Body composition assessment: method comparison of DXA, InBody BIA and manual measurements

Background & Aims: The proportion and distribution of fat mass (FM) and fat-free mass (FFM) are key health indicators. Traditionally, anthropometric measures such as height, weight, body mass index (BMI), waist circumference (WC) and hip circumference (HC) have been utilized. However, bioimpedance analysis (BIA) is increasingly used in population-based studies. Our aims were to compare BIA with Dual-energy X-ray Absorptiometry (DXA), assess the impact of standardized versus unstandardized BIA, and to compare BIA with manual waist and hip measurements. Methods: The body composition of 316 people was measured using BIA, DXA, and manually measured anthropometrics. In addition, BIA was assessed using both standardized and unstandardized procedures in 46 individuals. Results are presented as intra class correlation coefficients (ICC) with 95% confidence intervals (CI). Results: BIA and DXA showed excellent agreement for total FM and FFM (ICC ≥ 0.97), but lower agreement for FM in arms (ICC 0.69) and legs (ICC 0.76). BIA also showed good agreement with manual measures for WC, HC and waist-to-height ratio (WHtR) (all ICC ≥ 0.85), but less so for waist-to-hip ratio (WHR) (ICC 0.53). Stratified analyses revealed similar agreement across sex and age groups, but lower consistency for FM in arms and trunk with increasing BMI. Comparisons of standardized versus non-standardized BIA procedures indicated excellent agreement for all body composition metrics (ICC ≥ 0.94) with only minor differences in estimates. Conclusions: BIA is a feasible and cost-effective alternative to DXA for body composition measurements showing strong overall agreement, especially for whole-body and regional FFM estimates. Although it is less accurate for regional fat distribution, BIA remains valuable for population-based studies. Standardized protocols are recommended for individual follow-ups, but our findings indicate that non-standardized procedures still provide very good precision, making BIA ideal for large-scale health surveys.