Substance use and substance abuse occur at much higher levels in populations with mental health problems than in the general population, but the reasons for this are hard to untangle. Does poor mental health cause substance use, does substance abuse increase the risk of mental health problems, or are both affected by earlier life experiences? In order to try and untangle causality, methods such as Mendelian randomization can be used, where genetic predictors are used as proxies for confounded associations.