HFmrEF: Heart Failure with Mildly Reduced Ejection Fraction

2. What It Is

HFmrEF is defined by an ejection fraction of 40-49% in the setting of the clinical syndrome of heart failure: symptoms, signs, and objective evidence of cardiac dysfunction or raised filling pressures 5 / Solid . This category was formally introduced by the 2016 ESC guidelines under the name “heart failure with mid-range ejection fraction,” and was retained with the name changed to “mildly reduced” in 2021.

Before 2016, the cardiology literature effectively split heart failure into two categories: HFrEF (EF below 40%) and HFpEF (EF above 50%), and patients with EF 40-49% were classified arbitrarily to one category or the other, or simply omitted from trials with strict EF cutoffs.

The 2022 AHA/ACC/HFSA guidelines adopted the same categorization. The cutoffs are now standardized: HFrEF below 40%, HFmrEF 40-49%, HFpEF 50% and above.

HFmrEF has two principal subtypes:

1. HFrEF recovering: A patient whose EF was previously below 40% and has improved with treatment into the 40-49% range. This is termed HFimpEF (heart failure with improved EF) in some classification systems. This patient was treated with the full HFrEF protocol and improved; the question is whether to continue, taper, or modify therapy.
2. De novo HFmrEF: A patient who presents with EF 40-49% without a prior measurement below 40%. This patient may have primary diastolic-predominant dysfunction, moderate systolic impairment, or an intermediate phenotype with elements of both.

The prevalence of HFmrEF in major registries: the Swedish Heart Failure Registry (SwedeHF) reported EF 40-49% in 13% of heart failure patients 5 / Solid . The ESC Heart Failure Long-Term Registry found 12% of enrolled patients in this range 5 / Solid .

3. The Mechanism

HFmrEF is mechanistically heterogeneous. This is the key clinical reality that complicates both understanding and treatment.

In the HFrEF-recovering patient, the mechanism was systolic dysfunction (neurohormonal overactivation, myocardial fibrosis, chamber dilation) that has been partially reversed by the four-drug protocol. The neurohormonal pathways are still active at a lower level; stopping medications may result in EF re-decline. Studies that assessed GDMT withdrawal in HFimpEF patients showed that discontinuing beta-blockers or sacubitril-valsartan resulted in EF decline in 30-40% of patients within one year 5 / Solid .

In de novo HFmrEF, the pathophysiology may lean toward diastolic features (impaired relaxation, reduced compliance, raised filling pressures) with mild superimposed systolic dysfunction, or it may represent mild systolic failure before the full HFrEF phenotype develops. Coronary artery disease, hypertension, and diabetes are the most common comorbidities, similar to HFpEF but with more evidence of active myocardial damage.

Biomarker data: in the GUIDE-IT trial and other registries, NT-proBNP levels in HFmrEF patients fall between HFrEF and HFpEF values, consistent with an intermediate degree of wall stress. Troponin levels are similarly intermediate. CMR data shows that LGE patterns in de novo HFmrEF include mid-wall enhancement (suggesting non-ischemic cardiomyopathy), subendocardial enhancement (ischemic), and no LGE (possible diastolic-predominant disease). No single pattern predominates.

The trajectory of HFmrEF matters clinically. EF in this zone is not stable: it can worsen toward HFrEF or improve toward HFpEF over time. A patient with de novo EF 44% who is not on appropriate therapy may track downward. A patient recovering from HFrEF on full guideline-directed therapy may track upward.

4. How We Diagnose

The diagnosis of HFmrEF follows the same framework as HFrEF and HFpEF: symptoms of heart failure plus objective evidence of cardiac dysfunction with EF 40-49%. Practically, the workup includes:

Echocardiography: Required for EF classification. The EF must be assessed carefully: variability in echocardiographic EF measurement is approximately 5-10 percentage points between studies and operators. A patient with EF 42% at one center may measure 48% at another. CMR provides more accurate and reproducible EF measurement when classification at the boundary matters clinically. Repeat echocardiography (or CMR) after 90 days of optimized therapy is appropriate to determine the trajectory of EF.

