DXA-derived hip shape is associated with hip fracture: a longitudinal study of 38 123 UK Biobank participants

Despite advancements in fracture prediction tools and osteoporosis management, hip fractures remain a significant consequence of bone fragility, carrying a 22% 1-yr mortality rate. Hip geometric measures (GMs) have been associated with fracture risk; however, their strong correlation hinders the identification of independent influences, leaving their relative predictive value unclear. Statistical shape modeling (SSM) provides a more holistic assessment of hip shape compared to using predetermined GMs. This study aimed to evaluate whether SSM-derived hip shape from DXA scans can predict hip fracture, independently of individual GMs. Previously, we applied SSM to left hip DXA images in UK Biobank-a large prospective cohort with linked hospital records-generating 10 orthogonal hip shape modes (HSMs) that explained 86% of shape variance. Additionally, FN width (FNW), femoral head diameter (FHD), and hip axis length (HAL) were derived from these DXAs. In the current analysis, Cox proportional hazard models, adjusted for age, sex, height, weight, BMD, and GMs (FNW, HAL, and FHD), were used to examine the longitudinal associations between each HSM and first incident hospital diagnosed hip fracture. A Bonferroni adjusted p-value threshold (p < .004) was used to account for the 13 exposures. Among the 38 123 participants (mean age 63.7 yr; 52% female; mean follow-up 5 yr), 133 (0.35%) experienced subsequent hip fracture. HSM2, characterized by a narrower FNW, a higher neck shaft angle, and reduced acetabular coverage, showed a strong association with hip fracture risk (HR: 1.32, 95% CI: 1.11-1.58, p: 1.47 × 10-3), which persisted after full adjustment (1.30, 1.09-1.55, 3.27 × 10-3). There was no evidence for an association with other HSMs. These findings suggest that DXA-derived hip shape is associated with hip fracture risk independently of BMD and GMs. Incorporating global hip shape into fracture risk assessment tools could enhance prediction accuracy and inform targeted interventions.