Background Migrant and refugee women are at risk of perinatal depression due to multiple stressors experienced before, during and after the migration trajectory. In low-income settings, continued hardships following resettlement and limited access to mental health services may pose additional challenges. Social support has consistently been identified as protective against perinatal depression. This study assesses the associations between three different forms of social support - received, perceived and partner support - and perinatal depression among migrant and refugee women living on the Thai-Myanmar border.

Methods We conducted a cohort study on the Thai-Myanmar border of women recruited in their first trimester of pregnancy. Depression status was assessed using a clinical interview in the first, second and third trimesters and at one month post-partum. Received support, perceived support and partner support were measured in the third trimester. Associations between social support and perinatal depression were assessed using logistic regression with separate models for migrants and refugees. A series of multivariable regression models were built using stepwise estimation with demographic, socio-economic, migration, obstetric and psychosocial variables sequentially added to the model.

Results Of 568 women participating in the study, 451 (233 migrants; 218 refugees) had complete data for social support measures and were included in the current analysis. The prevalence of perinatal depression was 38.6% in migrants and 47.3% in refugees. Migrants reported higher levels of received, perceived and partner support than refugees. In the final model, after controlling for all other variables, higher levels of received support remained significantly associated with a lower likelihood of perinatal depression in migrants (adjusted odds ratio 0.82; 95% CI 0.68–0.99). In the final model for refugees, all three social support measures were dropped from the model. Among both migrants and refugees, a previous history of depression and experiences of trauma were strongly associated with perinatal depression after controlling for all other variables.

Conclusion Our findings highlight the importance of received social support to perinatal depression in migrant women on the Thailand-Myanmar border. Experience of trauma and prior depression also strongly predicted perinatal depression. The perinatal period offers a valuable window of opportunity to ask women about their mental health. Our results suggest women should also be asked about support networks available, trauma and past episodes of depression. Future research should focus on testing community-level and public policy interventions to nurture support networks for migrant and refugee women in their resettlement destinations.