority over semaglutide in terms of greater glycaemic improvement, more pronounced weight loss and favourable tolerability as recently emerged from the SURMOUNT-5 trial.26

Established Cardio-renal Benefits of GLP-1 RAs

The cardiorenal protective effects of GLP-1RAs and dual GLP-1/GIP in T2D, obesity and CKD are now well established across landmark randomised controlled trials (RCTs).26–28 Trials looking at CV outcomes in T2D provided the first signs of the benefits of GLP-1RAs in HF. Although not specifically powered to detect differences in HF outcomes, RCTs testing injectable GLP-1RAs in T2D have shown a significant overall reduction in the risk of HF hospitalisation.29 However, in the context of HF, the evidence is strikingly heterogeneous across ejection fraction categories, a pattern that carries important pathophysiological implications.

In HF with reduced ejection fraction (HFrEF), results have been largely neutral or inconclusive. The LIVE and FIGHT trials demonstrated no benefit of liraglutide on clinical outcomes or LV function in stable HFrEF.30 This lack of efficacy likely reflects the distinct pathophysiology of HFrEF, where neurohormonal activation and systolic dysfunction predominate over the metabolic and inflammatory pathways that incretin-based therapies most effectively target.31 Notably, a prespecified analysis of the SELECT trial showed that semaglutide reduced major adverse CV events and HF-related outcomes across the full spectrum of ejection fraction categories, although the interpretation of this finding in HFrEF remains debated.32

By contrast, evidence in HF with mildly reduced ejection fraction (HFmrEF) and HFpEF is substantially more consistent and mechanistically coherent.

The STEP-HFpEF and STEP-HFpEF DM trials showed that 2.4 mg semaglutide once weekly significantly improved HF symptoms, physical limitation and exercise capacity compared with placebo in patients with HFpEF and obesity, including those with T2D.33 Alongside these functional gains, the STEP-HFpEF programme documented meaningful reductions in biomarkers of inflammation and congestion, as well as favourable changes in cardiac structure and diastolic function in an echocardiographic substudy.34 Regarding hard outcomes, the largest pooled individual patient-level analysis of semaglutide in HFpEF to date, encompassing 3,743 patients from the SELECT, FLOW, STEP-HFpEF and STEP-HFpEF DM trials, showed that semaglutide reduced the composite of CV death or worsening HF events by 31% and worsening HF events alone by 41%.35 Cardiovascular death alone was not significantly reduced, although the analysis was probably underpowered due to the small number of events.

The evidence for tirzepatide in HFpEF is equally compelling, with the SUMMIT trial showing that tirzepatide lowered the risk of a composite of death from CV causes or worsening HF by 38% compared to placebo.36 How quickly all these benefits were achieved in people with HFpEF has been recently published, with semaglutide reaching time to first clinical benefit within nearly 125 days.37 Moreover, tirzepatide significantly improved Kansas City Cardiomyopathy Questionnaire (KCCQ) scores and 6-minute walk distance in patients with HFpEF and obesity.36 A mechanistic secondary analysis of SUMMIT provided important insights into the pathways mediating these effects, showing that tirzepatide reduced circulatory volume-pressure overload, blood pressure (BP), systemic inflammation, including a significant decline in high-sensitivity CRP and markers of CV-renal injury.38

The magnitude of improvement in patient-reported outcomes observed with higher-dose semaglutide and tirzepatide has appeared greater than that reported in other pharmacological HFpEF trials, including PARAGON-HF, EMPEROR-Preserved and TOPCAT.29

A systematic review and a reconstructed individual patient data meta-analysis comparing semaglutide and tirzepatide in HFmrEF/HFpEF, bfound that oth these molecules lowered morbidity/mortality without evidence of heterogeneity.37 Complementing the evidence from these trials, a large real-world cohort study by Krüger et al. confirmed that semaglutide and tirzepatide were both associated with a risk reduction of >40% in the composite of hospitalisation for HF or all-cause mortality compared with a placebo proxy in patients with cardiometabolic HFpEF with tirzepatide showing no meaningful benefit over semaglutide.39 Similarly, a real-world analysis from the TriNetX Global Collaborative Research Network found no significant differences in terms of clinical outcomes between semaglutide and tirzepatide among patients with obesity and HFpEF.40

