Statistical modelling of determinants of child stunting using secondary data and Bayesian networks: a UKRI Global...

Area of Science:

• Global Health
• Pediatrics
• Epidemiology

Background:

• Child stunting is a significant global health issue with complex, poorly understood causal pathways.
• Numerous factors are implicated in child stunting, necessitating a comprehensive approach to identify key determinants.

Purpose of the Study:

• To investigate the causal relationships between various determinants and child stunting.
• To elucidate the mechanisms and pathways contributing to child stunting through advanced analytical methods.

Main Methods:

• Utilizing data from national health surveys in India, Indonesia, and Senegal, supplemented by evidence reviews.
• Employing causal Bayesian networks (BN) to model interdependent causal relationships and visualize them in a directed acyclic graph.
• Generating conditional probability distributions to quantify the strength of direct causality between determinants and child stunting.

Main Results:

• The Bayesian network model will provide evidence for the causal role of different determinants in child stunting.
• The analysis will identify critical evidence gaps in the understanding of child stunting.
• The model will facilitate in-depth interrogation of the existing evidence base on child stunting.

Conclusions:

• Bayesian networks offer a robust framework for understanding the complex etiology of child stunting.
• This approach supports the integration of diverse data sources and expert knowledge for a holistic view of stunting determinants.
• The findings will inform targeted interventions and future research directions for child stunting prevention.