HFpEF: Heart Failure with Preserved Ejection Fraction

2. What It Is

HFpEF is heart failure with a preserved ejection fraction, defined as an EF above 50% (some older definitions used 45% as the lower threshold; current guidelines use 50%). The diagnosis requires the clinical syndrome of heart failure (symptoms: dyspnea, fatigue, reduced exercise capacity; signs: raised filling pressures, pulmonary congestion) in the setting of an EF above 50% and evidence of diastolic dysfunction or raised LV filling pressures on imaging 5 / Solid .

The critical distinction: the heart in HFpEF pumps with normal force but fills abnormally. The LV is stiff. During diastole, when the LV should relax and fill with blood from the left atrium, it relaxes slowly and incompletely. To fill a stiff ventricle, the left atrium must generate higher pressure to push blood across the mitral valve. This raised filling pressure is transmitted backward into the pulmonary vasculature, causing pulmonary venous hypertension and the classic symptom of exertional dyspnea.

The patient is typically:

• Female (HFpEF is 60-70% women) 5 / Solid
• Older (median age at diagnosis approximately 72 years)
• Obese (BMI above 30 in the majority of patients)
• Hypertensive (present in 80-90% of HFpEF patients)
• Diabetic (present in 40-60%)
• With atrial fibrillation (present in 40-65%)

This phenotypic cluster has led to the concept of HFpEF as a systemic, metabolic, inflammatory disorder rather than purely a structural cardiac disease. Epicardial and perivascular fat, systemic inflammation, and metabolic dysregulation all contribute to myocardial stiffness through mechanisms distinct from those in HFrEF.

The HFA-PEFF score and the H2FPEF score were developed to standardize the diagnosis of HFpEF in ambulatory patients, where the presentation is often less overt than in hospitalized patients 4 / Promising . An H2FPEF score of 6 or above has a probability of HFpEF exceeding 90%.

3. The Mechanism

Diastolic dysfunction

Diastole has two phases: active relaxation (isovolumic relaxation and early filling) and passive compliance. Active relaxation requires energy; it is an ATP-dependent process by which calcium is pumped back into the sarcoplasmic reticulum after systole. In HFpEF, active relaxation is prolonged and incomplete. The SERCA2a pump (sarcoplasmic/endoplasmic reticulum Ca2+-ATPase 2a) is downregulated, calcium clearance is slow, and the LV remains partially contracted at the time it should be filling.

The second mechanism is reduced compliance: the myocardial extracellular matrix is stiffer due to excess collagen deposition and collagen cross-linking. This is driven by chronic hypertension, aging, and possibly by circulating inflammatory mediators. The titin protein (the giant spring-like cytoskeletal protein that controls myofilament compliance) is stiffer in HFpEF: it is expressed predominantly as the N2B isoform rather than the more compliant N2BA isoform, and it is underphosphorylated by protein kinase G (PKG), which would normally reduce its stiffness. Reduced NO-cGMP-PKG signaling in the microvascular endothelium of HFpEF patients is now considered a central pathway in the pathophysiology 4 / Promising .

The systemic inflammation hypothesis

The Paulus-Tschope hypothesis (2013) proposed that in HFpEF, multiple comorbidities (obesity, diabetes, hypertension, chronic kidney disease, atrial fibrillation) produce a state of chronic systemic inflammation with raised levels of TNF-alpha, IL-6, and ST2. This inflammation reaches the coronary microvascular endothelium, reduces endothelial NOS activity, decreases cGMP production in adjacent cardiomyocytes, reduces PKG activity, and results in titin hypophosphorylation and myocardial stiffness. This model explains why HFpEF is fundamentally different from HFrEF in biology and why drugs targeting the neurohormonal axis (ACE inhibitors, beta-blockers) do not work in HFpEF: those pathways are not the primary drivers.

Exercise intolerance: the hemodynamic story

At rest, many HFpEF patients have near-normal hemodynamics. The abnormality becomes unmasked with exercise. During exercise, the normal heart increases stroke volume via preload (Frank-Starling) while heart rate rises and filling time shortens. In HFpEF, the stiff LV cannot increase volume substantially with exercise, stroke volume augmentation is blunted, and the only mechanism to increase cardiac output is heart rate. But if filling time is short (tachycardia) and the LV fills poorly even at rest, heart rate-dependent augmentation is limited. The result: exercise-induced pulmonary venous hypertension, reduced peak VO2, and the exertional dyspnea the patient experiences. Exercise challenge with Doppler echocardiography or invasive hemodynamics during exercise stress testing is now used to unmask exertional HFpEF in patients with unexplained dyspnea and normal resting parameters.

