Bioelectrical impedance analysis (BIA) is gaining popularity as a tool for body composition assessment. Although BIA has been studied and validated in different populations, age groups, and clinical settings, including critically ill patients, there are concerns about BIA reproducibility and reliability for different device types and postures. This study aimed to evaluate the reliability of BIA using different devices, postures, and lead types. Cross-sectional observational data were collected from 74 healthy volunteers (32 women, 42 men). We used two types of devices, three types of postures (standing, sitting, and lying), and two lead types (clamp lead and adhesive lead) to measure the whole-body phase angle (phA) at a single frequency of 50 kHz. The measurements were validated using the intraclass correlation coefficient (ICC) and Bland-Altman plot analysis. All phA measurements recorded using the two types of devices, three different postures, and two types of leads were equivalent (mean ICC = 0.9932, 95% confidence interval (CI) 0.9905-0.0053, p < 0.001). The average mean difference in phA was 0.31 (95% CI 0.16-0.46). The largest phA value was measured using BWA with an adhesive-type lead in the supine position. There were no differences between the standing and sitting positions. We compared the consistency and reliability of phA using two devices, two lead types, and three postures. Seven different phA were interchangeable in healthy volunteers.
Keywords: Bland–Altman; bioelectrical impedance analysis; comparison; intraclass correlation coefficient; phase angle; reliability.