Risk factors for incident vertebral fractures in osteoporosis pharmacotherapy: a 2-year, prospective, observational study.

Abstract

To identify predictors for incident fractures in patients on pharmaceutical treatment for osteoporosis by a secondary analysis of the Japanese Osteoporosis Intervention Trial protocol number 4 (JOINT-04), which was a 2-year, randomized, parallel-group, controlled trial of minodronate and raloxifene in women with primary osteoporosis. This was a prospective, observational study using JOINT-04 data, in which biomarkers, such as undercarboxylated osteocalcin (ucOC), N-telopeptide of type 1 collagen, tartrate-resistant acid phosphatase 5b (TRACP-5b), bone alkaline phosphatase, homocysteine, and pentosidine in blood, and physical functions, such as the timed up and go test and one-leg standing test with eyes open (OLST), and the fall risk index, were measured. The relationships of incident morphometric vertebral fractures during the treatment period, as well as prevalent vertebral fractures, and baseline data were analyzed. The full analysis set of the JOINT-04 included 3247 patients (1623 in the minodronate group and 1624 in the raloxifene group). The hazard ratio (95% confidence interval) for incident vertebral fractures over 2years of pharmacotherapy, adjusted for confounders, was 0.93 (0.90-0.96) for ucOC, 1.15 (1.08-1.23) for TRACP-5b, 1.02 (1.01-1.03) for pentosidine, 0.91 (0.88-0.94) for the OLST, and 1.27 (1.01-1.60) for the fall risk index, which were all independent predictors. Evaluating fracture risk for patients with osteoporosis considering these potential risk factors for fracture in addition to the established risk factors may be useful when starting pharmaceutical treatment.