Maternal and neonatal predictors of high Edinburgh Postnatal Depression Scale score on postpartum day 7: a retrospective cohort study

Objective To evaluate the association between Edinburgh Postnatal Depression Scale (EPDS) scores on postpartum day 7 and maternal demographic, sociodemographic, obstetric, complication-related, and neonatal factors.Methods This retrospective cohort study included 1,032 women who delivered at a tertiary referral hospital in 2025. Patients were stratified into low-risk (EPDS <13, n = 893) and high-risk (EPDS ≥13, n = 139) groups. Demographic, sociodemographic, obstetric, neonatal, and laboratory variables were compared, and independent predictors of high-risk status were identified by multivariable logistic regression.Results Of 1,032 patients, 13.5% (n = 139) screened high-risk. On univariable analysis, any prior abortion was significantly less frequent in the high-risk than in the low-risk group (5.0% vs 11.5%, p = 0.020).Higher education (university 30.2% vs 21.6%, p = 0.041) and employment (41.7% vs 26.9%, p < 0.001) were more frequent in the high-risk group. On multivariable analysis, prior abortion count was the only independent inverse predictor (aOR = 0.536, 95% CI 0.299–0.959, p = 0.036); gestational diabetes did not reach significance (aOR = 0.311, p = 0.053). Sociodemographic variables were analyzed univariately and were not entered into the multivariable model. No other variables differed significantly.Conclusions In this Turkish obstetric population, a higher prior abortion count was independently associated with a lower likelihood of high EPDS risk on postpartum day 7. This unexpected inverse association is hypothesis-generating and should be interpreted with caution, given the small number of events, the inability to distinguish loss subtypes, and the multiple comparisons performed. Univariate associations of higher education and employment with EPDS risk diverge from Western literature and warrant prospective multicenter confirmation.