Treatment for mental illness often focuses on changing cognitive patterns (for example, cognitive behavioural therapy). There is much evidence to suggest that cognition is different in those with mental illness but whether change in cognition is a causal risk factor has not yet been established. Classic observational studies of cognitive patterns and mental illness do not get around the problems of reverse causation and residual confounding. That is to say that the change in cognition might instead emerge as a result of the mental illness, or both might be the result of other unmeasured factors.

One way to get around this is using Mendelian randomisation. Here we take genetic variants associated with the trait of interest: cognition and use them to naturally randomise individuals to levels of the exposure. Therefore, analogous to a randomised control trial, we can make conclusions about whether or not the relationship is causal. In this study, we will be looking at the cognitive traits of emotion recognition, working memory, cognitive styles and impulsivity. This research could inform the development of cognitive training tasks as interventions for mental illness.