Response to Letters Regarding "Evidence Supporting the Existence of a Distinct Obese Phenotype of Heart Failure With Preserved Ejection Fraction”

We thank Drs. Koutsampasopoulos, Jin-shan, Carbone and colleagues for their interest in our recent publication characterizing cardiovascular structure, function, and reserve capacity in patients with the obese phenotype of heart failure with preserved ejection fraction (HFpEF).¹ Koutsampasopoulos et al. point out that diabetes was more prevalent in obese HFpEF and that this might confound interpretation. To address their concern, we have performed additional sub-analyses. We observed no differences between obese HFpEF with (n=33) and without diabetes (n=66) in biventricular structure or function, ventricular interaction, rest and exercise hemodynamics, or exercise capacity (all p>0.2). This indicates that the observed differences in the obese HFpEF group are not simply ascribable to diabetes alone.

Jin-shan and Xue-bin point out the high prevalence of hypertension and diabetes in patients with obese HFpEF in our study.¹ The authors propose a study in obese HFpEF patients with no hypertension or diabetes as compared to non-obese HFpEF. While this would be ideal, it would be difficult, as these comorbidities are highly prevalent in patients with obesity as well as HFpEF. The authors question whether obesity causes a transition from HFpEF to HF with reduced EF. The cross-sectional nature of our study does not allow us to address this question, but previous studies have reported that this transition is uncommon unless coronary disease is present.² Prospective longitudinal studies are needed to delineate this question further.

Carbone and colleagues state their belief that obesity is not a specific phenotype of HFpEF, but rather a co-existing condition. Our data do not support this belief.¹ The authors imply that our study cohort was biased, but it was drawn from consecutive patients referred to our laboratory, reflecting what is seen in the community. The authors suggest that adiposity is not correlated with cardiac parameters, but previous studies have shown that adiposity and weight gain are directly associated with increases in ventricular stiffening, which is known to underlie progression to HFpEF.^{3, 4} Importantly, our data show that directly-measured pulmonary capillary wedge pressure (PCWP) was significantly correlated with body mass in the obese HFpEF subjects, a relationship that was not seen in the non-obese HFpEF group (Figure 2).¹ We observed even stronger correlations between exercise PCWP and weight (r=0.44, p<0.0001) as well as BMI (r=0.50, p<0.0001). These data provide compelling evidence that adiposity does affect cardiac function in obese HFpEF.

Carbone et al. point out that we did not include a control group with obesity but no HFpEF. In response, we have analyzed data from obese patients free of HF who underwent the same evaluation (BMI ≥35 kg/m², n=13). We found that these obese control patients displayed normal pulmonary artery pressures (mean 19±3 mmHg) and normal PCWP (9±4 mmHg). Right ventricular size, function and total heart volume did not differ in obese patients without HFpEF from non-obese controls, and evidence of pericardial restraint was absent. These data do not support Carbone and colleagues’ assertion that obesity per se is associated with right heart failure.

We observed greater biventricular remodeling, impaired pulmonary vasodilation and enhanced ventricular interdependence in patients with HFpEF and obesity.¹ A phenotype can be defined as a set of observable traits and characteristics that are unique, and as such may be targeted therapeutically. Indeed, one of the primary reasons to characterize different phenotypes in HFpEF is to provide rationale to conduct trials testing novel, individualized therapies. Impairments in ventricular function were similar in obese and non-obese HFpEF, indicating that these are shared features that are not specific to obese HFpEF, but common to all patients with HFpEF.¹ However, greater ventricular remodeling, pulmonary vascular dysfunction and enhanced ventricular interdependence were specific to the obese HFpEF cohort. Therefore, future trials testing therapies targeted to these derangements might enroll patients with the obese HFpEF phenotype. Conversely, because ventricular systolic and diastolic function was similarly impaired in obese and non-obese HFpEF, our data suggest that therapies targeting ventricular impairments common to all HFpEF phenotypes hold promise for patients with obesity-related HFpEF as well as non-obese HFpEF.¹ An important implication of this finding is that obese patients with HFpEF should not be excluded from trials testing interventions targeting these more pervasive mechanisms, as is too often the case in the current era.^{5, 6}