Fan-beam densitometry of the growing skeleton: are we measuring what we think we are?

Magnification error in fan-beam densitometers varies with distance from the X-ray source to the bone measured and might obscure bone mineral changes in the growing skeleton. Magnification was examined by scanning aluminum rods of different shapes (square, rectangular, solid round, and hollow round) at four distances above the X-ray source in two orientations, with rods aligned parallel (SI) and perpendicular (ML) to the longitudinal axis of the scanning table. Measured area (cm(2)) decreased linearly with distance above the X-ray source for all rods in the SI orientation (p < 0.005). Measured mineral content (g) decreased linearly with distance but only for SI round rods (p < 0.0001) and for ML hollow round rods (p < 0.005). Area and mineral content decreased 1.6-1.8% per centimeter above the source for round rods. Measured mineral density (g/cm(2)) decreased linearly with distance from the source only for ML hollow round rods (p < 0.005). Variation in area, mineral content, and mineral density measurements was 6.6-6.9%, 6.9-7.5%, and 1.9-2.3%, respectively, for SI round rods. Magnification errors of this magnitude are problematic for clinical studies using fan-beam densitometry. Particularly in pediatric subjects, increases in soft tissue during normal growth could increase a bone's distance from the fan-beam source and result in apparent reductions in area and bone mineral content.