Network associations between MoCA domain scores and fried frailty phenotype symptoms among nursing home residents: a mixed graphical model study

OBJECTIVES: This study applied network analysis to clarify the interplay between cognitive domains and frailty symptoms among nursing home residents, and to identify highly connected and bridge nodes that may inform future longitudinal and intervention research. METHOD: We assessed a stratified random sample of 452 Chinese nursing home residents, mean age 80.89 ± 7.27 years, using the Montreal Cognitive Assessment (MoCA) and the Fried Frailty Phenotype (FFP). A mixed graphical model with EBIC-LASSO regularization was estimated. Centrality was indexed by strength and expected influence (EI), and bridge nodes were quantified by bridge strength and two-step bridge expected influence (BEI). Network stability was examined with nonparametric and case-drop bootstrapping. RESULTS: Exhaustion was identified as the most central frailty symptom (strength = 2.25, EI = 2.46), showing strong associations with slowness (edge weight = 0.81) and low physical activity (edge weight = 0.95). Both delayed recall (bridge strength = 1.32) and weight loss (BEI = 2.14) emerged as key bridge symptoms linking frailty and cognition. CONCLUSION: Exhaustion, delayed recall, and weight loss emerged as central and bridge symptoms within the network. These findings provide potential directions for future intervention research and may help inform personalized care strategies for nursing home residents.