Heart failure with normal natriuretic peptide levels: more fat, and that is the main problem

This commentary refers to ‘Heart failure with preserved ejection fraction in patients with normal natriuretic peptide levels is associated with increased morbidity and mortality’, by F.H. Verbrugge et al., https://doi.org/10.1093/eurheartj/ehab911 and the discussion piece ‘Mixing the fat with water: are normal natriuretic peptide heart failure with preserved ejection fraction patients just more obese?’, by A. Olivella et al., https://doi.org/10.1093/eurheartj/ehac192. We thank Olivella et al. for their interest in our study.1 The inverse relationship between excess body fat and natriuretic peptide (NP) levels is well established. Indeed, in patients undergoing bariatric surgery, body weight reductions of ∼25% following surgery are coupled with three-fold increases in both brain natriuretic peptide (BNP) and its degradation product N-terminal proBNP (NTproBNP).2 Adipocytes clear BNP, but not NTproBNP, suggesting that excess body fat predominantly inhibits the production of NP rather than enhancing its degradation. Natriuretic peptide release from cardiomyocytes is triggered by increased wall stress. Patients with obesity display greater external constraint on the heart, due to epicardial fat deposition from without, and cardiomegaly from within, increasing coupling of the heart with the pericardium, collectively resulting in a lower transmural distending pressure for any measured intravascular pressure. Indeed, compared with non-obese patients, obese patients have a greater pulmonary arterial wedge pressure but similar left ventricular transmural filling pressure for the same NTproBNP.3 This may lead to systematic underestimation of congestion using NP levels, which then leads to underrecognition (and undertreatment) of heart failure with preserved ejection fraction (HFpEF) among patients with obesity, as well as inadequate representation in clinical trials, as the authors point out. Two alternative biomarkers, adrenomedullin and endothelin-1, were found to display less sensitivity to body mass, suggesting that these markers may be more robust for diagnosis in those with normal NP, but further study is required.3 A recent study has shown how increases in stressed blood volume (SBV) play an important role in obesity-related HFpEF.4 Stressed blood volume refers to a functional volume that increases with venoconstriction to augment venous pressure, and the ratio of SBV to total blood volume can be taken as a measure of venous capacitance. As the ratio of SBV to total blood volume increases, cardiac filling pressures will rise out of proportion to overall volume status, providing for a mechanism by which filling pressures rise even in the absence of blood volume expansion. This ratio is increased in patients with HFpEF and was shown to be directly related to body mass, suggesting that these patients with normal NP/obese HFpEF qualify for consideration as a specific phenotype, and display unique mechanisms underlying haemodynamic perturbations. While most patients with HFpEF and normal NP are obese, one in five patients with normal NP in our study were not obese, so there are clearly other factors involved in this phenotype, and further study is required. Although clinical outcomes in HFpEF with normal NP were better than high NP, this should not be interpreted as evidence in favour of a milder phenotype, or protective effect of obesity (i.e. the so-called obesity paradox). Observational registries showing this paradox are impossibly confounded, and a recent Mendelian randomization study not subject to this survival bias has demonstrated an increased risk for both HF and mortality in patients with obesity, with no protective effect.5 As patients with obesity-related HFpEF develop HF sooner in their life, when they are less frail, they have seemingly better outcomes than older and frailer patients with a similar severity of HF, but we should not underestimate the clinical importance of HFpEF in patients with normal NP levels. Conflict of interest: none declared.