A new anthropometric model for body composition estimation: Comparison with a bioelectrical impedance consumer device

BACKGROUND: The present study refers to an anthropometric model, Dahlmann-Body-Analysis (DBA), based on formerly developed weight-height-frame tables. Including the anthropometric variable abdomen circumference (AC), a further differentiation into muscularity and fat mass could be realized. This enables to calculate the individual percentage fat mass (%FM), providing a cost-effective method for epidemiological studies. The present work sets out to investigate, whether %FM computed by the DBA model compares to BIA measurements, notably under conditions of heavy obesity. SUBJECTS/METHODS: In 103 adults (37 males, 66 females, age 42.5 ± 12.5 years, BMI 38.2 ± 4.8 kg/m²) %FM was estimated by a tetrapolar BIA device and compared to results derived by the DBA system. Bland-Altman and simple linear regression analyses were used to determine agreement between methods. RESULTS: The mean %FM estimates of men (women) ± standard deviation were 36.1 ± 4.8 (49.7 ± 4.5) for BIA and 36.7 ± 4.2 (49.1 ± 4.9) for DBA measurements. Pearson correlation coefficients (r) and Lin's concordance correlation coefficient (CCC) were r = 0.86 and CCC = 0.84 for men and r = 0.85 and CCC = 0.83 for women, respectively. Bland-Altman plot showed limits of agreement between DBA and BIA %FM that ranged from 5.5% to -4.3% for men and 4.6% to- 5.7% for women, respectively. Correlations between values measured by both methods were high and the observed confidence interval (SD of the difference between DBA and BIA result multiplied by 2) was low. No systematic error was found. The DBA system overestimates FM by 0.63 ± 4.98 (2SD) % for men and underestimates FM by -0.56 ± 5.28 (2SD) % for women, respectively, compared to the BIA results. The result for both genders is -0.066 ± 5.17 (2SD) %FM. Over all, there is a strong concordance and reproducibility between the DBA and BIA data sets. CONCLUSIONS: The implementation of the abdomen circumference (AC) into the DBA model as a proxy for body fat (%FM) resulted in a strong concordance with BIA measurements. These findings indicate that the DBA model may reflect the body shape of severely obese white European patients with regard to body composition.