Objective To establish a simple model for arteriosclerosis （AS） screening to provide a viable tool for the timely identification of AS risk among residents aged 40 ~ 65 years. Methods Data were obtained from the Sleep and Chronic Diseases Program in Fuquan City. The original dataset was divided into a training subset and a validation subset （80%：20%）. LASSO and logistic regression models were used to screen variables， perform multivariate regression analyses. Internal validation was performed using the Bootstrap method. Nomogram Plot was constructed， and risk score thresholds were determined based on ROC curves to classify high-risk populations. Results RS Model was established to include age， gender， napping， sleep efficiency， sleep disorders， hypertension and diabetes， with AUC = 74.80% and a model risk score threshold = 84.20. PHC Model was established to include age， gender， napping， sleep efficiency， systolic blood pressure， fasting blood glucose， and pulse variables， with AUC = 82.80% and a risk score threshold of 78.00. Decision curves showed that both models performed well in terms of calibration and actual benefits for health management. Conclusion The two AS screening models exhibit acceptable accuracy and differentiation. Therefore， it can be applied in residents' self-health management and in primary care organizations' screening work in a large scale.