Brain Structure and the Prediction of Outcome in First-Episode Schizophrenia-Spectrum Disorders and Affective Psychosis: A Population-Based Study

Abstract

Background: MRI studies of the prediction of outcome among subjects with first-episode psychosis (FEP) have brought misleading evidence to discussion. Cognitive impairment in subjects with psychosis is receiving particular attention, and may be the symptom dimension most associated with outcome in those patients. Methods: Structural MRI on a FEP (N=96) sample with subjects with schizophrenia-spectrum disorders (N=55) and affective psychoses (N=41). Outcome evaluation after a median period of one year consisted on PANSS, cognitive measures (verbal fluency and digit spans) and disability evaluation (WHO-DAS). Lateral ventricles, studied using a region of interest approach, and regional GM, analyzed using VBM with SPM, at the baseline entered statistics as predictive of outcome at the follow-up. Results: Frontal and temporal cortices GM volume were associated with outcome measures, in particular with the performance o cognitive tasks, in the overall FEP Group and in the subgroups (schizophrenia and affective psychoses). Particularly, the affective psychosis subgroup showed more robust associations of GM volumes with outcome measures than the schizophrenia-spectrum subgroup. Furthermore, temporal horns measures were negatively correlated with digit spans’ performances in the FEP group and in the subgroups. None of these findings could be attributed to confounding factors, such as outcome measures at baseline and antipsychotic intake. Conclusions: Brain structure at the moment of the first-episode of psychosis of patients with schizophrenia and subjects with affective psychoses was associated with outcome, particularly cognitive measures. The predominance of findings on frontal-temporal regions is compatible with the presence of a fronto-temporal disconnectivity underlying psychoses. Keyword(s): Schizophrenia, Bipolar Disorder, Neuroimaging, Outcome, Cognition