Biomarkers: NT-proBNP or BNP measurement establishes the degree of neurohormonal activation. An NT-proBNP above 1,000 pg/mL in a stable outpatient with EF 44% suggests significant hemodynamic stress and argues for aggressive therapy initiation.

Diastolic assessment: Because HFmrEF shares features with both HFrEF and HFpEF, echocardiographic diastolic parameters should be obtained: E/e’, left atrial volume index, and tricuspid regurgitation velocity. This helps define the degree of diastolic contribution to the syndrome.

Coronary assessment: CAD is common in HFmrEF. Coronary evaluation should follow the same logic as in HFrEF: unexplained mild systolic dysfunction warrants ruling out ischemic etiology.

Genetics: For patients with de novo HFmrEF without clear cause, genetic testing is appropriate under the same framework as HFrEF.

5. The Evidence

SGLT2 inhibitors in HFmrEF

The DELIVER trial (dapagliflozin) enrolled patients with EF above 40%, including the HFmrEF subgroup (EF 40-49%). In the pre-specified HFmrEF subgroup, dapagliflozin reduced the composite of worsening heart failure or cardiovascular death by 24% (HR 0.76, 95% CI 0.63-0.92) 5 / Solid . The benefit was numerically slightly larger in the HFmrEF subgroup than in the HFpEF subgroup.

The EMPEROR-Preserved trial (empagliflozin) enrolled patients with EF above 40%. In the EF 40-49% subgroup, empagliflozin reduced the primary endpoint by 29% (HR 0.71, 95% CI 0.58-0.87) 5 / Solid . The EMPEROR-Pooled analysis confirmed consistent benefit across the EF 40-49% range.

Both agents have now received FDA approval for heart failure with EF above 40%, which encompasses HFmrEF.

MRA evidence: TOPCAT and beyond

The TOPCAT spironolactone trial enrolled patients with EF above 45%, which partially overlaps HFmrEF. The Americas subgroup showed HR 0.82 for the primary outcome with spironolactone 4 / Promising . Spironolactone is commonly used in HFmrEF, particularly in the HFrEF-recovering patient whose prior regimen included an MRA and who continues to benefit from it.

ARNI evidence: PARAGON-HF subgroup

PARAGON-HF enrolled patients with EF above 45%. In the pre-specified EF 45-57% subgroup, sacubitril-valsartan showed nominally significant benefit (HR 0.78, 95% CI 0.64-0.95) 4 / Promising . For patients with EF in the lower end of the HFmrEF range (40-50%), sacubitril-valsartan extrapolation from the PARADIGM-HF and PARAGON-HF data is reasonable. The 2022 guidelines give sacubitril-valsartan a Class 2b recommendation in the EF 41-49% range.

EMPEROR-Pooled: the full spectrum analysis

The EMPEROR-Pooled meta-analysis combined EMPEROR-Reduced (HFrEF) and EMPEROR-Preserved (HFpEF/HFmrEF) data: 9,718 patients. Empagliflozin consistently reduced the composite of cardiovascular death or hospitalization for heart failure across the entire EF spectrum, with no evidence of heterogeneity by EF subgroup 5 / Solid 01094-3). The flat hazard ratio across EF categories supports treating the entire heart failure spectrum with SGLT2 inhibitors regardless of EF classification.

What does not have evidence in HFmrEF

ACE inhibitors, ARBs, and beta-blockers do not have specific RCT evidence in de novo HFmrEF. Their use is typically extrapolated from HFrEF trials (for patients with EF close to 40% or recovering from HFrEF) or from HFpEF/hypertension data (for blood pressure management). This is a genuine evidence gap acknowledged in both the 2021 ESC and 2022 AHA guidelines.

