Zhu Jue, Ye Youchun, Liu Xuan, Chen Yichen, Chen Lu, Lin Yi, Wang Qiming, Zhang Jing
Department of Gynecology and Obstetrics, Women and Children's Hospital of Ningbo University, Ningbo, Zhejiang, China.
Department of Gynecology and Obstetrics, The Affiliated People's Hospital of Ningbo University, Ningbo, Zhejiang, China.
Front Med (Lausanne). 2024 Aug 23;11:1407034. doi: 10.3389/fmed.2024.1407034. eCollection 2024.
Perinatal depression (PND) affects approximately 15%-20% of women. This study aimed to determine the incidence of PND and identify risk factors.
A prospective study was conducted at the Affiliated People's Hospital of Ningbo University. The Edinburgh Postnatal Depression Scale (EPDS) was used to screen for PND. Classification models were constructed using Extreme Gradient Boosting (XGBoost), Logistic Regression (LR), Random Forest (RF), and Support Vector Machine (SVM), and the optimal model was selected.
Between March 2019 and August 2021, a total of 485 participants completed all valid questionnaires. Depression was observed in 75 (15.5%), 47 (9.7%), 25 (5.2%), 94 (19.4%), 85 (17.5%), and 43 (8.9%) cases during the first trimester, the second trimester, the third trimester, 1 week postpartum, 6 months postpartum, and 12 months postpartum, respectively. During the prenatal period, factors such as monthly income, employment status, marital status, and thyroid function significantly impacted depression. Additionally, factors including monthly income, employment status, marital status, parity, and unintended pregnancy were found to affect the likelihood of developing postpartum depression. XGBoost was chosen for its accuracy (0.9097) and precision (0.9005) in predicting prenatal depression, as well as for its accuracy (0.9253) and precision (0.9523) in predicting postpartum depression.
In conclusion, the incidence of depression varies throughout the perinatal period, with different factors influencing prenatal and postpartum depression.
围产期抑郁症（PND）影响约15%-20%的女性。本研究旨在确定PND的发病率并识别风险因素。
在宁波大学附属人民医院进行了一项前瞻性研究。采用爱丁堡产后抑郁量表（EPDS）筛查PND。使用极端梯度提升（XGBoost）、逻辑回归（LR）、随机森林（RF）和支持向量机（SVM）构建分类模型，并选择最优模型。
在2019年3月至2021年8月期间，共有485名参与者完成了所有有效问卷。在孕早期、孕中期、孕晚期、产后1周、产后6个月和产后12个月，分别观察到75例（15.5%）、47例（9.7%）、25例（5.2%）、94例（19.4%）、85例（17.5%）和43例（8.9%）抑郁症病例。在孕期，月收入、就业状况、婚姻状况和甲状腺功能等因素对抑郁症有显著影响。此外，发现月收入、就业状况、婚姻状况、产次和意外怀孕等因素会影响产后抑郁症的发生可能性。XGBoost因其在预测产前抑郁症方面的准确性（0.9097）和精确性（0.9005），以及在预测产后抑郁症方面的准确性（0.9253）和精确性（0.9523）而被选中。
总之，围产期抑郁症的发病率在整个围产期各不相同，不同因素影响产前和产后抑郁症。
