Estimation of absolute fracture risk among middle-aged and older men and women: the EPIC-Norfolk population cohort study

Abstract

While estimates of relative risks associated with risk factors such as age and bone mineral density (BMD) may be of interest for etiologic and comparative purposes, clinical questions such as who might benefit most from preventive interventions or BMD monitoring depend on estimates of absolute fracture risk. The European prospective investigation into cancer (EPIC)-Norfolk study included 25,311 participants (11,476 men) aged 4,079 years in 1993-1997. All participants were followed for osteoporotic fractures to March 2007. Ten-year absolute risk of fracture in men and women were calculated using the baseline survivor function in multivariable Cox proportional-hazards models adjusting for age, sex, history of fractures, body mass index, smoking, and alcohol intake. In comparison of those without history of fracture versus those with history of fracture, the 10-year absolute risk of any fracture in men ranged from 1.0 vs. 1.2% at age 40 years to 3.0 vs. 4.4% at age 75 years. The respective estimates in women ranged from 0.7 vs. 1.0% at age 40 years to 9.3 vs. 17.2% at age 75 years. Statistically significant interaction between age and sex was found (P < 0.001), which contributed to the differences in predicted absolute fracture risks for men and women at different ages. Our study shows the need for population-specific data to develop efficient well calibrated algorithms for assessment of fracture risk. The interaction observed between sex and age points to the need for further prospective studies among men.