Comparison of trabecular bone score and hip structural analysis with FRAX^® in postmenopausal women with type 2 diabetes mellitus

Purpose: To evaluate (a) the performance in predicting the presence of bone fractures of trabecular bone score (TBS) and hip structural analysis (HSA) in type 2 diabetic postmenopausal women compared to a control group and (b) the fracture prediction ability of TBS versus Fracture Risk Calculator (FRAX^®) as well as whether TBS can improve the fracture prediction ability of FRAX^® in diabetic women. Methods: Eighty diabetic postmenopausal women were matched with 88 controls without major diseases for age and body mass index. The individual 10-year fracture risk was assessed by FRAX^® tool for Europe–Italy; bone mineral density (BMD) at lumbar spine, femoral neck and total hip was evaluated through dual-energy X-ray absorptiometry; TBS measurements were taken using the same region of interest as the BMD measurements; HSA was performed at proximal femur with the HSA software. Results: Regarding variables of interest, the only significant difference between diabetic and control groups was observed for the value of TBS (median value: 1.215; IQR 1.138–1.285 in controls vs. 1.173; IQR 1.082–1.217 in diabetic; p = 0.002). The prevalence of fractures in diabetic women was almost tripled than in controls (13.8 vs. 3.4 %; p = 0.02). The receiver operator characteristic curve analysis showed that TBS alone (AUC = 0.71) had no significantly lower discriminative power for fracture prediction in diabetic women than FRAX major adjusted for TBS (AUC = 0.74; p = 0.65). Conclusion: In diabetic postmenopausal women TBS is an excellent tool in identifying fragility fractures.