Our regression analyses take into account change in both Colorado and NMMS states prior to and during the time of medical marijuana commercialization. In those analyses, we also controlled for demographic factors which might differ between states and contribute to state-level differences. Those results support the notion that in Colorado changing perception of risk (12-17 year olds, p=0.005; those 26 years and older, p=0.01), availability of marijuana (those 26 years and older, p=0.07) and past-year marijuana use disorder prevalence (12-17 year olds, p=0.07), show at least trend deviation from the temporal patterns seen in NMMS.
