Long-term prognostic significance modified body mass index in patients with decompensated acute heart failure: Insight from the OPAR study

Abstract

Abstract Background/Introduction Malnutrition is a well-known poor prognostic factor in patients with acute decompensated heart failure (ADHF). Although conventional body mass index (cBMI) and serum albumin level are commonly used as indicators of malnutrition, these values are affected by fluid balance. The recently proposed modified BMI (mBMI, serum albumin [g/l] × cBMI [kg/m2] ) may be a better prognostic marker, but its usefulness in patients with ADHF remains unclear. Purpose This study aimed to investigate the prognostic significance of mBMI in patients hospitalized for ADHF. Methods The OPAR study is a single-center registry enrolling 605 consecutive patients who were admitted for ADHF with survival discharge between October 2011 and October 2017. The present study population was consisted of 573 patients admitted for ADHF who had a value of cBMI and serum albumin at discharge. The eligible patients divided into the following four groups according to the mBMI quartiles: lowest quartile (Q1, mBMI <600, n = 144), second quartile (Q2, 600≤ mBMI <715, n = 144), third quartile (Q3, 715≤ mBMI <842, n = 142), and highest quartile (Q4, mBMI ≥842, n = 143). The primary outcome measure was all-cause death, and the second outcome measures were cardiac death, non-cardiac death, and heart failure rehospitalizations. Cumulative incidences of death and heart failure rehospitalizations were estimated by the Kaplan-Meier and Gray’s method, respectively. Cox proportional hazards regression models were used to identify patients at risk of all-cause death to calculate the multivariable-adjusted hazard ratios (HRs) and 95% confidence intervals (CIs). The predictive power of mBMI for all-cause mortality was assessed using receiver operator characteristic curve analysis. Results During a median follow-up period of 4.68 years (interquartile range, 2.16–6.72 years), the cumulative 5-year incidences of all-cause death substantially increased in the lower mBMI quartiles (Q1: 60.7% [95% CI, 52.5%–68.9%], Q2: 53.9% [95% CI, 45.7%–62.5%], Q3: 38.8% [31.1%–47.7%], and Q4: 21.3%[15.3%–29.2%], log-rank P <0.001). After adjustment for the potential confounders, the excess risk of Q1 relative to Q4 for all-cause death tended to be higher (adjusted HR, 1.44 [95% CI, 0.94–2.20], P = 0.09). The predictive power of mBMI was modest with a C-statistic of 0.65 (95%CI, 0.61–0.70). The cumulative 5-year incidence of cardiac death and non-cardiac death also increased in the lower mBMI quartiles (Q1: 29.1% and 44.5%, Q2: 18.9% and 43.1%, Q3: 21.3% and 22.3%, and Q4: 7.2% and 15.2%, respectively, log-rank P <0.001 for both). The cumulative 5-year incidences of heart failure rehospitalizations did not significantly differ across the mBMI quartiles (P = 0.35). Conclusions The mBMI at discharge is a simple and potentially helpful marker for the prediction of long-term mortality in patients hospitalized for ADHF.Primary outcomeSecondary outcome