There are also a range of risks, both genetic and environmental, correlated with parental separation that remain unmeasured. It will be important for future research to examine more proximal risks from both parental substance dependence and separation in samples of sufficient size to parse potentially mediated or moderated effects. In the present study, statistical power was limited, including power to examine separately maternal versus paternal substance dependence and, in some cases, alcohol versus cannabis dependence. Lastly, participants are almost exclusively European ancestry, reflecting the predominantly Caucasian population from which twin parents were ascertained. Replication in a more diverse sample is critical, as is cross-national replication.