Moreover, although there is a lack of RCTs in HFpEF that are specifically powered for kidney outcomes, a recent retrospective, real-world TriNetX-based target trial emulation analysis reported a 56% relative reduction in major adverse kidney events with tirzepatide in HFpEF.41

Weight-independent Effects of GLP-1 RAs

Weight reduction is a significant mediator of benefits observed with GLP-1RAs. However, the clinical effects in HFpEF are unlikely to be explained by weight loss alone. They may derive from a combination of weight-dependent and weight-independent mechanisms, involving systemic metabolism, inflammation, the biology of adipose tissue, vascular function and myocardial remodelling.42

A post-hoc analysis of the SUMMIT trial suggested that tirzepatide produced greater improvements in 6-minute walk distance in patients with obesity-related HFpEF and higher baseline BMI, and greater on-treatment weight loss was associated with more pronounced improvements in HF symptoms. Notably, however, the reduction in rates of HF events or CV death with tirzepatide was consistent across baseline BMI categories.43

In the STEP-HFpEF and STEP-HFpEF DM trials, participants with more severe HF, such as those with higher N-terminal pro-B-type natriuretic peptide (NT-proBNP) levels, worse New York Heart Association (NYHA) functional class, AF or the need for loop diuretics, experienced greater improvements in HF-related health status with semaglutide, despite achieving a magnitude of weight loss comparable to that observed in patients with less severe disease.44–47

In the STEP-HFpEF echocardiography substudy, the degree of weight loss with semaglutide correlated with reductions in LA volume, while improvements in E wave velocity and E/e′ did not parallel weight loss.34,42

In addition, both the SUMMIT trial for tirzepatide and the STEP HFpEF programme for semaglutide significantly reduced systemic hs-CRP levels, with similar magnitude of CRP reduction across BMI strata and weight loss.42,48

Indeed, experimental and translational data indicate that GLP-1RAs suppress pro-inflammatory cytokine signalling, in part through inhibition of the nuclear factor-κB (NF-κB) pathway, including TNF-α and IL-6, thereby attenuating chronic low-grade inflammation that contributes to myocardial and vascular dysfunction.29,49

The anti-inflammatory actions of GLP-1Ras are potentially clinically meaningful also in the context of MASLD associated with HFpEF. By reducing IL-6 production, NF-κB inhibition directly influences hepatic CRP synthesis, linking cardiac, hepatic and systemic inflammatory pathways.42 In the Phase III ESSENCE trial, semaglutide improved histological liver outcomes and similar benefits were observed with tirzepatide in the Phase II SYNERGY-NASH trial.50,51 A recent meta-analysis confirmed that GLP-1RAs reduced liver fat deposition and improved histological steatosis without worsening fibrosis in MASLD.20

GLP-1RAs have also been shown to exert favourable effects on the lipid profile, likely mediated by reduced hepatic lipid synthesis, increased fatty acid oxidation and modulation of lipoprotein metabolism.29

Beyond a reduction in total body weight, changes in body composition further support the concept that the CV benefits of GLP-1RAs are not solely weight-dependent. In the SELECT trial, semaglutide reduced CV events regardless of the magnitude of weight loss, with reductions in waist circumference explaining only about one-third of the observed benefit, although there was a relationship between reduction in waist circumference and CV outcomes.42

GLP-1RAs promote not only overall fat loss, but also favourable adipose tissue redistribution, including clinically meaningful reductions in EAT thickness, which may contribute directly to improved cardiac mechanics.21,42,52 Importantly, GLP-1 receptor activation may also reverse the pro-inflammatory phenotype of adipocytes independently of weight loss, limiting their pathogenic paracrine effects on the myocardium.36

