The deleterious effects of low-dose corticosteroids on bone density in patients with polymyalgia rheumatica

The beneficial effects of corticosteroid therapy in the treatment ofrheumatic diseases may be offset by the occurrence of corticosteroid-related osteoporosis. This problem may be overcome by using low-dosecorticosteroids; however, the dose of corticosteroids that is bothefficacious and skeletal sparing is uncertain. Therefore, the aim of thisstudy was to determine whether low-dose prednisolone treatment results inbone loss and modifies bone turnover. Nineteen patients (12 female, sevenmale) suffering from polymyalgia rheumatica received 10 mg or less daily,given in reducing dosage, with a range of 2.5-10 mg and an average of6.0+/-0.2 mg daily (+/-S.E.M.). Prior to the commencement of therapy and atregular intervals during treatment, bone mineral density (BMD) using dualX-ray absorptiometry and circulating biochemical and hormonal determinantsof bone turnover were measured. The patients were followed for 14.4+/-1.6months (range 6-27). They were compared to 19 age-matched controls. Despitea mean exposure dose of 6 mg/day and disease remission, BMD decreased inthe patients at the lumbar spine (2.6+/-0.8%, P < 0.01), femoral neck(2.9+/-1.5%, P=0.06), Ward's triangle (5.5+/-2.9%, P=0.06) and thetrochanter (4.3+/-1.9%, P < 0.05). Total body bone mass decreased by50+/-19 g in the first 6 months (P < 0.02), and by 39+/-30 g in theremaining 8 months of follow- up [not significant (NS)]. In the first 6months, BMD decreased at the lumbar spine (1.7+/-0.9%, P = 0.06). From 6months to the end of follow- up, BMD decreased by 8.5+/-3.5% at Ward'striangle (P < 0.05) and by 4.8+/-2.5% at the femoral neck (P=0.08). Thefall in BMD correlated with the cumulative prednisolone dose attrabecular-rich regions (trunk r=-0.72, P < 0.001; ribs r=-0.53, P <0.05). Bone resorption, assessed by urinary cross-laps, was 54.7% higherthan controls before treatment was started (P < 0.05) and decreased by23.5+/-7.1% in the first month of treatment when the mean prednisolone dosewas 9.1 mg/day, range 5-10 (P < 0.0001). Serum osteocalcin was notsuppressed by disease before treatment, decreased by 27.4+/-5.1% during thefirst month of treatment (P < 0.001), remained suppressed while thedaily dose of prednisolone was > 5 mg/day, but returned to baselinebelow this dose. Serum parathyroid hormone was 19.3% lower in the patientsthan controls at baseline (NS), and increased by 46.1% (P < 0.05) butwas no higher than controls at any time. Muscle strength increased by20-60% (P < 0.05 to < 0.01). Prophylaxis should be considered inpatients receiving > or = 5 mg/day prednisolone daily as bone loss is 2-to 3-fold expected rates. Earlier trabecular bone loss may predispose tospine and rib fracture; later cortical bone loss may predispose to hipfractures. Doses of prednisolone of < 5 mg daily may be skeletalsparing, but may not be efficacious.