Bioelectric impedance vector analysis (BIVA) in hospitalised children; predictors and associations with clinical outcomes

Background Clinical use of bioelectric impedance is limited by variability in hydration. Analysis of raw bioelectric impedance vectors (BIVA), resistance (R), reactance (Xc) and phase angle (PA) may be an alternative for monitoring disease progression/treatment. Clinical experience of BIVA in children is limited. We investigated predictors of BIVA and their ability to predict clinical outcomes in children with complex diagnoses. Methods R, Xc and PA were measured (BODYSTAT Quadscan 4000) on admission in 108 patients (4.6–16.8 years, mean 10.0). R and Xc were indexed by height (H) and BIVA-SDS for age and sex calculated using data from healthy children. Potential predictors and clinical outcomes (greater-than-expected length-of-stay (LOS), complications) were recorded. Results Mean R/H-SDS was significantly higher (0.99 (SD 1.32)) and PA-SDS lower (−1.22 (1.68))) than expected, with a wide range for all parameters. In multivariate models, the Strongkids risk category predicted R/H-SDS (adjusted mean for low, medium and high risk = 0.49, 1.28, 2.17, p = 0.009) and PA-SDS (adjusted mean −0.52, −1.53, −2.36, p = 0.01). BIVA-SDS were not significantly different in patients with or without adverse outcomes. Conclusions These complex patients had abnormal mean BIVA-SDS suggestive of reduced hydration and poor cellular health according to conventional interpretation. R/H-SDS was higher and PA-SDS lower in those classified as higher malnutrition risk by the StrongKids tool. Further investigation in specific patient groups, including those with acute fluid shifts and using disease-specific outcomes, may better define the clinical role of BIV.