The risk of attention deficit hyperactivity disorder in children exposed to maternal smoking during pregnancy - a re-examination using a sibling design.

Background: Conventional cohort studies have consistently shown that exposure to maternal smoking in pregnancy is associated with about twice the risk of attention deficit hyperactivity disorder (ADHD) in the offspring. However, recent studies using alternative designs to disentangle the effect of social and genetic confounders have suggested that confounding may account for the association. In this study we aimed to estimate the association by a sibling design.

Methods: We used a design with half and full siblings in a Danish national register-based cohort on all singletons born between January 1991 and December 2006 and followed until January 2011. Data were available for 90% (N = 968,665) of the singleton live births in the period. We used the combination of the International Classification of Diseases (10th version) diagnosis of hyperkinetic disorder (HKD) and ADHD medication to identify children. We used sibling-matched (conditional) Cox regression to control social and genetic confounding.

Results: Using conventional cohort analyses, we found the expected association between pregnancy smoking and offspring ADHD (adjusted HR 2.01, 95% CI 1.94-2.07). In the sibling analysis, however, we did not detect such a strong association (adjusted HR 1.07, 95% CI 0.94-1.22). There was no difference between results for half- and full sibling analyses. The link between pregnancy smoking and low birth weight remained robust in the sibling design (adjusted OR 1.68, 95% CI 1.33-2.12).

Conclusions: We found no support for prenatal smoking as a strong causal factor in ADHD. Our findings suggest that the strong association found in most previous epidemiological studies is likely to be due to a strong link between maternal smoking and maternal ADHD genetics or shared family environment. Pregnant women should still be encouraged to stop smoking because of other risks, but we have no reason to believe that this would reduce the risk of ADHD in the offspring.