Aging
Nine-year frailty transitions in older cancer survivors in the National Health and Aging Trends Study (NHATS) Allison Musty* Allison Musty Kathryn Callahan Heidi Klepin Amresh Hanchate Jennifer Gabbard Charles Semelka Emilie D Duchesneau
Background
Older cancer survivors are at risk of frailty. Most research focuses on frailty transitions around cancer diagnosis and treatment; transitions during survivorship are understudied. We described longitudinal frailty transitions in older cancer survivors.
Methods
Using Rounds 1-9 (2011-2019) of NHATS, a nationally representative cohort of adults 65+, we analyzed 1303 older adults with a self-reported cancer history. Frailty was assessed annually using two measures common in research: (1) the frailty phenotype (robust: 0, prefrail: 1-2, and frail: 3-5 of the following symptoms: exhaustion, low physical activity, shrinking, weakness, slowness) and (2) a 40-item deficit accumulation frailty index (FI; robust: <0.15, prefrail: 0.15-0.24, frail: ≥0.25). Multistate Markov models estimated 1-year probabilities of transitioning across frailty states and death, pooling all rounds. Multiple imputation and inverse probability weights accounted for missing data and attrition.
Results
In Round 1, 35% were robust, 47% prefrail, and 18% frail by the frailty phenotype vs. 37%, 25%, and 38% by the FI. Transition probabilities differed by frailty measure (Figure). For both measures, prefrail and frail individuals had a higher one-year probability of death than robust individuals. Using the frailty phenotype (panel A), transitions between adjacent states (e.g., robust to prefrail) were common, with greater probabilities of improvement (e.g., frail to prefrail: 27%) compared to the FI (e.g., frail to prefrail: 9%). The FI (panel B) showed fewer transitions across frailty states and more stability for frail individuals.
Conclusions
Frailty is dynamic in older cancer survivors and transitions vary by frailty measure. Frailer individuals face elevated mortality risks compared to robust individuals. These findings highlight the importance of frailty measurement selection and the need for targeted interventions to address frailty progression and improve outcomes in this population.
