The intra- and inter-instrument reliability of DXA based on ex vivo soft tissue measurements

Objective: Comparison of ex-vivo soft tissue measurements using the GE/Lunar pencil (DPX-L; GE/Lunar Co., Madison, WI) and fan beam (Prodigy dual-energy X-ray absorptiometers (DXA) GE/Lunar Co.). Research Methods and Procedures: Intra-instrument reliability was assessed by repeatedly scanning soft tissue phantoms for lean tissue (water) and fat tissue (methanol) using one DPX-L and two identical Prodigy DXAs at fast, medium, and slow scan modes. For each machine, 10 scans of each phantom were performed at each scan speed. The number of scans per instrument totaled 60. Data were analyzed using ANOVA to ascertain whether scan speed affected the intra-instrument reliability and to test whether soft tissue measurements differed among instruments. Percentage fat (phantom density) was the outcome variable. Results: Intra-instrument reliability, expressed as coefficient of variation, ranged between 0.7% and 5.2% for the DPX-L and 0.4% and 4.5% for the Prodigy, with the lowest coefficients of variation observed when scanning the fat tissue phantom. Scan speed also affected the intra-instrument reliability (p < 0.01). Furthermore, differences in the measurement of percentage body fat for both the lean and fat tissue phantoms were observed among all three absorptiometers (all p < 0.01). After adjusting for scan speed, differences persisted for all three instruments. Discussion: Intra- and inter-instrument reliability of DXA machines, even those from the same manufacturer, remains unpredictable. Thus, when measuring body composition using DXA, it is important to consider that even in the absence of measurement bias, the use of different DXA machines, particularly when using a variety of speed settings, will increase the residual error around the true value.