Stability and variability of Edinburgh Postnatal Depression Scale networks among postpartum women: Insights from the...

This study aimed to clarify the network structure of the Edinburgh Postnatal Depression Scale (EPDS) in postpartum women using data from the Japan Environment and Children's Study (JECS), a large-scale nationwide birth cohort.

Network analysis was performed on EPDS data from 89,151 participants at 1 and 6 months postpartum. Gaussian graphical models (GGM) were employed to estimate the direct relationships between EPDS items; predictability measures were used to assess the degree of variance in each item that could be explained by the network. Subgroup analyses were conducted for groups with high (≥9) and low (<9) EPDS scores based on the Japanese cutoff for postpartum depression screening.

The EPDS network structure was well-preserved from 1 month to 6 months postpartum. In both periods, "Sad" and "Anxious" demonstrated high centrality, playing key roles in the psychological network. Strong associations between "Hard to sleep" and "Self-harm" were observed in the high-score group at 1 month postpartum, but not at 6 months. Overall predictability was lower in the high-score group, suggesting greater involvement of external factors such as social support, economic conditions, and childcare burden.

This study provides the first insights into the EPDS network structure among postpartum women. Anxiety, sadness, and sleep disturbances emerged as key symptoms, underscoring the need for early intervention. Lower predictability in the high-score group suggests the influence of external factors, emphasizing the need for comprehensive support programs. Future research should consider other populations and time points, such as pregnancy and beyond 1 year postpartum.