Causal diagrams for research about childhood-onset disabilities

Directed acyclic graphs (DAGs) are increasingly used to clarify assumptions, identify sources of bias, and structure reasoning about causal pathways across the health sciences. In developmental medicine, where causes often span the preconception to postnatal periods, DAGs offer a systematic way to navigate complexity. This review introduces foundational DAG concepts for clinicians and researchers in childhood-onset disability, with an emphasis on accessibility and applied relevance. We review examples involving cerebral palsy, autism, and attention-deficit/hyperactivity disorder, showing how DAGs support confounder control, effect estimation, and study design. The figures throughout the review use a consistent, clinically grounded example to walk readers through concepts like mediation, backdoor paths, and collider bias. Beyond modeling rigor, DAGs help foster collaboration across disciplines and communicate causal structure to families and individuals with lived experience. We also show how DAGs can support intervention prioritization by identifying strategic leverage points using network measures such as node centrality and graph characteristics. Finally, we emphasize the importance of drawing DAGs before data collection, when their guidance is most actionable.