O008 - BIOIMPEDANCE ANALYSIS-DERIVED BODY COMPOSITION ESTIMATES IN CRITICALLY ILL PATIENTS: DAY-TO-DAY RELIABILITY AND CLINICAL CORRELATES

O008

BIOIMPEDANCE ANALYSIS-DERIVED BODY COMPOSITION ESTIMATES IN CRITICALLY ILL PATIENTS: DAY-TO-DAY RELIABILITY AND CLINICAL CORRELATES

W. Wasyluk^1,*, R. Fiut², A. Zwolak¹, A. Wójcik-Załuska², W. Dąbrowski³

¹Department of Internal Medicine and Internal Nursing, ²Department of Clinical Physiotherapy, ³I Clinic of Anaesthesiology and Intensive Therapy, Medical University of Lublin, Lublin, Poland

Rationale: Bioimpedance analysis (BIA) may assess body composition in intensive care unit (ICU) patients, but fluid shifts may distort derived estimates. We evaluated day-to-day reliability of estimates from two BIA devices.

Methods: In this prospective single-centre observational study, 43 adults underwent morning BIA on 4 consecutive ICU days using the Medical Body Composition Analyzer 525 (mBCA 525; seca, Germany) and the Body Composition Monitor (BCM; Fresenius, Germany). Day-to-day agreement of fat-free mass (FFM), fat mass (FM), skeletal muscle mass (SMM) and lean tissue mass (LTM) was assessed with intraclass correlation coefficients (ICC) and Spearman correlations. Clinical correlations of the BIA-derived parameters were also assessed.

Results: Complete measurements were available at 145/172 (84.3%) time points for mBCA 525 and 143/172 (83.1%) for BCM; partial data were obtained at 15/172 (8.7%) and 10/172 (5.8%), respectively. Day-to-day agreement was high for mBCA 525-derived FFM, FM and SMM (ICC 0.94, 0.90, 0.91), whereas BCM-derived LTM and FM showed moderate and high reproducibility (ICC 0.59 and 0.87). For both devices, short-term changes in body composition estimates were strongly related to changes in total body water (TBW). PhA correlated strongly with the intracellular water-to-extracellular water ratio (ICW/ECW) (rho=0.82, p<0.001), whereas associations with Acute Physiology Score (APS), inflammatory markers and ICU mortality were weak or inconsistent.

Conclusion: In ICU patients, BIA-derived body composition estimates were reproducible across consecutive days, but short-term changes were strongly influenced by body water changes, limiting interpretation as true tissue change. PhA was more closely related to ICW/ECW than to illness severity, inflammatory markers or ICU mortality, suggesting potential usefulness for hydration assessment in ICU patients.

Disclosure of Interest: None declared