There has also been a great deal of interest in delineating whether and how environmental factors moderate measured genetic predispositions for conduct disorder. In an early example of this, Caspi and colleagues (2002) showed that genetic variation in MAOA interacted with harsh physical discipline to predict antisocial behavior. Although there is much enthusiasm for this type of approach (Caspi and Moffitt, 2006), the study of candidate gene-by-environment interaction (cG × E) in psychiatry has also been met with skepticism and criticism out of concern for Type I error (Duncan and Keller, 2011), particularly in view of the difficulty in identifying genes and genetic variants associated with psychiatric outcomes such as conduct disorder (Dick et al., 2015). There are many studies of cG × E for conduct disorder (Veroude et al., 2015). However, given the controversy surrounding cG × E research and the concern for Type I error, we review here selected examples that have used a hypothesis-driven approach to select the environment, and that have built upon large-scale gene identification results to select the relevant genes/genetic variants.
