Cortical and trabecular bone microarchitecture as an independent predictor of incident fracture risk in older women and men in the Bone Microarchitecture International Consortium (BoMIC): a prospective study

Abstract

Background: Although areal bone mineral density (aBMD) assessed by dual-energy x-ray absorptiometry (DXA) is the clinical standard for determining fracture risk, most older adults who sustain a fracture have T scores greater than -2·5 and thus do not meet the clinical criteria for osteoporosis. Importantly, bone fragility is due to low BMD and deterioration in bone structure. We assessed whether indices of high-resolution peripheral quantitative CT (HR-pQCT) were associated with fracture risk independently of femoral neck aBMD and the Fracture Risk Assessment Tool (FRAX) score.

Methods: We assessed participants in eight cohorts from the USA (Framingham, Mayo Clinic), France (QUALYOR, STRAMBO, OFELY), Switzerland (GERICO), Canada (CaMos), and Sweden (MrOS). We used Cox proportional hazard ratios (HRs) to estimate the association between HR-pQCT bone indices (per 1 SD of deficit) and incident fracture, adjusting for age, sex, height, weight, and cohort, and then additionally for femoral neck DXA aBMD or FRAX.

Findings: 7254 individuals (66% women and 34% men) were assessed. Mean baseline age was 69 years (SD 9, range 40-96). Over a mean follow-up of 4·63 years (SD 2·41) years, 765 (11%) participants had incident fractures, of whom 633 (86%) had femoral neck T scores greater than -2·5. After adjustment for age, sex, cohort, height, and weight, peripheral skeleton failure load had the greatest association with risk of fracture: tibia HR 2·40 (95% CI 1·98-2·91) and radius 2·13 (1·77-2·56) per 1 SD decrease. HRs for other bone indices ranged from 1·12 (95% CI 1·03-1·23) per 1 SD increase in tibia cortical porosity to 1·58 (1·45-1·72) per 1 SD decrease in radius trabecular volumetric bone density. After further adjustment for femoral neck aBMD or FRAX score, the associations were reduced but remained significant for most bone parameters. A model including cortical volumetric bone density, trabecular number, and trabecular thickness at the distal radius and a model including these indices plus cortical area at the tibia were the best predictors of fracture.

Interpretation: HR-pQCT indices and failure load improved prediction of fracture beyond femoral neck aBMD or FRAX scores alone. Our findings from a large international cohort of men and women support previous reports that deficits in trabecular and cortical bone density and structure independently contribute to fracture risk. These measurements and morphological assessment of the peripheral skeleton might improve identification of people at the highest risk of fracture.

Funding: National Institutes of Health National Institute of Arthritis Musculoskeletal and Skin Diseases.

Figure 1.

Hazard ratios (HR) and 95% confidence intervals (CI) represent associations between femoral neck (FN) aBMD (DXA) and HR-pQCT indices at the tibia (top) and radius (bottom) and incidence of any fracture. HRs are expressed per one standard deviation change in the bone measure in the expected direction of increased fracture risk. Black squares indicate HRs adjusted for age, sex, height, weight and cohort. Red circles indicate HRs additionally adjusted for FN aBMD. Blue inverted triangles indicate HRs adjusted for cohort and FRAX score (for major osteoporotic fracture). Cumulative incidence of (any) fracture was 765/7,254 (11%). Numerical data are available in Supplementary Table 8. * Incident fracture includes all sites except skull, face, sternum, finger, and toe and pathological fracture FN aBMD=femoral neck areal BMD (DXA), TtBMD=total (integral) volumetric bone mineral density, CtBMD=cortical volumetric bone mineral density, CtTMD=cortical tissue mineral density, CtAr=cortical area, CtAr/TtAr=cortical area proportion, CtTh=cortical thickness, CtPo=cortical porosity, TbBMD=trabecular volumetric bone mineral density, TbN=trabecular number, TbTh=trabecular thickness, TbSp=trabecular separation, μFEA_FL=Micro finite element analysis failure load