A few findings emerged from our analyses that are inconsistent with prior studies. First, we note that the correlations between parental antisocial behavior and alcohol problems measures were unexpectedly low in the CDP sample (r = 0.15). This association not as strong as other reports in the literature (e.g., Krueger et al., 2002). We speculate that the low correlation may be attributable to restriction of range in the CDP parental antisocial behavior and alcohol problems measures, which can attenuate correlations. Second, we did not find evidence for main effects of GABRA2 on adolescent externalizing in either the CDP or FinnTwin12 samples. This could be attributable to a couple of factors. Genetic associations for GABRA2 tend to be more robust in clinically-ascertained samples compared to community-based samples (Irons et al., 2014). Previous work in the CDP sample has found evidence for association between GABRA2 and an elevated persistent trajectory of externalizing behavior across adolescence and emerging adulthood (Dick et al., 2009). However, that study used a latent class approach to test for genetic association with a relatively extreme phenotype; in contrast,