The obesity-HFpEF phenotype

Adipose tissue, particularly epicardial and visceral fat, contributes to HFpEF through mechanical compression of the pericardium and heart, local inflammatory cytokine production, and systemic insulin resistance. Patients with obesity-HFpEF have higher filling pressures at any given level of exercise, more impaired cardiac reserve, and higher plasma BNP (though BNP is paradoxically lower in obese patients per unit of LV pressure due to fat-mediated clearance). The recognition of obesity as a primary driver of HFpEF, rather than a comorbidity, opened the door to weight-loss interventions as a treatment strategy.

4. How We Diagnose

Diagnosing HFpEF is harder than diagnosing HFrEF, because the echocardiogram EF is normal, which falsely reassures. The key is to look for evidence of raised filling pressures despite a normal EF.

Echocardiographic diastolic assessment

The ASE/EACVI 2016 guidelines for diastolic dysfunction assessment define four parameters:

1. Mitral annular early diastolic velocity (e’): measured by tissue Doppler imaging. Low e’ (septal e’ below 7 cm/s or lateral e’ below 10 cm/s) indicates impaired active relaxation.
2. E/e’ ratio: E is the peak early transmitral inflow velocity (peak filling velocity); e’ is the tissue Doppler velocity. E/e’ above 14 correlates with raised mean pulmonary capillary wedge pressure.
3. Left atrial volume index: a chronically raised LA filling pressure enlarges the left atrium. LAVI above 34 mL/m2 is abnormal.
4. Peak TR velocity: above 2.8 m/s suggests raised pulmonary artery systolic pressure, a consequence of chronically raised LA and pulmonary venous pressure.

When three of four criteria are above threshold, diastolic dysfunction is classified as Grade III (most severe, raised filling pressures). When criteria are mixed, exercise or provocative testing is needed.

Biomarkers

NT-proBNP above 125 pg/mL (or BNP above 35 pg/mL) in the outpatient setting is supportive of raised filling pressures. Notably, BNP is lower in obese patients, so the threshold may not be met even when hemodynamic congestion is present. High-sensitivity cardiac troponin T (hs-cTnT) is chronically raised in HFpEF, reflecting ongoing cardiomyocyte stress.

Exercise testing

Cardiopulmonary exercise testing (CPET) with measurement of peak VO2 and VE/VCO2 slope quantifies exercise capacity objectively. Peak VO2 below 14 mL/kg/min is the threshold for advanced heart failure device therapies. In HFpEF, peak VO2 is typically 12-18 mL/kg/min (moderately reduced).

Exercise echocardiography or exercise right heart catheterization can unmask exertional elevation of pulmonary capillary wedge pressure above 25 mmHg that is not present at rest, confirming the diagnosis in ambiguous cases.

5. The Evidence

The long failure: trials that did not work

The history of HFpEF trials is a history of negative results, and those negatives are informative.

PEP-CHF (2006): Perindopril (ACE inhibitor) in elderly HFpEF patients showed no significant reduction in the primary composite of all-cause mortality or unplanned heart failure-related hospitalization at one year 5 / Solid . Enrollment challenges limited the trial, but the directional signal was absent.

CHARM-Preserved (2003): Candesartan (ARB) in HFpEF reduced heart failure hospitalizations modestly (HR 0.84, p = 0.017) but had no effect on cardiovascular mortality 5 / Solid 14187-8).

I-PRESERVE (2008): Irbesartan (ARB) in HFpEF: no effect on primary composite of all-cause mortality or cardiovascular hospitalization 5 / Solid .

TOPCAT (2014): Spironolactone (MRA) in HFpEF: no significant reduction in the primary composite of cardiovascular death, cardiac arrest, or heart failure hospitalization overall (HR 0.89, p = 0.14). However, there was marked geographic heterogeneity: patients enrolled in Russia and Georgia (who appeared not to have genuine HFpEF based on very low event rates and BNP levels) diluted the result. In the Americas subgroup, spironolactone reduced the primary outcome by 18% (HR 0.82, p = 0.026) 4 / Promising . Spironolactone is widely used in clinical practice for HFpEF based on this signal, particularly in patients with EF 40-60%.

