Risk factors for osteoporotic fractures in persons with spinal cord injuries and disorders.

ABSTRACT

UNLABELLED Clinical risk factors for fracture were explored among Veterans with a spinal cord injury. At the end of 11 years of follow-up, the absolute risk of fracture was approximately 20 %. Among the clinical and SCI-related factors explored, a prior history of fracture was strongly associated with incident fracture. INTRODUCTION: Few studies to date have comprehensively addressed clinical risk factors for fracture in persons with spinal cord injury (SCI). The purpose of this study was to identify risk factors for incident osteoporotic fractures in persons with a SCI that can be easily determined at the point of care. METHODS: The Veteran's Affairs Spinal Cord Dysfunction Registry, a national database of persons with a SCI, was used to examine clinical and SCI-related risk factors for fracture. Incident fractures were identified in a cohort of persons with chronic SCI, defined as SCI present for at least 2 years. Cox regression models were used to estimate the risk of incident fractures. RESULTS: There were 22,516 persons with chronic SCI included in the cohort with 3365 incident fractures. The mean observational follow-up time for the overall sample was 6.2 years (median 6.0, IQR 2.9-11.0). The mean observational follow-up time for the fracture group was 3.9 years (median 3.3, IQR 1.4-6.1) and 6.7 years (median 6.7, IQR 3.1-11.0) for the nonfracture group. By the end of the study, which included predominantly older Veterans with a SCI observed for a relatively short period of time, the absolute (i.e., cumulative hazard) for incident fractures was 0.17 (95%CI 0.14-0.21). In multivariable analysis, factors associated with an increased risk of fracture included White race, traumatic etiology of SCI, paraplegia, complete extent of SCI, longer duration of SCI, use of anticonvulsants and opioids, prevalent fractures, and higher Charlson Comorbidity Indices. Women aged 50 and older were also at higher risk of sustaining an incident fracture at any time during the 11-year follow-up period. CONCLUSIONS: There are multiple clinical and SCI-related risk factors which can be used to predict fracture in persons with a SCI. Clinicians should be particularly concerned about incident fracture risk in persons with a SCI who have had a previous fracture.