Frailty in older adults is dynamic and shaped by clinical, functional, and social factors, yet longitudinal transitions in community populations are not well characterised. We examined one-year changes in frailty and identified factors associated with improvement and progression.

We performed a longitudinal analysis of community-dwelling adults aged ≥65 years from the population-based Queenstown Health District cohort. Baseline (Wave 1, 2023–2024) demographic, clinical, and lifestyle data were collected, with reassessment at one year (Wave 2). Frailty was measured using the Clinical Frailty Scale. Ordinal logistic regression modelled overall direction of frailty transition, while multinomial logistic regression identified factors associated with improvement or worsening relative to stability. Alluvial plots illustrated transitions across frailty states over time.

Among 1,153 participants (mean age 74.5 ± 6.6 years; 55.8% female), frailty remained stable in 627 (54.4%), improved in 211 (18.3%), and worsened in 315 (27.3%) over one year. Visualisation demonstrated predominant stability with bidirectional movement between frailty categories. In ordinal analyses, diabetes (OR 1.39, 95% CI 1.06–1.83), smoking (OR 1.89, 95% CI 1.19–3.03), and iADL impairment (OR 1.56, 95% CI 1.01–2.41) were associated with worsening, while higher physical activity was protective (OR 0.70, 95% CI 0.51–0.98). Multinomial models showed stroke history associated with improvement (OR 4.23, 95% CI 1.46–12.25), whereas iADL impairment reduced the likelihood of improvement (OR 0.48, 95% CI 0.28–0.82).

Frailty was largely stable but exhibited meaningful bidirectional transitions. Functional impairment and modifiable behaviours were key predictors, underscoring the importance of repeated assessment and function-focused prevention to mitigate frailty progression.