Analysis of risk factors and construction of a prediction model for sarcopenia in Chinese elderly population

WU Xiaojuan; WANG Meiping; ZENG Jingbo

doi:10.3969/j.issn.1002-266X.2025.06.007

Abstract：Objective To analyze the risk factors for sarcopenia among the Chinese elderly population and to devel-op a predictive model,thereby providing a basis for the prevention of sarcopenia.Methods We obtained the 2015 fol-low-up data of the China Health and Retirement Longitudinal Study(CHARLS),and selected individuals aged 60 years and older as the study population.Sarcopenia was diagnosed based on the criteria established by the Asian Working Group for Sarcopenia(AWGS)in 2019.Demographic characteristics(age,sex,marital status,and educational level),physical measurements[blood pressure,height,weight,waist circumference,handgrip strength,5-meter walking speed,five-time chair stand test,and body mass index(BMI)],health-related behaviors(smoking,alcohol consumption,and sleep dura-tion),health status[Center for Epidemiologic Studies Depression Scale(CESD)score and comorbid chronic conditions],and serum biochemical markers[fasting blood glucose,blood urea nitrogen,creatinine(Cr),C-reactive protein,total cho-lesterol,triglycerides,high-density lipoprotein cholesterol,low-density lipoprotein cholesterol,uric acid,cystatin C(Cys C),and glycated hemoglobin]were collected.Least Absolute Shrinkage and Selection Operator(LASSO)regression was used to screen the potential predictors of sarcopenia,Logistic regression analysis was used to determine the independent risk factors and we constructed a predictive model.Model accuracy was assessed using calibration curves,while model per-formance was evaluated through receiver operating characteristic(ROC)curve analysis and decision curve analysis(DCA).Results A total of 2 938 elderly individuals aged 60 years and above were included,among whom 447(15.2％)were diagnosed with sarcopenia.Elevated CESD score and history of falls were identified as risk factors,while higher Cr/Cys C ratio and higher BMI were protective factors.A predictive model was developed based on CESD score,history of falls,Cr/Cys C ratio,and BMI,with the following calibrated equation:-39.43+0.03×(CESD score)+0.50×(fall history)-4.45×(Cr/Cys C)+4.73×(BMI)+1.35×(Cr/Cys C)2-0.13×(BMI)2.The area under the ROC curve(AUC)for the training and validation sets was 0.90 and 0.91,respectively,indicating good predictive perfor-mance.The Hosmer-Lemeshow test yielded P-values>0.05,and the calibration curves demonstrated good model fit and high calibration accuracy.Decision curve analysis indicated the model had good clinical utility.Conclusions Elevated CESD score and history of falls are significant risk factors for sarcopenia in the elderly,whereas higher Cr/Cys C ratio and BMI serve as protective factors.The predictive model based on these four variables shows strong predictive power and can aid in the identification of individuals at high risk for sarcopenia.