RESUMO
BACKGROUND: Perinatal depression has attracted increasing attention. However, a detailed investigation of the network structure of depression is still lacking. We aim to examine the similarities and differences between the Edinburgh Postnatal Depression Scale (EPDS) and Patient Health Questionnaire (PHQ-9) from a network perspective. METHODS: A cross-sectional study was conducted from August 2020 to March 2022. We followed the STROBE checklist to report our research. Pregnant women (n = 2484) were recruited. All participants completed the EPDS and PHQ-9. We mainly used network analyses for statistical analysis and constructed two network models: the EPDS and PHQ-9 models. RESULTS: The detection rates of prenatal depression measured by the EPDS and PHQ-9 were 30.2 % and 28.2 %, respectively. In the EPDS network, the EPDS8 'sad or miserable' node (strength = 1.2161) was the most central node, and the EPDS10 'self-harming' node (strength = 0.4360) was the least central node. In the PHQ-9 network, the PHQ4 'fatigue' node (strength = 0.9815) was the most central node, and PHQ9 'suicide' was the least central symptom (strength = 0.5667). For both models, 'sad' acted as an important central symptom. CONCLUSIONS: Psychological symptoms may be more important in assessing depression using the EPDS, while physical symptoms may be more influential in assessing depression using the PHQ-9. For both the EPDS and PHQ-9, "sad" was an important central symptom, suggesting that it may be the most important target for further maternal depression interventions in the future.