Comparison of body composition assessment across body mass index categories by two multifrequency bioelectrical impedance analysis devices and dual-energy X-ray absorptiometry in clinical settings

Background InBody-770 and SECA mBCA 515 are multifrequency bioelectrical impedance analysis (BIA) devices, which are commonly used in the clinic to assess fat-free mass (FFM) and body fat (BF). However, the accuracy between devices in clinical settings, across different body mass index (BMI) groups remains unclear. Methods Body composition for 226 participants (51% men, aged 18–80 years, BMI 18–56 kg/m²) was assessed by two commercial multifrequency BIA devices requiring standing position and using eight-contact electrodes, InBody 770 and SECA mBCA 515, and compared to results from dual-energy X-ray absorptiometry (DXA). Measurements were performed in a random order, after a 3 h fast and no prior exercise. Lin’s-concordance correlation and Bland–Altman analyses were used to compare between devices, and linear regression to assess accuracy in BF% across BMI groups. Results We found strong correlation between DXA results for study population BF% and those obtained by InBody ( ρ c = 0.922, 95% confidence interval (CI) 0.902, 0.938) and DXA and SECA ( ρ c = 0.940, CI 0.923, 0.935), with 95% limits of agreements between 2.6 and −8.9, and 7.1 and −7.6, respectively. BF% assessment by SECA was similar to DXA (−0.3%, p = 0.267), and underestimated by InBody (−3.1%, p < 0.0001). InBody deviations were largest among normal weight people and decreased with increasing BMI group, while SECA measurements remained unaffected. Conclusions Both BIA devices agreed well with BF% assessment obtained by DXA. Unlike SECA, InBody underestimated BF% in both genders and was influenced by BMI categories. Therefore, in clinical settings, individual assessment of BF% should be taken with caution.