Prediction of hip fracture by high-resolution peripheral quantitative computed tomography in older Swedish women

Abstract

The socioeconomic burden of hip fractures, the most severe osteoporotic fracture outcome, is increasing and the current clinical risk assessment lacks sensitivity. This study aimed to develop a method for improved prediction of hip fracture by incorporating measurements of bone microstructure and composition derived from HR-pQCT. In a prospective cohort study of 3028 community-dwelling women aged 75-80, all participants answered questionnaires and underwent baseline examinations of anthropometrics and bone by DXA and HR-pQCT. Medical records, a regional x-ray archive, and registers were used to identify incident fractures and death. Prediction models for hip, major osteoporotic fracture (MOF), and any fracture were developed using Cox proportional hazards regression and machine learning algorithms (neural network, random forest, ensemble, and Extreme Gradient Boosting). In the 2856 (94.3%) women with complete HR-pQCT data at 2 tibia sites (distal and ultra-distal), the median follow-up period was 8.0 yr, and 217 hip, 746 MOF, and 1008 any type of incident fracture occurred. In Cox regression models adjusted for age, BMI, clinical risk factors (CRFs), and FN BMD, the strongest predictors of hip fracture were tibia total volumetric BMD and cortical thickness. The performance of the Cox regression-based prediction models for hip fracture was significantly improved by HR-pQCT (time-dependent AUC; area under receiver operating characteristic curve at 5 yr of follow-up 0.75 [0.64-0.85]), compared to a reference model including CRFs and FN BMD (AUC = 0.71 [0.58-0.81], p < .001) and a Fracture Risk Assessment Tool risk score model (AUC = 0.70 [0.60-0.80], p < .001). The Cox regression model for hip fracture had a significantly higher accuracy than the neural network-based model, the best-performing machine learning algorithm, at clinically relevant sensitivity levels. We conclude that the addition of HR-pQCT parameters improves the prediction of hip fractures in a cohort of older Swedish women.

Keywords: HR-pQCT; bone micro-architecture; fracture prediction; hip fracture; machine learning; osteoporosis.

Figure 1

HR-pQCT parameters and incident fracture risk. The associations between HR-pQCT parameters and incident fracture risk are presented according to fracture category (hip, MOF, and any fracture) and HR-pQCT site (ultra-distal and distal tibia and ultra-distal radius and distal radius). The HRs (95% CI) per 1 SD increase in the HR-pQCT parameter are presented for models with 3 levels of adjustments, Model 1 (top line): adjusted for age and BMI. Model 2 (middle line): additional adjustments for CRFs (previous fracture, parental hip fracture, current smoking, oral glucocorticoids, rheumatoid arthritis, secondary osteoporosis, and alcohol intake). Model 3 (bottom line): additional adjustment for FN BMD. Abbreviations: HR, hazard ratio; Tt.Ar, total area; Ct.Ar, cortical area; Tb.Ar, trabecular area; Tt.BMD, total volumetric BMD; Ct.BMD, cortical volumetric BMD; Tb.BMD, trabecular volumetric BMD; Ct.Th, cortical thickness; Ct.Pm, cortical perimeter; Tb.N, trabecular number; Tb.1/N.SD, inhomogeneity of trabecular network; Tb.Th, trabecular thickness; Tb.Sp, trabecular separation; Ct.Po, cortical porosity; MOF, major osteoporotic fracture; CRFs, clinical risk factors. The associated data are presented in Table S2.