Abstract The BIA primary result sheets as a rule contain one-dimensional graphical scales with a selected area of normal values. In 1994, Piccoli et al. proposed BIVA, an alternative form of BIA data presentation, where two bioimpedance parameters are considered simultaneously as tolerance ellipses: resistance and reactance normalized to height. The purpose of this study is to develop an approach to data analysis in body composition bioimpedance research in two-dimensional representations. The data of 1.124.668 patients aged 5 to 85 years who underwent a bioimpedance study in Russian Health Centers from 2009 to 2015 were used. Statistical programming in the R Studio environment was carried out to estimate two-dimensional distribution densities of pairs of body composition parameters for each year of life. The non-Gaussian distribution is found in most parameters of bioimpedance analysis of body composition for most ages (Lilliefors test, p-value << 0.0001). The slices of the actual two-dimensional distribution pairs of body composition parameters had an irregular shape. The authors of the article propose using the actually observed distribution for populations where numerous bioimpedance studies have already been carried out. Such technology can be called two-dimensional bioimpedance analysis of human body composition (2DBIA). The 2DBIA approach is clearer for practitioners and their patients due to the use of body composition parameters in addition to electrical impedance parameters.

Based on data of the cross-sectional bioimpedance study of 819,808 male and female visitors of health centers (representing about 0,6% of the Russian population) by the BIA meters ABC-01 «Medas», an estimates of overweight and obesity prevalence in Russia, and the diagnostic efficiency of the WHO criteria for obesity depending on sex and age, were obtained. An average 15% frequency of incorrect measurements was observed which emphasizes the importance of the development and implementation of the training program on BIA measuring technique and data interpretation.

Here, we report first results of the large-scale ongoing bioelectrical impedance body composition study in Russians. By the end of 2012, 216 out of 800 Russian Health Centres submitted raw bioimpedance data on 844,221 adults and children aged 5-80 years, representing nearly 0.6% of the Russian population, who were accessed cross-sectionally using the same type of bioimpedance meter, ABC-01 Medas. Estimates of overweight, obesity, and normal weight obesity prevalence in the general population, as well as characteristics of diagnostic sensitivity and specificity of the conventional WHO BMI-based criteria of obesity depending on age are obtained. The smoothed reference centile curves for percentage fat mass are constructed, and localized cut-offs for fatness and thinness are provided that can be used both at the individual and epidemiological levels.