Assessment of the factor structure of the EPDS in Chinese perinatal women: a longitudinal study using multigroup confirmatory factor analysis

Introduction Perinatal depression poses significant risks to the mental health of perinatal women, affecting both their well-being and their ability to care for themselves and infants. This study investigates the factor structure and reliability of the Edinburgh Postnatal Depression Scale (EPDS) across multiple time points in perinatal women in China. Methods A total of 849 perinatal women participated in this study, with data collected at six time points: early, mid, and late pregnancy, as well as 1, 3, and 6 months postpartum. Parallel analysis and exploratory factor analysis were conducted to determine the factor structure of the EPDS. Internal consistency was assessed using Cronbach’s alpha. Multigroup confirmatory factor analysis was performed to assess measurement invariance between the antepartum (T1–T3) and postpartum (T4–T6) groups, and to assess the fit of model across the perinatal period. Results Three-factor models fit best for the EPDS during the perinatal period when analyzed at each time points and across the perinatal period. When the same item assignment was applied to the antepartum and postpartum groups, a three-factor model for the EPDS fit well [χ2(df = 48) = 231.92, p < 0.001; CFI = 0.97, RMSEA = 0.06, TLI = 0.96]. The EPDS showed a Cronbach’s alpha of 0.84. Discussion The 3-factor model of EPDS showed good reliability, internal consistency, and measurement invariance. Our findings suggested that the EPDS can effectively capture symptom variability in perinatal depression, supporting its use as a monitoring tool throughout both antepartum and postpartum.