Can the Frailty of Older Adults in China Change? Evidence from a Random-Intercept Latent Transition Profile Analysis

Abstract

Background: A major aspect of caring for older adults in the medical field is addressing their health risks. The term "frailty" is generally used to describe the changes in health risks of older adults. Although there is considerable heterogeneity in the Chinese older adult population who are classified as frail, there remain few relevant studies. Furthermore, there is a lack of research on the frailty status transitions of older Chinese adults at different time points. This research intends to determine the frailty status and category of older adults according to their physical, psychological, social, and cognitive function domains, and on this basis, to investigate changes in their frailty states.

Methods: This article studied 2791 respondents who were over 60 years old (n = 2791; 53.2% were women) from the Chinese Longitudinal Healthy Longevity Survey (CLHLS) follow-up survey on factors affecting the health of older adults in China. In this article, the frailty variables include self-reported health, social function, mental health, cognitive function, functional limitations, and morbidity status. Random-intercept latent transition profile analysis (RI-LTPA) was used to divide older adults into different subgroups, and then an in-depth analysis of the state transitions was conducted.

Results: The latent profile analysis revealed that the evaluation results of the frailty state of older adults showed obvious group heterogeneity. Each fitting index supported four latent states, which were named according to the degree of the symptoms (i.e., multi-frailty, severe socially frailty, mild socially frailty, and relatively healthy frailty). Based on the categorical probability and the probability of transition, it can be concluded that most of the samples belonged to the healthy population, and the health status had generally improved across the four time points. The relatively healthy frailty group and the severe socially frailty group have relatively strong stabilities. The multi-frailty group and the mild socially frailty group had the highest probability of joining to the relatively healthy frailty group. Strengthening social interactions among older adults and promoting their participation in social activities can significantly improve their frail state.

Conclusions: This study supplements related research on frailty. Firstly, it deepens the meaning of frailty, which is defined based on four aspects: physical, psychological, cognitive, and social functioning. Secondly, it divides frailty into different sub-categories. Frailty is discussed from the perspective of longitudinal research, which can provide practical adjustment suggestions for older adult nursing intervention systems and measures in China.

Keywords: cognitive function; frailty; functional limitation; longitudinal study; mental health; morbidity status; older adults; random-intercept latent transition profile analysis (RI-LTPA); self-reported health; social function.

Figure 1

Conditional probability of each item of the frailty state at T1 (2005). Note: HN = self-reported health now; HP = self-reported health compared with past year; SF = social function; MH = mental health; CF = cognitive function; FL = functional limitation; MS = morbidity status. The proportion of four profiles: C1 = 6.84%; C2 = 6.99%; C3 = 13.44%; C4 = 72.73%.

Figure 2

Conditional probability of each item of the frailty state at T2 (2008). Note: HN = self-reported health now; HP = self-reported health compared with past year; SF = social function; MH = mental health; CF = cognitive function; FL = functional limitation; MS = morbidity status. The proportion of four profiles: C1 = 8.89%; C2 = 6.20%; C3 = 11.18%; C4 = 73.73%.

Figure 3

Conditional probability of each item of the frailty state at T3 (2011). Note: HN = self-reported health now; HP = self-reported health compared with past year; SF = social function; MH = mental health; CF = cognitive function; FL = functional limitation; MS = morbidity status. The proportion of four profiles: C1 = 11.93%; C2 = 7.27%; C3 = 13.40%; C4 = 67.40%.

Figure 4

Conditional probability of each item of the frailty state at T4 (2014). Note: HN = self-reported health now; HP = self-reported health compared with past year; SF = social function; MH = mental health; CF = cognitive function; FL = functional limitation; MS = morbidity status. The proportion of four profiles: C1 = 6.63%; C2 = 12.15%; C3 = 18.05%; C4 = 63.17%.