6. The Patient Experience

The patient experience of HFmrEF mirrors either HFrEF or HFpEF depending on which direction the patient’s history runs. For the patient recovering from HFrEF, the trajectory is often positive: EF improved, symptoms reduced, energy returning. The key risk is therapeutic complacency: because the patient “feels better,” medication discontinuation or dose reduction becomes tempting. The Halliday TRED-HF trial data showed EF re-decline in 40% of patients who had their HFrEF medications withdrawn after achieving EF above 50%. The lesson applies proportionally to those with EF recovering through the 40-49% zone.

For de novo HFmrEF, the experience is often that of HFpEF: dyspnea, fatigue, reduced exercise tolerance, with the frustration of a “borderline” label that neither triggers aggressive treatment nor provides a clear prognosis. Patients in this zone frequently receive fewer medications and less intensive follow-up than HFrEF patients, despite evidence of comparable symptom burden and hospitalization risk.

The 30-day readmission rate for HFmrEF hospitalization is approximately 20%, similar to HFrEF and HFpEF. The mortality at one year is approximately 10-15%, higher than the general population but lower than HFrEF at initial presentation 5 / Solid .

Cardiac rehabilitation for HFmrEF: the evidence base for exercise training is primarily from HFrEF trials (HF-ACTION) and HFpEF trials. For HFmrEF specifically, there are no large dedicated trials, but the physiological rationale is strong and the risk is low. Cardiac rehabilitation programs in Illinois accept patients with EF 40-49% under the heart failure diagnosis coding. Available at Carle Foundation Hospital (Urbana), Northwestern Memorial (Chicago), and OSF Saint Francis Medical Center (Peoria).

7. Decisions and Trade-Offs

Should HFmrEF be treated like HFrEF?

The answer depends on trajectory and etiology. For a patient recovering from documented HFrEF on the full four-drug protocol, the answer is clear: maintain the drugs. Do not withdraw them because EF improved. The improvement is likely because of the drugs.

For de novo HFmrEF, the decision is more nuanced. The guideline framework:

• SGLT2 inhibitor: start, Class 2a, evidence-based.
• Diuretic for congestion: use as needed.
• ARNI (sacubitril-valsartan): consider, Class 2b, especially if EF 40-49% and EF trending downward or patient was previously in HFrEF range.
• Beta-blocker: consider, particularly if there is coexisting hypertension, tachycardia, or AF with rapid ventricular response, or if the patient is recovering from HFrEF.
• MRA: consider, Class 2b, particularly if spironolactone was previously started for HFrEF.

The practical answer for many clinicians: treat HFmrEF with the same drugs as HFrEF when the EF is 40-44% or when the trajectory is unclear or downward. Treat more like HFpEF when the EF is 45-49% with no prior HFrEF history and the clinical picture looks more like diastolic dysfunction.

Serial EF monitoring

Because HFmrEF EF can move in either direction, serial echocardiography is appropriate: every six months in the first year after diagnosis, annually once stable. CMR may be preferable for borderline EF cases where reproducible measurement matters for device decisions.

ICD timing in HFmrEF

ICD is recommended for primary prevention only when EF remains below 35% on guideline-directed therapy. HFmrEF (EF 40-49%) does not meet the ICD threshold. However, for a patient who previously had EF below 35% and has now recovered to the 40-49% range, the ICD that was implanted at EF below 35% should not be removed unless shared decision-making supports that choice. The device serves as protection if the EF declines again.

LBBB and CRT in HFmrEF

CRT requires EF below 35% per guidelines. For HFmrEF patients with LBBB and EF 40-44%, CRT is not guideline-recommended at that EF. If the EF was recently above 35% and the patient has LBBB, this is a discussion for the electrophysiology/heart failure team at a center experienced in complex device management.

