Relationship between NT-proBNP levels and BMI on clinical outcomes in heart failure: attenuation of the obesity paradox at higher disease severity

Background

There is increasing incidences of obesity and HF (heart failure) worldwide. The obesity paradox, where overweight and obese HF patients demonstrate improved survival has remained controversial. The N-terminal pro-BNP (NT-proBNP), a biomarker of cardiac wall stress, strongly reflects HF severity and may confound BMI–mortality relationships.

Purpose

We evaluated whether the obesity paradox persists after stratifying patients by NT-proBNP tertiles.

Methods

We analyzed 4529 adults hospitalized for HF across eight tertiary hospitals in Singapore (2016–2024). BMI was categorized using Asian thresholds: Underweight (<18.5 kg/m²), Normal (18.5–23), Overweight (23–27.5), and Obese (≥27.5). NT-proBNP values were divided into rank-based tertiles representing low (T1), mid (T2), and high (T3) cardiac stress. The most recent body mass index (BMI) prior to an event (or last available) was used. The primary endpoint was all-cause mortality. Survival was examined using Kaplan–Meier analysis and Cox proportional hazards models adjusted for age, sex, and NYHA class.

Results

In the overall cohort, higher BMI was independently associated with better survival. Obese (HR 0.73, p<0.001) and overweight patients (HR 0.88, p=0.024) demonstrated significantly lower mortality than normal-weight HF patients, consistent with the obesity paradox. Underweight trended toward higher mortality but was not significant. After stratifying by NT-proBNP tertiles, BMI effects weakened substantially:

●T1 (Low NT-proBNP): Overweight and obese were not significant; only underweight remained associated with higher mortality (HR 1.70, p=0.007).

●T2 (Mid NT-proBNP): Only overweight was modestly protective (HR 0.82, p=0.038).

●T3 (High NT-proBNP): No BMI category remained significantly associated with mortality (all p>0.22).

Kaplan–Meier curves showed narrowing separation of BMI groups across increasing NT-proBNP tertiles, and adjusted Cox models confirmed loss of significance for obesity-related protection.

Conclusions

The obesity paradox is clearly present in the unstratified HF population, but its effect is markedly attenuated and largely disappears after accounting for NT-proBNP–defined disease severity. These findings suggest that the apparent survival advantage of higher BMI may be driven by underlying differences in baseline cardiac stress rather than true protective effects of adiposity. NT-proBNP should be incorporated when evaluating BMI–prognosis relationships in HF.

Survival by BMI in Cohort and NT-proBNP For image description, please refer to the figure legend and surrounding text. Hazards for BMI in Cohort and NT-proBNP For image description, please refer to the figure legend and surrounding text.