Heart failure with preserved ejection fraction

Article written and reviewed by Dr Luca Moretti

Heart failure with preserved ejection fraction, or HFpEF, is a clinical syndrome with symptoms or signs of heart failure, a left ventricular ejection fraction of at least 50%, and demonstration of a structural or functional abnormality causing elevated filling pressures or congestion. The simple association of dyspnea and a normal ejection fraction is insufficient: the cardiac origin must be documented, and pulmonary diseases, anemia, obesity, deconditioning, and precapillary pulmonary hypertension must be distinguished. “Preserved” describes a percentage, not intact cardiac function. Longitudinal contraction, relaxation, compliance, chronotropic reserve, ventricular-arterial coupling, and right heart function may be severely impaired; a small ventricle may eject a normal percentage but an insufficient stroke volume. HFpEF is a multisystem and phenotypically heterogeneous syndrome. Visceral obesity, hypertension, diabetes, kidney disease, atrial fibrillation, and aging are frequent, but amyloidosis, hypertrophic cardiomyopathy, valvular diseases, constriction, and high-output conditions are specific causes or phenocopies requiring their own pathways.

Prevalence increases with age and obesity and frequently affects women, without being exclusive to either sex. The burden arises from repeated hospitalizations, severe limitation, and interaction with the kidneys, lungs, and muscle; mortality includes a substantial proportion of noncardiovascular causes, but the risk of cardiac worsening remains high. Therapy is no longer limited to diuretics; SGLT2 inhibitors reduce events across the preserved spectrum; finerenone reduced the burden of worsening at ejection fractions of at least 40%. In the adipose phenotype, semaglutide improves symptoms and capacity, and tirzepatide reduced worsening events, supporting obesity treatment as an intervention targeting a cause; the ACC 2026 consensus emphasizes phenotype-based management. Diagnosis must identify the dominant mechanism, and therapy integrates decongestion, blood pressure and rhythm control, management of cardiometabolic and kidney disease, exercise, and treatment of specific causes.

Etiology, pathogenesis, and pathophysiology

Chronic hypertension increases afterload and stimulates concentric hypertrophy, fibrosis, and stiffness. Concurrent arterial stiffness amplifies late systolic load and makes blood pressure sensitive to small volume changes; during exercise, inadequate vasodilation increases work and filling pressure. Aging modifies the matrix, titin, endothelium, and beta-adrenergic reserve and increases comorbidities, but HFpEF is not an inevitable consequence of age. Female sex is more often associated with concentric geometry, vascular stiffness, and smaller dimensions, while men may have more coronary artery disease; overlap is substantial. Diabetes promotes glycation, lipotoxicity, and mitochondrial and microvascular dysfunction. Kidney disease adds sodium retention, anemia, inflammation, and uremic toxins, creating a cardiorenal cycle.

Visceral and epicardial obesity is not merely a comorbidity. It increases plasma volume and demand, compresses the pericardium, promotes inflammation and sleep apnea, reduces natriuretic peptides, and alters muscle; adipose HFpEF often features elevated right-sided pressures, greater interdependence, and disproportionate physical limitation. Epicardial adipose tissue may exert mechanical and paracrine effects; intravascular volume increases more than necessary relative to lean mass. For this reason, intentional weight loss improves hemodynamics and symptoms, while involuntary loss with sarcopenia is unfavorable. Systemic inflammation and microvascular dysfunction reduce nitric oxide availability, protein kinase G, and titin phosphorylation, increasing cardiomyocyte stiffness; this model does not explain every case, but links comorbidities and fibrosis.

Diastole requires ATP-dependent active relaxation and passive compliance. Ischemia and altered calcium handling slow relaxation; hypertrophy, titin, collagen, and infiltration increase stiffness; atrial pressure rises to maintain filling and is transmitted to the pulmonary circulation. Many patients have normal resting pressures but a pathological rise during exercise; limited diastolic reserve, tachycardia, and shortened filling unmask the phenotype; this explains inconclusive baseline tests despite true dyspnea. Systolic function is not truly normal: longitudinal strain and contractile reserve may be reduced, while ejection fraction remains preserved because of concentric geometry and the circumferential contribution. Stroke volume and maximal cardiac output are often limited.