PARAGON-HF (2019): Sacubitril-valsartan in HFpEF (EF above 45%): did not significantly reduce the primary composite of total hospitalizations for heart failure plus cardiovascular death (RR 0.87, p = 0.059, just missing significance). A pre-specified subgroup analysis showed benefit in patients with EF 45-57% (the mildly reduced or lower end of preserved) and in women 4 / Promising . The FDA did not approve sacubitril-valsartan for HFpEF based on this trial.

The breakthrough: SGLT2 inhibitors

EMPEROR-Preserved (2021): 5,988 patients with symptomatic heart failure and EF above 40%, randomized to empagliflozin 10 mg versus placebo. Empagliflozin reduced the composite of cardiovascular death or hospitalization for worsening heart failure by 21% (HR 0.79, 95% CI 0.69-0.90) 5 / Solid . This was the first positive mortality-relevant trial in HFpEF. The benefit was consistent across EF subgroups and in diabetics and non-diabetics.

DELIVER (2022): 6,263 patients with heart failure and EF above 40%, randomized to dapagliflozin 10 mg versus placebo. Dapagliflozin reduced the composite of worsening heart failure or cardiovascular death by 18% (HR 0.82, 95% CI 0.73-0.92) 5 / Solid . Both agents were subsequently pooled: the EMPEROR-Preserved and DELIVER combined analysis confirmed consistent benefit across the EF spectrum above 40%.

Mechanism in HFpEF: SGLT2 inhibitors reduce LV stiffness through interstitial fluid mobilization, reduce epicardial fat, reduce oxidative stress, and improve endothelial NOS-cGMP-PKG signaling in the microvascular endothelium, the same pathway that is disrupted in HFpEF pathophysiology. They also reduce afterload, lower NT-proBNP, and reduce LA volume.

The FDA approved dapagliflozin for heart failure regardless of EF in 2022. Empagliflozin similarly received broad heart failure labeling. Both are now Tier 1A recommendations in the 2022 AHA/ACC guidelines for HFpEF 5 / Solid .

The obesity breakthrough: STEP-HFpEF (2023)

STEP-HFpEF: 529 patients with HFpEF and BMI above 30, randomized to semaglutide 2.4 mg weekly (Wegovy) versus placebo. The primary endpoint was change in Kansas City Cardiomyopathy Questionnaire-Clinical Summary Score (KCCQ-CSS) and change in 6-minute walk distance. At 52 weeks, semaglutide improved KCCQ-CSS by 16.6 points versus 8.7 for placebo (p < 0.001) and improved 6-minute walk distance by 21.5 m (p < 0.001). It also reduced NT-proBNP by 22%, CRP by 43%, and weight by 13.3 kg 5 / Solid . The STEP-HFpEF DM trial in patients with obesity-HFpEF and diabetes showed similar results 5 / Solid .

These results established obesity treatment as a disease-modifying therapy in the obesity-HFpEF phenotype, not just a comorbidity to manage. The magnitude of symptom improvement with semaglutide was larger than seen with most other drugs in HFpEF.

SUMMIT (2024): Tirzepatide (GIP/GLP-1 dual agonist, Zepbound/Mounjaro) in obesity-HFpEF: the trial showed significant improvement in KCCQ-CSS, 6-minute walk distance, and a reduction in the composite of worsening heart failure or cardiovascular death 5 / Solid ). This extends the GLP-1 class benefit to tirzepatide and further reinforces obesity as a therapeutic target.

Sex differences in HFpEF

Women represent 60-70% of HFpEF cases. The mechanism likely involves the fact that female hearts hypertrophy concentrically (increase wall thickness without increasing cavity size) in response to pressure overload, whereas male hearts eccentric remodel (dilate). Concentric hypertrophy impairs diastolic filling more than eccentric dilation. Women also have smaller coronary microvascular caliber, higher microvascular resistance, and may have a stronger inflammatory phenotype driving myocardial stiffness.

