SAT-528 BMI and Relationship to Prior Fracture in the Super Obese

Background: Super obesity, defined as BMI ≥ 50 kg/m2, comprises approximately 1.4% percent of the adult U.S. population and has been rising significantly over the past two decades. While obesity is considered to be protective against fracture, existing literature suggests that obesity affects fracture risk in a site-specific manner. In particular, higher risk of fracture has been observed at the lower leg/ankle and upper arm/shoulder, while lower risk of fracture has been observed at the hip, wrist, and spine. It is not clear if these site-specific relationships continue indefinitely as BMI reaches the super obese levels. Thus, the purpose of this study was to examine associations of BMI and history of fracture in a large academic center with a high rate of super obesity. Methods: Subjects at least 40 years old with BMI ≥50 were selected using the electronic medical record (EMR). History of prior fracture was identified via ICD codes and reviewed for accuracy. Comorbid conditions were also derived from ICD codes and used to derive the Charlson Comorbidity Index (Quan Modification) as a composite comorbidity index. Subjects with history of bariatric surgery were excluded. Clavicle fractures were excluded in multivariate modeling due to small number. Results: Of 6,181 subjects with super obesity, 145 subjects (2.3%) were identified as having prior history of fracture. These 145 subjects were of similar age, gender, and race-ethnicity, but lower BMI (53.3±4.1 vs. 54.2±5.2, p=0.039) than those without fracture history. Among these 145 subjects, a total of 172 fractures were identified. Lower leg/ankle fractures were most prevalent (30.2%), followed by forearm (15.7%), upper arm/shoulder (13.4%), and foot (11.6%). Of the 161 fractures where information regarding trauma severity was available, 62.7% were from minimal trauma (e.g. fall from standing height), 18.6% were from minor trauma (e.g. fall from multiple steps), and 18.6% were from major trauma (e.g. MVA, assault). In a multivariate model of lower leg and upper arm fractures only (controlling for age, gender, and race), higher BMI showed no association with prior fracture (OR 1.00 per 5 kg/m2, p=0.98, 95% CI 0.78-1.26). Interestingly, subjects with a higher comorbidity index had increased odds of prior fracture (OR 1.23 per 1 point increase, p