Substance abuse and criminality in the offspring of substance abusing parents: Children-of-siblings analyses in nationwide register data

Parental substance abuse (SA) is associated with elevated risks for substance use problems and antisocial behavior in the offspring. The children-of-twins design has been used to distinguish direct parental effects from... [ view full abstract ]

Parental substance abuse (SA) is associated with elevated risks for substance use problems and antisocial behavior in the offspring. The children-of-twins design has been used to distinguish direct parental effects from genetic confounding but studies have been based on just two different datasets of twin parents and relatively small samples. Using data from Swedish nationwide registries, we conducted children-of-siblings analyses in offspring of monozygotic and dizygotic twin, full sibling, and half-sibling parents. Parental SA was identified from medical and legal registries as diagnoses of alcohol/drug use disorders and alcohol/drug related criminal convictions. Offspring SA and any criminal convictions were identified from the same registries. Individual-level analyses were conducted among nearly 2.5 million offspring, born 1958-1995, of whom 15% had paternal and 5% had maternal SA registrations. Both maternal and paternal SA more than doubled the risk for offspring SA (maternal hazard ratio [HR]=2.81 [95% CI: 2.77-2.86]; paternal HR=2.51 [2.49-2.54]) and criminality (maternal HR=2.20 [2.17-2.23]; paternal HR=2.07 [2.05-2.09]). Associations were attenuated in stratified analyses among full cousins suggesting genetic confounding. Using data from cousins and siblings from the different extended families (n=1.6 million) we conducted quantitative genetic structural equation modeling to estimate the proportions of variance and covariance for parental SA and the offspring outcomes. Additive genetic effects were estimated to explain 42% (95% CI: 28-56%) and 46% (36-55%) of the variance in maternal and paternal SA, respectively, and between 37% (31-44%) and 53% (47-58%) of the variance in the offspring outcomes. Only additive genetic effects significantly contributed to the intergenerational associations between paternal SA and the offspring outcomes as well as maternal SA and offspring criminality. In contrast, the association between maternal and offspring SA was due to genetic (54% [95% CI: 25-84%]), extended family environmental (17% [2-31%]) and nuclear family environmental (29% [12-45%]) components, suggesting also potential environmental transmission.