The obesity paradox in heart failure: insights from a nationwide multi-ethnic Asian cohort
S M Jauhari, S M L Tan, M M Yu, Y C Lim, P L F Ng, P Chai, T L W Li, J S Y Ong, R Cherian, M Y Y Chan, J H S Chew, R C C Wong, W Q LinAbstract
Background
The obesity paradox—where higher body mass index (BMI) is associated with better outcomes in heart failure (HF)—has been described in Western cohorts, but remains insufficiently characterised in Asian populations, which exhibit distinct body composition and BMI distributions. We therefore aim to evaluate the association between BMI categories and composite outcome of all-cause mortality or HF re-hospitalisation in a large multicentre HF cohort in Singapore.
Methods
This study included 7,963 patients, including 571 HF with preserved ejection fraction (HFpEF) and 7,392 HF with reduced ejection fraction (HFrEF) patients, hospitalized for HF across eight tertiary hospitals in Singapore between 2016 and 2024. BMI was categorised according to Asian cut-offs: underweight (<18.5 kg/m2), normal (18.5–22.9 kg/m2), overweight (23.0–27.4 kg/m2) and obese (≥27.5 kg/m2). The primary endpoint was a composite outcome of all-cause mortality and HF re-hospitalisation. Kaplan–Meier curves were used to compare survival across BMI categories for the overall HF population as well as HFrEF and HFpEF subtypes. Cox proportional hazards models were adjusted for age, sex and NYHA class.
Results
BMI demonstrated a strong and graded association with composite outcome (Figure 1; overall log-rank p<1x10-15). In the overall cohort, when compared with normal BMI (Figure 2): obese patients had 92.3% lower risk (HR 0.0766, 95% CI 0.063–0.093); overweight patients had 67.5% lower risk (HR 0.325, 95% CI 0.277–0.382) and underweight patients had >2.6-fold higher risk (HR 2.61, 95% CI 2.37–2.87) of composite outcomes. In HFrEF, the obesity paradox was even more pronounced (obese: HR 0.0735; overweight: HR 0.333; underweight: HR 2.38, all p<1x10-51). In HFpEF, obesity remained protective (HR 0.36, 95% CI 0.18–0.72), while underweight was associated with double the risk (HR 2.01, 95% CI 1.05–3.85), and overweight was neutral (HR 0.93). Across all analyses, underweight status consistently conferred the highest risk, while overweight and obesity were protective.
Conclusion
In this contemporary multi-ethnic HF cohort, we observed a robust and consistent obesity paradox, particularly among HFrEF patients. Obese and overweight patients had substantially lower risk of all-cause mortality and HF re-hospitalisation, whereas underweight status was a strong independent predictor of adverse outcomes. These findings highlight the prognostic importance of nutritional status among HF patients, where underweight status could potentially represent underlying sarcopenia, frailty and cachexia. Collectively, these insights support the need for tailored risk stratification and management strategies involving assessment and optimisation of nutritional and BMI status in this population.For image description, please refer to the figure legend and surrounding text.For image description, please refer to the figure legend and surrounding text.