Comparison of methods for estimating body fat percentage using skinfolds, InBody, and DXA: A pilot study

Abstract

Accurate determination of body fat percentage (%BF) is crucial for health assessment, sports diagnostics, and body composition monitoring in both clinical and field settings. Dual-energy X-ray absorptiometry (DXA) is widely regarded as a reference method because of its ability to distinguish fat, lean, and bone tissue with high precision, but its limited accessibility and cost reduce its routine applicability in schools, sports clubs, and practical field environments. Consequently, more accessible methods such as skinfold measurements and bioelectrical impedance analysis (BIA) are frequently used. However, both methods may be influenced by methodological, device-specific, and population-specific sources of bias. The aim of this pilot study was to compare the agreement and systematic deviations among an innovative 10-site skinfold method, the InBody 770 BIA device, and DXA in an adult sample. Ten participants (6 men, 4 women; age 29–62 years) underwent three body fat assessments on the same day under standardized fasted conditions: 1) 10-site skinfold measurement, 2) InBody 770 measurement, and 3) DXA scanning. Data were analyzed using Pearson correlation, Bland–Altman analysis, and Deming regression. Skinfold measurements significantly correlated with DXA (r = 0.78, p = 0.007) and showed a mean systematic difference of −2.45%. InBody 770 also correlated strongly with DXA (r = 0.96, p = 0.001), but showed a more pronounced underestimation of %BF (InBody − DXA = −3.98%). Deming regression confirmed distortion relative to the reference in both field methods. The 10-site skinfold method appears to be a transparent, standardized, and practically applicable field approach. The findings should be interpreted as pilot data that identify systematic bias and provide a basis for the design of larger validation studies.