Living DAGs for Epidemiology: The Example of Cerebral Palsy Etiology

Details

Cerebral Palsy (CP) is an acquired movement disorder that typically results from a perinatal brain injury. However, this definition is greatly oversimplified, as CP has been known to result from heterogeneous causes not only across the pre-, peri-, neo-, and postneonatal periods, but along diverse causal pathways within each. In 2023 an international working group convened with the goal of drawing a comprehensive causal directed acyclic graph (DAG) for CP etiology, adapting a methodology originally developed for NASA’s Human System Risk Board to manage 29 spaceflight health risks. Over the two years that followed, the group constructed an initial DAG for CP, with 106 nodes and 378 edges.
This talk will review the working group’s experience, showcase the DAG, and discuss the uses of this “Living DAG” for communication and scientific inquiry, as well as how this can generalize to many areas in Epidemiology.

Robert J. Reynolds, PhD MS MPH PStat®

Associate Professor of Aerospace Medicine, UCF College of Medicine
Director of Research, Mortality Research & Consulting Inc.

Robert J. Reynolds is a spaceflight epidemiologist and biostatistician specializing in causal inference and small-n analytics. He is Associate Professor of Aerospace Medicine at UCF College of Medicine and Director of Research at Mortality Research & Consulting, Inc. He previously served as Visiting Data Scientist at NASA’s Human Health and Performance Directorate, where he introduced causal diagramming to the agency’s Human System Risk Board. His current work in causal inference focuses on guiding scientific inquiry and interventions using network analytics to gain structural insight from DAGs.