Associations of bioelectrical impedance and anthropometric variables among populations and within the full spectrum of malnutrition

This study aimed to evaluate body composition variability assessed by bioimpedance in relation to nutritional status assessed by anthropometry in children and adolescents living in countries characterised by contrasting nutritional conditions. The sample comprised 8614 children (4245 males; 4369 females), aged 3 to 19 years, from Nepal (477 children), Uganda (488 children and adolescents), UK (297 children and adolescents) and US (7352 children and adolescents). Height-for-age (HAZ) and body mass index-for-age (BAZ) z-scores were calculated according to WHO growth references. Specific bioelectrical impedance vector analysis (BIVA) was used to evaluate body composition variability. In each population sample, the relationship of HAZ and BAZ with bioelectrical outcomes was analysed by confidence ellipses and cubic spline regression, controlling for sex and age. The participants from Uganda and Nepal were more affected by undernutrition, and those from US and UK by obesity. In all groups, phase angle and specific vector length were weakly associated with HAZ, with null or opposite relationships in the different samples, whereas they were positively associated with BAZ. The stronger association was between vector length, indicative of the relative content of fat mass, and BAZ in the UK and US samples. Confidence ellipses showed that the relationships are more strongly related to phase angle in Nepalese and Ugandan samples. Bioelectrical values were more strongly associated with BAZ than HAZ values in all population samples. Variability was more related to markers of muscle mass in Ugandan and Nepalese samples and to indicators of fat mass in UK and US samples. Specific BIVA can give information on the variability of body composition in malnourished individuals.