Corticosteroid effects on proximal femur bone loss

Abstract

Prolonged high‐dose corticosteroid therapy is known to result in an increased risk of osteoporotic fracture. Reductions in bone density have been demonstrated at the distal radius and lumbar spine in patients receiving corticosteroids; however there have been few studies of bone density in the hip (the most important site of osteoporotic fracture) in this context. To examine the effect of corticosteroids on the hip we measured bone mineral density (BMD) by dual‐photon absorptiometry at three sites in the proximal femur as well as the lumbar spine in 32 patients aged 18–77 years who had been treated with corticosteroids (mean daily prednisone dose 12.7 mg) for up to 23 years. BMD was compared with the expected values using age regressions in normal subjects. BMD was significantly reduced in the femoral neck, Ward's triangle, and the trochanteric region (p < 0.001 all sites). In the lumbar spine BMD was also significantly reduced (p < 0.001). We also measured BMD serially in 29 patients receiving corticosteroids. BMD measurements were made in 12 patients who had already been treated with long‐term corticosteroids at the time of first BMD measurement (chronic group) and from the commencement of corticosteroid therapy in 17 patients (acute group). The mean (± SEM) change in BMD (g/cm² per year) in the lumbar spine and femoral neck were 0.006 ± 0.006 and –0.021 ± 0.007, respectively, for the chronic group and –0.02 ± 0.005 and –0.039 ± 0.006 for the acute group. The rate of bone loss from the lumbar spine was significantly greater in the acute than the chronic group (p < 0.01), and a similar trend was seen in the femoral neck (p = 0.07). These data suggest that bone loss with corticosteroids is most rapid soon after starting treatment and occurs to a similar extent from the proximal femur to the lumbar spine. The bone loss at the femoral neck suggests an increased risk of fracture at that site, particularly in older subjects on long‐term corticosteroid therapy.