Incident fracture associated with increased risk of mortality even after adjusting for frailty status in elderly Japanese men: the Fujiwara-kyo Osteoporosis Risk in Men (FORMEN) Cohort Study.

Abstract

While fractures reportedly increase the risk of mortality, frailty may complicate this association, generating a false-positive result. We evaluated this association after adjusting for pre-fracture levels of frailty. We examined 1998 community-dwelling ambulatory men aged ≥65years at baseline in the Fujiwara-kyo Osteoporosis Risk in Men Study for frailty status as represented by activities of daily living (ADL), physical performance tests (grip strength, one-foot standing balance with eyes open, timed 10-m walk), and laboratory sera tests. Participants were then followed for 5years for incident clinical fractures and death. Effects of incident fracture on death were determined by Cox proportional hazards model with the first fracture during follow-up as a time-dependent predictor and with frailty status indices as covariates. We identified 111 fractures in 99 men and 138 deaths during the follow-up period (median follow-up, 4.5years). Participants with incident fractures did not have significantly worse frailty statuses, but did show a significantly higher cumulative mortality rate than those without fractures (p=0.0047). Age-adjusted hazard ratio (HR) of death for incident fracture was 3.57 (95% confidence interval: 2.05, 6.24). When adjusted for physical performance, this decreased to 2.77 (1.51, 5.06), but remained significant. The HR showed no significant change when adjusted for laboratory test results (3.96 (2.26, 6.94)). Exclusion of deaths within the first 24months of follow-up did not alter these results. Incident clinical fracture was associated with an elevated risk of death independently of pre-fracture levels of frailty in community-dwelling elderly men.