Frailty Risk Patterns and Mortality Prediction in Community-Dwelling Older Adults: A 3-Year Longitudinal Study

Objectives: Frailty is a heterogeneous syndrome with distinct patterns. This study aimed to identify frailty risk patterns and their predictive value for mortality in older adults. Design: Prospective longitudinal study. Setting and Participants: Data were obtained from a 2017 survey of 609 independently mobile adults aged 65 years and older in suburban Japan, focusing on those at risk for at least 1 frailty dimension. Methods: Frailty assessments were extracted from the Kihon checklist, and subgroups were identified using latent class analysis. Associations between frailty patterns and 3-year mortality were assessed using Kaplan-Meier survival analysis and Cox proportional hazards modeling. Results: Three frailty patterns were identified: "high risk of cognitive impairment" (76.0%), "moderate risk of cognitive, physical, and oral dysfunction" (14.3%), and "high risk of cognitive, physical, and functional decline" (9.7%). We recorded 52 deaths during a mean follow-up time of 25.7 months (standard deviation: 12.6) and a median follow-up time of 26.5 months. Kaplan-Meier analysis showed significant survival differences among the groups (log-rank: P < .001). Compared with the high risk of cognitive impairment group, the moderate risk of cognitive, physical, and oral dysfunction group had a 145% higher mortality risk (adjusted hazard ratio, 2.45; 95% confidence interval, 1.22-4.90), while the high risk of cognitive, physical, and functional decline group exhibited a 220% higher risk of mortality (adjusted hazard ratio, 3.20; 95% confidence interval, 1.53-6.70). Conclusions and Implications: The findings reveal the heterogeneity of frailty among community-dwelling Japanese older adults, with a high prevalence of cognitive impairment risk. The subgroup with risk of cognitive, physical, and functional decline had the highest mortality risk, highlighting the need for multidimensional assessment and intervention. (c) 2024 Post-Acute and Long-Term Care Medical Association.