Principal Component Analysis of EPDS Questions to Identify Trends in Depressive Symptoms Among At-Risk Populations [29K]

    loading  Checking for direct PDF access through Ovid



The Edinburgh Postnatal Depression Scale (EPDS) is a validated screening tool for peripartum depression. We seek to characterize a subset of questions most predictive of peripartum depression in patients with a history of anxiolytic/antidepressant or substance use.


We performed a case-control study of pregnant women who received prenatal care in an academic center. Pregnancy records were reviewed for patient demographics, medical and social history. The EPDS scores of subjects with a history of mood disorder (anxiolytic/antidepressant use, untreated depression) or substance use (ethanol, illicit drug use) were compared to controls to determine likelihood of probable depression (score >12) using a chi square test. Principal component analysis (PCA) was used to identify the individual EPDS survey items and question themes that were commonly associated with depressive symptoms.


297 subjects were reviewed. Subjects in the mood disorder and substance use groups were more likely to have EPDS scores >12 than controls in both the antepartum (P=.001, P=.030, respectively) and postpartum (P=.016, P=.008). PCA of EPDS scores for the mood disorders group identified 3 themes (anhedonia, anxiety, sadness) that accounted for 59% and 67% of variance in the antepartum and postpartum scores, respectively. PCA of EPDS scores in the substance use group identified 2 themes that accounted for 58% variance in the antepartum and 66% in the postpartum scores.


Patients with a history of mood disorder or substance use are at increased risk of peripartum depression and may have characteristic symptom complexes. Recognition of patterns in at-risk populations may facilitate early intervention.

Related Topics

    loading  Loading Related Articles