Cross-Lagged Panel Analysis between Physical Frailty, Cognitive Function, and Falls by Sex

Abstract

Objectives: This study investigated the reciprocal longitudinal relationships between physical frailty (PF), cognitive function (CF), and falls among community-dwelling older adults, according to sex. The study proposed hypotheses that present PF and CF will affect the occurrence of falls 2 years later.

Design: Secondary data analysis using the first (T1, 2016-2017) and second (T2, 2018-2019) waves of the Korean Frailty and Aging Cohort Study (KFACS).

Setting and participants: A total of 2318 community-dwelling older adults aged 70-84 years in South Korea; mean (SD) age: 75.72 (3.83) years; 47.7% men.

Methods: PF and CF were measured with the modified version of the Fried Frailty Phenotype and the Korean version of the Mini-Mental State Examination, respectively. The number of falls were assessed. Multigroup cross-lagged panel analysis was used.

Results: The results showed that relationships between PF, CF, and falls were maintained over time through an autoregressive effect. PF at T1 had a statistically significant longitudinal relationship with fall experience at T2 [standardized regression coefficient (β) = 0.087, 95% CI 0.045-0.129; P < .001], and fall experience at T1 had a significant longitudinal relationship with PF at T2 (β = 0.041, 95% CI 0.006-0.076; P = .020). There was no statistically significant relationship between CF and fall experience. PF and CF had statistically significant reciprocal longitudinal relationships (all P < .001). Based on sex, there was a statistically significant longitudinal relationship between fall experience at T1 and PF at T2 for men only (β = 0.063, 95% CI 0.012-0.114; P = .015).

Conclusions and implications: Findings highlight that health care providers should plan fall prevention programs through early intervention for PF improvement along with improvement and maintenance of CF. Specifically, even if older men are currently healthy and have a low risk of falls, it is important to prevent future fatal PF through prior interventions, such as risk activities attention and concerns about falls.

Keywords: Accidental falls; aged; cognition; cross-lagged panel analysis; fall; frailty.