Directed acyclic graphs for causal inference of the impact of maternal and child helminth infections on child growth

Date

2026-05-01

Author

Gabain, Isobel L.
Davies-Kershaw, Hilary
Rosenstock, Todd S.
Yet, Barbaros
Rudge, James W.
Webster, Joanne P.

Metadata

Show full item record

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Item Usage Stats

55
views

0
downloads

Cite This

As of 2025, an estimated 150.2 million children under 5 years globally were stunted-falling more than two standard deviations below the World Health Organization (WHO) Child Growth Standards median. Helminth infections represent a potentially significant yet under-addressed contributor to childhood stunting. Our aim was to elucidate the potential causal relationship between helminth infections and childhood stunting, including the mediating role of low birth weight (LBW). Three causal directed acyclic graphs (DAGs) were constructed using evidence synthesis for constructing DAGs (ESC-DAGs) methodology: (i) maternal infection, (ii) infant infection during exclusive breastfeeding (<6 months old), and (iii) child infection post-weaning (6-24 months old). Minimally sufficient adjustment sets (MSAS) were derived from our DAGs. All three DAG models included core adjustments for geographical location, healthcare access and water, sanitation and hygiene factors. The maternal infection model additionally incorporated maternal age, socioeconomic status (SES) and maternal education as key confounders. For the child models, both age groups included adjustments for child sex, age, birth weight and parental education. The model for infants <6 months of age additionally incorporated breastfeeding status, while the model for children post-weaning (6-24 months of age) included SES alongside the other shared variables. These MSAS provide researchers with standardized confounder adjustment frameworks, helping to resolve inconsistencies in prior literature stemming from incomplete confounder control and advancing methodologically rigorous causal inference.

URI

https://hdl.handle.net/11511/119345

Collections

Graduate School of Informatics, Article