A key goal of psychiatric epidemiology is to identify causal risk factors for psychiatric disorders that, when modified through an intervention, will lower incidence of disease and decrease burdens of symptoms in patients with the disease. Observational studies have identified several potential risk factors that are associated with mental illness, but identifying those that are causal rather than associated is challenging because of confounding, selection bias, reverse causation, or pleiotropy (ie, a single gene may influence multiple unrelated phenotypic traits).
Enda M. Byrne, Jian Yang, Naomi R. Wray. Inference in Psychiatry via 2-Sample Mendelian Randomization—From Association to Causal Pathway?. JAMA Psychiatry. Published online November 01, 2017. doi:10.1001/jamapsychiatry.2017.3162