Maternal Smoking during Pregnancy and Offspring Severe Mental Illness: Evidence of Familial Confounding

Population-based studies have demonstrated associations between exposure to maternal smoking during pregnancy and offspring bipolar disorder and schizophrenia. Although these findings could be interpreted to support the... [ view full abstract ]

Population-based studies have demonstrated associations between exposure to maternal smoking during pregnancy and offspring bipolar disorder and schizophrenia. Although these findings could be interpreted to support the hypothesis that maternal smoking during pregnancy causally increases risk of severe mental illness in offspring, such studies may have been unable to fully rule out the possibility of familial confounding. At least some of the covariation between smoking and schizophrenia is explained by familial factors, suggesting that transmission of shared liability for both outcomes (e.g., via passive gene-environment correlation) could confound a putative environmental impact of exposure to maternal smoking. Indeed, one prior sibling comparison study failed to support an association between exposure to smoking during pregnancy and offspring schizophrenia spectrum disorders after accounting for factors shared by siblings (Meier et al., 2017, American Journal of Psychiatry, 174, 187). In the present study, we used family-based comparisons and Swedish population register data on 1,680,219 offspring born 1983-2001 to examine associations between maternal self-reported smoking during pregnancy and several indices of severe mental illness drawn from clinical diagnoses. Replicating prior population-based studies, offspring with moderate (hazard ratio [HR], 1.25; 95% CI, 1.19-1.30) and high (HR, 1.51; 95% CI, 1.44-1.59) maternal smoking during pregnancy exposure had greater rates of severe mental illness than did unexposed offspring. In contrast, however, sibling comparisons were weaker in magnitude and did not reach statistical significance (moderate HR, 1.09; 95% CI, 0.94-1.26; high HR, 1.14; 95% CI, 0.96-1.35). Cousin comparisons, associations with specific diagnoses (e.g., bipolar disorder, schizophrenia spectrum disorders), and other sensitivity analyses supported the overall pattern of results. These findings provide evidence that familial confounding explains much of the population association between maternal smoking during pregnancy and severe mental illness in offspring.