Heart Failure and Body Habitus: What Would Be the Confounding Factors of Outcome Prediction in Obese Patients with Heart Failure?

Obesity has been known as one of the major confounding factors in the prediction of clinical outcomes in various cardiovascular diseases, including acute myocardial infarction¹⁾ and heart failure (HF).²⁾ Recently, increasing prevalence and changes in diagnostic criteria of obesity are precipitating a number of issues in HF, including the “obesity paradox” in clinical outcomes²⁾ and an inverse correlation with brain natriuretic peptide (BNP) level.³⁾ Body mass index (BMI) was reported to affect the concentration of BNP in patients with HF by regulating secretion and clearance. A study of 3,389 patients from the Framingham study group reported that BNP was lower in obese patients due to increased clearance from the higher concentration of natriuretic peptide (NP) receptor-C (NPR-C) developed in adipose tissue, and this may contribute to their susceptibility to hypertension and hypertension-related disorders.⁴⁾ However, soon another group reported that N-terminal pro-B type NP (NT-proBNP) was also lower in obese patients.⁵⁾,⁶⁾ Given that this peptide may not be cleared via NPR-C, unlike BNP, the increased clearance of BNP is not a proper explanation for low BNP levels in obesity anymore. The mechanism is explained with the glycosylation of proBNP. ProBNPs are composed of 108 amino acids and processed by corin and furin at the site between amino acids 76 and 77. When proBNP is cleaved into BNP and NT-proBNP, BNP is an active form of NP. However, obesity is associated with more glycosylation of proBNP than normal conditions, especially at threonine 71.⁷⁾ As a consequence, the processing of proBNP cleavage can be hindered by the glycosylation, which leads to the low serum level of BNP and NT-proBNP. This phenomenon might also affect clinical practice. There are opinions about lowering BNP cut-off value in the diagnosis of HF in obese patients.⁸⁾

Nevertheless, it is controversial whether this lower BNP level can affect the predictability of BNP in terms of clinical outcomes such as mortality or rehospitalization. In this edition of the journal, Kim et al.⁹⁾ investigated the different impacts of NT-proBNP level on the clinical outcomes of HF patients according to a BMI cut-off of 23 kg/m². The authors reported that lower BMI group showed greater discriminative power of NT-proBNP level in mortality prediction. In a brief post-hoc analysis of PARADIGM-HF trial with 8,217 chronic HF patients, Nadruz et al.¹⁰⁾ showed the attenuated ability of NT-proBNP to predict the primary outcome of cardiovascular death and HF hospitalization in moderately or severely obese patients. Kim et al.⁹⁾ also mentioned the study in the discussion section. However, the BMI value of the high BMI group seemed to be lower in the study of Kim et al.⁹⁾ than in that of Nadruz et al.¹⁰⁾ Since only very high BMI such as >35 kg/m² was associated with less prognostication of BNP in the study of Nadruz et al.¹⁰⁾, it is not easy to declare consistency between the 2 studies. To date, however, most of the studies were carried out in patients with obesity only and the impact of overweight on the BNP level has not been investigated well. Moreover, this kind of study might be particularly difficult because obesity lowers the NT-proBNP level itself and also affect positively on clinical outcomes according to the “obesity paradox”.²⁾

In future studies, the types of obesity and HF should be taken into account. A recent study conducted in Asia showed that the lean-fat patients showed the worst HF clinical outcomes even with low BMI.¹¹⁾ Even with the same BMI value, patients with central obesity showed more adverse outcomes than overall lean patients. Type of HF, such as HF with preserved ejection fraction (HFpEF) and HF with reduced ejection fraction (HFrEF), may also affect the interaction between BMI and BNP level. A recent post hoc analysis of Treatment of Preserved Cardiac Function Heart Failure With an Aldosterone Antagonist trial showed that there is a bimodal increase of mortality in low- and high-end of BMI group.¹²⁾ Regarding mortality, NP was stronger risk factor regardless of the BMI group. Regarding rehospitalization, however, the outcome was more easily affected by BMI. In the end, the composite outcome seemed to be driven by higher rehospitalization. The study included HFpEF patients only and showed very different outcomes from HFrEF studies. In conclusion, HF studies involving obese patients require careful interpretation as the results of them are subject to change according to ejection fraction, fat distribution, and severity of obesity.