Abstract
BACKGROUND: Longitudinal studies are needed to examine the association between maternal depression, trauma and childhood mental health in conflict-affected settings. AIMS: To examine maternal depressive symptoms, trauma-related adversities and child mental health by using a longitudinal path model in conflict-affected Timor-Leste. METHOD: Women were recruited in pregnancy. At wave 1, 1672 of 1740 eligible women were interviewed (96% response rate). The final sample comprised 1118 women with complete data at all three time points. Women were followed up when the index child was aged 18 months (wave 2) and 36 months (wave 3). Measures included the Edinburgh Postnatal Depression Scale, lifetime traumatic events and the Child Behaviour Checklist. A longitudinal path analysis examined associations cross-sectionally and in a cross-lagged manner across time. RESULTS: Maternal depressive symptom score was associated with child mental health (cross-sectional association at wave 2, β = 0.35, P < 0.001; cross-sectional association at wave 3, β = 0.33, P < 0.001). The maternal depressive symptom score at wave 1 was associated with child mental health at wave 2 (β = 0.12, P < 0.001), and the maternal depressive symptom score at wave 2 showed an indirect association with child mental health at wave 3 (indirect standardised coefficient 0.23, P < 0.001). There was a time-lagged relationship between child mental health at wave 2 and maternal depression at wave 3 (β = 0.08, P = 0.02). CONCLUSIONS: Maternal depressive symptoms are longitudinally associated with child mental health, and traumatic events play a role. Maternal depression symptoms are also affected by child mental health. Findings suggest the need for skilled assessment for depression, trauma-informed maternity care and parenting support in a post-conflict country such as Timor-Leste.