Utility of bioelectrical impedance vector analysis in hospitalized polymorbid patients with heart failure

Abstract

Rationale: To evaluate bioelectrical impedance vector analysis (BIVA) with raw bioelectrical data as prognostic markers in polymorbid heart failure (HF) patients hospitalized with acute decompensation, where conventional nutritional tools are limited. Methods: Prospective cohort study of patients admitted with HF to Internal Medicine Department of the Insular-Materno Infantil Universitary Hospital of Gran Canaria (Dec 2022–Apr 2023). Clinical data, comorbidities, functional status (Barthel Index), and nutritional risk (MNA-SF) were collected. BIVA at 50 kHz assessed raw BIVA parameters:resistance (R), reactance(Xc), phase angle (PhA). Outcomes included length of stay, in-hospital mortality, and mortality during follow-up (Dec 31, 2024). The association between raw BIVA parameters and in-hospital mortality was assessed using linear regression. For variables showing a significant relationship, ROC curve analysis was conducted to identify predictive cutoff points. The Mann–Whitney U test was used to evaluate mortality differences based on these cutoffs. Survival was analyzed using Cox regression and Kaplan–Meier curves. Results: 32 patients (78% women), mean age 81.6 years, were included. Women were older and had lower functional scores. Average hospital stay was 15.5 days. In-hospital mortality was 15.7%; overall mortality during follow-up was 43%. Xc and SPA were significantly associated with in-hospital mortality (Xc: p < 0.001; SPA: p = 0.012). Xc < 31.9 predicted mortality with 72% sensitivity and 100% specificity (AUC = 0.88). Reduced Xc was associated with higher in-hospital mortality (p = 0.006) and long-term mortality (HR 5.99; p = 0.003). Conclusion: Older HF patients often present with complex clinical profiles that limit standard nutritional assessments. Raw BIVA data, especially reactance (Xc), enables quick bedside risk stratification and shows potential as a prognostic tool in this population. Disclosure of Interest: None declared