Key predictors of postpartum depression and anxiety symptoms among mothers in Kilifi, Kenya: a machine learning approach

The burden of maternal postpartum depression and anxiety is disproportionately high in sub-Saharan Africa (SSA), yet the use of advanced analytical methods to capture the complex interplay of variables influencing these conditions remains underexplored. To apply machine learning (ML) methods to predict depressive and anxiety symptoms in postpartum mothers and to identify key and actionable predictors.

Predicting off-track development in infants aged 0–6 months in low-resource settings using machine learning

The earliest years of life, particularly from birth to age 3, represent a critical window for brain development, characterized by plasticity and sensitivity to environmental influences, both positive and negative. If foundational skills are not adequately nurtured during this sensitive period, children may face a permanent loss of developmental potential. This can lead to long-term negative consequences across the life course, including lower educational attainment, reduced social mobility, limited economic achievement, and poorer health outcomes. One of the aims of the Sustainable Development Goals is to increase the proportion of children under 5 years of age who are developmentally on track in terms of health, learning, and psychosocial well- being.

Depressive Symptoms and Associated Factors Among Kenyan Health Care Workers

The mental health of health care workers (HCWs) is an important global concern, particularly in low- and middle-income countries (LMICs) such as Kenya, where unique stressors may place HCWs at disproportionate risk for depression.¹ However, depression prevalence and risk factors in this population are not well understood. To address this gap, we evaluated the prevalence of depressive symptoms among Kenyan HCWs and identified associated demographic, psychological, and workplace factors.

Data Resource Profile: Harmonisation of a multimodal dataset to evaluate adolescent mental health in rural South Africa

Mental health disorders among adolescents and young adults in Africa are a growing concern, with most cases remaining undiagnosed or untreated due to limited resources. As the youth population increases, mental health issues are expected to rise, emphasising the need for targeted interventions. However, the lack of longitudinal data hinders researchers from understanding how social, behavioural, and clinical factors interact, which is essential for developing effective interventions. This study aimed to create a mental health data resource accessible to mental health researchers across Africa.

Synthetic data generation of health and demographic surveillance systems data: a case study in a low- and middle-income country

This study explored the effectiveness of open-source generative models in generating realistic, privacy-preserving synthetic datasets of a Health and Demographic Surveillance System (HDSS) dataset obtained from rural Kenya, as a proof of concept for low- and middle-income countries (LMICs). Using 3 open-source models—CTGAN, TableGAN, and CopulaGAN—synthetic data were created from the real data while maintaining important characteristics like missing values and outliers. The quality of the synthetic data was evaluated to determine whether it maintained the statistical properties present in the real data and provided privacy protection. Among the models tested, CTGAN performed the best, producing synthetic data that closely resembled the real data while preserving data privacy. In contrast, CopulaGAN and TableGAN were less successful, with TableGAN completely failing to generate realistic synthetic data. This study shows that CTGAN can be a valuable tool for creating synthetic data in LMICs, making real datasets more accessible and preserving privacy.

Application of machine learning in early childhood development research: a scoping review