AUTHOR=Guo Xiuyan , Shi Chunhong 

TITLE=Risk prediction model of physical frailty for a rural older population: a cross-sectional study in Hunan Province, China

JOURNAL=Frontiers in Public Health

VOLUME=Volume 13 - 2025

YEAR=2025

URL=https://www.frontiersin.org/journals/public-health/articles/10.3389/fpubh.2025.1525580

DOI=10.3389/fpubh.2025.1525580

ISSN=2296-2565

ABSTRACT=IntroductionPhysical frailty is a common medical syndrome characterized by low muscle strength, low endurance, and reduced physiological function that leads to significantly negative health outcomes in older adults. This study investigated the risk variables among rural older adults in Hunan Province, China, and developed a physical frailty prediction model to inform policymaking to enhance their health and well-being.MethodsThis study was conducted from July 22 to September 3, 2022. A total of 291 participants were recruited using stratified cluster random sampling from five large villages in Hunan Province. Frailty screening was performed based on the Fatigue, Resistance, Ambulation, Illnesses, and Loss of Weight (FRAIL) scale, Geriatric Depression Scale 15-item version (GDS-15), Falls Efficacy Scale-International (FES-I), and Mini Nutrition Assessment-Short Form (MNA-SF). A logistic regression analysis was performed to identify the predictive factors for physical frailty and develop a physical frailty prediction model based on the area under the receiver operating characteristic (ROC) curve (AUC), sensitivity, specificity, and Youden index.ResultsThe physical frailty prevalence among rural older adults in Hunan Province was 21.31% (n = 62). Household income and expenditure [odds ratio (OR): 1.826, 95% confidence interval (CI): 1.142–2.918], physical exercise frequency (OR: 1.669, 95% CI: 1.137–2.451), depressive symptoms (OR: 9.069, 95% CI: 3.497–23.516), and fear of falling (OR: 3.135, 95% CI: 1.689–5.818) were identified as significant predictors of physical frailty in rural older individuals. The AUC for the frailty predictive model was 0.860 (95% CI: 0.805, 0.914). The sensitivity and specificity at the optimal cutoff value were 80.6 and 76.0%, respectively, with a Youden index of 0.566.ConclusionThe prediction model constructed in this study demonstrated promise as a potential tool for evaluating physical frailty risk in older adults, which can contribute to healthcare providers’ screenings for high-risk populations. Further multidimensional and experimental intervention studies should be conducted to prevent the occurrence and delay the progression of physical frailty in older adults.