Construction Of Prediction Models For Mental Disorders In Mid-late Pregnant Women

ObjectiveTo investigate risk factors for pregnancy-related anxiety, anxiety and depression and build the prediction models for these mental disorders in mid-late pregnant women.MethodsThere were two parts in this study:The first part:Using the methods of literature review and expert counseling, related factors of pregnant women’s mental health were preliminarily screened and the questionnaire was formed; Six hundred women in the second or third trimester pregnancy who underwent prenatal routine examination in the hospital were recruited. Pregnancy-related anxiety, anxiety and depression of these women were analyzed. Prediction models were constructed through single factor analysis and multi-factor Logistic regression analysis.The second part:Another three hundred and forty-six mid-late pregnant women were recruited and filled in the questionnaire which was consisted of related factors based on the first part; Receiver operating characteristic curve (ROC) analysis, sensitivity and specificity were used to verify the prediction models.Results(1) Five hundred and fifty one of 600 participants completed the questionnaire; Women with positive pregnancy-related anxiety, anxiety and depression were 19.4%,6.0% and 11.3%, respectively; There were certain correlations among the three, with the highest degree of correlation between anxiety and depression (r=0.692, P<0.001) and the lowest between pregnancy-related anxiety and depression (r=0.384, P<0.001).(2) According to the single-factor analysis and multi-factor Logistic regression analysis, the risk factors of pregnancy-related anxiety were young age, unwanted pregnancy, inadequate preparation for pregnancy, low self-image satisfication, vulnerable personality style, life event during pregnancy, negative coping style; The risk factors of anxiety were young age, average and difficult family economic income, low satisfaction of economic position at home, poor physical health condition before pregnancy, life event during pregnancy, vulnerable personality style; The risk factors of depression were young age, low marital satisfacation, low satisfaction of economic position at home, uplanned pregnancy, life event during pregnancy, vulnerable personality style, low self-efficacy.(3) Prediction models were verified according to the 342 completed data; For pregnancy-related anxiety, the sensitivity was 63.1%, specitivity was 73.3% and the area under the ROC curve (AUC) was 0.682-0.726; For anxiety, the sensitivity was 78.1%, specitivity was 73.9% and AUC was 0.760-0.837; For depression, the sensitivity was 69.4%, specitivity was 76.8% and AUC was 0.731-0.840.ConclusionsThe logistic regression models had certain prediction value for pregnancy-related anxiety, anxiety and depression in mid-late pregnant women, which could provide evidence reference for identifying high risk pregnant women with mental disorders and development of interventions in clinic.