Abstract: Objective: To identify distinct frailty classes among middle-aged and older adults with atrial fibrillation and to analyze their associated factors, in order to inform the development of personalized care strategies. Methods: A total of 431 middle-aged and older adults with atrial fibrillation were recruited through convenience sampling in a tertiary grade A hospital in Nanning from February 2024 to February 2025. Frailty was assessed using the Fatigue, Resistance, Ambulation, Illness, Loss of weight scale. Data on cognitive function, psychological status, and nutritional status were collected. Latent class analysis was employed to identify frailty subgroups, and multinomial Logistic regression was used to analyze the associated factors. Results: Latent class analysis identified three distinct frailty classes, "Fatigue-Function-Limited" (39.4%), "High Reserve-Low Risk" (53.4%), and "Comorbid-Cachectic" (7.2%). Multinomial Logistic regression analysis revealed that mild cognitive impairment, hypertension, malnutrition, depressive symptoms, persistent atrial fibrillation, and anxiety symptoms were identified as independent associated factors for membership in the latent frailty classes. Conclusion: Frailty in middle-aged and older adults with atrial fibrillation exhibits significant heterogeneity. Healthcare professionals should formulate targeted and personalized stratified nursing interventions based on the unique characteristics and associated factors of each class to prevent or mitigate the progression of frailty.

Key words: atrial fibrillation; frailty; latent class analysis; personalized care