A Longitudinal Study from Early Pregnancy to Postpartum: The Moderating Effect of Marital Support on the Development of Perinatal Depression and the Construction of a Risk Prediction Model

Objective To investigate the moderating effect of marital support on the development of perinatal depression and to construct a risk prediction model for perinatal depression, providing important references for clinical practice. Methods This study adopted a longitudinal design, including 1,520 pregnant women who visited West China Second University Hospital, Sichuan University between January 2023 and December 2024. Data were collected at three time points: early pregnancy (T1), late pregnancy (T2), and postpartum (T3). The Dyadic Adjustment Scale (DAS) was used to assess marital support, and the Edinburgh Postnatal Depression Scale (EPDS) was used to assess depressive symptoms. Data analysis methods included univariate analysis, logistic regression, cross-lagged models, and prediction model construction. Results The mean EPDS scores at early pregnancy, late pregnancy, and postpartum were 7.45 ± 3.57, 7.02 ± 3.51, and 7.99 ± 4.17, respectively. The prevalence of depression showed a U-shaped trend, with 16.5% in early pregnancy, 15.0% in late pregnancy, and 18.5% postpartum. The results of the univariate analysis indicated that the marital support scores (including consensus, positive interaction, marital satisfaction, and total marital score) in the depressed group were all significantly lower than those in the non-depressed group ( P < 0.05 ), while the conflict score was significantly higher than that in the non-depressed group ( P < 0.05 ). Cross-lagged model analysis indicated that marital support in early pregnancy significantly reduced the risk of depression in late pregnancy ( β=-0.24, P = 0.003 ), while depression in late pregnancy significantly reduced marital support postpartum ( β=-0.18, P = 0.020 ). Multivariate logistic regression analysis showed that three dimensions of marital support (including consensus, positive interaction, marital satisfaction) were protective factors ( OR < 1, P < 0.05 ) in all three stages, whereas conflict was a risk factor ( OR > 1, P < 0.05 ). The performance of the postpartum depression prediction model based on marital support was the highest ( AUC = 0.82 ), indicating its excellent performance. Conclusion Marital support plays a significant moderating role in the development of perinatal depression. High marital support reduces the risk of depression, while low support increases the risk. The postpartum depression prediction model based on marital support performs well, providing important references for clinical interventions in perinatal depression.