Left Ventricular Ejection Fraction Thresholds Defining Heart Failure Upheld

Incident rates decrease for most adverse clinical outcomes among patients with heart failure (HF) as left ventricular ejection fraction (LVEF) rises, approaching a threshold of about 40% to 50% with little change above this level. Current LVEF thresholds that define HF with mildly reduced EF are supported by these results, according to findings published in Circulation.

Investigators sought to assess if obvious thresholds in patient characteristics at baseline and inflection points for clinical outcomes during follow-up exist for LVEF among participants with HF, considering high-normal (supranormal) levels of LVEF and including differences according to sex.

They conducted an analysis with pooled data assessing HF from more than 33,000 participants enrolled in 6 randomized controlled trials that included patients with preserved and reduced EF. The pooled cohort median follow-up duration was 35.4 months. The investigators used Poisson regression models to assess the relationship between the incidence of all-cause mortality and HF hospitalization and LVEF.

Mean (SD) LVEF in women was 47.6% (16.5%) and 36.8% (14.2%) in men. Mean age ranged from 61.6 to 69.4 years among patients with LVEF of less than 50% and between 71.2 and 71.8 years among patients with LVEF of greater than 50% (mean 10-year age difference between patients with LVEF ≤20% vs LVEF >70%, 62 vs 72 years). Patients with LVEF of less than 50% were predominantly men (ranging 62.2% to 80.7%) and patients with LVEF of greater than 50% were predominantly women (range, 52.2% to 68.9%). Most patients were White in all LVEF categories, increasing from 67.1% (lowest LVEF category) to 84.4% (highest LVEF category).

Estimated glomerular filtration rate, ischemic pathogenesis, and N-terminal pro-B-type natriuretic peptide (NT-proBNP) decreased as LVEF increased. Additionally, the prevalence of atrial fibrillation and diabetes, systolic blood pressure, body mass index, age, and the proportion of women increased as LVEF increased. The investigators noted no meaningful changes in characteristics as LVEF increased beyond 50% except the proportion of women continued to increase, and NT-proBNP and ischemic pathogenesis continued to decrease. As LVEF increased, comorbidities such as hypertension, diabetes, and cancer increased, while myocardial infarction, previous percutaneous catheter intervention, and coronary artery bypass graft decreased (greatest differences in hypertension and myocardial infarction, most significant in the 40% to 50% LVEF range).

As LVEF increased, (other than non-cardiovascular death) incidence of clinical outcomes decreased (LVEF inflection point about 35% for HF hospitalization; about 40% for pump failure death; about 50% for all-cause mortality and cardiovascular death). Incidence rates had insignificant further decline beyond these thresholds.

There was no evidence that patients with supranormal LVEF faced worse outcomes or of a J-shaped relationship between LVEF and death. According to available echocardiographic data, patients with supranormal LVEF had no structural differences suggestive of amyloidosis (supported by NT-proBNP levels).

Study limitations include lack of generalizability to the general population with HF, and thresholds and inflection points are not accurately defined by a statistically established method.

“Current left ventricular ejection fraction thresholds defining heart failure phenotypes are supported by this analysis of clinical outcomes,” the study authors wrote. “Incidence rates for most clinical outcomes decreased as left ventricular ejection fraction increased to a threshold of around 40% to 50% but did not change much thereafter.”

Disclosure: Some study authors declared affiliations with biotech, pharmaceutical, and/or device companies. Please see the original reference for a full list of authors’ disclosures.