Simon Stewart, PhD, reported at the European Society of Cardiology Heart Failure Discoveries virtual meeting.,
This analysis from the ongoing National Echocardiography Database of Australia (NEDA) included 499,153 men and women who underwent echocardiography in routine clinical practice for a variety of indications, with more than 3 million person-years of follow-up.
This study broke new ground. There is surprisingly little information from routine clinical practice to describe the spectrum and prognostic importance of left ventricular ejection fraction (LVEF). Indeed, most data have come from clinical trials in patients with heart failure with reduced ejection fraction (HFrEF), in which women are traditionally underrepresented. By comparison, the NEDA analysis included 237,046 women in routine care, noted Dr. Stewart, a National Health and Medical Research Council of Australia Senior Principal Research Fellow at Torrens University in Adelaide.
Among the novel findings in the new NEDA analysis: an LVEF below 50% was more than twice as common in men than women, occurring in 17.6% and 8.3%, respectively. Also, women had a higher average LVEF: 64.2%, compared with 59.5% in men. The overall 1- and 5-year all-cause mortality rates in the half-million participants were 5.8% and 18.4%.
Cardiovascular-related mortality occurred in 7.1% of women in median of 5.6 years of follow-up and in 8.1% of men with 5.5 years of follow-up.
All-cause and cardiovascular mortality rates followed a J-shaped curve, with the clear nadir occurring at an LVEF of 65%-69.9% in both women and men. But for LVEF values outside the nadir, a striking sex-based difference was present. Cardiovascular mortality, when adjusted for body mass index, age, heart rate, valvular heart disease, E-wave velocity, and other potential confounders, wasn’t significantly different between men whose LVEF was 65%-69.9% and those with an LVEF of 45%-64.9%. It started climbing in earnest only at an LVEF below 45%. In contrast, women with an LVEF of 45%-54.9% had a statistically significant twofold increased cardiovascular mortality rate compared to those in the nadir. Moreover, women with an LVEF of 55%-59.9% showed a trend in the same unwanted direction.
High LVEF, higher mortality in women
Dr. Stewart drew attention to an inflection point in the mortality curve for women whereby mortality began climbing at LVEF values of 70% or more. Values in that high range were documented in 72,379 women and 51,317 men.
He noted that the NEDA finding of an increasing mortality risk at LVEFs of at least 70%, especially in women, is similar to a recent report from another big data study, this one involving more than 200,000 patients who underwent echocardiography in routine clinical practice in the Geisinger health system in Pennsylvania. The investigators found in this retrospective study that during a median of 4 years of follow-up after echocardiography, the adjusted risk for all-cause mortality followed a U-shaped curve. The nadir of risk occurred in patients with an LVEF of 60%-65%, with a 1.71-fold increased risk at an LVEF at 70% or more and a near-identical 1.73-fold increased risk at an LVEF of 35%-40%. In this study, however, which was less than half the size of the NEDA analysis, the U-shaped LVEF/mortality curve applied to both men and women. Similar findings were seen in a validation cohort of nearly 36,000 patients from New Zealand (Eur Heart J. 2020 Mar 21;41:1249-57).
The investigators predicted that in addition to the existing categories of HFrEF, heart failure with preserved ejection fraction (HFpEF), and the more recently proposed heart failure with midrange ejection fraction (HFmrEF), their results “may herald the recognition of a new phenotype characterized by supranormal LVEF,” with a moniker of HFsnEF.