In EMPEROR-Preserved, the benefit of empagliflozin was present in both sexes but the absolute risk reduction was numerically larger in women. In PARAGON-HF, the signal for sacubitril-valsartan reached nominal significance in women even when the overall trial did not. These sex differences are important to report: a cardiologist treating a 72-year-old woman with HFpEF, hypertension, and obesity should be more confident in SGLT2i benefit, consider sacubitril-valsartan if EF is 45-57%, and have a lower threshold for obesity pharmacotherapy.

6. The Patient Experience

The HFpEF patient experience is defined by a specific frustration: the heart looks “normal” on most tests, symptoms are attributed to aging, obesity, or deconditioning, and years pass before the correct diagnosis is made. The average time from symptom onset to diagnosis of HFpEF is 2-5 years in many health systems.

The symptoms are less dramatic than acute HFrEF decompensation, at least initially. The patient notices that the walk from the car to the store requires a pause. That picking up a grandchild leaves her breathing hard. That she cannot sleep flat since she started waking up coughing. These symptoms are real; they represent documented impairment in peak VO2 and quality of life. The KCCQ-CSS score, a validated 23-item questionnaire measuring heart failure symptom burden, is typically 20-30 points lower than age-matched controls in HFpEF patients.

Hospitalizations for HFpEF look like HFrEF hospitalizations: fluid overload, pulmonary edema, the need for IV diuresis. The 30-day readmission rate is similar to HFrEF (approximately 20-25%). But the drivers of HFpEF hospitalization are often triggers: uncontrolled hypertension, new-onset atrial fibrillation, dietary salt excess, or NSAID use (which causes sodium and water retention and raises filling pressures acutely).

The medication burden is real. The daily regimen for a patient with HFpEF and obesity includes an SGLT2 inhibitor, often a diuretic (furosemide or torsemide for congestion control), blood pressure management (typically an ACE inhibitor or ARB, possibly adding spironolactone for resistant hypertension and additional filling pressure reduction), and now potentially semaglutide or tirzepatide for weight management. Polypharmacy management, cost, and insurance prior authorization for GLP-1 agents are genuine barriers.

The cost of semaglutide (Wegovy) is approximately $1,350/month without insurance. Tirzepatide (Zepbound) is similarly priced. Prior authorization requirements vary by insurer, and many plans require documented failure of multiple weight-loss interventions before approval. In Illinois, Medicaid does not consistently cover these agents. For patients in rural central Illinois (Decatur, Bloomington, Springfield), access to an endocrinologist or obesity medicine specialist familiar with HFpEF-directed weight management requires referral to Carle Foundation Hospital in Urbana or the University of Illinois Health system.

Atrial fibrillation is both a consequence and driver of HFpEF. Raised LA pressure from chronic diastolic dysfunction enlarges the left atrium and promotes atrial fibrillation. Atrial fibrillation then removes the atrial contribution to LV filling (the “atrial kick”), which is especially important in a stiff ventricle that depends on late diastolic filling. The result is acute hemodynamic deterioration. Rate control is essential; rhythm control may be especially beneficial in HFpEF with AF. The CASTLE-AF trial data (HFrEF-specific; 10.1056/NEJMoa1707855) cannot be directly extrapolated to HFpEF, but rhythm control with cardioversion or catheter ablation in HFpEF-AF is a reasonable strategy for symptomatic patients.

7. Decisions and Trade-Offs

Which drug to start first

The 2022 AHA/ACC guidelines give SGLT2 inhibitors a Class 2a recommendation for HFpEF (moderate-strength, reasonable to use). There is no Class 1 indication for any specific drug in HFpEF as of this writing, reflecting the more modest absolute risk reduction compared to HFrEF. In practice, dapagliflozin or empagliflozin should be started in all patients with HFpEF who tolerate them and do not have eGFR below 25 mL/min/1.73m2.

Diuretics (furosemide, torsemide, bumetanide) are used for congestion control but have not been shown to affect mortality in HFpEF. They are the primary tool for symptom management and should be uptitrated to achieve euvolemia.

MRAs (spironolactone, eplerenone) carry a Class 2b recommendation in HFpEF (may be reasonable). The Americas subgroup of TOPCAT provides the primary supporting data. They are reasonable to use, particularly in patients with EF 40-55% and persistent symptoms despite SGLT2i.

