Abstract
BACKGROUND: The relationship between frailty and dementia is unclear and there are very few population-based studies regarding this issue in China. The purpose of this study is to estimate the association between frailty and incident dementia in China, and to explore different effects of frailty established by three definitions of frailty on dementia incidence. METHODS: A five-year prospective cohort study was carried out in 2022 participants aged 65 years and over in urban and rural sites in Beijing, China. The participants were interviewed by trained community primary health care workers from 2004 to 2009. Frailty was defined using modified Fried frailty phenotype, physical frailty definition, and multidimensional frailty definition. Dementia was diagnosed using the 10/66 dementia criterion for calculating cumulative incidence. Both competing risk regression models and Cox proportional hazards models were applied to examine the associations between frailty at baseline and five-year cumulative incidence of dementia. RESULTS: At the end of follow-up the five-year cumulative incidence rates of dementia with frailty and without frailty defined by the modified Fried frailty were 21.0% and 9.6%, those defined by the physical frailty were 19.9% and 9.0%, and those defined by the multidimensional frailty were 22.8% and 8.9%, respectively. Compared with non-frail participants, frail people had a higher risk of incident dementia using multidimensional frailty definition after adjusting covariates based on competing risk regression model (HR = 1.47, 95% CI 1.01~2.17) and Cox proportional hazards model (HR = 1.56, 95% CI 1.07~2.26). The association between frailty and incident dementia was statistically significant in participants in the upper three quartiles of age (aged 68 years and over) using the multidimensional frailty definition based on the competing risk regression model (HR = 1.61, 95% CI 1.06~2.43) and Cox proportional hazard model (HR = 1.76, 95% CI 1.19~2.61). CONCLUSIONS: Multidimensional frailty may play an inherent role in incident dementia, especially in the people aged over 68, which is significant for distinguishing high risk people and determining secondary prevention strategies for dementia patients.