A critical evaluation of bioimpedance spectroscopy analysis in estimating body composition during GH treatment : comparison with bromide dilution and dual X-ray absorptiometry

Objective: To compare estimates by Bioimpedance Spectroscopy Analysis (BIS) of extracellular water (ECW), fat mass (FM), and fat free mass (FFM) against standard techniques of bromide dilution and DXA during intervention that causes significant changes in water compartments and body composition. Methods: Body composition analysis using BIS, bromide dilution, and DXA was performed in 71 healthy recreational athletes (43 men, 28 women; aged 18-40 years; BMI 24±0.4 kg/m2) who participated in a double-blinded, randomized, placebo-controlled study of growth hormone (GH) and testosterone treatment. The comparison of BIS with bromide dilution and DXA was analyzed using linear regression and the Bland-Altman method. Results: At baseline, there was a significant correlation between BIS and bromide dilution derived estimates for ECW, and DXA for FM and FFM (p<0.001). ECW by BIS was 3.5±8.1 % lower compared to bromide dilution, while FM was 22.4±26.8 % lower and FFM 13.7±7.5 % higher compared to DXA (p<0.01). During treatment, the change in ECW was similar between BIS and bromide dilution, whereas BIS gave a significantly greater reduction in FM (19.4±44.8 %) and a greater increase in FFM (5.6±3.0 %) compared to DXA (p<0.01). Significant differences in body composition estimates between the BIS and DXA were observed only in men, particularly during the treatment that caused greatest change in water compartments and body composition. Conclusion: In healthy adults, bioimpedance spectroscopy is an acceptable tool for measuring ECW, however BIS overestimates FFM and substantially underestimates FM compared to DXA.