Construction and Validation of an Individualized Efficacy Prediction Model for Transcranial Direct Current Stimulation Treatment of Adolescent Depression Based on Nomogram

TIAN Yuxia; WANG Juan; YANG Xiaoyan; SUN Yan

doi:10.3969/j.issn.1004-437X.2026.14.017

Abstract：Objective To construct and verify an individualized efficacy prediction model of transcranial direct current stimulation(tDCS)for adolescent depression based on nomogram,so as to assist clinicians in early identification of high-risk groups with poor efficacy and provide a basis for optimizing individualized treatment plans.Methods A total of 241 adolescent depression patients admitted to the Second People's Hospital of Shangqiu from January 2023 to April 2025 were selected.After 2 weeks of tDCS treatment,they were divided into good group and poor group according to clinical efficacy.The clinical data of the two groups were compared.The influencing factors of tDCS efficacy in adolescent depression patients were analyzed by LASSO and logistic.The nomogram prediction model was constructed.The receiver operating characteristic(ROC)curve,calibration curve and decision curve were used to evaluate the efficacy of the prediction model.Results After 2 weeks of tDCS treatment,176 cases(73.03％)were good and 65 cases(26.97％)were poor in 241 patients with adolescent depression.The course of disease,Hamilton depression scale(HAMD)score,Montreal cognitive assessment scale(MoCA)score,neutrophils,vitamin D-binding protein(VDBP)and high-sensitivity C-reactive protein(hs-CRP)levels of the two groups were compared,and the differences were statistically significant(P＜0.05).Five factors including course of disease,HAMD score,MoCA score,serum VDBP and serum hs-CRP were selected by LASSO into multivariate analysis.Logistic analysis showed that course of disease,HAMD score,serum VDBP and serum hs-CRP were risk factors for the efficacy of tDCS in adolescent depression patients,and MoCA score was a protective factor(P＜0.05).The ROC curve showed that the area under the curve of the model for predicting the efficacy of tDCS in patients with adolescent depression was 0.823(95％CI:0.762-0.885).The prediction results of the model in the calibration curve were highly consistent with the actual observation results,and when the threshold probability was in the range of 0.01-0.80,the application of the model could produce better clinical net benefit.The external validation showed that the nomogram prediction model had a coincidence rate of 92.54％with the actual efficacy of tDCS,and the Kappa value was 0.811(95％CI:0.708-0.935).Conclusion The nomogram model based on course of disease,HAMD score,MoCA score,serum VDBP and hs-CRP to predict the efficacy of tDCS in adolescent depression has high prediction accuracy,calibration and clinical practicability,which is helpful to assist clinical early identification of high-risk groups with poor efficacy.