Given the high expression of GIP receptors in epicardial adipocytes, dual agonism with tirzepatide may provide additional anti-inflammatory and cardiometabolic benefits beyond those achieved with GLP-1 receptor agonism alone.23,25,26,53 Supporting these mechanistic considerations, an analysis by Packer et al. from the SUMMIT trial showed that tirzepatide achieved a similar reduction in pericardial fat and left ventricular mass for patients both with and without T2D.53 Moreover, preclinical evidence by Fan et al. supports the concept that GLP-1RAs may reverse myocardial hypertrophy and attenuate cardiac fibrosis through the cAMP/protein kinase A/Ras homolog (Rho) A/Rho-associated protein kinase 2 signalling pathway.54

Given the recognised role of EAT-driven inflammation and fibrotic remodelling in the development of VAC impairment in HFpEF, the anti-inflammatory and EAT-reducing effects of incretin-based therapies provide a plausible mechanistic basis for the favourable modulation of VAC that goes beyond their effects on weight loss. This hypothesis is supported by the evidence showing reduced systemic inflammation and improved cardiac parameters and vascular function with GLP-1RA therapies, as discussed above. Collectively, these changes are mechanistically consistent with improvement in both ventricular and arterial stiffness. However, no study has directly evaluated the effects of GLP-1RAs on VAC by measuring Ea/Eas ratio, pulse wave velocity or other indices in patients with HFpEF. Prospective studies incorporating these assessments represent an important unmet research need to determine whether improvements in VAC contribute mechanistically to the clinical benefits observed with these therapies.

Further, GLP-1RAs may increase energy expenditure, potentially involving thermogenesis and modulation of mitochondrial function in various tissues.29 Although direct evidence in HFpEF is lacking, preclinical and clinical data suggest that GLP-1RAs may improve cardiac metabolism by enhancing glucose oxidation and reducing circulating free fatty acids. Weight loss itself also restores myocardial metabolic flexibility, suggesting that the metabolic benefits of GLP-1RAs reflect both direct drug effects and weight-loss-mediated optimisation of the use of cardiac substrate.42 Moreover, improved insulin sensitivity and consequent enhanced glycaemic control may help alleviate myocardial damage linked to chronic hyperglycaemia, especially in people with T2D.29

Some evidence shows that GLP-1RAs directly affects cardiomyocytes. GLP-1 receptor is predominantly expressed in the sinoatrial node in the atrium, where its activation increases cAMP levels and may potentially influence the electrophysiological properties of atrial tissue.29 A direct effect of semaglutide on human atrial myocardium specifically in patients with HFpEF has been recently described, providing mechanistic support for the early and weight-loss independent improvement in HF symptoms observed in STEP-HFpEF.55

Of note, cardiac myocytes, pericytes and vascular endothelial cells also express GIP receptor (GIPR), supporting potential direct weight-independent, organ-specific GIP signalling too. GIPR is pharmacologically complex, with a wider range of physiological functions including regulation of glucagon release, vascular function, bone formation, adipose tissue perfusion and lipid storage.56 In a recent Mendelian randomisation study by Emanuelsson et al., a pathway-specific genetic score for GIP/GIPR signalling was associated with substantially greater relative reductions in major adverse CV events and HF compared with a polygenic BMI score, which supports additional mechanisms relevant to CV risk reduction achieved by the GIP pathway.57 The GIP/GIPR score was also associated with lower high sensitivity (hs)-CRP.

The kidney represents another target organ in which the weight-independent effects of GLP-1RAs may also contribute to clinical benefit, with important implications for HFpEF progression and prognosis.58 Nephroprotective actions of GLP-1RAs include neurohormonal modulation, attenuation of renal inflammation and oxidative stress, improved renal perfusion and reduced albuminuria.29 Reduction of pararenal fat depot has been proposed as a contributing mechanism of GLP-1RAs, paralleling the analogous role of EAT reduction in cardiac benefit.15

