OBJECTIVE To analyze the influencing factors of adverse drug reactions （ADRs） related to Semaglutide injection， construct a risk prediction model， and provide a reference for individualized treatment.METHODS Patients diagnosed with type 2 diabetes and/or obesity or overweight who also had at least one weight-related complication and were treated at our hospital for the first time with Semaglutide injection from April 2024 to December 2025 were followed up for 6 months to analyze the incidence of related ADRs. Using a 7∶3 stratified random sampling method， the enrolled patients were divided into a training set （216 cases） and a validation set （93 cases）. With the occurrence of ADRs as the dependent variable， univariate analysis was performed on the training set； the least absolute shrinkage and selection operator （LASSO） regression analysis was used to screen predictive variables in the training set， and multivariate Logistic regression analysis was employed to identify independent risk factors for the occurrence of semaglutide-related ADRs. A risk prediction model was constructed based on the regression coefficients obtained from the multivariate Logistic regression analysis， and a nomogram was generated. The predictive performance of the model was evaluated in both the training and validation sets using receiver operating characteristic （ROC） curves， decision curve analysis （DCA）， and calibration curves.RESULTS A total of 309 patients were enrolled， of whom 165 （53.40%） experienced semaglutide-related ADRs； the time to onset was 4.20 （2.04， 6.53） days after initiation of treatment， with gastrointestinal discomfort being the most common manifestation. ADRs were predominantly mild to moderate in severity， and 15 patients （4.85%） discontinued treatment due to intolerance of the ADRs. Univariate analysis and LASSO regression analysis based on the training set identified seven predictive variables： gender， duration of diabetes， baseline fasting blood glucose （FBG）， baseline glycated hemoglobin （HbA1c）， concomitant use of basal insulin， concomitant use of acarbose， and concomitant use of antiplatelet drugs. Results of multivariate Logistic regression analysis showed that female gender， baseline FBG≥11.1 mmol/L， baseline HbA1c≥9.0%， and concomitant use of acarbose were independent risk factors for the occurrence of semaglutide-related ADRs （odds ratios of 2.11， 18.11， 0.11， and 0.10， respectively； P ＜0.05）. Evaluation of the constructed risk prediction model showed that the areas under the ROC curves for the training and validation sets were 0.82 and 0.81， respectively； within the corresponding threshold probability ranges （0-0.80 and 0-0.75）， the clinical net benefit of using this model was greater than that of the “full intervention” or “no intervention” strategies. The unadjusted and bias-corrected calibration curves for both sets coincided with the ideal calibration curves （ P values for the Hosmer-Lemeshow test were all greater than 0.05）.CONCLUSIONS Semaglutide-related ADRs occur predominantly in the early stages of treatment， with gastrointestinal discomfort as the main clinical manifestation. Female gender and baseline FBG≥11.1 mmol/L are independent risk factors for ADRs， while baseline HbA1c≥9.0% and concomitant use of acarbose are independent protective factors. The developed risk prediction model demonstrates good discrimination， consistency and clinical utility， and can be used to predict the risk of ADRs associated with Semaglutide injection in clinical practice.