Depression during pregnancy and postpartum poses significant risks to both maternal and child well-being. The underlying biological mechanisms are unclear, but epigenetic variation could be exploited as a plausible candidate for early detection. We investigated whether DNA methylation signatures are associated with antenatal depressive symptoms (ADS) and whether early alterations in methylation patterns could be used to predict postpartum depressive symptoms (PDS). 201 pregnant women in early pregnancy, without a prior history of depression disorders, from the STratification of Risk of Diabetes in Early Pregnancy study were recruited. Using the Patient Health Questionnaire-9 (PHQ-9), 92 women were identified with ADS, while 109 served as controls. Edinburgh Postnatal Depression Scale (EPDS) was used to assess PDS during 6-12 weeks after delivery. The dataset was split into 80 % for training and testing and 20 % for validation, to discern potential CpGs for ADS using a support vector machine classifier. Analysis revealed 591 CpGs significantly associated with ADS, from which a panel of 7 CpGs was identified to discriminate between ADS and controls with high sensitivity and specificity (AUC: 0.85 in test, 0.73 in validation). Pathway analysis highlighted involvement in inositol phosphate metabolism, notch, and calcium signaling. The same 7 CpGs predicted PDS with an AUC of 0.76 (95 % CI: 0.66-0.87). Integration of CpG data with patient-reported information significantly enhanced PDS prediction. Our study identified DNA methylation signatures that could potentially differentiate ADS from controls and predict PDS. This suggests potential for developing a CpG panel for diagnostic and preventive strategies for perinatal depression.
Förderungen Department of Health Research (DHR), Govt. of India Indian Council of Medical Research (ICMR) MRC-DBT Newton fund DBT-Wellcome Trust India Alliance DBT-Wellcome Trust India Alliance Intermediate Clinical & Public health Fellowship