Finally, GLP-1RAs have shown to provide clinically meaningful BP reduction which is relevant in HFpEF management and prognosis. Whether this effect may be related to weight-independent mechanisms of BP regulation remains unclear. While a recent meta-analysis suggests that the BP-lowering effect of these drugs may be primarily a downstream consequence of weight loss, emerging evidence supports the pleiotropic effects of GLP-1RAs across the vasculature, kidney, heart and central nervous system.59 These include modulation of autonomic balance, with a shift away from sympathetic overactivity towards increased parasympathetic tone; improvements in renal haemodynamics through the promotion of natriuresis and reduction of intraglomerular pressure; and enhanced vascular function via increased bioavailability of nitric oxide, leading to vasodilation and potentially improved myocardial perfusion.29 Notably, early haemodynamic changes observed during GLP-1RA therapy, such as reductions in BP, aldosterone levels and increased natriuresis that occur before any significant weight loss may further support weight-independent mechanisms.42

Clinical Implications of the Use of GLP-1RAs in HFpEF and the Cardio-renal-metabolic Continuum

The convergence of robust trial evidence and mechanistic coherence positions GLP-1RAs as transformative therapeutic agents across the cardio-renal-metabolic continuum (Figure 1). Their benefits, encompassing intentional weight reduction and improvements in glycaemic control, BP, insulin sensitivity, hepatic steatosis, systemic inflammation and cardiorenal outcomes, collectively broaden the therapeutic applicability of this drug class to patients across the full spectrum of metabolic dysregulation. In support of this, a recent meta-analysis has shown that GLP-1RAs were associated with lower rates of HF events and CV death across a wide range of cardio-renal-metabolic conditions, including HFpEF, obesity, T2D and CKD, with benefits observed irrespective of the specific combination of comorbidities.60

Figure 1: A Mechanistic-to-clinical Framework for Patient Selection and Therapeutic Targets of Glucagon-like Peptide-1 Receptor Agonists in Heart Failure with Preserved Ejection Fraction.

BP = blood pressure; CAP = controlled attenuation parameter (fibroscan); CKD = chronic kidney disease; COPD = chronic obstructive pulmonary disease; EAT = epicardial adipose tissue; eGFR = estimated glomerular filtration rate; GLP-1RA = glucagon-like peptide-1 receptor agonist; hs-CPR = high sensitivity C-reactive protein; IR = insulin resistance; LA = left atrium; Lp(a) = lipoprotein(a); MASLD = metabolic dysfunction-associated steatotic liver disease; MetS = metabolic syndrome; NT-proBNP = N-terminal pro-B-type natriuretic peptide; NYHA = New York Heart Association; RAS = renin-angiotensin system; T2D = type 2 diabetes; VAC = ventricular-arterial coupling; WHR = waist-to-height ratio. Created in BioRender. Merolla A (2026) https://BioRender.com/22eg11k.

This evidence has important clinical implications. Individual disorders within the cardio-renal-metabolic syndrome, rarely occur in isolation as they actively interact through shared pathophysiological mechanisms and mutually reinforce disease progression.2 Once established, this self-perpetuating cycle is difficult to reverse using conventional single-target therapeutic approaches.

In contrast, the pleiotropic profile of GLP-1RAs offers the opportunity to intervene simultaneously across several interconnected disease pathways. Initiation of therapy for a single indication within the continuum may therefore generate clinically meaningful collateral benefits across the broader comorbidity burden. For example, treatment initiated for obesity may also improve glycaemic control, lower BP, reduce renal hyperfiltration, decrease hepatic and epicardial adiposity and attenuate systemic inflammation. This concept carries strategic relevance across two complementary clinical scenarios, including both real-world treatment and potential prevention approaches in HFpEF. On the one hand, in the context of the global epidemics of obesity and metabolic syndrome, which are expected to drive a parallel rise in HFpEF prevalence, early implementation of GLP-1RAs in high-risk individuals may offer an opportunity to delay or prevent the onset of HFpEF, thereby shifting the therapeutic paradigm from disease management to earlier interception of disease.61 On the other hand, the same multi-target profile of GLP-1RAs confers meaningful practical advantages in patients with established HFpEF, who are typically older, frequently frail and burdened by multimorbidity requiring multiple concomitant therapies. In this population, polypharmacy is often clinically necessary, yet carries well-recognised risks, including adverse drug interactions, reduced tolerability and impaired adherence.62 By simultaneously improving several cardiometabolic domains, GLP-1RAs may reduce the need for additional medications targeting BP, lipid profile, glycaemia or related metabolic complications, ultimately simplifying treatment plans.63

