Beside areal bone mineral density (aBMD), evaluation of fragility fracture risk mostly relies on global microarchitecture. However, microarchitecture is not a uniform network. Therefore, this study aimed to compare local structural weakness to global microarchitecture on whole vertebral bodies and to evaluate how local and global microarchitecture was associated with bone biomechanics.
From 21 human L3 vertebrae, aBMD was measured using absorptiometry. Parameters of global microarchitecture were measured using HR-pQCT: trabecular bone volume fraction (Tb.BV/TV _global ), trabecular number, structure model index and connectivity density (Conn.D). Local minimal values of aBMD and Tb.BV/TV were identified in the total (Tt) or trabecular (Tb) area of each vertebral body. "Two dimensional (2D) local structural weakness" was defined as Tt.BMD _min , Tt.BV/TV _min and Tb.BV/TV _min . Mechanical testing was performed in 3 phases: 1/ initial compression until mild vertebral fracture, 2/ unloaded relaxation, and 3/ second compression until failure.
Initial and post-fracture mechanics were significantly correlated with bone mass, global and local microarchitecture. Tt.BMD _min , Tt.BV/TV _min , Tb.BV/TV _min , and initial and post-fracture mechanics remained significantly correlated after adjustment for aBMD or Tb.BV/TV _global (p < 0.001 to 0.038). The combination of the most relevant parameter of bone mass, global and local microarchitecture associated with failure load and stiffness demonstrated that global microarchitecture explained initial and post-fracture stiffness, while local structural weakness explained initial and post-fracture failure load (p < 0.001).
Local and global microarchitecture was associated with different features of vertebral bone biomechanics, with global microarchitecture controlling stiffness and 2D local structural weakness controlling strength. Therefore, determining both localized low density and impaired global microarchitecture could have major impact on vertebral fracture risk prediction.