For initiation of marijuana use, common environment explained the largest proportion of variance (c2 = 47%), although genetic factors also played a significant role (a2 = 35%). For quantity of marijuana use, both genetic (64%) and unique environmental (36%) factors played a role. The confidence intervals around both these estimates included 0; however, they could not both be dropped from the model simultaneously without a significant deterioration in fit. The beta coefficient was high (α = 0.88), with the upper limit of the confidence interval approaching 1 (upper CI = 0.99), again suggesting considerable overlap of liabilities for initiation and progression. Of the 77% of the variance of progression that was explained by factors influencing initiation, 47% was due to shared environmental factors, 35% by genetic factors and 18% by non-shared environmental factors.
