Background: Frailty is a dynamic process with transitions over time.

Objectives: To examine frailty transitions and their relationship to utilization of health services. 

Methods: Frailty status using the Vulnerable Elders Survey (VES-13) was determined for 608 community-dwelling older people interviewed in a 2008 national survey and for 281 re-interviewed in 2014. The effect of frailty on death 6 years later was assessed using Cox proportional hazards analysis. Participants were divided into four groups based on their frailty transition. Demographic, functional and health characteristics were compared between the four groups using the Kruskal-Wallis and paired t-test. The independent association between the four frailty groups and health services utilization was assessed using logistic regression. 

Results: Between 2008 and 2014, 24% of 608 participants were lost to follow-up, 9% were non- frail, 37% were frail, and 30% died. The Cox ratio showed that 86% of the non-frail in 2008 were alive 6 years later vs. 52% of the frail (hazard ratio 3.5, confidence interval 2.2–5.4). Frailty transitions in the 281 participants interviewed at both time points revealed that 19% stayed non-frail, 22% became frail, 22% stayed frail and 37% became more frail. Becoming frail, staying frail or becoming more frail compared to staying non-frail was independently associated with a greater risk for requiring help on a regular basis, having a formal caregiver, and requiring home care. 

Conclusions: Any transition away from the non-frail state increased the use of health care services. Interventions to target early transition to frailty should be encouraged.