Atrial fibrillation reflects pressure and fibrosis but worsens the syndrome by eliminating atrial contraction and increasing heart rate; the atrium becomes a decisive organ: compliance, reservoir function, and atrial myopathy influence pressure, mitral regurgitation, and thromboembolic risk. Chronotropic incompetence prevents an increase in cardiac output; conversely, an excessive heart rate shortens diastole. Beta-blockers may be useful for specific indications but worsen capacity in patients with marked incompetence, requiring individualization; atrial or ventricular functional mitral regurgitation and tricuspid regurgitation increase pressures and congestion. The valve may be a primary cause, consequence, or amplifier and must be quantified under appropriate loading conditions.

Chronically increased pulmonary pressure causes vascular remodeling and may produce combined post- and precapillary hypertension. The afterload-sensitive right ventricle dilates and develops tricuspid regurgitation; right heart dysfunction is one of the main determinants of mortality and systemic congestion. During exercise, venous return and pulmonary pressure increase; a vascular bed that cannot be recruited prevents accommodation of cardiac output. Impaired diffusion and effusions contribute to dyspnea; inadequate peripheral vasodilation and arterial stiffness increase load. Ventricular-arterial coupling sometimes appears preserved at rest but loses reserve during activity.

Skeletal muscle shows reduced capillary and mitochondrial density, fatty infiltration, and a shift toward glycolytic fibers; these abnormalities cause insufficient peripheral oxygen extraction and early lactate production; capacity may improve with training even without modifying diastolic function. Anemia and iron deficiency reduce oxygen transport and metabolism. Obesity and deconditioning increase the energy cost of movement and ventilation, making dyspnea a multiorgan outcome; renal and splanchnic congestion activates retention and inflammation. Kidney function may deteriorate because of venous pressure even with normal cardiac output, and the response to diuretics depends on tubular adaptation.

ATTR amyloidosis frequently presents as apparent HFpEF with thick walls, carpal tunnel syndrome, spinal stenosis, neuropathy, or low blood pressure. AL amyloidosis requires urgent hematological assessment; noninvasive ATTR diagnosis is possible only after exclusion of a monoclonal component according to validated algorithms. Hypertrophic cardiomyopathy, Fabry disease, sarcoidosis, and hemochromatosis have their own treatments and risks. Constrictive pericarditis mimics myocardial stiffness but shows respiratory interdependence and intrathoracic pressure dissociation; pericardiectomy may be definitive. Aortic stenosis, mitral regurgitation, congenital heart disease, and high-output states are not subsumed under the HFpEF label; the phenocopy must be named and treated.

Clinical manifestations

Exertional dyspnea is the main symptom and reflects increased pulmonary pressure, an insufficient chronotropic response, vascular stiffness, and peripheral abnormalities. In early stages it appears only at high intensities; with progression, daily activities and then rest become symptomatic. Orthopnea, nocturnal dyspnea, and cough indicate more advanced congestion; hypertensive pulmonary edema may develop rapidly because of afterload and redistribution, with little weight change and a nondilated ventricle. Fatigability and reduced endurance may exceed dyspnea, particularly with chronotropic incompetence, obesity, anemia, and sarcopenia.

Edema, weight gain, jugular venous distention, ascites, and early satiety indicate right-sided and systemic involvement; crackles may be absent in the chronic state; ultrasound B-lines detect subclinical congestion but are not specific to a cardiac origin. Atrial fibrillation may present with palpitations or a sudden loss of capacity, but is often asymptomatic in older adults; a rapid response precipitates edema; an excessively slow response or excessive beta-blockade limits cardiac output. Angina may result from epicardial coronary artery disease or microvascular dysfunction. Syncope requires investigation for aortic stenosis, arrhythmia, pulmonary hypertension, amyloidosis, or hypotension and is not a typical benign symptom.

Examination may show hypertension, obesity, jugular venous distention, a fourth heart sound, murmurs, or edema. An apex beat without signs of dilation and absence of a third heart sound do not exclude the syndrome; orthostatic blood pressure is important in frail patients and those on multiple medications. Signs of amyloidosis include bilateral carpal tunnel syndrome, biceps rupture, lumbar stenosis, neuropathy, and hypotension; macroglossia and periorbital purpura point more toward AL. Hypertrophic phenotypes may have a dynamic murmur and family history. Pulmonary hypertension and right heart dysfunction produce an accentuated P2 component, a parasternal heave, a v wave, a pulsatile liver, and ascites.

