Influencing Factors of High-risk Population for Postpartum Depression Based on Latent Profile Mode
Objective The latent profile model(LPM)was used to search for the optimal classification of post-partum depression,understand the proportion of each latent category in the total population,and explore the factors influencing the population at high risk of postpartum depression.Methods The Edinburgh postnatal depression scale(EPDS)was used to investigate postpartum depression in 529 cases selected from the depart-ment of obstetrics of four hospitals in Yinchuan and Zhongwei cities,and the LPM was used to analyze the opti-mal classification of potential profiles of postpartum depression,and Logistic regression was used to analyze the influencing factors of potential categories of postpartum depression.Results The result of LPM support the classification model that postpartum depression was divided into two potential categories.According to the score characteristics,the two potential categories were named as low-risk type[(3.500±2.477)points,accounting for 58.0％of the total population]and high-risk type[(11.550±3.189)points,accounting for 42.0％of the total popu-lation].Logistic regression analysis of postpartum depression showed that the relationship between moth-er-in-law and daughter-in-law(OR=2.088,95％CI:1.055-4.132,P=0.035),prenatal depression(OR=3.587,95％CI:1.456-8.839,P=0.006)were risk factors for high risk type of postpartum depression.Husband's educational level(OR=0.306,95％CI:0.108-0.864,P=0.025)and puerpera age(OR=0.283,95％CI:0.129-0.623,P=0.002)were protective factors of the high-risk types of postpartum depression.Conclusion The 2 potential categories were the optimal classification model for postpartum depression,and the mother-in-law and daughter-in-law relationship,prenatal depression,husband's education level affected on the type of high risk of postpartum de-pression,and prenatal depression was the most strongly associated with postpartum depression.
postpartum depressioninfluencing factorlatent profile model
万方数据
维普
