Analyses for cocaine and opioids were conducted independently; separate GWAS were performed for each outcome. Although individuals exposed to both opioids and cocaine contributed to both analyses, cocaine exposure/dependence did not inform analyses of opioids and vice versa. Time-to-dependence (TD) was defined as the interval between self-reported age at first opioid/cocaine use and the age at first self-reported endorsement of three or more DSM-IV OD/CD criteria within the same month. Individuals who used opioids or cocaine at least once but never became dependent were censored at their age at interview. No differentiation was made between exposure to or dependence on prescription vs. illicit opioids, but the majority (65%) of users in the Yale-Penn sample reported heroin as their most heavily used opioid.
