%0 Journal Article
%T Development of a prediction model for frailty among older Chinese individuals with type 2 diabetes residing in the community.
%A Du J
%A Zhang D
%A Chen Y
%A Zhang W
%J Public Health Nurs
%V 0
%N 0
%D 2024 Aug 5
%M 39101656
%F 1.77
%R 10.1111/phn.13377
%X METHODS: The study employed a retrospective survey of 458 older individuals with T2D residing in a Chinese community, conducted between June 2020 and May 2021, to develop a predictive model for frailty. Among the participants, 83 individuals (18.1%) were diagnosed with frailty using modified frailty phenotypic criteria. The predictors of frailty in this community-dwelling older population with T2D were determined using least absolute shrinkage and selection operator (LASSO) regression and multivariable logistic regression. These predictors were utilized to construct a nomogram. The discrimination, calibration, and medical usefulness of the prediction model were assessed through the C-index, calibration plot, and decision curve analysis (DCA), respectively. Additionally, internal validation of the prediction model was conducted using bootstrapping validation.
RESULTS: The developed nomogram for frailty prediction predominantly incorporated age, smoking status, regular exercise, depression, albumin (ALB) levels, sleep condition, HbA1c, and polypharmacy as significant predictors. Our prediction model demonstrated excellent discrimination and calibration, as evidenced by a C-index of 0.768 (95% CI, 0.714-0.822) and strong calibration. Internal validation yielded a C-index of 0.732, further confirming the reliability of the model. DCA indicated the utility of the nomogram in identifying frailty among the studied population.
CONCLUSIONS: The development of a predictive model enables a valuable estimation of frailty among community-dwelling older individuals with type 2 diabetes. This evidence-based tool provides crucial guidance to community healthcare professionals in implementing timely preventive measures to mitigate the occurrence of frailty in high-risk patients. By identifying established predictors of frailty, interventions and resources can be appropriately targeted, promoting better overall health outcomes and improved quality of life in this vulnerable population.