Preferential osteopenia in women with osteoporotic fractures

One-hundred thirty-six women with spontaneous vertebral crush fractures had bone density measurements by photon absorptiometry. Their absolute and relative extent of osteopenia were expressed as Z scores with respect to values from normal white women. Absolute Z scores were used to express the mean bone mass of women with crush fractures vis-a-vis normal women at peak bone mass; relative Z scores were used to express the same means vis-a-vis age-matched controls. The skeletal sites measured included the distal radius, lumbar spine, and three sites on the femur (neck, Ward's triangle, and trochanter). The absolute deficit for the five sites ranged from -1.66 to -2.60. The relative deficit was close to 1 SD for all sites. A discriminant analysis revealed that the spine and femur were predictive for vertebral crush fractures but the radius was not. Comparison with a previously reported hip-fracture population (n = 31) showed that the low bone density of the spine separated the two fracture groups best. We propose the following: (a) vertebral-fracture patients have extensive generalized osteoporosis including the hip with a preferential deficit in the spine and (b) hip-fracture patients have generalized osteoporosis with preferential osteopenia of the hip and do not have a greater mineral deficit in the spine or radius than age-matched normal women. It may be that if a single site is to be selected for screening for spine and hip fractures, the greatest value would be in measurement of the femur, although ideally the skeletal site of interest would be measured. Prospective studies are needed to confirm the utility of measurements at the spine and hip in prediction of future fractures.