Furthermore, to date, the approach for finding genes underlying the heritability of cannabis initiation was to focus on the ‘ever/never used’ dichotomy at the expense of the age at which one initiates (i.e., age at onset). Yet, age at onset is a complex trait (Visscher et al. 2001), subject to the influences of both environmental and genetic factors (Lynskey et al. 2003), and may serve as an important proxy for heavy use. Initiation of cannabis use before age 18 is predictive of both experimentation with other drugs (Agrawal et al. 2006; Lynskey et al. 2006), and of escalated drug use (e.g., Lynskey et al. 2003). Among those initiating in adolescence the risk of progression to symptoms of abuse and dependence is higher relative to the general population (i.e., 17 vs. 9 %, respectively; Volkow et al. 2014). Given its relevance as a predictor for escalated use, our second aim is to perform a genomewide search for GVs that give rise to individual differences in age at onset. To model age at onset as a function of genotype we will apply