Current Limitations and Unresolved Questions

Treatment Discontinuation

Treatment persistence is a critical determinant of sustained therapeutic benefit.37 Despite the favourable efficacy and safety profile of GLP-1RAs discontinuation rates remain a clinically relevant challenge in clinical practice.58,64 The most frequently reported adverse events are gastrointestinal, including nausea, vomiting and diarrhoea, typically mild to moderate in severity and predominantly occurring during the dose-escalation phase.65 Nevertheless, these symptoms are the principal driver of early discontinuation of treatment and their inadequate management remains a major barrier to long-lasting clinical benefit.37

As expected, withdrawal from GLP-1RAs is followed by substantial weight regain, often proportional to the magnitude of the initial weight loss, together with partial or complete reversal of the cardiometabolic improvements achieved during treatment.66 A post-hoc analysis of the SURMOUNT-4 trial showed that greater weight regain within 1 year of tirzepatide withdrawal was associated with a greater reversal of improvements in cardiometabolic parameters, including body weight, BP, lipids and HbA_1c, highlighting that the metabolic benefits of these agents are dependent on their continued use.67

The implications of treatment withdrawal may be particularly relevant in HFpEF. Recurrence of adiposity would be expected to promote re-expansion of visceral and epicardial fat depots, re-establishing the pro-inflammatory adipokine milieu that contributes to myocardial stiffness, impaired diastolic reserve and adverse cardiac remodelling.12 However, no dedicated discontinuation studies have yet been conducted in patients with HFpEF, and the CV consequences of withdrawal from GLP-1RAs in this specific population remain largely undefined. Moreover, semaglutide has been associated with higher discontinuation rates in patients with HFpEF compared with tirzepatide, which may partly reflect the complementary receptor engagement conferred by dual GLP-1/GIP agonism, potentially translating into better tolerability and greater treatment durability.37 However, direct head-to-head data evaluating long-term adherence and persistence remain limited.

Taken together, these considerations suggest that, because HFpEF and its principal metabolic drivers, particularly obesity, are chronic and progressive conditions, GLP-1RAs should be viewed as long-term disease-modifying therapies. Whether indefinite treatment is necessary for all patients or whether individualised maintenance strategies could preserve efficacy while improving tolerability, reducing cost and enhancing adherence, remains an important unresolved question that warrants dedicated prospective and real-world investigation.

Long-term Effect of GLP-1RAs on Hard Outcomes in HFpEF

Despite the encouraging evidence available to date, important gaps in hard CV outcomes in HFpEF remain. Although recent meta-analyses support a beneficial role for GLP-1RAs in this setting as previously discussed, the contributing trials were not primarily designed with CV death or HF hospitalisation as their principal endpoint. In the SELECT trial, STEP-HFpEF and STEP-HFpEF DM studies, analyses of HF-related outcomes in participants with HFpEF were prespecified within the statistical analysis plans, whereas corresponding analyses in the FLOW trial were conducted post hoc.35 The SUMMIT trial was based on two primary endpoints which were a composite of adjudicated death from CV causes or a worsening HF event and the change in KCCQ score.36

In addition, the relatively short follow-up duration of major RCTs investigating GLP-1RAs in HFpEF may be insufficient to fully capture the long-term impact of these therapies on hard outcomes, recurrent HF hospitalisation and HFpEF progression.35

Accordingly, adequately powered, statistically robust, large-scale RCTs with extended follow-up are needed. Equally important is the generation of evidence in primary prevention, a setting that current RCTs have not adequately addressed. The ongoing SURMOUNT-MMO trial will provide important data on the effects of tirzepatide on multiple outcomes, including HF events, in individuals with overweight or obesity but without T2D, across both primary and secondary CV prevention.68 Beyond this, future research should clarify whether the cardiometabolic benefits of GLP-1RAs extend to individuals living with overweight but not obesity, or whether any therapeutic effects may also be observed in patients with no excess adiposity.

