IEPA 11

Chantal Michel1,2, Nina Schnyder1, Petra Walger3, Maurizia Franscini4, Benno G. Schimmelmann1,5, Frauke Schultze-Lutter1,6; 1University Hospital of Child and Adolescent Psychiatry and Psychotherapy, University of Bern, Switzerland, 2Developmental Clinical Psychology Research Unit, Faculty of Psychology and Educational Sciences, University of Geneva, Switzerland, 3University Hospital of Child and Adolescent Psychiatry and Psychotherapy, Cologne, Germany, 4University Hospital of Child and Adolescent Psychiatry, Zurich, Switzerland, 5University Hospital of Child and Adolescent Psychiatry, University Hospital Hamburg-Eppendorf, Hamburg, Germany, 6Department of Psychiatry and Psychotherapy, Medical Faculty, Heinrich-Heine-University, Düsseldorf, Germany
               
In the early detection of psychosis, neurocognitive predictors have been suggested to enhance predictive accuracy of clinical high risk (CHR) criteria. While mainly sample-dependent means of adult samples were used so far, a recent study of an adult sample used neurocognitive deficits defined according to test norms in order to facilitate individual prediction. Yet, data on child and adolescent samples are missing. We investigated the discriminative power of neurocognitive deficits defined according to norms in 8-17-year-olds. 160 CHR outpatients (AtRisk; mean age=15.02±2.20, 39% male), 270 non-psychotic inpatients (ClinS; mean age=14.46±2.43, 38% male) and 220 subjects of a general population sample (GPS; mean age=13.91±2.78, 48% male) had been assessed with a neurocognitive battery, including a verbal fluency (VF) test, the Digit-Symbol Test, TMT A and B, the Auditory Verbal Learning Test (AVLT) and the Subject Ordered Pointing Task. GPS were slightly younger than AtRisk and ClinS (Chi2(2)=7.656, p=0.022); no differences were found with regard to gender and premorbid IQ. Compared to ClinS and GPS, AtRisk more frequently exhibited deficits according to norms in verbal memory (AVLT learning capacity; 22.4% vs. 10.7%; OR=2.4, 95% CI: 1.3-4.6) and VF (48.8% vs. 34.1%; OR=1.8, 95% CI: 1.1-3.0), while ClinS and GPS did not differ. Partly in line with findings from adult samples, deficits in verbal memory and VF might be specifically associated with a CHR state in children and adolescents – even when compared to a more severely ill inpatient group. Yet, these findings need further examination in larger samples and longitudinal studies.