Abstract:

Objective　To analyze the correlation between depression with psychotic symptoms and childhood abuse and construct a prediction model. Methods　A total of 600 patients with depression admitted to Chang'an Hospital and Dali Shanda Psychiatric Hospital from February 2021 to February 2024 were selected as the study objects, including 208 males and 392 females, aged 21 to 50 (35.11±4.88) years. The patients were divided into a psychotic symptom group (65 cases) and a non-psychotic symptom group (535 cases) according to the 90-item Symptom Check List (SCL-90) score. General data of the patients were collected and compared through the electronic medical record system, including demographic characteristics (gender, age, body mass index, duration of depression, marital status, education, employment, family history of depression, and childhood abuse) and clinical characteristics (age of onset of depression, depression with anxiety, and depression with suicidal tendency). Independent sample t test and χ² test were used for statistical analysis. Multivariate logistic regression analysis was used to analyze the influencing factors of depression patients with psychotic symptoms, and a prediction model was constructed. The receiver operating characteristic curve (ROC) was drawn by MedCalc software, and the Hosmer-Lemeshow goodness of fit test was used to verify the difference between the prediction model and the actual observation results. Results　The age of onset of depression in the psychotic symptom group was lower than that in the non-psychotic symptom group [(27.11±3.94) years old vs. (31.19±4.01) years old]; the proportions of family history of depression, childhood abuse, depression accompanied by suicide tendency were higher than those in the non-psychotic symptom group [20.00% (13/65) vs. 6.73% (36/535), 41.54% (27/65) vs. 24.30% (130/535), 46.15% (30/65) vs. 29.35% (157/535)] (all P<0.05). Multivariate logistic regression analysis showed that family history of depression (OR: 2.483, 95%CI: 1.133-5.445), childhood abuse (OR: 2.234, 95%CI: 1.250-3.995), low age of onset of depression (OR: 0.794, 95%CI: 0.740-0.853), and depression with suicidal tendency (OR: 1.935, 95%CI: 1.093-3.428) were risk factors for psychotic symptoms in patients with depression (all P<0.05). ROC results showed that the area under the curve of the prediction model was 0.780 (95%CI: 0.744-0.812), the sensitivity was 80.00%, and the specificity was 64.11%. Conclusions　Depression patients with psychotic symptoms are closely related to family history of depression, childhood abuse, age of onset of depression, and depression with suicidal tendency. The predictive model is effective and can help medical staff identify psychotic symptoms early.

Key words:

Depression,  , Psychotic symptoms,  , Childhood abuse,  , Influencing factors, Prediction model

摘要：

目的　分析抑郁症伴精神病性症状与儿童期虐待的相关性并构建预测模型。方法　选取2021年2月至2024年2月长安医院和大荔善达精神专科医院收治的600例抑郁症患者作为研究对象，男208例，女392例，年龄21～50（35.11±4.88）岁。根据90项症状清单（SCL-90）评分将患者分为精神病性症状组（65例）和非精神病性症状组（535例）。通过电子病历系统收集患者一般资料并比较，包括人口统计学特征（性别、年龄、体重指数、抑郁症病程、婚姻情况、受教育情况、就业情况、抑郁症家族史，儿童期受虐情况）、临床特征（抑郁症起病年龄、抑郁症伴焦虑、抑郁症伴自杀倾向）。采用独立样本t检验和χ²检验进行统计学分析。采用多因素logistic回归分析抑郁症患者伴精神病性症状的影响因素，构建联合预测因子，采用MedCalc软件绘制受试者操作特征曲线（ROC），Hosmer-Lemeshow拟合优度检验，验证预测模型与实际观察结果的差异。结果　精神病性症状组抑郁症起病年龄低于非精神病性症状组［（27.11±3.94）岁比（31.19±4.01）岁］，有抑郁症家族史、儿童期有虐待、抑郁症伴有自杀倾向占比均高于非精神病性症状组［20.00%（13/65）比6.73%（36/535）、41.54%（27/65）比24.30%（130/535）、46.15%（30/65）比29.35%（157/535）］（均P<0.05）。多因素logistic回归分析显示，有抑郁症家族史（OR：2.483，95%CI：1.133～5.445）、儿童期有虐待（OR：2.234，95%CI：1.250～3.995）、抑郁症起病年龄低（OR：0.794，95%CI：0.740～0.853）、抑郁症伴有自杀倾向（OR：1.935，95%CI：1.093～3.428）均是抑郁症患者伴精神病性症状的独立危险因素（均P<0.05）。ROC结果显示，预测模型曲线下面积为0.780（95%CI：0.744～0.812），灵敏度为80.00%，特异度为64.11%。结论　抑郁症患者伴精神病性症状与抑郁症家族史、儿童期受虐情况、抑郁症起病年龄、抑郁症伴自杀倾向密切相关，预测模型效能良好，可帮助医务人员早期识别精神病性症状。

关键词:

抑郁症, 精神病性症状, 儿童期虐待, 影响因素, 预测模型