2011 Volume 50 Issue 8 Pages 817-824
Objective The incidence and risk factors for symptomatic vertebral fracture were analyzed in glucocorticoid-treated male patients.
Methods This was an observational cohort study at Shimoshizu National Hospital in Japan. Analyzed were 161 male patients newly treated with high-dose glucocorticoid (≥20 mg/day prednisolone equivalent) (initial age: 53.5±16.9, initial glucocorticoid dose: 38.9±12.9 mg/day (0.66±0.23 mg/kg/day), follow-up time: 70.4±52.5 months) and 33 male patients with no glucocorticoid (initial age: 52.7±13.0, follow-up time: 76.4±62.7 months). The vertebral fracture was determined by x-rays.
Results Symptomatic vertebral fractures occurred more frequently in the high-dose glucocorticoid group (21.1%) than in the no glucocorticoid group (3.0%). Using Cox model, the adjusted hazard ratio (HR) for the high-dose glucocorticoid group was 8.16 (95% confidence interval: 1.09-60.86) relative to the no glucocorticoid control group. In the high-dose glucocorticoid group, Kaplan-Meier analyses demonstrated that the incidence of fractures in the patients with glucocorticoid dose increase was significantly higher in comparison with those with no glucocorticoid dose increase. Cox model demonstrated that the risk was independently higher in every 10-year increment of initial age with HR 1.58 (1.18-2.13), in every 10 mg increment of initial dose of prednisolone with HR 2.03 (1.43-2.88), in every dose increase of glucocorticoid increase with HR 3.63 (2.04-6.46), and with each 1-gram decrease of cumulative dose of glucocorticoid with HR 0.88 (0.84-0.93).
Conclusion In male patients, high-dose glucocorticoid causes a significantly high prevalence of symptomatic vertebral fractures, and the independent risk factors are age, initial glucocorticoid dose, glucocorticoid dose increase, and decrease of cumulative glucocorticoid dose.