Sacubitril-valsartan carries a Class 2b recommendation for patients with HFpEF and EF below normal (typically interpreted as EF 45-55%), particularly women. The PARAGON-HF subgroup data, combined with the well-established HFrEF data, supports this consideration.

GLP-1 receptor agonists (semaglutide, tirzepatide) are not yet formally guideline-recommended as cardiac therapies, but the STEP-HFpEF data is compelling. Patients with obesity-HFpEF (BMI above 30) and intolerable symptoms should be discussed for GLP-1 therapy, either under the endocrinology umbrella (for obesity) or with explicit recognition of the cardiac benefit data.

The diuretic equilibrium problem

HFpEF patients are particularly vulnerable to diuretic over-treatment. Because the stiff LV depends on adequate preload to fill, aggressive diuresis lowers preload and can precipitate low-output symptoms (fatigue, lightheadedness) even as it removes congestion. The patient reports feeling “worse” on diuretics because their functional capacity drops when preload is reduced. Finding the euvolemic “sweet spot” requires careful titration: the goal is absence of JVD, absence of crackles, and a KCCQ score that improves, not worsens, on the diuretic dose.

Rate vs rhythm control in HFpEF with AF

In HFrEF with persistent AF, CASTLE-AF showed a 38% mortality benefit from catheter ablation over medical rate control. The same trial did not include HFpEF patients. The AFFIRM trial and the 2022 EAST-AFNET 4 trial data support early rhythm control in AF patients with cardiovascular disease for multiple reasons including QOL and possibly stroke prevention. For HFpEF patients with AF, rhythm control is a reasonable strategy, with cardioversion first (to assess whether sinus rhythm improves symptoms) before proceeding to ablation. Centers offering this in Illinois: Northwestern Memorial (Chicago), University of Chicago, Carle Foundation Hospital (Urbana), and OSF HealthCare (Peoria).

When to refer for right heart catheterization

Invasive hemodynamic assessment with right heart catheterization (RHC) is appropriate when: (1) the diagnosis is uncertain despite echocardiographic workup, (2) exercise-induced pulmonary hypertension is suspected, (3) the patient is being considered for clinical trial enrollment, or (4) pre-transplant or LVAD evaluation is needed. At rest, many HFpEF patients have a PCWP that is only mildly raised (15-20 mmHg). With exercise, PCWP may rise to 30-40 mmHg. Exercise RHC at an experienced center unmasks this hemodynamic abnormality in 30-40% of patients with unexplained dyspnea and normal resting echo.

Clinical Synthesis

HFpEF is the emblematic heart disease of the modern American lifestyle. Its substrate is decades of hypertension, obesity, physical inactivity, and metabolic dysfunction. It is predominantly a women’s disease, and it is predominantly underdiagnosed and undertreated. The retired schoolteacher in Decatur with “normal” echo results who stopped playing with her grandchildren is not unusual; she is representative.

The clinical mission intersects with HFpEF at multiple levels.

Upstream: a structured cardiovascular assessment at age 45-55, with formal diastolic function parameters included in the echocardiogram protocol, can identify diastolic dysfunction Grade II or above before symptoms begin (asymptomatic diastolic dysfunction). This is Stage B HFpEF. Identifying it allows aggressive treatment of hypertension, reduction of obesity, and initiation of SGLT2i consideration before the first hospitalization.

Once diagnosed: structured remote monitoring provides the ongoing follow-up HFpEF requires. Weight monitoring, blood pressure optimisation, diuretic dose adjustment, and atrial fibrillation rate and rhythm management are calibration tasks that cannot be adequately managed in annual cardiology visits. Remote monitoring tools (connected scales, BP cuffs, wearables) feed data into a proactive management protocol that catches decompensation before hospitalisation.

For detection and staging, a formal cardiovascular assessment with diastolic function quantification is the entry point. For ongoing management, remote monitoring closes the gap between clinic visits. For patients with severe obesity-related HFpEF who may benefit from semaglutide or tirzepatide, a cardiologist can facilitate referral to obesity medicine specialists or cardiometabolic programs at major Illinois academic centres.

The teacher in Decatur went three years without the correct diagnosis. Two weeks on empagliflozin and a spironolactone titration, combined with a structured low-sodium dietary plan, improved her KCCQ score by 14 points. She is back to the stairs.