Objective:To investigate the influencing factors for sarcopenia in hospitalized patients,to construct a risk prediction model for sarcopenia in elderly hospitalized patients,and to provide a quantitative tool for early screening and intervention of sarcopenia based on the integration of multi-dimensional indicators.Methods:A retrospective analysis was performed for the data of 2105 elderly patients who were hospitalized in The First Affiliated Hospital of Chongqing Medical University from March 2016 to June 2023.The least absolute shrinkage and selection operator analysis was used for the screening of variables,and the logistic regression analysis was used to investigate the influencing factors for sarcopenia.A predictive model was constructed,and internal and external validation was performed.The Shapley additive explanations model was used to analyze feature contributions,and a nomogram model was constructed to visualize and interpret the results.Results:The 1259 patients from March 2016 to December 2020 were randomly divided into a training set with 882 patients and an internal test set with 377 patients at a ratio of 7∶3,and the 846 patients from January 2021 to June 2023 were established as the external validation set.A total of 489 cases of sarcopenia(55.44%)were detected in the training set.The logistic regression analysis based on the training set showed that asthenia,dependence on Activity of Daily Living,malnutrition,and in-creasing age were risk factors for sarcopenia(odds ratio[OR]>1,P<0.05),and male sex,normal body mass index,and overweight were protective factors against sarcopenia(OR<1,P<0.05).The model had an AUC of 0.876(95%CI=0.854-0.899)in the training set,0.883(95%CI=0.849-0.918)in the internal test set,and 0.750(95%CI=0.717-0.783)in the external validation set,suggesting that the model had good performance.The decision curve analysis showed that the nomogram model had a good clinical value.Conclu-sion:The predictive model for sarcopenia has good performance and holds promise for clinical application.