Within- and between-day reliability of laboratory-based body composition assessment methods

This study examined the within- and between-day reliability of commonly employed laboratory-based body composition assessment methods. Eighteen healthy adults reported to the laboratory on two occasions, separated by 1-2 d, and were assessed with multiple methods under standardised conditions. These included two dual-energy X-ray absorptiometry (DXA) scanners with manual or automated regions of interest, air displacement plethysmography (ADP) with measured or estimated thoracic gas volume (TGV), two multi-frequency bioelectrical impedance analysers (MFBIA), single-frequency BIA (SFBIA) and bioimpedance spectroscopy. Within-day reliability was established via immediate test-retest assessments at the first visit (technical error) and between-day reliability by repeating assessments at the second visit (technical plus biological error). Within- and between-day measures were evaluated using technical error of measurement (TEM) and intraclass correlation coefficients (ICC). For body fat percentage, within-day TEM ranged from 0·03 to 0·70 %, with ICC of 0·993-1·000, while between-day ranged from 0·37 to 1·24 %, with ICC of 0·965-0·998. For fat and fat-free mass (FM and FFM), within-day TEM ranged from 0·02 to 0·53 kg, with ICC of 0·998-1·000 for FFM and 0·992-1·000 for FM. Between-day TEM ranged from 0·26 to 0·90 kg for FM and FFM, with ICC of 0·995-0·999 for FFM and 0·955-0·998 for FM. Across metrics, DXA demonstrated the best reliability, followed by one MFBIA and ADP with predicted TGV. Overall, laboratory-based body composition devices provide excellent within- and between-day reliability under standardised conditions, though observed differences in error may influence method selection when tracking small body composition changes.