Combined use of two frailty tools in predicting mortality in older adults

Abstract

We aimed to verify the combined use of two frailty tools in predicting mortality in older adults. We used the data of 10,276 Japanese older adults (aged ≥ 65 years) who provided valid responses to two frailty assessment tools in a mail survey in Japan's Kyoto‒Kameoka Prospective cohort study. Frailty status was categorized into four groups depending on the validated frailty screening index and Kihon Checklist, respectively: Non-frailty (n = 5960), Physical frailty (n = 223), Comprehensive frailty (n = 2211), and Combination (n = 1882) groups. Mortality data were collected between July 30, 2011, and November 30, 2016. We assessed the relationship between frailty status and all-cause mortality risk using multivariate Cox proportional hazards models. During a median follow-up of 5.3 years, we recorded 1257 deaths. After adjusting for confounders, the Combination group had the highest mortality risk compared with the other groups [Non-frailty: reference; Physical frailty: hazards ratio [HR], 0.99 (95% confidence interval [CI] 0.58 to 1.70); Comprehensive frailty: 1.91 (1.63 to 2.23); Combination: 2.85 (2.44 to 3.22)]. People who are positive for frailty in both instruments have a higher risk of death than those who are positive to one model.