Limited bone loss due to corticosteroids; a systematic review of prospective studies in rheumatoid arthritis and other diseases.

Abstract

To clarify the relation between changes in bone density, the treated disease, and dose of corticosteroids prescribed. MEDLINE database (1966-95) and bibliographic searches selected cohorts of patients with rheumatoid arthritis (RA) and non-RA patients, studied by reliable serial bone density measurements. Two randomized controlled trials in early RA found greater lumbar bone loss after corticosteroid treatment (pooled effect size at 6 mo 3.9%; 95% CI: 1.9, 6.0%). The other studies included 66 patients with RA taking mean 7 mg prednisone/day; 371 "untreated" RA patients; and 216 non-RA patients taking mean 20 mg prednisone/day. Lumbar bone mass changed (weighted mean) 0.0% (-0.6, 0.7%) per year in steroid treated RA, -0.6% (-0.9, -0.2%) in untreated RA, and -4.7% (-5.2, -4.3%) in non-RA. Femoral neck changed -3.0% (-4.2, -1.8%), -0.7% (-1.0, -0.3%), and -1.5% (-2.5, -0.4%), respectively. In RA, most bone was lost in the first half year, and in early or uncontrolled disease. In patients with RA bone loss is limited, influenced by the interaction of disease characteristics and low dose corticosteroid therapy. In contrast, non-RA patients taking higher doses of corticosteriods may loss clinically relevant amounts of bone (i.e., > 5%) within one year.