Lack of consistent information concerning the pathophysiology of corticosteroid-related bone loss may be due to coexisting independent factors that influence bone mineral density (BMD). For example, the disease being treated may increase bone turnover and cause bone loss, and its severity may influence the dose of corticosteroids chosen. Similarly, disease remission due to the treatment or disease progression despite treatment may influence bone turnover and the rate of bone loss. The hormonal changes purportedly responsible for reduced bone formation or increased bone resorption may be the result of the disease, not the corticosteroids.

To determine the pathophysiology of corticosteroid-related bone loss, we conducted a controlled, prospective study in men with no systemic illness treated with corticosteroids to reduce antisperm antibodies. We measured BMD using dual x-ray absorptiometry and circulating biochemical and hormonal determinants of bone turnover in 9 men before and during prednisolone treatment and in 10 age-matched controls. The results were expressed as the mean ± sem.

There were no differences in BMD between the two groups at baseline. The patients received 50 mg prednisolone daily for 3.7 ± 0.6 months (range, 1–6). BMD decreased by 4.6 ± 0.8% at the lumbar spine (P = 0.0007), by 2.6 ± 0.6% at the trochanter (P = 0.004), and by 4.8 ± 1.9% at the Ward’s triangle (P < 0.04). The decrease in lumbar spine BMD correlated with the cumulative dose of corticosteroids (r = −0.49; P = 0.03). Serum osteocalcin and skeletal alkaline phosphatase decreased by 28.5 ± 15.5% (P = 0.08) and 24.2 ± 8.6% (P < 0.03), respectively. The decrease in lumbar spine BMD correlated with the decrease in osteocalcin (r = −0.48; P < 0.02). Serum testosterone and sex hormone-binding globulin decreased by 28.6 ± 4.4% (P < 0.003) and 28.5 ± 8.3% (P < 0.007), respectively. The testosterone/sex hormone-binding globulin ratio did not change. The decrease in total testosterone correlated with the decrease in osteocalcin (r =− 0.40; P = 0.05). There were no detectable changes in urinary C-telopeptide, serum PTH, or serum calcium. Estradiol decreased by 23.5 ± 11.4% (P < 0.003).

Corticosteroid therapy results in rapid bone loss, probably due to reduced bone formation. Neither increased bone resorption nor secondary hyperparathyroidism appears to contribute to the rapid bone loss. Whether the reduction in bone formation may be partly mediated by changes in sex steroids remains unclear.