NYHA class describes limitation but is influenced by obesity and comorbidities; KCCQ quantifies symptom frequency, limitation, and quality of life; the six-minute walk test measures integrated function. In the adipose phenotype, weight reduction may improve KCCQ and walking distance to a clinically meaningful degree; cardiopulmonary exercise testing identifies reduced peak VO2, ventilatory inefficiency, the chronotropic response, and sometimes peripheral limitation. Adding exercise hemodynamics clarifies uncertain cases. Frailty and cognitive impairment may manifest as falls, inactivity, and loss of independence. Reported symptoms must be assessed within the functional context, not only with echocardiography.

Acute decompensations may be triggered by uncontrolled blood pressure, atrial fibrillation, ischemia, infection, NSAIDs, kidney dysfunction, or excess sodium; redistributive congestion requires greater attention to afterload, while accumulation-related congestion requires sodium removal; they often coexist. Worsening treated in an outpatient setting with increased diuretics has prognostic value. Repeated events indicate insufficient therapy, an unrecognized phenotype, or right heart and renal progression. Shock is less frequent than in HFrEF but may occur in infarction, valvular regurgitation, advanced amyloidosis, or right heart failure; a preserved ejection fraction does not guarantee adequate cardiac output.

Investigations and diagnosis

Clinical probability arises from symptoms, factors, and findings. ECG may show hypertrophy, fibrillation, ischemia, or low voltages; a normal ECG does not exclude HFpEF. BNP or NT-proBNP provides support, but obesity reduces levels, and a proportion of hemodynamically confirmed HFpEF may have values below the usual thresholds. Atrial fibrillation, age, and kidney disease increase peptides, requiring contextual thresholds and interpretation; an elevated value does not distinguish HFpEF from valvular disease, pulmonary hypertension, or amyloidosis. Complete blood count, kidney function, electrolytes, liver function, glucose, HbA1c, TSH, iron status, and urine testing investigate mimics and comorbidities. Persistent troponin elevation identifies risk and may suggest ischemia or infiltration.

Echocardiography documents an ejection fraction ≥50% and assesses hypertrophy, atrial dilation, reduced e’, elevated E/e’, pulmonary pressure, valves, and right heart function; no isolated diastolic parameter has sufficient accuracy; multiparametric integration is essential. Reduced longitudinal strain indicates subclinical systolic dysfunction and may suggest amyloidosis if it shows apical sparing, but the pattern is not pathognomonic. Atrial volume reflects chronic pressure exposure, except in fibrillation or valvular disease; diastolic stress testing evaluates E/e’, tricuspid regurgitation velocity, and capacity during exercise. Image quality and heart rate may limit reliability, and intermediate results require hemodynamic assessment.

The H2FPEF score uses obesity, treated hypertension, fibrillation, pulmonary pressure, age, and E/e’; HFA-PEFF integrates functional, morphological, and biomarker domains. They are probability tools, not definitions, and may incorporate characteristics that vary with treatment or population; a high probability supports diagnosis; a low probability suggests alternatives; an intermediate probability requires functional tests. The score must not be used to avoid investigation for amyloidosis or valvular disease; right heart catheterization with exercise is the reference standard in uncertain cases: a pathologically elevated pulmonary capillary wedge pressure during exercise demonstrates the mechanism. Accuracy of wedge pressure, zeroing, and cardiac output is indispensable.

Standard cardiopulmonary exercise testing distinguishes ventilatory inefficiency, chronotropic incompetence, and peripheral limitation, but does not always identify pressure; the invasive combination measures the pressure-flow relationship and reserve; an abnormal response may emerge before high absolute thresholds are reached. Magnetic resonance imaging measures mass and volumes and characterizes fibrosis and infiltration. Late enhancement, T1, and extracellular volume point toward amyloidosis, HCM, Fabry disease, or myocarditis; the result must be linked to the clinical picture. Coronary CT or ischemia tests are selected according to probability and symptoms; microvascular dysfunction may require coronary reserve measurements, but the finding alone does not prove that all dyspnea is cardiac.