HFmrEF in Common Comorbidities

HFmrEF and type 2 diabetes

The overlap between HFmrEF and type 2 diabetes is substantial: approximately 35-40% of HFmrEF patients have comorbid diabetes. This creates a particularly important clinical opportunity, because SGLT2 inhibitors address both conditions simultaneously. For the diabetic patient with HFmrEF on metformin, adding empagliflozin or dapagliflozin is both a cardiovascular and glycemic intervention. The Class 2a recommendation for SGLT2i in HFmrEF exists independently of diabetes status, but the indication is reinforced in the diabetic patient.

GLP-1 receptor agonists (semaglutide, liraglutide) have cardiovascular outcome data in diabetes and obesity, but dedicated heart failure trial data for HFmrEF is limited. Semaglutide’s STEP-HFpEF trial showed benefit in HFpEF with obesity; whether similar benefit extends to HFmrEF is under study. For now, in the HFmrEF patient with diabetes and obesity, semaglutide for weight and glycemic management is reasonable alongside, not instead of, the heart failure drug regimen.

HFmrEF and atrial fibrillation

AF is present in approximately 25-30% of HFmrEF patients and is associated with worse outcomes in this population. Rate control in HFmrEF-AF: target resting heart rate below 80 bpm. Rhythm control in HFmrEF-AF: the EAST-AFNET 4 trial demonstrated early rhythm control reduces cardiovascular outcomes in AF with cardiovascular disease broadly, which includes heart failure. For HFmrEF specifically, whether rhythm control translates to EF improvement is plausible but not definitively proven.

Beta-blocker dosing in HFmrEF-AF: use cautiously, target rate without inducing bradycardia that reduces forward output. Digoxin is a third-line option for rate control when beta-blockers are not tolerated and the heart rate target is not met.

HFmrEF and chronic kidney disease

CKD complicates HFmrEF management in several ways. Diuretic resistance increases as GFR falls below 30 mL/min/1.73m². SGLT2i efficacy extends to eGFR as low as 20 mL/min/1.73m² for the heart failure indication (dapagliflozin) per updated label data. ACE inhibitor or ARB dosing requires monitoring for hyperkalemia as renal function declines. Sacubitril-valsartan is generally avoided when eGFR is below 30 due to limited safety data and hyperkalemia risk.

For HFmrEF patients with progressive CKD approaching dialysis, a multidisciplinary discussion involving nephrology and heart failure teams should address volume management, medication adjustments, and the impact of dialysis initiation on cardiac filling pressures and hemodynamics. Cardiorenal syndrome in this population is bidirectional: impaired cardiac output reduces renal perfusion, and worsening renal function limits the dose of the cardiac medications that might improve cardiac output.

Clinical Synthesis

HFmrEF represents one of the most common places where the gap between guideline-directed and actual care is widest. It is the zone where the label “borderline” becomes a clinical excuse for inaction. The truck driver from Peoria with EF 44% is the archetypal patient who falls through the protocol gap.

This clinical approach to HFmrEF is to refuse the “borderline” framing. An EF of 44% with symptoms of heart failure is heart failure. It deserves an SGLT2 inhibitor. It deserves serial monitoring. It deserves a structured follow-up plan that tracks EF trajectory and adjusts medication strategy based on whether the EF is moving up or down.

A structured cardiovascular assessment is valuable in this population specifically because it performs serial echocardiographic measurement, including diastolic parameters, rather than treating each study as a one-time snapshot. A patient whose EF tracked from 52% to 46% to 44% over three years is telling a story of progressive systolic dysfunction that demands early intervention, not watchful waiting.

Structured remote monitoring is appropriate for established HFmrEF patients, providing the serial biomarker monitoring, medication tracking, and proactive adjustment that prevents the trajectory from continuing downward.

For rural patients in central Illinois, Peoria is the nearest tertiary center with advanced heart failure services. Stop Dying Early facilitates that referral when needed, reducing the friction that often delays access.

The “borderline heart” is not a lesser heart. It is a heart whose signal deserves equal clinical respect.