Femoral stress is prominently associated with fracture risk in children: The Generation R Study.

Bone modeling is an important process in the growing skeleton. An inadequate bone modeling in response to mechanical loads would lead some children to develop weaker bones than others. The resulting higher stresses in the bones would render them more susceptible to fracture. We aimed to examine the association between femoral stress (FS) derived from structural parameters and BMD in relation to incident fractures in children. Bone stress was evaluated at the medial femoral neck, a skeletal site subject to large forces during normal locomotion. This study comprises 1840 children from the Generation R Study, with whole body and hip DXA scans at a mean age of 6.01 years. Hip structural analysis (HSA) was used to measure femur geometry for the FS calculation. Data on fractures occurring over the following 4 years after the DXA assessment were obtained by questionnaire. Incident fracture was observed in 7.6% of the participating children. Cox-multivariate regression analysis, described as hazard ratios (HR), showed that after adjustment for sex, ethnicity, age, weight and lean mass fraction, there was a significant increase in the risk of incident fracture for every standard deviation (SD) decrease in total body BMD (HR: 1.35, 95% CI 1.05-1.74, p-value = 0.021), femoral neck BMD (HR: 1.31, 95% CI 1.09-1.58, p-value = 0.005) and narrow neck BMD (HR: 1.39, 95% CI 1.14-1.68, p-value = 0.001). Whereas, every increment of one SD in femoral stress resulted in 1.33 increased risk of incident fractures (HR: 1.33, 95% CI 1.13-1.57, p-value = 0.001). This association remained (borderline) significant after the adjustment for DXA derived BMD measurements. Our results show that increased bone stress may underlie greater susceptibility to traumatic fractures in children (partially independent of BMD) and underscore the utility of hip DXA scans for the assessment of paediatric bone health and specifically fracture risk.