Abstract: Objective  To construct a risk prediction model for sarcopenia in maintenance hemodialysis (MHD) patients based on the data from meta-analysis and to verify the predictive efficacy of the model.  Methods   Literature about the risk factors for sarcopenia in MHD patients published by CNKI, China Biology Medicine disc, VIP, Wanfang, Pub Med, Embase, Web of Science were retrieved . The search period was from January 2010 to April 2022. The related articles were manually retrieved and rescreened according to the inclusion and exclusion criteria, and meta-analysis was then performed. A risk prediction model using the data from meta-analysis for the presence of sarcopenia in MHD patients was constructed. A total of 134 MHD patients admitted to Nantong Haimen People's Hospital from March 2022 to March 2023 were retrospectively recruited as a validation set. Receiver operating characteristic (ROC) curve and Hosmer-Lemeshow test were used to evaluate the efficacy of the model to predict sarcopenia in the validation set.  Results  A total of 20 studies were found from the databases. Meta-analysis showed that age ≥60 years old (RR=1.781, 95% CI:1.661～4.623,P＜0.001), higher C-reactive protein (RR=1.663, 95% CI:1.234～4.923, P＜0.001), low physical activity (RR=1.488, 95% CI:1.214～3.853,P＜0.001), history of diabetes (RR=1.559, 95% CI:1.222～2.732, P＜0.001), longer dialysis vintage (RR=1.101,95% CI:1.041～1.161,P＜0.001), male (RR=1.354, 95% CI: 1.142～2.958, P＜0.001), and higher modified quantitative subjective global assessment (MQSGA) score (RR=1.523, 95% CI:1.323～3.266, P＜0.001) were the risk factors, while higher body mass index (BMI) (RR=0.553, 95% CI:0.122～0.986, P＜0.001) and high plasma phosphorus (RR=0.586, 95% CI:0.443～0.777, P＜0.001) were the protective factors for sarcopenia in MHD patients. Based on the results of the meta-analysis, a risk prediction model for sarcopenia in MHD patients was constructed. ROC curve analysis showed that the area under the curve (AUC) of the model for prediction of sarcopenia in the validation set was 0.849 (95% CI: 0.768～0.930), the optimal truncation value was 35 scores, the sensitivity was 70.0%, and the specificity was 93.7%. Hosmer-Lemeshow test showed χ2=8.97 and P=0.201.  Conclusion  The risk prediction model we constructed has better prediction and discrimination abilities for the presence of sarcopenia in MHD patients. This prediction model may be clinically useful.

Key words: Maintenance hemodialysis, Sarcopenia, Risk prediction model