Abstract P021: NT-proBNP Predicts Incident Heart Failure Among Individuals in All BMI Categories

Abstract

Background: The N-Terminal of pro-Brain Natriuretic Peptide (NT-proBNP) is a potent predictor of incident HF in the general population. Persons with obesity, despite their predisposition to developing HF, are known to have low NT-proBNP levels. The utility of NT-proBNP for HF prediction among individuals with obesity is therefore unclear.

Hypothesis: We hypothesized that NT-proBNP levels may predict incident HF differently across body mass index (BMI) categories.

Methods: We followed 12,509 ARIC participants free of prior HF at baseline (Visit 2, 1990-1992) with BMI ≥18.5 kg/m2 and available NT-proBNP. BMI (kg/m2) was categorized as normal (18.5-24.9), overweight (25-29.9), obese (30-34.9) and severely obese (≥35). The primary outcome was incident HF (1,986 events during a mean follow-up of 15.8 years). The potential effect modification of obesity on the NT-proBNP-HF relationship was assessed on both relative and absolute scales using multivariable Cox regression models. We also evaluated whether NT-proBNP improves HF prediction beyond traditional risk factors within the ARIC HF Risk Score similarly across BMI categories.

Results: As expected, higher BMI was associated with lower NT-proBNP (r=-0.09, p<0.0001). Within each BMI category, higher NT-proBNP levels were significantly and similarly associated with increased relative risk for HF (Figure 1A). However, given the higher incidence of HF associated with obesity, the absolute risk difference associated with higher NT-proBNP levels was greatest among individuals with severe obesity (Figure 1B). In all BMI categories, NT-proBNP significantly improved the C-statistic beyond the ARIC HF Risk Score (increases of 0.026 [0.009-0.043] and 0.033 [0.015-0.052] for normal weight and severely obese, respectively).

Conclusions: Despite the inverse relationship between NT-proBNP and obesity, NT-proBNP provides significant prognostic information regarding HF risk in all BMI categories, supporting its use for HF prediction even among those with obesity.