Accuracy of DXA in estimating body composition changes in elite athletes using a four compartment model as the reference method

Abstract

Background: Dual-energy x-ray absorptiometry (DXA) provides an affordable and practical assessment of multiple whole body and regional body composition. However, little information is available on the assessment of changes in body composition in top-level athletes using DXA. The present study aimed to assess the accuracy of DXA in tracking body composition changes (relative fat mass [%FM], absolute fat mass [FM], and fat-free mass [FFM]) of elite male judo athletes from a period of weight stability to prior to a competition, compared to a four compartment model (4C model), as the criterion method. Methods. A total of 27 elite male judo athletes (age, 22.2 ± 2.8 yrs) athletes were evaluated. Measures of body volume by air displacement plethysmography, bone mineral content assessed by DXA, and total-body water assessed by deuterium dilution were used in a 4C model. Statistical analyses included examination of the coefficient of determinant (r2), standard error of estimation (SEE), slope, intercept, and agreement between models. Results. At a group level analysis, changes in %FM, FM, and FFM estimates by DXA were not significantly different from those by the 4C model. Though the regression between DXA and the 4C model did not differ from the line of identity DXA %FM, FM, and FFM changes only explained 29%, 36%, and 38% of the 4C reference values, respectively. Individual results showed that the 95% limits of agreement were -3.7 to 5.3 for %FM, -2.6 to 3.7 for FM, and -3.7 to 2.7 for FFM. The relation between the difference and the mean of the methods indicated a significant trend for %FM and FM changes with DXA overestimating at the lower ends and underestimating at the upper ends of FM changes. Conclusions. Our data indicate that both at group and individual levels DXA did not present an expected accuracy in tracking changes in adiposity in elite male judo athletes. © 2010 Santos et al; licensee BioMed Central Ltd.