Midlife risk factors predict long-term hip fracture risk in women: a 35-yr follow-up

Abstract Hip fractures in the aging population are a substantial burden for health care. This study investigates whether risk factors presented in midlife could predict long-term hip fracture risk in the coming decades. The study cohort consisted of 11 169 Finnish women who completed a health survey in 1994, establishing the baseline risk factors. These included all FRAX variables except femoral neck bone mineral density (BMD). The national registry data was used to identify hip fractures and deaths. The follow-up baseline was set at the age of 55 and was terminated by hip fracture, death, or the end of the follow-up period, whichever occurred first. The follow-up age ranged from 55 to 90 yr, with a mean follow-up time of 29.5 yr (range 0.1-35.0 yr). Long-term risk estimates were performed for each risk factor individually over a cumulative time of one-year intervals, using Cox proportional hazard regression models. In total, 801 hip fractures were observed during the 35 yr. Several midlife risk factors appear to sustain their relevance later in life, predicting long-term hip fracture risk in women. The significant risk factors included (hazard ratio minimum to maximum) previous fracture (HR 1.5-4.4), parental hip fracture (HR 1.3-2.8), regular smoking (HR 1.8-4.1), glucocorticoid use (HR 1.7-5.3), rheumatoid arthritis (HR 2.2-4.8), poor self-reported health (HR 1.0-4.3), and polypharmacy (HR 1.1-2.7). The findings suggest that several risk factors identified in midlife demonstrate persistent associations with the long-term risk of hip fractures. Early detection and prevention of these factors may help to reduce the risk in later life.