Prevalence of substance involvement and average ages of onset are shown in Table 1, separately by race/ethnicity. For all substance use variables except cannabis, AA twins were at reduced likelihood of initiation, compared to EA twins (log-rank test p<0.05). By twins’ age 18, 75% of AA parents and 38% of EA parents had separated, with AA twins experiencing earlier time to parental separation relative to EA twins (X21=439.06, p<0.0001). Either or both parents were more likely to be coded positive for alcoholism in AA compared to EA families (42% versus 35%, X21=0.42, p<0.01). Log-rank tests of differences in substance involvement by zygosity were significant in EA families only, such that DZ twins were at increased likelihood of initiation across substance class, compared to MZ twins (p<0.05). While zygosity was unrelated to parental separation in either AA or EA families, more EA parents of DZ twins were coded positive for parental alcoholism than EA parents of MZ twins (X21=4.38, p<0.05). Given differences by zygosity in both timing of substance involvement and parental alcoholism, zygosity was included as an additional covariate in adjusted Cox models.
