We performed dual-energy x-ray absorptiometry (craniocaudal and lateromedial views) on 10 pairs of humeruses, radiuses, femurs, and tibias from dogs, using an alignment jig, to determine the homotypic bone mineral density variations of long bones. The bones were divided into 3 regions: proximal, middle, and distal parts of the diaphysis. The bone mineral density of cortical bone, medullary bone, and total bone was determined. Of 160 homotypic comparisons, 21 indicated significant (P less than or equal to 0.05) differences between right and left bones at either a region or location. These differences were observed most frequently in the craniocaudal view and were probably secondary to positioning errors. Evaluation of elliptical bones, such as the radius, also indicated that, when the thicker dimension, such as the lateromedial view of the radius was measured, the bone mineral density of regions-of-interest in that view was greater than that in the opposite view (ie, craniocaudal view of the radius). This study validates the concept of using the contralateral limb as the control condition in orthopedic studies of dogs, particularly when evaluating the densitometric properties of long bones. This study also emphasizes the importance of accurate positioning to prevent inadvertent alteration of bone mineral density results when using dual-energy x-ray absorptiometry. Dual-energy x-ray absorptiometry is particularly susceptible to positioning errors, because it converts a 3-dimensional structure into a 2-dimensional image.

As more sophisticated research is performed to refine fracture fixation techniques for horses, it is important that normal values for the geometric properties of the bones of the appendicular skeleton be determined and that suitable controls be available. We evaluated the geometric properties of total bone width, cortical bone width, and medullary canal/trabecular bone width measured from 2 radiographic projections of equine long bones (humerus, radius, third metacarpal bone, femur, tibia, and third metatarsal bone) obtained from a general population of horses. Measurements were performed on slices separated by intervals equal to 5% of the bone's length. Slices were then grouped into 5 regions: proximal epiphysis, proximal part of the metaphysis, diaphysis, distal part of the metaphysis, and distal epiphysis. Results validated use of the contralateral bone as a control for assessing experimental models or clinical cases. Of 858 homotypic slice comparisons between left and right bones, significant (P less than or equal to 0.05) differences were detected in 31 (3.6%) of the comparisons. Of 168 homotypic region comparisons, significant differences were observed in 3 (1.8%) of the comparisons. The greatest variation between left and right bones was observed in metaphyseal regions, areas with bony protuberances, and regions with prominent bone superimposition. At a power of 0.8 for the statistical tests performed in this study, the mean homotypic variation of bones in each region is < 5.8% for the proximal epiphysis, 11.3% for the proximal part of the metaphysis, 6.8% for the diaphysis, 12.2% for the distal part of the metaphysis, and 5.2% for the distal epiphysis.