When amyloidosis is suspected, serum and urine immunofixation and light-chain testing are performed before or together with bone-tracer scintigraphy. A gammopathy prevents noninvasive ATTR diagnosis without further clarification because AL may show uptake; diagnostic delay is particularly dangerous in AL. Constriction is investigated with respiratory echocardiographic assessment, CT/MRI, and catheterization; features include septal bounce, flow variations, and ventricular discordance; obesity and abdominal pressure may mimic interdependence. Valvular diseases and hypertrophic cardiomyopathies follow dedicated pathways; the HFpEF label must not conceal a correctable cause.

The final diagnosis describes the dominant phenotype: adipose-cardiometabolic, hypertensive, atrial fibrillation, ischemic, pulmonary hypertension/right-sided, infiltrative, or valvular; these categories overlap and serve to organize therapy, not to create rigid subtypes. Monitoring includes weight, blood pressure, rhythm, kidney function, potassium, congestion, KCCQ, and capacity; echocardiography is repeated when symptoms, valves, or right heart function change; a stable ejection fraction does not mean stable disease. The differential diagnosis remains open over time: “HFpEF due to hypertension” may reveal ATTR or aortic stenosis as it evolves. Red flags must be reassessed.

Treatment and prognosis

Diuretics control congestion and symptoms; the dose must remove sodium without excessively reducing preload in a stiff ventricle; weight, jugular veins, kidney function, electrolytes, and blood pressure guide management. Diuretic resistance may require sequential blockade, with close monitoring. SGLT2 inhibitors are a foundational therapy: empagliflozin and dapagliflozin reduce the composite of cardiovascular death or worsening, mainly through fewer heart failure events, independently of diabetes. Benefit extends to high ejection fractions, although attenuated at the extremes in some analyses; initiation requires assessment of volume status, eGFR, genital infections, and ketoacidosis risk. The initial filtration rate dip is expected and does not require withdrawal if the patient is stable.

Finerenone reduced total worsening events and cardiovascular death in FINEARTS-HF, with a greater effect on events. It is a nonsteroidal mineralocorticoid receptor antagonist; hyperkalemia and kidney function require monitoring, and prescribing must follow applicable indications and approvals. Spironolactone in TOPCAT did not reduce the overall primary endpoint but reduced hospitalizations, with substantial regional heterogeneity; in ESC 2026, mineralocorticoid receptor antagonists are recommended in symptomatic heart failure regardless of ejection fraction. ARNI in PARAGON-HF did not meet the primary endpoint, but analyses support benefit at lower ejection fractions and in selected women. ARBs may reduce hospitalizations in some patients and are useful for blood pressure and kidney disease. Beta-blockers are not a universal HFpEF therapy: they are used for ischemia, arrhythmias, and blood pressure, avoiding excessive bradycardia and chronotropic incompetence.

Hypertension must be controlled with drugs that address comorbidities and tolerability. Overly aggressive targets with orthostatic symptoms or low cardiac output are harmful; blood pressure crises require prevention through adherence and volume management; atrial fibrillation requires anticoagulation according to risk, individualized rate control, and consideration of rhythm control. Maintenance of sinus rhythm may improve capacity in patients dependent on atrial systole, but advanced atrial disease and duration reduce success. Coronary artery disease and valvular disease are treated according to their own indications; specific pulmonary vasodilators are not routinely recommended in hypertension due to left heart disease.

In the obesity-related phenotype, intentional weight loss reduces load and inflammation. Semaglutide 2.4 mg improved KCCQ, weight, and walking distance in the STEP-HFpEF trials with and without diabetes; tirzepatide in SUMMIT reduced the composite of cardiovascular death or worsening and improved health status. These results do not automatically extend to nonobese or cachectic patients. Body composition, muscle mass, gastrointestinal tolerability, and frailty must be monitored; protein nutrition and resistance exercise help preserve lean mass. Bariatric surgery may improve hemodynamics and risk in selected candidates, but requires multidisciplinary assessment; treatment of obstructive sleep apnea, diabetes, and kidney disease completes management of the phenotype.

Supervised exercise improves peak VO2 and quality of life by acting on vessels and muscle; rehabilitation combines aerobic exercise, strength, balance, and education; deconditioning must not be mistaken for a reason to avoid activity. Iron deficiency is corrected according to evidence and context; anemia requires investigation of the cause. Vaccinations, smoking cessation, nutrition, and review of NSAIDs and edema-inducing drugs reduce precipitants. Sodium and fluid restrictions are personalized; a home plan for weight, symptoms, and diuretics helps recognize worsening, but must not generate unsupervised fluctuations in frail patients.

