Relationship between frailty status and risk of death in the elderly based on frailty index analysis

Objective: To analyze the relationship between frailty status and the risk of death in the elderly based on the frailty index (FI). Methods: Data from a prospective cohort study conducted between 2005 and 2015 in elderly people of an urban community in Beijing were analyzed. The variables related to health and frailty status based on the 2005 baseline survey and death as outcome variables collected in 2015 were used. A FI model was used to evaluate the correlation between FI and mortality in the elderly people in different age groups was analyzed. Cox regression was applied to evaluate the influence of FI on the risk of death, and Kaplan-Meier curves was used to show the survival rate of different frailty levels in the elderly adults. Results: Of the 1 301 elderly people included in the analysis, 403 died during 2005-2015, with the 10-year mortality rate of 31.0%(403/1 301). The mortality rate of the elderly increased with the increase of FI, but, with the increase of FI value, the rate of mortality increased slowly. The limit value of FI causing death was around 0.70, indicating any new health problem might cause death at this value. Cox regression analysis showed that higher FI was associated with higher risk for death (HR=1.143, 95%CI: 1.034-1.248, P=0.000), and FI was more significantly associated with death than age (HR=1.143 vs. HR=1.048, t=5.827, P=0.000). With the increase of age, the effect of frailty on the risk of death decreased (HR=1.179 to HR=1.120). Kaplan-Meier curves showed that the survival rate of the elderly in all age groups decreased with the increase of frailty (Log-rank=317.812, 354.203, 247.258, all P=0.000). The survival time between different frailty levels in the elderly were significantly different, except for the elderly adults aged ≥80 years with severe frailty level (0.4≤FI<0.5, FI≥0.5, P=0.368). Conclusions: Compared with other evaluation tools of frailty, FI model can better reflect the frailty status of the elderly in communities in Beijing and has a high sensitivity in predicting adverse outcomes such as mortality. In the intervention of frailty in the elderly, focusing on relatively young elderly might be more effective in reducing the adverse outcomes caused by frailty.