Improved fracture prediction by identifying and integrating novel risk factors in fracture risk assessment.

Abstract

Abstract Osteoporosis causes significant morbidity and healthcare costs through fragility fractures, with substantial gaps remaining in fracture risk assessment, particularly in older women. This thesis presents four studies from the SUPERB cohort (3,028 women aged 75-80, Gothenburg 2013-2016) examining: (1) type 2 diabetes (T2D) impact on fracture risk; (2) bone marrow adiposity assessment using HRpQCT; (3) FRAXplus® risk prediction enhancement; and (4) population-based screening effectiveness. Results: T2D women (n=294) showed higher BMD but increased fracture risk (HR 1.26, 95% CI 1.04-1.54) with physical function impairment as a key associated factor. Bone marrow fat fraction independently predicted fractures (HR 1.20-1.24 per SD), mediated through femoral neck BMD. FRAXplus adjustments improved risk stratification with upward probability adjustments for most women, with 11.2% reclassification above intervention thresholds and a net reclassification improvement of 4.82 (95% CI: 1.87-7.77%). Population screening achieved 5-fold (HR 5.00, 95% CI 4.39-5.74) increased treatment rates in 443 osteoporotic women and significantly reduced fracture risk compared to non-osteoporotic controls (n=2,585).

Conclusions: Physical function impairment explains increased T2D fracture risk despite higher BMD. Bone marrow adiposity represents a novel HRpQCT-accessible biomarker. FRAXplus enhances risk prediction accuracy. Population screening effectively increases treatment uptake and reduces fracture risk in real-world settings.