Reevaluating obesity definitions and their association with hospital length of stay in older adults hospitalized for decompensated heart failure
M Czapla, C S Lee, M Jedrzejczyk, B Uchmanowicz, I UchmanowiczAbstract
Background
The prognostic relevance of obesity in patients hospitalized for decompensated chronic heart failure (HF) remains unclear. Traditional metrics such as body mass index (BMI) and waist circumference (WC) may not adequately reflect the clinical consequences of excess adiposity in this population.
Purpose
To compare the prognostic value of three obesity definitions—BMI-based, WC-based, and clinical obesity (as defined by The Lancet Diabetes & Endocrinology Commission) for hospital length of stay (LOS) in older adults with decompensated HF, and to identify independent predictors of prolonged hospitalization within each obesity category.
Methods
This prospective study included 250 patients aged ≥60 years hospitalized for decompensated chronic HF. Anthropometric measurements were obtained after clinical stabilisation and before discharge. Two WC thresholds were analysed: definition 1 (≥94 cm men / ≥80 cm women) and definition 2 (≥102 cm men / ≥88 cm women). Obesity was defined as: (1) BMI ≥30 kg/m² (classical obesity), (2) WC above threshold (central obesity), and (3) clinical obesity combining anthropometric and disease-based criteria. Length of stay was analysed using negative-binomial regression. Models included demographic, clinical, laboratory, nutritional, and frailty-related variables as potential determinants (age, sex, NYHA class, NT-proBNP, haemoglobin, renal function, C-reactive protein, presence of hypertension, diabetes, coronary artery disease, atrial fibrillation, nutritional status, and frailty indicators).
Results
Obesity prevalence differed across definitions: 36.8% (BMI-based), 64.8% (WC-based), and 50.0% (clinical). None of the obesity definitions were independently associated with LOS. In the clinical obesity model, NT-proBNP was the only independent predictor (IRR 1.004, 95% CI 1.001–1.007; p = 0.004). For classical obesity, NYHA class IV (IRR 2.36, 95% CI 1.32–4.19; p = 0.004), NT-proBNP (IRR 1.004, 95% CI 1.001–1.007; p = 0.004), and smoking ≥41–50 pack-years (IRR 1.95, 95% CI 1.10–3.43; p = 0.022) were independent predictors of longer hospitalization. In central obesity, haemoglobin (IRR 0.92, 95% CI 0.86–0.99; p = 0.029) and NT-proBNP (IRR 1.004, 95% CI 1.002–1.006; p < 0.001) remained significant. Detailed multivariable regression results are presented in Table 1.
Conclusions
In older adults hospitalized for decompensated HF, none of the obesity definitions, including the recent clinical obesity construct were independently related to LOS. Instead, disease severity indicators, particularly NT-proBNP and NYHA class, dominated prognostic performance. These findings emphasize the limited value of static obesity classifications in the inpatient setting and highlight the need for dynamic, pathophysiology-driven risk stratification in HF.Multivariable regression resultsFor image description, please refer to the figure legend and surrounding text.