Analysis Of Influencing Factors And Prediction Of Falls Among Rural Older Adults In China Based On A Nomogram Model

Objective To explore the factors influencing falls among older adults in rural China and to construct a nomogram prediction model. This study aims to provide a scientific basis for identifying high-risk populations and implementing targeted fall prevention interventions. Methods The stratified multi-stage cluster random sampling method was employed, selecting one city each from the northern, central, and southern regions of Anhui Province. From each city, one county was randomly selected, and within these counties, a total of 18 villages were randomly chosen. Potential participants, identified through local household registration records and recruited via village committees, completed face-to-face interviews that incorporated a structured questionnaire, physical measurements, and environmental observations. A total of 1546 older adults were included. Inclusion criteria were: (1) aged ≥60 years; (2) local residents for at least six months;capability for effective communication.. Exclusion criteria included: (1) severe physical or mental conditions preventing participation; (2) being bedridden. A fall was explicitly defined as "an unexpected event where the participant comes to rest on the ground, floor, or lower level," excluding instances due to sudden paralysis, stroke, or violent collision. These participants were randomly divided into a training set (1208 individuals) and a validation set (338 individuals) in an 8:2 ratio. Univariate analysis was conducted using the Mann-Whitney U test and Kruskal-Wallis H test, and variables with a p-value < 0.05in the univariate analyses were included in the initial multivariate binary logistic regression model. Backward stepwise selection was then employed to identify independent predictors.. A nomogram model was subsequently developed based on these factors. Results From the univariate and multivariate analyses of the training set, five variables were identified: age, anxiety, frailty, living style, and frequency of coarse grain consumption. These variables were incorporated into the nomogram model, which exhibited an area under the ROC curve (AUC) of 0.722, indicating good discriminative ability. The calibration curve demonstrated high calibration accuracy. Internal validation of the nomogram model using the validation set yielded an AUC of 0.703, reflecting high discriminative ability, and the Hosmer-Lemeshow test result of P=0.08 indicated no significant deviation between predicted and observed probabilities, suggesting good calibration. Conclusion The constructed nomogram, incorporating age, anxiety, frailty, living style, and coarse grain consumption frequency, serves as a practical tool for predicting fall risk among rural older adults. It provides significant value for identifying high-risk populations and implementing targeted interventions.