The study is also highly sensitive to the way drug use disorders are classified by behavioral health professionals, particularly due to the changing criteria of various DSM editions. I’d want to look at the underlying dataset to see if they’ve properly accounted for exogenous effects like that.