OA37 Regional variations in adiposity is a strong predictor of fragility fractures in a large cohort: a cross sectional study

Abstract

Abstract Background/Aims BMI is a core component of fracture risk estimation tools; however, it does not consider the body composition of patients. Emerging research is highlighting that total body fat percentage (TBF%) may be a superior predictor of fracture. Increases in TBF% is thought to be protective of hip fracture while associated with major fragility fractures. However, we have yet to see whether regional variations (abdominal vs hip) in fat deposition is even still a better predictor of fracture. Methods Patients were referred to a DEXA scanner in the northwest of England from 2004-2019. A multivariate logistic regression analysis reporting odds ratios was conducted looking at the fat percentage at the left and right hip (FPLH% and FPRH%), as well as abdominally (FPA%), as predictors of fracture. Results were adjusted for smoking, gender, BMI, steroid use and left femoral t-score. Statistical significance was set at p < 0.05 with STATA V12 used as the statistical package. Results 31,547 patients (5,390 males and 26,156 females) were included in our analysis. The mean age was 64.9 (95% CI 64.8-65.1), BMI 27.1 (95% CI 27.0-27.2), weight 71.1kg (95% CI 70.9-71.3) and height 162cm (95% CI 162.0 cm-162.1cm). 11,875 fractures were reported with 1,375 (11.6%) hip fractures. Table 1 demonstrates the results of our analysis. Conclusion The FPLH% was associated with both hip and all major fragility fractures, while FPA% and FPRH% was only associated with major fragility fractures. We would have expected to see a protective effect for hip fractures, considering our patients were clinically overweight; however, this was not observed. Hence, we report in the literature for the first time that FPLH% is predictive for both hip and major fragility fractures even in patients who are clinically overweight. Our approach of considering regional variations in body fat seems to be superior to methods currently used. Hence, we make the recommendation to conduct further research to validate our findings, as well as investigating ways to incorporate such methods into future iterations of fracture risk calculators. Disclosure H. Amin: None. A. Khan: None. Z. Sultan: None. M. Bukhari: None.