儿童青少年抑郁症状的风险预测模型:一项基于NHIS数据的临床研究

Risk prediction model for depressive symptoms in children and adolescents:A clinical study based on NHIS data

目的探讨儿童青少年抑郁症状发生的主要危险因素,并构建列线图模型。方法选取2022年美国健康访谈调查(NHIS)收集的符合条件的数据进行分析,数据主要包含儿童和家庭的人口学特征。采用LASSO分析进行儿童青少年抑郁症状的预测因素初步筛选,利用logistic回归进行多因素分析,最后用筛选的独立危险因素构建抑郁症状预测列线图。结果有抑郁症状组的年龄为(12.7±3.3)岁,无抑郁症状组的年龄为(11.7±3.5)岁。Logistic回归分析显示,女性、年龄稍大、参加体育活动较少、被种族歧视、曾与有精神疾病或严重抑郁者同住、曾受过或目睹过暴力、曾患孤独症、曾患多动症、健康状况不良为儿童青少年抑郁症状发生的主要危险因素(P<0.05)。儿童青少年抑郁症状的列线图模型的C-index为0.71;校正曲线的预测值和实际值基本吻合;模型的ROC曲线下面积为0.712。结论性别、年龄、参加体育活动、种族歧视、曾与有精神疾病或严重抑郁者同住、曾受过或目睹过暴力、孤独症、多动症、健康状况是儿童青少年抑郁症状发生的主要影响因素,儿童青少年抑郁症状列线图预测模型的临床应用价值较高。

Objective To explore the main risk factors for the occurrence of depressive symptoms in children and adolescents,and to construct a nomogram model.Methods The eligible data collected by the 2022(American)National Health Interview Survey(NHIS)were selected for analysis,and the data mainly included the demographic characteristics of children and families.Least absolute shrinkage and selection operator(LASSO)analysis was used to preliminarily screen the predictors of depressive symptoms in children and adolescents,logistic regression was used for multivariate analysis,and finally,the nomogram of predictive symptoms of depressive symptoms was constructed with the screened independent risk factors.Results The average age of the group with depressive symptoms was(12.7±3.3)years,and the average age of the group without depressive symptoms was(11.7±3.5)years.Logistic regression analysis showed that female gender,older age,less physical activity,racial discrimination,living with people with mental illness or severe depression,having experienced or witnessing violence,having autism,having attention deficit and hyperactivity disorder(ADHD),and poor health were the main risk factors for depressive symptoms in children and adolescents(P<0.05).The C-index of the nomogram model of depressive symptoms in children and adolescents was 0.71,the predicted value of the calibration curve was basically consistent with the actual value,and the area under the ROC curve of the model was 0.712.Conclusion Gender,age,participation in physical activities,racial dis-crimination,living with people with mental illness or severe depression,having experienced or witnessed violence,autism,ADHD,and health status are the main influencing factors for the occurrence of depressive symptoms in children and adolescents.