RESUMO
OBJECTIVE: The aims of this study were to validate BOD POD in a wide sample of healthy and independent Mexican elderly men and women subjects using the 4 compartment (4C) model as the reference method, and to evaluate the assumptions of the densitometric two-compartment (2C) model. DESIGN: Cross-sectional study designed to assess body composition and validation of a method based on 2C model (BOD POD). SETTING: Urban and rural regions of Sonora, Mexico. SUBJECTS: Two hundred and two free-living subjects >or=60 years old were completed in this study. METHODS: Body density and body fat were measured by the BOD POD, total body water by deuterium dilution and total body bone ash by dual energy X-ray absorptiometry. Body composition was determined using Baumgartner's equation. RESULTS: Percent body fat by the 4C model was 31.2 and 42.5% in men and women, respectively (P<0.001). Group mean accuracy of body fat by BOD POD against that of the 4C model showed an effect of sex (P<0.001), but not the method (P=0.27). Results of individual accuracy showed no significant difference with the identity line and the slope was significantly different from zero or a slope similar to one. Precision assessed by model R (2) was high for all subjects and for men and women by separate. The standard error of the estimate was low for all and for men and women by separate. Bland and Altman analysis showed no significant bias. CONCLUSION: The BOD POD technique is a valid and reliable method compared to the 4C model and it could be applied in subjects with similar physical and anthropometric characteristics to subjects of this study.
