The ENIGMA-EOP collaboration aims to identify structural phenotypic markers that robustly discriminate adolescents with early-onset psychosis (EOP) from healthy controls through mega- or meta-analysis of magnetic resonance imaging (MR) data. Through larger samples we will obtain sufficient power to detect the brain structural correlates, overcome some of the clinical heterogeneity and characterize the developmental trajectories.

Multiple linear regression was used to investigate structural brain differences in two Scandinavian adolescent EOP cohorts (altogether 50 patients; ages 12.1-18.3 years (mean 16.4 years), 60% female; 68 controls; ages 12.0-18.8 years (mean 16.2 years), 62% female) acquired on two different 3 T GE MRI scanners. The statistical analysis included site as a covariate in addition to age, sex and intracranial volume (ICV). The results are presented by p-values, Cohens's-d effect size and with an indication of directionality. MRI scans were processed following the ENIGMA (http://enigma.ini.usc.edu/) structural image processing protocols using FreeSurfer (Fischl 2012) version 5.3.0 to measure subcortical brain volumes.

Preliminary results show significant or trend-significant group effects on right amygdala (P = 0.001, d = 0.33, patients < controls), total grey matter volume (P = 0.037, d = 0.21, patients < controls), ICV (P = 0.028, d = 0.22, patients < controls) and third ventricle (P = 0.067, d = 0.19, patients > controls). Sub-analyses in the two individual groups show overlapping findings in right amygdala. Previously reported enlarged lateral and 4th ventricles, and caudate, from a similar Scandinavian adolescent EOP cohort (Juuhl-Langseth, 2012) were not replicated.

There is a need for larger subject samples in EOP to better capture disease mechanisms. Research groups interested in participating can join ENIGMA-EOP through: http://enigma.ini.usc.edu/ongoing/enigma-eop-working-group/.

The authors have not supplied their declaration of competing interest.