"However, the study authors caution that their analysis doesn’t account for health attitudes in different states that might explain the association." As the ar

"However, the study authors caution that their analysis doesn’t account for health attitudes in different states that might explain the association." As the article suggests, correlation does not equal causation. Looking at this map: https://en.wikipedia.org/wiki/File:Map-of-US-state-cannabis-laws.svg https://en.wikipedia.org/wiki/File:Map-of-US-state-cannabis-... one can imagine numerous other reasons that might lead to differential rates of painkiller (ab)use between states. Socioeconomic status is fairly clear to me (states with higher status are more likely to have medical marijuana), other correlations are a bit more of a can of worms.