For analyses involving potential moderation of the effect of ALDH2 polymorphism by elder sibling alcohol use (gauged by the elder sibling combined alcohol problem index) and peer deviance (table S3), results were less consistent than those involving moderation by parental alcohol use. These inconsistencies may reflect limitations in the measurements available for elder sibling alcohol use and peer deviance. Notably, since not every participant had an older sibling, fewer observations were available for analyses assessing the effect of elder sibling drinking, reducing power to detect moderating effect upon the ALDH2 polymorphism. Moreover, although alcohol-related peer influences have previously been shown to moderate the heritability of alcohol use (Dick et al., 2007; Agrawal et al., 2010), it is possible that measures of adopted familial alcohol use may exert a more readily apparent moderating influence upon ALDH2 genotype effect than peer deviance, because they are specific measures of environmental exposure to alcohol use, whereas the scale of deviant peer behavior that we used reflected the proportion of peers engaging in a wider variety of externalizing behaviors. Because this ALDH2 polymorphism influences alcohol
