Changes in Frailty among Community-Dwelling Older Chinese Adults and its Predictors: Evidence from a Two-Year Longitudinal Study

Abstract Background: Understanding the characteristics related to frailty transitions will allow for better future health practice and healthcare strategies. We evaluated the changes in frailty among community-dwelling older adults and to examine the predictors of the changes in frailty. Methods: A total of 4050 community residents aged ≥ 60 years were recruited in 2015 with follow-up after 2 years. At baseline, a multiple deficits approach was used to construct the Frailty Index (FI) according to the methodology of FI construction, and sociodemographic characteristics and lifestyles were also collected. The transitions in frailty between baseline and 2-year follow-up were evaluated. Multinomial logistic regressions were used to examine associations between predictors and the changes of frailty, adjusting for all of the covariates. Results: Of all of the 3988 participants at baseline, those with frailty status of robust, prefrail, and frail were 79.5%, 16.4%, and 4.1%, and these changed to 68.2%, 23.0%, and 8.8% after 2 years with 127 deceased and 23 dropped out. Twelve kinds of transitions from baseline of the three frailty statuses to four outcomes at follow-up (including death) significantly differed within each of gender and age group, as well between genders and age groups. Among these, 7.8% of prefrail or frail elders improved, 70.0% retained their frailty status, and 22.2% of robust or prefrail elders worsened in frailty status. In multivariable models, age was significantly associated with changes in frailty except for in the frail group; higher educational level and working predicted a lower risk of robust worsening. Of the lifestyle predictors, no shower facilities at home predicted a higher risk of robust worsening; more frequent physical exercise predicted a lower risk of robust worsening and a higher chance of frailty improvement; more frequent neighbor interaction predicted a lower risk of robust worsening and prefrail worsening; and more frequent social participation predicted a higher chance of prefrail improvement. Conclusions: The status of frailty was reversible among community-dwelling elderly, and sociodemographic and lifestyle factors were related to changes in frailty. These findings help health practitioners to recognize susceptible individuals in a community and provide health promotional planning to target aged populations.