Reliable DXA post-scan regional analysis feasible: study

A recent study suggests the potential of performing dual-energy X-ray absorptiometry (DXA) post-scan regional analysis with very high reliability and low measurement error. However, manual review remains essential.

In this study, DXA measurements were performed with a total body scan in 86 athletes (64.0 percent females) from 12 sports. Two independent testers did a post-scan analysis based on the manufacturer’s guidelines. The researchers then evaluated intra- and inter-tester reliability using the intra-class correlation coefficient (ICC), mean differences, paired samples t-tests, and Bland-Altman plots.

A nearly perfect intra- and inter-tester reliability was noted (ICC: 0.998‒1.000) across all regions and components (fat mass [FM], lean soft tissue [LST], and bone mineral content [BMC]). The mean differences were consistently <1 percent for all regions and components, except for left arm FM (1.4 percent) in inter-tester reliability.

Inconsistencies were common between automatically defined regions of interest (ROIs) and manufacturer guidelines, with chest line corrections (separating the arm from the trunk) required in 91.9 percent of scans. Inter-tester ICC for regional changes (n=52) ranged from 0.668 to 0.981, with wider limits of agreement than cross-sectional estimates.

“DXA post-scan regional analysis can be performed with very high reliability and low measurement error, but manual review is essential,” the researchers said. “When monitoring changes over time, baseline and follow-up scans should be analysed by the same evaluator.”