Predictive validity of body fat percentage by bioimpedance compared with deuterium oxide dilution in Costa Rican schoolchildren.

OBJECTIVE

The purpose of this study was to verify the validity of body fat percentage (BF%) by bioelectrical impedance analysis (BIA) in a sample of Costa Rican schoolchildren.

METHODS

The sample consisted of 52 male (mean age 8.1 ± 0.9 years) and 49 female (mean age 7.5 ± 1.0 years) schoolchildren. Deuterium oxide dilution was the reference method used as the "gold standard" to determine the fat mass (FM) expressed as BF%. Linear regression analyses and paired sample t-tests were used to test the association and mean differences between BIA and deuterium oxide dilution BF%. Concordance between BIA and deuterium oxide dilution BF% was determined by Lin's concordance correlation coefficient. Measurement of agreement between the two methods was analyzed using the Bland-Altman procedure.

RESULTS

Significant associations between BIA and deuterium oxide dilution BF% were found in males (R = 0.97, R  = 0.95, P < .0001) and in females (R = 0.98, R  = .96, P < .0001). The measurements of BF% (mean ± SD) were not significantly different between males (BIA 31.1 ± 7.6% vs. deuterium oxide dilution 31.0 ± 7.4%) and females (BIA 25.9 ± 7.9% vs. deuterium oxide dilution 26.3 ± 7.8%) by t-test. Lin's concordance correlation coefficient identified an almost perfect strength of agreement between males and females. Bland-Altman plots showed that the Bodystat measured similar BF% compared to the deuterium oxide dilution in both males and females.

CONCLUSIONS

BIA Bodystat presented high agreement with BF% measured by deuterium oxide dilution; therefore BIA Bodystat is recommended for estimating BF% in Costa Rican schoolchildren.