Amyloidosis therapy depends on the type: tafamidis stabilizes transthyretin in appropriate indications, while AL requires urgent control of the clone. HCM, Fabry disease, and sarcoidosis have specific therapies; pericardiectomy may cure selected constriction. Atrial shunt devices and other technologies remain experimental or selective; REDUCE LAP-HF II did not demonstrate overall benefit. Hemodynamic monitoring of pulmonary pressure may reduce events in selected patients with previous hospitalizations if the system responds to data. ICD and CRT are not indicated for HFpEF on ejection fraction alone, but may be indicated for arrhythmias, cardiomyopathy, or pacing; assessment remains etiological.

Prognosis varies with phenotype. Right heart dysfunction, combined pulmonary hypertension, kidney disease, fibrillation, hospitalizations, troponin, peptides, frailty, and low capacity increase risk. Obesity may create an apparent “paradox” because of confounding and unintentional weight loss; it does not justify leaving adiposity untreated; noncardiovascular causes contribute substantially to mortality, making multisystem management essential. Tumors, infections, pulmonary disease, and frailty must be prevented or treated without losing control of the cardiac condition. Refractory symptoms, repeated hospitalizations, and right heart dysfunction require advanced assessment and integrated palliative care. Transplantation is rare but possible in selected patients with severe physiological impairment; preserved ejection fraction does not exclude terminal disease.

Complications

Acute congestive decompensations are frequent and manifest as hypertensive pulmonary edema or systemic accumulation. Venous redistribution may produce respiratory failure without substantial weight gain; vasodilation and blood pressure control then take on a greater role alongside diuresis. Cardiorenal syndrome results from venous pressure and preexisting kidney disease; a rise in creatinine must be interpreted with volume status and response; leaving congestion to protect a single value may worsen outcomes. Hyponatremia, hypokalemia, and hyperkalemia complicate diuretics, RAASi, and MRA; correction must avoid unnecessary permanent withdrawals and rapid sodium changes.

Atrial fibrillation increases pressure and stroke risk; tachycardias and bradycardias reduce reserve; the atrial arrhythmia may become permanent with remodeling and atrial functional mitral regurgitation, aggravating the cycle. Pulmonary hypertension and right heart dysfunction cause tricuspid regurgitation, ascites, hepatic congestion, and low cardiac output; right heart progression is often the most important prognostic transition. Venous thromboembolism and pulmonary embolism may precipitate right heart failure; anticoagulation is guided by the indication, not by isolated HFpEF.

Hepatic congestion produces cholestasis and fibrosis; acute hypoperfusion causes hypoxic hepatitis. Intestinal edema reduces absorption and promotes malnutrition, even in an obese patient. Sarcopenia and sarcopenic obesity worsen capacity, falls, and tolerance of weight-management therapies; rapid weight loss requires distinction among diuresis, fat reduction, and muscle loss. Iron deficiency, anemia, and kidney disease increase dyspnea and hospitalizations; treatment requires diagnosis of the cause and cannot rely on hemoglobin alone.

Unrecognized amyloidosis progresses with conduction blocks, arrhythmias, hypotension, and low cardiac output; an incidental gammopathy may delay the ATTR/AL distinction; late diagnosis reduces the benefit of specific therapies. Aortic stenosis and mitral and tricuspid regurgitation may worsen and must not be attributed simply to age. Serial assessment and the Heart Team define the timing of intervention. Drugs may cause orthostatic symptoms, bradycardia, kidney injury, or dehydration; review must favor those with benefit and eliminate duplication or obsolete indications.

Frailty, cognitive decline, depression, and isolation increase readmissions and loss of independence; the caregiver is part of the plan for weight, medications, and warning signs, but also needs support. Repeated hospitalizations cause deconditioning and may lead to dependence. Early mobilization, rehabilitation, and delirium prevention reduce the nonhemodynamic harm of hospitalization; terminal progression may occur with severely elevated pulmonary pressure, right heart failure, kidney dysfunction, and cachexia while maintaining a normal ejection fraction. Advance care planning and palliative care prevent the echocardiographic percentage from concealing severity.