Determining causality in relation to early risk factors for ADHD: The case of breastfeeding duration

The association between breastfeeding and reduced likelihood of child attention-deficit hyperactivity disorder (ADHD) has been repeatedly shown in observational studies and meta-analyses. However, what is the causal direction of this association? There are several reasons to hypothesize that breastfeeding duration causally affects the potential development of child ADHD, one of them, for example, being that breastfeeding has been associated directly with white matter development in infants and toddlers. Conversely, early temperament in children who will have ADHD may disrupt breastfeeding attempts by the mother, which implies that breastfeeding is more likely to be the result of ADHD. The present study evaluates the causal direction of the ADHD-breastfeeding association using one of the largest, well characterized cross-sectional samples (n=829 children aged 7-13years). Principles of direction dependence analysis (DDA) are applied to make statements about the directionality of the causal flow that generates the association. Results suggest that child ADHD is more likely to be the cause of reduced breastfeeding, and not the other way around. This causal model has been confirmed for two different ADHD measures (parent and teacher-based ADHD ratings) and proved quite robust under various scenarios of covariate adjustment.