Unaddressed perinatal mental health issues, such as anxiety and depressive symptomatology, are significant public health concerns and can have detrimental consequences for mother and child. Early identification of women at risk for perinatal mental health disorders is essential for effective intervention, yet screening tools remain limited. Less is known about positive mental health as well. Hence, this study aims to identify predictors of poor and positive perinatal mental health. We harmonized data from 4 international multi-ethnic cohorts comprising over 1800 women and split them into 7:3 training-testing sets. We used a semi-supervised machine learning algorithm to construct a prediction tool identifying high anxiety (STAI ≥ 40) and depressive symptoms (EPDS ≥10) and positive mental health during pregnancy. Tenfold cross-validation was employed to tune the model. An additional 132 preconception data from pregnant women were used to evaluate model performance. The gradient boosting model (GBM) demonstrated the best predictive performance for both high depressive and anxiety symptoms (AUC= 0.840 and 0.883 respectively). Partner connection, sleep disturbances, and maternal age were found to be common predictors across models. Additionally, high depressive symptoms were predicted by paternal attachment, while childhood trauma and affective disposition predicted high anxiety symptoms. The top predictors of positive mental health were partner connection, childhood trauma, maternal attachment, sleep disturbances, and affective disposition. Findings demonstrated that the validated models were capable of identifying both risks and protective factors for perinatal maternal mental health, bringing valuable insights for perinatal care planning.

Authors:
Santhi Ponmudi, Institute for Human Development and Potential, Agency for Science, Technology and Research, Singapore
Shi Ying Ng, National University of Singapore, Singapore
Roscoe Lai, Institute for Human Development and Potential, Agency for Science, Technology and Research, Singapore
Helen Chen, KK Women’s and Children’s Hospital, Singapore
Tina Montreuil, McGill University, Canada
Tuong Vi Nguyen, McGill University, Canada
Sylvana M. Côté, University of Montreal, Canada
Kieran J. O’Donnell, Yale School of Medicine, United States
Shiao Yng Chan, National University Hospital, Singapore
Michael J. Meaney, McGill University, Canada
Michelle Z.L Kee, Institute for Human Development and Potential, Agency for Science, Technology and Research, Singapore

About the Presenter(s)
Santhi Ponmudi is currently a Senior Research Officer at A*STAR IHDP in Singapore. Her current research interests is in investigating the complexities of perinatal mental health.

See this presentation on the full schedule – Tuesday Schedule

Bookmark