Comment No (Score 4, Informative) 217
I don't think so. The JAMA article http://archinte.jamanetwork.co... does look at longitudinal effects but the 25% figure comes from comparing states with and without. From the abstract:
States with medical cannabis laws had a 24.8% lower mean annual opioid overdose mortality rate (95% CI, 37.5% to 9.5%; P=.003) compared with states without medical cannabis laws.
The common way to statistically analyse the effect of one variable is to model as many variables as the data allows and run a regression to isolate the effect of the target variable.
It may be that there are other problems with the study (e.g. correlations between the variables assumed to be independent) but this isn't one of them.