I was interested to read the report by Gaasch et al on the distribution of left ventricular (LV) ejection fractions (EFs) in patients with chronic heart failure (HF). Specifically, they used trial data to test whether there is a continuum of HF from patients with normal EFs (HF with normal EF [HFNEF]) to those with reduced EFs (HF with reduced EF [HFREF]). Because the investigators found that the EF distribution was suggestive of a bimodal distribution, they concluded that there is no continuum.

This study misses the point and is misleading. Whether a patient with HFNEF progresses to HFREF depends on the cause of the HF and the community in which the patient lives. Acute myocardial infarction appears to be a potent stimulus to ventricular remodeling, and patients will pass relatively quickly after infarction from HFNEF to HFREF with typical LV dilatation. Hypertension (with or without diabetes mellitus) is a much less powerful stimulus to LV chamber enlargement, and the initial response is LV hypertrophy, and if blood pressure is controlled, patients will not progress to HFREF with dilated remodeled left ventricles. However, it is common to see patients in Africa or Asia with severe HF typical of systolic HF or HFREF whose only risk factor is hypertension. Presumably, these patients passed through HFNEF (or diastolic HF with LV hypertrophy) but gradually developed typical HFREF because their hypertension was not managed properly. This is the continuum. Obviously, in the United States and Europe, there is less progression, because generally hypertension is better managed. Thus, the distribution of LV EFs will depend on the prevalence of risk factors, the cause of HF, and the health care system. The response to treatment with standard HF drugs that primarily target LV remodeling will depend on whether remodeling has taken place or not, and the failure of these drugs in HFNEF is not surprising.

Finally the argument against the term “diastolic HF” is purely semantic. Diastolic HF implies that the cause is due primarily or entirely to LV diastolic dysfunction. Clearly this is not true, as HFNEF and HFREF have mixtures of systolic and diastolic LV abnormalities to varying degrees, but they represent different stages in the continuum of HF.