Repository logo

Examining the factors that moderate and mediate the effects on depression during pregnancy and postpartum



Journal Title

Journal ISSN

Volume Title




Degree Level



Background: Antenatal depression is relatively a new area of study compared to postnatal depression and the depth and sophistication of this research is yet developing. For instance, very little is reported on the specific role of the risk factors as moderators and mediators to explain the variability in the magnitude of exposure and the causal pathway for depression during pregnancy. Moderators are those variables that are not modifiable (e.g., ethnicity, and gender), or have qualitative character or nominal in nature, and could also often be antecedent to other independent variables (e.g., behavioural and psychosocial) and depression. Mediators are those variables that may be better able to describe the pathway that connects a predictor to an outcome and intervention can be designed targeting mediators as they are causally related to the outcome. This thesis will address this gap in research and provide empirical evidence to increase the understanding of the role of each identified risk factors that could potentially influence maternal mental health interventions. Methods: In this thesis, I have used the Feelings in Pregnancy and Motherhood (FIP) study. This was a longitudinal study and 649 pregnant women participated in the study. Women were interviewed three times over the course of their pregnancy and the immediate postpartum. Depression status was assessed by the Edinburgh Postnatal Depression Scale (EPDS); sociodemographic characteristics, psychosocial and behavioural information were collected at each time point. Depression status in late pregnancy and postpartum were the two outcomes of interest. Non-modifiable sociodemographic risk factors were considered as moderators. Behavioural and psychosocial variables were considered as mediators. Moderators and mediators were tested through series of regression analysis. Results: In modeling moderating effects in late pregnancy, low income women who were in poor marital relationships (β=1.54; p<0.05) and partnered women (married or common law) who reported having used recreational drugs (β= -1.62; p<0.05) were more likely to be depressed. Young mothers with low social support (β= 1.04; p=0.15) and Aboriginal mothers with low social support (β= 1.12; p=0.17) were also almost significantly noted to have depression symptoms in late pregnancy. In mediating analysis for late pregnancy, psychosocial mediators such as stress, social support, and marital satisfaction, and behavioural factors, such as smoking and recreational drug use exerted partial or full mediating effect for depressive symptoms in women in late pregnancy. In moderating analysis for postpartum, Aboriginal women who had never exercised in late pregnancy were found to be depressed at postpartum compared with non Aboriginal mothers who did not exercise. In looking at mediating effects in postpartum, smoking at late pregnancy exerted full mediating effects for ethnicity and marital satisfaction pathways, and partial mediating effects for age, education, and stress pathways in predicting depression in postpartum period among mothers. Conclusion: Depression, particularly during pregnancy and in postpartum, is a top priority for women themselves, their families, care providers, and society in general. This study found that characteristics of women or their psychosocial or behavioural experiences could have specific effects such as either a mitigating or exacerbating role, or a mediating role, in depression in late pregnancy or in postpartum. This information could be strategically used by clinicians or by health promotion professionals to either target or provide tailored programs to women who might experience depression during pregnancy and postpartum.



Moderating role, Mediating role, Antenatal depression, Postpartum depression, Maternal depression, Longitudinal study



Master of Science (M.Sc.)


Community Health and Epidemiology


Community and Population Health Science


Part Of