Beyond this, no dedicated RCTs have been designed for hard renal outcomes specifically in patients with HFpEF treated with GLP-1 or dual GIP/GLP-1 agonists. Future HFpEF trials should therefore incorporate prespecified outcomes such as estimated glomerular filtration rate (eGFR) decline, progression to end-stage kidney disease or renal death.

Long-term Effects on Body Composition and Non-cardiac Endpoints

Despite the emerging cardiometabolic benefits of GLP-1RAs in HFpEF, important concerns remain regarding their long-term effects on body composition, particularly the potential loss of lean mass in older, frail and patients with multiple comorbidities. This population is markedly underrepresented in RCTs despite accounting for a substantial proportion of individuals with HFpEF in routine clinical practice.1,69

The significant weight loss associated with these therapies is typically accompanied by reductions in fat mass and lean mass which, together with appetite suppression and the potential reduction in energy intake, may raise concerns regarding detrimental impact on muscular health. These issues are particularly relevant in older adults and in patients with pre-existing sarcopenic obesity in whom eventual inadequate protein and energy intake may accelerate skeletal muscle loss and functional decline.70,71 For this reason, caution is advised when prescribing GLP-1RAs to patients who are at risk of sarcopenia, particularly those who are older, frail or multimorbid. Preventive strategies should include regular monitoring for excessive weight loss, dehydration, adequate protein intake and resistance or strength training, to minimise loss of muscle mass while promoting reduction in fat mass.72

Reassuringly, emerging evidence suggests that the proportion of lean mass loss observed with incretin-based therapies is broadly comparable to that seen with lifestyle interventions alone and intentional weight reduction does not appear to be associated with functional decline.73,74 Available data indicate that GLP-1RA-mediated weight loss may be associated with preservation or even improvement in muscle quality, strength and physical function in older individuals with obesity, despite reductions in absolute lean mass.75

Most recent evidence indicates that GLP-1RAs exert beneficial effects on muscle vasculature and glucose uptake in muscle, and preclinical data suggest potential favourable effects on muscle protein metabolism.76 The SURPASS-3 MRI substudy supports this concept, showing that tirzepatide treatment was associated with significant reductions in muscle fat infiltration, suggesting an overall improvement in muscle quality.77 Findings from the STEP-HFpEF programme reported that semaglutide-mediated weight loss was similar across frailty status but a greater improvement in HF-related symptoms was observed in individuals who were most frail. Moreover, semaglutide reduced frailty burden after a year of treatment.78 Dedicated prospective studies and real-world evidence are needed to better characterise body composition trajectories and functional outcomes in more heterogeneous and clinically representative HFpEF populations.

Given that BMI remains an imperfect surrogate measurement of adiposity, more robust measures, including waist circumference, waist-to-hip ratio and direct body composition assessment, should be incorporated into future trial design and eligibility criteria, enabling a more accurate evaluation of treatment effects across the full spectrum of body composition phenotypes. More broadly, this reflection extends to several non-cardiac endpoints that should be systematically captured in future HFpEF-focused incretin trials: lean mass and fat mass, quantified by bioelectrical impedance; liver outcomes, including hepatic fat fraction by MRI; non-invasive fibrosis measures, such as liver stiffness by elastography and liver histology, where feasible; and kidney function measures, including eGFR, albuminuria or cystatin-C trajectories, to better define these therapies’ potential for renal protection. The adoption of such a multi-organ endpoint framework would better reflect the systemic nature of HFpEF and align future trial design with the increasingly recognised cardio-renal-metabolic paradigm.

Phenotype-directed Therapy, Combination Strategies and Future Directions

Identifying Candidates for GLP-1RAs and Therapeutic Targets in HFpEF

HFpEF encompasses a highly heterogeneous clinical spectrum, with variation in adiposity burden, systemic inflammation, haemodynamic profile and coexistence of comorbidities such as insulin resistance, CKD, MASLD or sarcopenia. Many individuals with excess adiposity do not develop HFpEF and benefits of GLP-1RAs in this group are less clear.42 Of note, age, sex and further risk factors, including coronary artery disease, chronic obstructive pulmonary disease or sleep apnoea, for which tirzepatide has recently been shown to provide clinical benefit, as well as iron deficiency, valvular disease and polypharmacy influence disease severity and clinical response to therapy.79,80 Therefore, key questions remain regarding the optimal use of GLP-1RAs in HFpEF.

