| Objective:Postpartum depression(PPD),a major mental health disorder occurred after delivery,is a common complication of child bearing which can lead to maternal and fetal adverse consequences.It has become a significant public health problem in the world for its high prevalence and adverse impacts on the wellbeing of women and their infants.The signs and symptoms of PPD are complex and diverse,and the causes of PPD are unknown;however,the pathophysiology of PPD is thought to involve interactions between biological susceptibility and multiple risk factors.At present,insufficient understanding and inadequate attention was paid on PPD in China leading to short-and long-term adverse effects for numbers of pregnant women and their families.If PPD is to be prevented by public health or clinical interventions,or predicted,related factors need to be readily identified.This study focused on the prevalence and risk factors of PPD in Tianjin urban pregnant women,and the efficiency of predictive models at different stages.Methods:In 2009,Tianjin set up a computerized database of Maternal and Child Health Information System(MCHIS),consisting of records of antenatal examinations,delivery data and regular health examinations of children up to 6 years old,managed by the Tianjin Women and Children's Health Center.Briefly,22302 pregnant women were registered with a primary care hospital in the central city for pregnancy from 2010 to 2013.A total of 8972 women agreed and filled the Chinese Mainland version of EPDS,and the remaining 8842 women were included in the analysis after excluding 57 invalid response.In the MCHIS,all the health related information was collected using standard methods.The PPD was defined as having an EPDS score of 10 or more,and scores of 10-12,?13 indicated the presence of minor and major depressive symptoms,respectively.The mental health subscale of medical outcomes study 36-item short-form health survey(SF-36)was used to evaluate the condition of mental health during pregnancy.Odds ratios(ORs)and 95% confidence intervals(CIs)of potential correlated factors for minor and major depressive symptoms were obtained by General Logit Model in univariable and multivariable analysis.The data set was randomly stratified divided into tranining set and test set with a ratio of 7:3.Binary logistic regression was used to created predictive models for major depression with a method of optimal subsets and Akaike Information Criterion(AIC).Calibration was checked with Hosmer-Lemeshow test and accuracy rate using test data set,and discrimination was checked using receiver operating characteristic curve(ROC)and the areas under ROC(C index),sensitive,and specificity.Results:Mean age of the 8842 pregnant women was 28.4(standard deviation: 2.9)years with a mean of 10.3(standard deviation: 2.6)weeks at first registration.Depression symptoms over the past 7 days were collected at a median of 3(interquartile: 1 to 6)weeks postpartum.All the women got a mean of 3.68(standard deviation: 3.55)and a median of 3(interquartile: 1 to 6)scores of Chinese Mainland version of EPDS.Among the 8842 pregnant women,the prevalence of PPD,minor and major depression were 8.45%(95%CI: 7.87%-9.03%),6.99%(95%CI: 5.25%-8.73%),1.46%(95%CI: 0.07%-2.26%),respectively.In 2011,the prevalence of PPD in the six districts were Heping(13.48%),Hexi(11.01%),Nankai(10.30),Hedong(10.29%),Hongqiao(10.12%)and Hebei(8.61%),descendingly,and no difference was found among Nankai,Hedong and Hongqiao.In 2012,the prevalence of PPD in the six districts were Hongqiao(7.32%),Nankai(6.72),Hexi(6.35%),Heping(5.93%),Hedong(5.67%),and Hebei(5.17%),descendingly.Women with blood B had lower risk of minor and major postpartum depression as compared with those with other blood groups(adjusted OR: 1.27,95%CI: 1.02-1.58;adjusted OR: 1.34,95%CI: 1.04-1.79).Women with bad sleep quality before pregnancy had higher risk of minor and major postpartum depression as compared with those sleep well(adjusted OR: 2.68,95%CI: 1.68-4.25;adjusted OR: 2.24,95%CI: 1.07-6.35).Women with medium and high life stress before pregnancy had higher risk of minor and major postpartum depression as compared with those with little life stress(adjusted OR: 1.81,95%CI: 1.11-2.94;adjusted OR: 2.31,95%CI: 1.05-5.10).Women exposed to passive smoking during pregnancy had higher risk of minor postpartum depression as compared with those not-exposed to(adjusted OR: 1.31,95%CI: 1.04-1.65).Women with bad sleep quality during pregnancy had higher risk of minor postpartum depression as compared with those sleep well(adjusted OR: 1.32,95%CI: 1.04-1.67).Women with common life stress during pregnancy had higher risk of minor and major postpartum depression as compared with those with little life stress(adjusted OR: 1.49,95%CI: 1.14-1.96;adjusted OR: 1.43,95%CI: 1.07-1.67).Women with medium and high life stress during pregnancy had higher risk of minor and major postpartum depression as compared with those with little life stress(adjusted OR: 1.61,95%CI: 1.07-3.36;adjusted OR: 1.63,95%CI: 1.06-5.81).The risk of minor and major postpartum depression can be reduced by 5% when the scores of SF-36 mental health subscale increasing 1 point(adjusted OR: 0.95,95%CI: 0.91-0.97;adjusted OR: 0.95,95%CI: 0.91-0.98).Women with formula milk-feeding had higher risk of minor and major postpartum depression as compared with those with breast-feeding(adjusted OR: 1.42,95%CI: 1.05-2.12;adjusted OR: 2.39,95%CI: 1.57-3.64).Women resting at home after delivery had lower risk of minor postpartum depression as compared with those resting at other places(adjusted OR: 0.74,95%CI: 0.60-0.90).Women eating nutrition supplements after delivery had lower risk of minor and major postpartum depression as compared with those not ate(adjusted OR: 0.77,95%CI: 0.80-1.17;adjusted OR: 0.59,95%CI:0.32-0.79).Women having diseases after delivery had higher risk of minor and major postpartum depression as compared with those have not(adjusted OR: 2.58,95%CI: 1.71-3.91;adjusted OR: 3.59,95%CI:1.78-7.25).The area under ROC of first trimester predictive mode of PPD was 0.605(95%CI:0.583-0.627,P<0.0001),and the sensitivity,specificity,and accuracy rate were 0.623,0.515,and 49.7% respectively with a positive cutoff of ?0.0792.The area under ROC of second trimester predictive mode of PPD was 0.637(95%CI:0.612-0.662,P<0.0001),and the sensitivity,specificity,and accuracy rate were 0.642,0.558,and 59.4% respectively with a positive cutoff of ?0.0760.The area under ROC of postpartum predictive mode of PPD was 0.671(95%CI:0.653-0.688,P<0.0001),and the sensitivity,specificity,and accuracy rate were 0.654,0.565,and 62.7% respectively with a positive cutoff of ?0.0811.The P values of Hosmer-Lemeshow of the three models were 0.6646,0.9441 and 0.8962.Conclusions:The prevalence of PPD Tianjin urban pregnant women was lower than foreign population and other population in China.Bad sleep quality before and during pregnancy,high life stress before and during pregnancy,exposed to passive smoking during pregnancy,and having diseases after delivery were risk factors for PPD in Chinese women;and blood group B,breast-feeding,rest at home and eating nutrition supplement after delivery were protective factors for PPD in Chinese women.We can pay attention to improving the sleep quality of pregnant women,reducing the pressure of life,avoiding passive smoking,promoting breastfeeding and postpartum consumption of nutritional products in the development of prevention strategies.It is feasible to develop predictive models at pre-pregnancy,pregnancy and postpartum to estimate the prevalence of PPD.In the future research,more attention should be paid to collecting clinical characteristics of pregnant women who are screened positively to improve the efficacy and practicality of the predictive model based on the collection of information on lifestyle and pregnancy test. |
PDF Full Text Request