Associations between frailty trajectories and frailty status and adverse outcomes in community-dwelling older adults

Abstract

Background: The association between frailty and adverse outcomes has been clearly defined. Frailty is associated with age, but different frailty evolution patterns might determine the incidence of adverse outcomes at older ages. So far, few observational studies have examined how distinct frailty trajectories could be associated with differences in the risk of adverse events and assessing whether frailty trajectories could define risk of death, hospitalization, worsening, and incident disability better than one-off assessment. Our hypothesis is that prospective increases in frailty levels are associated with higher risk of adverse events compared with subjects that prospectively decreased frailty levels.

Methods: Participants' data were taken from the Toledo Study of Healthy Ageing. Frailty was evaluated using the Frailty Trait Scale 5 (FTS5), being 0 the lower (the most robust) and 50 the highest (the frailest) score. FTS5 scores at baseline and follow-up (median 5.04 years) were used to construct frailty trajectories according to group-based trajectory modelling (GBTM). Multivariate Cox proportional hazard and logistic regression models were used to explore associations between frailty status and trajectory membership and the adverse outcomes. Deaths were ascertained through the Spanish National Death Index. Disability was evaluated through the Katz Index. Hospitalization was defined as first admission to Toledo Hospital.

Results: Nine hundred and seventy-five older adults (mean age 73.14 ± 4.69; 43.38% men) were included. GBTM identified five FTS5 trajectories: worsening from non-frailty (WNF), improving to non-frailty (INF), developing frailty (DF), remaining frail (RF), and increasing frailty (IF). Subjects belonging to trajectories of increasing frailty scores or showing consistently higher frailty levels presented with an increased risk of mortality {DF [hazard ratio (HR), 95% confidence interval (CI)] = 2.01 [1.21-3.32]; RF = 1.92 [1.18-3.12]; IF = 2.67 [1.48-4.81]}, incident [DF (HR, 95% CI) = 2.06 (1.11-3.82); RF = 2.29 (1.30-4.03); IF = 3.55 (1.37-9.24)], and worsening disability [DF (HR, 95% CI) = 2.11 (1.19-3.76); RF = 2.14 (1.26-3.64); IF = 2.21 (1.06-4.62)], compared with subjects prospectively showing decreases in frailty levels or maintaining low FTS5 scores. A secondary result was a significant dose-response relationship between baseline FTS5 score and adverse events.

Conclusions: Belonging to trajectories of prospectively increasing/consistently high frailty scores over time are associated with an increased risk of adverse outcomes compared with maintaining low or reducing frailty scores. Our results support the dynamic nature of frailty and the potential benefit of interventions aimed at reducing its levels on relevant and burdensome adverse outcomes.

Keywords: Disability; Frailty; Mortality; Older adults; Trajectory modelling.