Bioimpedance vector analysis in stable chronic heart failure patients: Level of agreement between single and multiple frequency devices.

Abstract

The accuracy of estimating body composition compartments is critical in the clinical setting. Currently, there are different bioelectrical impedance analysis (BIA) devices available for obtaining raw BIA parameters. The aim of this study was to determine the level of agreement between multiple frequency (MF)-BIA and single frequency (SF)-BIA devices in obtaining raw BIA measurements (resistance (R), reactance (Xc), and phase angle (PhA)), as well as the agreement on the classification of hydration status and body cell mass by the bioelectrical impedance vector analysis (BIVA) method. This cross-sectional study included 406 outpatients with stable chronic heart failure (HF). The raw BIA measurements at 50kHz obtained by tetrapolar MF-BIA (Bodystat QuadScan 4000) were compared with those obtained by tetrapolar SF-BIA (RJL Quantum X). In addition, the patients were classified by their hydration status and body cell mass according to the BIVA method. Strong and significant correlations were observed between the two methods in all raw BIA variables (r≥0.90). Lin's concordance correlation coefficient (CCC) values were almost perfect for R (CCC=0.99; 95% CI 0.997 to 0.998), moderate for Xc (CCC=0.93; 95% CI 0.92 to 0.94), and poor for PhA (CCC=0.88; 95% CI 0.85 to 0.90). The agreement obtained in the two classifications (quadrants and hydration status) was >0.81. MF-BIA and SF-BIA demonstrated good agreement for measurement of the R parameter; however, the Xc and PhA parameters must be used carefully due to the previously reported variability. Likewise, the agreement in all classifications by the BIVA method was almost perfect.