Bioimpedance Spectroscopy in Critically Ill Children – Feasibility and Association with Protein Intake and Functional Status

Abstract

Abstract Objectives Bioimpedance spectroscopy (BIS) is used to assess body composition, but its application in critically ill children remains scant. In this pilot study, we evaluated the feasibility of BIS in patients admitted to a pediatric intensive care unit (PICU), and aimed to examine the association between body composition changes, macronutrient intake, and functional outcomes in this population. Methods Children requiring invasive mechanical ventilation with an expected length of stay of >5 days were enrolled. Functional status scale (FSS) was obtained from medical records and BIS measurements were completed within 3 days of admission; BIS was repeated weekly and FSS was repeated at PICU discharge. Daily clinical data and nutrient intake were recorded by dietitians. Nonparametric statistics were used to describe the characteristics of the cohort. Results Of the 20 patients enrolled to date, 18 (50% female, 56% medical, 44% surgical) had complete baseline data and 12 had follow-up BIS. Median (IQR) age was 4.1 (1.4, 7.8) years. Admission weight was 16.0 (11.2, 24.0) kg and mid-arm circumference Z-score was −0.3 (−1.8, 1.0). BIS measured total body fluid (TBF) was 11.0 (7.9, 17.5) liters, the ratio of extracellular to intracellular fluid (ECF: ICF) was 0.9 (0.8, 1.1), and fat free mass (FFM) was 15.1 (10.8, 23.9) kg. Age (r = 0.77, P < 0.001) and weight (r = 0.84, P < 0.001) were strongly correlated with TBF at enrollment, but not with ECF: ICF. Initial phase angle at 50 kHz (PA) was 2.8 (2.2, 3.8) and was negatively correlated with ECF: ICF (r = −0.79, P < 0.001). Average intake during PICU stay was 0.7 (0.4, 1.3) g protein/kg/day and 20 (10, 44) kcal/kg/day. After 1 week in PICU, median change in FFM among patients with 2 plausible measurements (n = 5) was −0.8 (−1.8, 0.5) kg. Median (IQR) length of stay (days) was 12.2 (6.9, 18.9) in the PICU, and 23.9 (14.7, 33.6) in the hospital. FSS improved from 15 (12, 22) to 9 (7, 11) at PICU discharge (P = 0.001). Intake and body composition measurements by BIS were not significantly correlated with FSS. Conclusions BIS measurements were feasible and ECF: ICF ratio was high in this population, possibly due to fluid shifts. FSS improved over the course of PICU admission. The interaction between nutrient intake, body composition, phase angle and functional status during pediatric critical illness remains unknown. Funding Sources ASPEN Rhoads Research Foundation.