Impact of body mass index on mortality in heart failure patients

BACKGROUND: Higher body mass index (BMI) is associated with incident heart failure (HF), but paradoxically associated with better prognosis, recognized as the obesity paradox in HF. However, the impact of BMI on detailed prognosis on HF and the mechanism of obesity paradox remain still unclear. MATERIALS AND METHODS: We researched consecutive 648 patients admitted for HF as follows: underweight (BMI < 18·5 kg/m(2) , n = 86), normal (18·5 ≤ BMI < 25, n = 380), overweight (25 ≤ BMI < 30, n = 147) and obese (30 ≤ BMI, n = 35) and compared the results from their laboratory tests and echocardiography. We also followed cardiac and all-cause mortality. RESULTS: Obese group had a higher prevalence of obesity-related comorbidity (hypertension, diabetes, dyslipidemia); however, tumour necrosis factor-α, adiponectin, troponin T and systolic pulmonary arterial pressure were higher in the underweight group than in the other groups (P < 0·05, respectively). Left and right ventricular systolic function did not differ among the groups. In the Kaplan-Meier analysis, cardiac and all-cause mortality progressively increased from obese to overweight, normal and underweight group. Importantly, in the Cox proportional hazard analyses after adjusting for known risk factors, BMI was an independent predictor of cardiac and all-cause mortality (P < 0·01, respectively) in HF patients. CONCLUSIONS: Body mass index was an independent predictor of cardiac death and all-cause mortality in HF patients. Furthermore, lower BMI was associated with higher circulating levels of tumour necrosis factor-α, adiponectin and troponin T and higher systolic pulmonary arterial pressure.