Abstract
BACKGROUND: Data on incidence and determinants of fracture risk among children living in non-Western settings are limited. The predictive value of radial quantitative ultrasound (QUS) for fracture risk in children remains uncertain. METHODS: We conducted secondary analyses of data from 8,348 schoolchildren aged 6-13 years living in Ulaanbaatar, Mongolia, who participated in a population-based randomised controlled trial of vitamin D supplementation with 3-year follow-up. Radial speed-of-sound (SOS) was measured at baseline in a subset of 1,456 children. Multivariable Poisson regression was used to estimate fracture incidence, identify determinants of fracture risk, and test whether radial SOS measurements predicted fracture risk. RESULTS: A total of 614 fractures occurred in 521 children over 24,719.3 person-years of follow-up. Overall fracture incidence was 24.9 per 1,000 person-years (95% CI 22.9 to 26.9). Fracture risk was higher in boys vs. girls (adjusted incidence rate ratio [aIRR] 2.10, 95% CI 1.77 to 2.49, P<0.0001) and in older vs. younger children (aIRR 1.18 per additional year of age, 95% CI 1.13 to 1.24, P<0.0001), peaking at age 14 in both boys (87.4 fractures per 1,000 boys per year, 95% CI 63.3 to 117.8) and girls (24.9 fractures per 1,000 girls per year, 95% CI 12.4 to 44.5, respectively). Fracture risk was also higher in children who smoked tobacco vs. those who did not (aIRR 2.20, 95% CI 1.09 to 4.43, P=0.03), while children whose parents were homeowners vs. non-homeowners had lower fracture risk (aIRR 0.83, 95% CI 0.69 to 0.99, P=0.048). Baseline radial SOS did not associate with fracture risk (IRR 0.99; 95% CI 0.99 to 1.00, P=0.31). CONCLUSIONS: Fracture incidence among schoolchildren in Mongolia is higher than in Western settings, particularly for boys. Male sex, older age and tobacco smoking were risk factors in this population, while parental home ownership (denoting higher socio-economic status) was protective. Baseline radial SOS did not predict fracture risk in this population.