Assessment of body composition by bioelectrical impedance in children and young adults is strongly age-dependent.

Body composition was measured in a group of 246 children and young adults, ranging in age from 7 to 25 years, by means of densitometry, body impedance (R) and anthropometry. Body fat percentage and fat-free mass (FFM) were calculated from body density using age-specific calculation formulas. From body impedance, FFM and body height, the specific impedance of the body (rho = R*FFM/height) was calculated in which body impedance is corrected for differences in the shape of the conductor. From age 10 onwards the specific impedance was positively related with age until age 13 in girls and age 16 in boys, after which it stabilized in boys at a significantly higher level than in girls. Based on the relationship of the specific impedance with age and sex, three age groups could be defined in which the relation between FFM and body impedance was analysed: age group I, boys and girls younger than 10 years; age group II, boys aged 10-15 years, girls aged 10-12 years; age group III, boys 16 years and older, girls 13 years and older. The regression equation in age group I had the same slope as the sex-specific regression equations in age group III. The regression equation in age group II had a steeper slope compared to the regression equations in age group I and III. Thus the relationship between FFM and body impedance was found to be slightly S-shaped, being identical for boys and girls until age 13, after which sex differences became apparent. For the different age categories the best prediction formula for the FFM from body impedance, sex, age and anthropometric variables was calculated. The prediction error of the age-specific regression equations was lower than the prediction error of the regression equation for the entire population.