可干预指标的七年级学生抑郁风险提示模型及随访情景OA

Objective:To address the resource-risk trade—off in school depression screening regarding high posi-tive rates and limited follow-up capacity,this study aims to construct an interpretable risk-flagging model for cur-rent elevated depressive symptoms in seventh-grade students and quantify threshold-workload scenarios.Methods:A total of 912 seventh-grade students were surveyed.Four modifiable psychological indicators were used to predict SDS≥53(non-diagnostic).A logistic regression model was trained(n=643)and validated in an independent test set(n=269)with threshold scenario analysis.Results:The overall logistic regression model was statistically signifi-cant(x2=207.644,P＜0.001).Consistent with prior hypotheses,executive dysfunction(OR=1.038)significantly and positively predicted depression risk,whereas mindfulness(OR=0.975)and positive psychological capital(OR=0.971)significantly and negatively predicted depression risk.The model showed good discrimination and calibration(AUC=0.835,95％CI 0.788-0.882;Brier score=0.169).Threshold scenario analyses indicated that at τ=0.268,follow-up would be required for 73.2％of students,with approximately 5.2 missed cases per 100;at τ=0.748,fol-low-up would be required for 23.8％of students,with approximately 30.1 missed cases per 100.Conclusion:The findings generally support the research hypotheses.Executive dysfunction,mindfulness,and psychological capital are important correlates of elevated current depressive symptoms in seventh-grade students.The model shows good discrimination and potential practical utility.Different thresholds correspond to different follow-up workloads and missed-case trade-offs,providing a quantitative reference for schools to implement stratified follow-up and fur-ther assessment under resource constraints.