Clinical applicability of bioelectric impedance to measure body composition in health and disease.

The applicability of bioelectric impedance analysis (BIA) as a measure of body composition in children and young adults in both health and disease was studied in 155 subjects (5 mo to 54 yr of age) who were healthy (n = 21), had cystic fibrosis (n = 16), or had end-stage liver disease with variable clinical ascites or edema (n = 62). BIA and phase angle measured at a frequency of 50 kHz between the wrist and contralateral ankle by use of a tetrapolar measuring technique was compared with fat-free mass (FFM) estimated from skin-fold (SK) thickness measurements (n = 57), and body cell mass (grams potassium) measured by total-body potassium (TBK) counting. In a subgroup of subjects with end-stage liver disease, BIA was compared with total-body water measured by deuterium dilution (n = 21). High levels of correlation were found in healthy subjects, cystic fibrosis patients, and patients with end-stage liver disease when impedance (height2/resistance) was used to predict TBK (r = 0.96, 0.96, 0.98, respectively), and SK was used to predict FFMs (r = 0.96, 0.99, 0.90, respectively), by linear regression analysis. However, less satisfactory relationships were found when the methods were more appropriately analyzed with an estimated limits of agreement procedure (1 SD = 8.5, 5.0, 27.7% [TBK] and 9.4, 6.7, 23.0% [FFMs], respectively). A poor level of technique agreement (1 SD = 14%) was found when this method was compared with total-body water measured by isotope dilution in patients with liver disease.(ABSTRACT TRUNCATED AT 250 WORDS)