| Description: |
A variety of laboratory-based methods have been utilized to assess body composition. However, technical and physiological variations can influence the estimates from these devices. PURPOSE: The purpose of this study was to assess the within- and between-day reliability of bioimpedance spectroscopy (BIS) and single-frequency bioelectrical impedance analysis (SFBIA), with varying electrode placement, for estimates of body fat percentage (BF%), fat-free mass (FFM), and fat mass (FM). METHODS: Eighteen healthy young adults (mean ± SD; age = 25.7 ± 7.5 years; BMI = 22.4 ± 2.3 kg/m2) completed two visits to the laboratory for BIS and SFBIA assessments, separated by 1-2 days. Participants remained supine during tests. For the first visit, duplicate assessments were performed using the same adhesive electrodes without removal (BIS same electrode [BISS], SFBIA same electrode [SFBIAS]). Another assessment was performed after replacing the electrodes (BIS different electrode [BISD], SFBIA different electrode [SFBIAD]). During the second visit, new electrodes were applied, and BIS and SFBIA measures were repeated. Test-retest reliability was established using the absolute technical error of the measurement (TEM) to determine the within-day and between-day errors for BF%, FFM, and FM. RESULTS: All within-day measures of body composition with the same electrode placement exhibited the lowest absolute TEMs (BF%: BISS = 0.06, SFBIAS = 0.03; FFM (kg): BISS = 0.04, SFBIAS = 0.02; FM (kg): BISS = 0.04, SFBIAS = 0.02) followed by within-day different electrode placement (BF%: BISD = 0.70, SFBIAD = 0.28; FFM: BISD = 0.53, SFBIAD = 0.19; FM: BISD = 0.53, SFBIAD = 0.19), and the largest errors were observed for between-day different electrode placement (BF%: BISD = 1.24, SFBIAD = 1.07; FFM: BISD = 0.90, SFBIAD = 0.85; FM: BISD = 0.81, SFBIAD = 0.73). CONCLUSION: Electrode placement is a more important source of within-day technical error for BIS and SFBIA technologies than the analyzers themselves. BIS exhibited greater error due to ... |