A growing body of research has sought to characterise distinct HFpEF phenotypes to better inform its clinical management. These include: cardiometabolic, left atrial myopathy, stiff vascular and pulmonary vascular.81 Many remain underrepresented in RCTs and observational studies.7,80

The evidence most strongly supports the use of GLP-1RAs in the metabolic phenotype of HFpEF with NYHA II–IV and obesity and/or T2D. HFpEF patients with recent HF decompensation may also benefit from these interventions, while efficacy in other populations with specific cardiomyopathies is unclear.29

Most recent evidence integrating high-dimensional, multi-modal data has uncovered a structured, continuous spectrum of phenotypic, genetic and outcome variation in HFpEF, which appears to reflect variable pathophysiological processes that coexist to a greater or lesser extent in each individual.82 In clinical practice, this is consistent with clinicians’ observations that patients typically present overlapping features of multiple phenotypes with variable severity.

Comprehensive screening for coexisting comorbidities should anchor clinical assessment as management often remains disease-centred and may overlook cumulative multi-organ dysfunction. This is particularly important because obesity may mask early HFpEF as NT-proBNP is frequently lower in obesity and eGFR may appear spuriously preserved due to glomerular hyperfiltration, together potentially delaying diagnosis.15,44 Similarly, conventional CV risk scores fail to capture excess adiposity, potentially underestimating risk.

Within this context, while refining patient selection remains important, an even more valuable step for improving patient care is the clear definition of the clinical objectives that should guide the use of these therapies in routine practice.

Therefore, we propose a pragmatic, patient-oriented, mechanistic-to-clinical framework linking organ-specific molecular and pathophysiological pathways of HFpEF to the corresponding clinical manifestations and biochemical or imaging markers and the mechanistic rationale and clinical evidence supporting GLP-1RA therapy across each domain (Figure 1). This framework is intended to guide patient selection and to help clinicians define concrete, organ-specific therapeutic targets for individual patients with HFpEF.

Looking ahead, refined characterisation of inflammatory activity and EAT biology, alongside circulating biomarkers – hs-CRP, IL-6, cystatin C, lipoprotein(a) – may improve risk stratification.7,83 Haemodynamic studies are also needed to quantify volume, filling pressure and VAC changes over time. The HERMES trial, testing ziltivekimab targeting IL-6 in individuals with HFpEF and elevated CRP, should clarify the value of targeted anti-inflammatory strategies, further advancing a precision medicine framework for HFpEF management.84

Importantly, incretin-based therapy should be regarded as a chronic, disease-modifying strategy requiring longitudinal follow-up to maximise benefit and minimise risk and unnecessary use of additional medications. Depending on the pattern of comorbidity, benefit may be reinforced by combination strategies, notably SGLT2i in T2D and/or CKD, as discussed below.

Combination Therapy: Current Potential Additive and Synergistic Strategies

Given the marked heterogeneity of HFpEF, no single intervention is likely to fully address the multiple interacting drivers of disease progression. Accordingly, combination strategies employing therapies with complementary mechanisms of action are likely to play a central role in the evolving treatment landscape of this condition.58,85 Among the most compelling current approaches is the potential combined use of GLP-1RAs and SGLT2i, whose complementary mechanisms provide a strong biological rationale for additive or synergistic benefit in HFpEF: while GLP-1RAs primarily target adipose inflammation, insulin resistance and systemic metabolic dysregulation, SGLT2i reduce preload through osmotic natriuresis, lower intraglomerular pressure, improve renal haemodynamics and shift myocardial substrate metabolism towards the body’s use of ketones.58 Supporting this rationale, Patel et al. reported that the addition of GLP-1RAs to ongoing SGLT2i therapy was associated with a lower risk of HF exacerbations, all-cause emergency department visits, new-onset arrhythmias, acute kidney injury and pulmonary hypertension in patients with HFpEF.86

