青少年抑郁障碍患者非自杀性自伤行为发生风险的列线图预测模型构建

Construction of a nomogram prediction model for the risk of non-suicidal self-injury behaviors in adolescents with depressive disorders

中文关键词:  青少年  抑郁障碍  非自杀性自伤  危险因素  预测模型  列线图

英文关键词:Adolescents  Depressive disorders  Non-suicidal self-injury  Risk factors  Prediction model  Nomogram

中文摘要:

      背景 非自杀性自伤(NSSI)已成为广泛关注的公共卫生问题。青少年抑郁障碍患者NSSI行为发生风险较高。既往研究多集中于NSSI行为的相关因素分析，而面向临床快速评估、兼具直观量化特征的可视化模型构建的研究较为有限。目的 考查青少年抑郁障碍患者NSSI行为的危险因素，并构建NSSI行为发生风险的列线图预测模型。方法 采用横断面研究，纳入2025年2月—12月于兰州大学第二医院心理卫生科就诊、符合DSM-5抑郁障碍诊断标准的青少年患者448例，按7:3随机分为建模组(n=313)和验证组(n=135)。采用SDS、SAS、BIS-11、CTQ-SF及PSQI进行评定。采用二元Logistic回归分析筛选建模组NSSI行为的独立预测因子，并构建列线图模型。采用ROC曲线、校准曲线和DCA分别评定该模型的区分度、准确度及临床应用价值。结果 二元Logistic回归分析显示，年龄较小、女生、SDS评分较高、BIS-11评分较高、CTQ-SF评分较高、PSQI评分较高是青少年抑郁障碍患者发生NSSI行为的危险因素(OR=0.816、2.671、1.131、1.030、1.064、1.102，P<0.05或0.01)。基于上述6个指标构建的列线图模型，HL检验显示拟合度较好(χ²=15.144，P=0.056)。内部验证及外部验证结果显示，建模组AUC为0.895，验证组AUC为0.891；校准曲线平均绝对误差分别为0.043和0.023；DCA显示当预测风险阈值大于0.11时，模型具有明确的临床净收益。结论 基于年龄、性别、抑郁症状严重程度、冲动性、童年创伤、睡眠质量6个指标构建的青少年抑郁障碍患者NSSI行为发生风险的列线图预测模型具有较高的区分度、准确度和临床应用价值。

英文摘要:

      Background Non-suicidal self-injury (NSSI) has become a public health concern of wide interest. Adolescents with depressive disorders are at higher risk of NSSI behaviors. Previous studies have mostly focused on analyzing factors associated with NSSI, while research on developing visual models suitable for rapid clinical assessment is relatively limited. Objective To investigate the risk factors for NSSI behaviors in adolescents with depression disorders, and to construct a nomogram prediction model. Methods A cross-sectional study was conducted. A total of 448 adolescent patients diagnosed with depressive disorders were enrolled and randomly divided into a modeling group (n=313) and a validation group (n=135) at a ratio of 7:3. Assessments were performed using SDS, SAS, BIS-11, CTQ-SF, and PSQI. Binary Logistic regression analysis was applied to screen independent predictors, and a nomogram was constructed. ROC curve, calibration curve, and decision curve analysis were used to evaluate the model. Results Binary Logistic regression analysis revealed that younger age, female sex, higher SDS score, higher BIS-11 score, higher CTQ-SF score, and higher PSQI score were risk factors for NSSI behaviors (OR=0.816, 2.671, 1.131, 1.030, 1.064, 1.102, P<0.05 or 0.01). The HL test showed a good fit (χ²=15.144, P=0.056). Model validation showed that the AUC was 0.895 for the modeling group and 0.891 for the validation group. The mean absolute errors were 0.043 and 0.023, respectively. DCA indicated that the model provided a clear net clinical benefit when the risk threshold was greater than 0.11. Conclusion The nomogram model based on six indicators for predicting NSSI risk in adolescent patients with depressive disorders exhibits good discrimination, calibration, and clinical application value.