Using a genetically informative design to examine the relationship between breastfeeding and childhood conduct problems

Katherine H. Shelton, Stephan Collishaw, Frances J. Rice, Gordon T. Harold, Anita Thapar

A number of public health interventions aimed at increasing the uptake of breastfeeding are in place in the United States and other Western countries. While the physical health and nutritional benefits of breastfeeding for the mother and child are relatively well established, the evidence for psychological effects is less clear. This study aimed to examine whether there is an association between breastfeeding and later conduct problems in children. It also considered the extent to which any relationship is attributable to maternally-provided inherited characteristics that influence both likelihood of breastfeeding and child conduct problems. A prenatal cross-fostering design with a sample of 870 families with a child aged 4-11 years was used. Mothers were genetically related or unrelated to their child as a result of assisted reproductive technologies. The relationship between breastfeeding and conduct problems was assessed while controlling for theorised measured confounders by multivariate regression (e.g. maternal smoking, education, and antisocial behaviour), and for unmeasured inherited factors by testing associations separately for related and unrelated mother-child pairs. Breastfeeding was associated with lower levels of conduct disorder symptoms in offspring in middle childhood. Breastfeeding was associated with lower levels of conduct problems even after controlling for observed confounders in the genetically related group, but not in the genetically unrelated group. In contrast, maternal antisocial behaviour showed robust associations with child conduct problems after controlling for measured and inherited confounders. These findings highlight the importance of using genetically sensitive designs in order to test causal environmental influences.