Moreover, the renin–angiotensin–aldosterone system (RAAS) represents another important therapeutic axis in HFpEF. Although dedicated RCTs specifically evaluating the combination of GLP-1RAs and RAAS in HFpEF is currently lacking, there is a biologically plausible and clinically meaningful rationale for its benefit. A recent network meta-analysis of eight RCTs, encompassing more than 25,000 patients with HFpEF and HFmrEF, found that the combination of mineralocorticoid receptor antagonists (MRAs), GLP-1RA and SGLT2i to standard care provides substantial benefit in reducing HF events.87

The potential for further synergy with antifibrotic strategies also warrants consideration. Neuen et al. estimated that triple combination therapy with GLP-1RAs, SGLT2i and the non-steroidal mineralocorticoid receptor antagonist finerenone could yield clinically meaningful gains in survival free from HF hospitalisation and CKD progression.88

These projections are promising but require prospective confirmation in HFpEF populations. Dedicated RCTs specifically designed to test combination regimens are needed to determine the true magnitude of benefit, as well as their safety, tolerability and optimal sequencing in patients with HFpEF.

Emerging Molecules and Future Therapeutic Horizons

Within the rapidlyevolving landscape ofcardiometabolic pharmacotherapy, several next-generation incretin-based and incretin-adjacent agents are emerging and may further expand therapeutic options for HFpEF.

These include: the triple GLP-1/GIP/glucagon-RA retatrutide, which has showed substantial metabolic efficacy with marked weight loss and broad improvements in glycaemic and cardiometabolic parameters; orforglipron, an oral non-peptide GLP-1RA that overcomes barriers related to injectable administration while providing meaningful reductions in body weight and cardiometabolic risk markers; ecnoglutide, a cAMP-biased GLP-1 analogue offering an alternative intracellular signalling profile with potential implications for efficacy and tolerability, although clinical advantages over existing agents remain to be established; fixed-dose combination strategies such as cagrilintide-semaglutide, which have provided superior weight reduction compared with GLP-1RA monotherapy; and peptide-antibody conjugates, including maridebart cafraglutide combining GLP1 agonism and GIP antagonism, representing a further promising class under early clinical evaluation.89

Notably, metabolic benefits have also been reported with GIP receptor antagonism, underscoring the complexity of GIP biology. This paradox can be potentially explained by tissue-specific signalling, receptor desensitisation or convergence of agonism and antagonism on shared appetite- and adiposity-regulating pathways. Mechanistically, GIP agonists appear to act via a GLP-1-independent, gamma-aminobutyric acid (GABA) pathway, while GIP antagonists may potentiate GLP-1-mediated hypophagia. However, GIP receptor antagonism remains experimental, with no clear evidence of CV benefit although further research should help clarify this paradox.90

Collectively, these agents illustrate a shift towards more potent and diversified incretin-based strategies, however, robust evidence on CV outcomes is still lacking. Specifically designed large-scale phase 3 trials in HFpEF populations will be essential to better clarify the role of these therapies in clinical practice.

Conclusion

Incretin-based therapies are supporting a paradigm shift in HFpEF, targeting both cardiometabolic drivers and downstream clinical consequences of the syndrome. While definitive evidence on hard outcomes remains forthcoming, the consistency of available data supports their integration into current clinical practice, with meaningful improvements in symptoms, quality of life and overall metabolic health. Strategies to enhance long-term adherence and preserve safety will need to be systematically integrated into routine care to maximise therapeutic benefit. Looking forward, a precision medicine approach, grounded in detailed phenotypic characterisation, biomarker profiling, accurate assessment of adiposity and body composition analysis, will be key to tailoring treatment across the heterogeneous HFpEF population. Mechanism-based combination strategies targeting metabolic, inflammatory, haemodynamic and fibrotic pathways, together with an expanding pipeline of next-generation incretin agents, are likely to define the next frontier in HFpEF management. Ultimately, these advances are not simply adding new treatments, but are reshaping how we understand and manage HFpEF, moving from a predominantly cardiac perspective to an integrated cardiometabolic paradigm.