Although more and more attention has been paid to the psychological consequences of the lockdown policy amongst pregnant women, the underlying mechanism linking the lockdown policy to maternal depression has not been studied in the context of China. This study aimed to explore the association between the lockdown policy and maternal depressive symptoms, and whether such association was mediated by internet use and/or family support.

This cross-sectional study used multi-stage sampling techniques in central and western China. Data were collected from 1266 pregnant women using a structtured questionnaire that measured internet use, family support, and depressive symptoms. The Patient Health Questionnaire-9 (PHQ-9) was used to measure depressive symptoms. Internet use was measured by length of usage and varierity of purpose for internet use. Family support was measureed by spousal support and parental support. The structural equation modelling was employed to conduct mediation analysis to test the specificity of the hypothetical paths.

Overall, 527 respondents (41.63%) presented depressive symptoms. The lockdown policy was negatively associated with depressive symptoms in pregnant women (β = − 0.925, 95% CI = −1.510, − 0.360). The impact of the lockdown policy on depressive symptoms was partially mediated by internet use (β = 1.589, 95% CI = 0.730, 2.807) and family support (β = − 0.162, 95% CI = − 0.341, − 0.017), accounting for 42.67% of the total effect.

The lockdown policy was generally associated with fewer depressive symptoms in pregnant women. The lockdown policy increased maternal depressive symptoms through increased internet use, but decreased maternal depressive symptoms through enhanced family support. The findings suggest that the psychological consequence of the lockdown policy may vary across different populations, and warrant the need to take into consideration the features of subgroups.