Cross calibration of the GE-Prodigy and iDXA for the measurement of total and regional body composition in adults

Introduction: Dual energy X-ray absorptiometry (DXA) body composition measurements are widely performed in both clinical and research settings, and enable the rapid and non-invasive estimation of total and regional fat and lean mass tissue. DXA upgrading can occur during longitudinal monitoring or study, therefore cross calibration of old and new absorptiometers is required. We compared soft tissue estimations from the GE Prodigy with the more recent iDXA, and developed translational equations to enable Prodigy values to be converted to iDXA values. Methodology: Eighty three men and women aged 20.1 to 63.3 years and with a BMI range of 17.0 to 34.4 kg.m-2 were recruited to the study. Fifty nine participants (41 women: 18 men) comprised the cross calibration group and 24 (14 women: 10 men) comprised the validation group. Total body Prodigy and iDXA scans were performed on each subject within 24 hours. Predictive equations for total and regional soft tissue parameters were derived from linear regression of the data. Results: Measures of lean and fat tissue were highly correlated (R2=0.95-0.99) but significant differences and variability between machines were identified. Bland Altman analysis revealed significant biases for most measures, particularly for arm, android and gynoid fat mass (12.3 to 22.7%). The derived translational equations reduced biases and differences for most parameters, although limits of agreement exceeded iDXA least significant change. Conclusion: Variability in soft tissue estimates between the Prodigy and iDXA were detected, supporting the need for translational equations in longitudinal monitoring. The derived equations are suitable for group but not individual analysis.