A Genome Wide Association Study of Psychosis Endophenotypes
Robin Murray
John Powell, Cathryn Lewis, Maria J Arranz, Dan Rujescu, Rene S Kahn, Steven C Bakker, Jim Van Os, Durk Wiersma, Roel A Ophoff, Don Linszen, Andrew McIntosh, Jeremy Hall, Shitij Kapur, David Collier, Stephen Lawrie.
Psychotic disorders affect 2% of the population and are within the seven leading causes of disability (WHO). Although psychosis is highly heritable and neuregulin 1 and dysbindin are promising candidate genes, no unambiguous risk variant(s) have been identified yet. Psychiatric diagnoses do not constitute optimal phenotypes for genetic research and we propose using endophenotypes as an alternative. Endophenotypes are quantitative biological traits characterising psychosis which are obtained by laboratory-based methods rather than clinical observations.
From twin and family studies we selected the most suitable psychosis endophenotypes:
- cognitive deficits affecting verbal memory, digit span and block design performance,
- neurophysiologic deficits in the P300 wave from the human electroencephalogram, and
- neuro-anatomical abnormalities in whole brain and lateral ventricular volumes.
Through our European collaboration we obtained DNA and clinical diagnoses from 7,098 individuals [1,470 patients, 2,515 relatives and 3,113 controls]. Of these a large proportion had their cognitive, neuro-anatomic and neurophyisologic endophenotypes ascertained. The USA Consortium On the Genetics of Schizophrenia, a comparable independent study, provides an opportunity for replications.
Understanding the relationship between susceptibility genes and endophenotypes of psychosis will help to clarify the pathophysiological mechanisms by which these genes have their effects. Ultimately, elucidation of susceptibility genes and their mechanism of action will advance the understanding and treatment of psychosis.