Kristine Krakauer



Exploring clinical, cognitive, and brain structural predictors of ´risk resolution´ in individuals at ultra-high risk for schizophrenia

Intensive research is being done in order to identify young individuals at risk of developing schizophrenia since early intervention may delay or even prevent overt disease. Specific criteria have been developed that combine subtreshold- or attenuated psychotic symptoms, genetic predisposition and a significant decline in functioning in order to identify individuals at ultra-high risk (UHR) of developing a psychosis. 

Early studies in these UHR individuals found that 30-50 % developed a psychosis within 1-2 years. However, more recent studies have reported a steady decline in transition rates to around 12 %. 

Until now no studies have investigated the large group of individuals who do not progress to psychosis but who remain in UHR or whose symptoms resolve. Identification of biological or functional predictors of ‘risk resolution’ in these individuals may serve as refined tools of risk stratification within the group of high risk patients. Ultimately, such tools may facilitate early and personalized intervention targeted towards individuals who are at continuous risk for developing schizophrenia.

We will try to characterize UHR individuals by testing cognitive, olfactory and brain structural domains that are known to be impaired in schizophrenia patients, and compare them to healthy controls. 

We will also try to find predictors of risk resolution in UHR individuals by correlating changes in tensor based morphometry and risk status at one year follow-up.  

Additionally we will use multivariate analysis (a support vector machine) to try to identity patterns of clinical, cognitive and brain structural markers, which can predict the risk resolution in UHR individuals at an individual level.

So far 40 individuals have been included and tested with MRI-scans, psychometric- and cognitive tests. One year follow up has been done on 20 UHR individuals. 

Responsible editor