In 161 ambulatory rheumatic disease patients receiving long-term prednisone therapy, diaphyseal mass (DM) and metaphyseal mass (MM) of the forearm were measured by single photon absorptiometry, and bone radiographs were reviewed when available. Multivariate analysis of treatment and patient characteristics demonstrated that glucocorticoid-induced osteopenia (defined as an elevated DM:MM ratio) and bone fractures occurred with similar frequency in patients of each sex, in whites and blacks, in patients with various rheumatic diseases, and in patients receiving different regimens of prednisone therapy. However, large cumulative doses of prednisone were associated with elevated DM:MM ratios as well as with bone fractures, and menopause or age 2250 years (males or females) was associated with bone fractures. We conclude that longterm therapy with various prednisone regimens results in glucocorticoid-induced osteopenia and fractures. This affect is cumulative, occurs in all patient groups, and results in more bone fractures in certain groups.It has long been recognized that long-term administration of oral glucocorticoids can result in a significant loss of bone mass (1-3). The degree of loss is more severe in areas of the skeleton with a high content of trabecular bone, such as vertebrae and ribs. It is less striking in the diaphyses of long bones, which consist primarily of compact cortical bone (4-6).We have previously used single photon absorptiometry to measure appendicular bone mineral mass in normal individuals and rheumatic disease patients (7-10). Measurements at diaphyseal and metaphyseal sites of the radius demonstrate a constant relationship between bone mineral mass at the diaphyseal site and that at the metaphyseal site, both in normal individuals and in most patients with conditions associated with osteopenia (for example, idiopathic osteoporosis). However, in osteopenia associated with glucocorticoid therapy, there is a greater loss of metaphyseal mass (MM) than diaphyseal mass (DM), and the DM:MM ratio increases (7-10). A combination of factors may account for this observation. First, the metaphyseal site contains a higher proportion of trabecular bone (>25%) than the diaphyseal site (-5%) (1 1). Second, the metaphyseal site has a greater surface area than the diaphyseal site, and endosteal resorption of cortical bone may be more easily measured at the metaphyseal site.Since measurement of DM:MM ratios provides a simple, inexpensive, and rapid method of identifying patients with glucocorticoid-induced osteopenia, we have measured these ratios in 161 patients with various rheumatic diseases treated with long-term glucocorticoid therapy. This report evaluates whether different glucocorticoid treatment regimens and patient characteristics are associated with glucocorticoid-induced osteopenia and bone fractures.