Epidemiology of mid-life fracture: Self-Reported and Hospital Episode Data from the UK Biobank

Abstract Fracture risk in mid-life (ages 35–65) is under-recognised despite its implications for long-term skeletal health. Using UK Biobank data, we aimed to characterize fracture epidemiology in this critical age group through two complementary approaches. We conducted two studies: a cross-sectional analysis of self-reported fractures (2006–2010) to estimate annualized incidence risk per 10,000 people, and a longitudinal cohort analysis using linked Hospital Episode Statistics (2001–2022) to calculate incidence rates per 10,000 person years, both stratified by sex, skeletal site, and 10-year age bands. Fracture incidence varied substantially by age and sex. Among women, risk increased from age 35 years and accelerated notably from the mid-40s, peaking at 246 per 10,000 people in the 56–65 age group. This female trajectory, emerging earlier than previously recognized, contrasts with men, whose highest fracture risk occurred in early mid-life (peak: 232 per 10,000 people, ages 35–45). Across both sexes and age bands, the most reported fracture sites were: “other” (including digits and facial bones), wrist, ankle, arm, leg, spine, and hip, in descending order. In HES-linked data, 43,572 fractures were identified. Incidence patterns mirrored those from self-report: higher early mid-life fracture rates in men, followed by a transition to female predominance from around age 45. This large-scale, dual-method analysis offers the first clear evidence that female fracture risk begins to rise from age 35, with a marked acceleration from 45 onward. These sex-specific trajectories in mid-life fracture incidence are not fully captured in current clinical models and indicate that further work is needed to determine whether earlier or tailored approaches to risk assessment could provide cost-effective benefit in reducing the burden of fracture in mid-life.