The aim of this paper was to investigate differences in brain structure volumes between schizophrenia and affective psychoses, and whether cumulative lifetime antipsychotic or benzodiazepine doses relate to brain morphology in these groups. We conducted two systematic reviews on the topic and investigated 44 schizophrenia cases and 19 with affective psychoses from the Northern Finland Birth Cohort 1966. The association between lifetime antipsychotic and benzodiazepine dose and brain MRI scans at the age of 43 was investigated using linear regression. Intracranial volume, sex, illness severity, and antipsychotic/benzodiazepine doses were used as covariates. There were no differences between the groups in brain structure volumes. In schizophrenia, after adjusting for benzodiazepine dose and symptoms, a negative association between lifetime antipsychotic dose and the nucleus accumbens volume remained. In affective psychoses, higher lifetime benzodiazepine dose associated with larger volumes of total gray matter and hippocampal volume after controlling for antipsychotic use and symptoms. It seems that in addition to antipsychotics, the severity of symptoms and benzodiazepine dose are also associated with brain structure volumes. These results suggest, that benzodiazepine effects should also be investigated also independently and not only as a confounder.

No abstract

IntroductionThe effects of long-term antipsychotic medication use on structural brain changes in psychoses are still unknown. Severity and duration of illness are key confounders when evaluating antipsychotic effects on brain morphology.ObjectivesUnderstanding the role of antipsychotic medication on brain morphology in psychoses.AimsTo analyze whether cumulative lifetime or current antipsychotic medication dose relates to brain morphology in schizophrenia and other psychoses at age of 43 years.MethodsForty-four schizophrenia cases and 35 with other psychoses from the Northern Finland Birth Cohort 1966 were scanned on a 1.5T GE Signa scanner and brain structures were extracted using volBrain automated volumetry system (http://volbrain.upv.es). Data of antipsychotic medication were collected from medical records and interviews. We used linear regression model to analyze the effect of antipsychotic medication on brain volumes and used intracranial volume and onset age as covariates. We also performed additional analyses adding psychotic symptoms (PANSS Total score) as a covariate.ResultsHigher lifetime and current dose associated to left lateral ventricle increase (b = 0.33, P = 0.033; b = 0.307, P = 0.042, respectively) and right and left accumbens decrease (b = −0.405, P = 0.013, b = −0.404, P = 0.010; b = −0.302, P = 0.027, b = −0.282, P = 0.036, respectively) in schizophrenia but not in other psychoses. When PANSS was added to the model, the findings remained regarding right and left accumbens, but not regarding left lateral ventricle.ConclusionsIt seems that antipsychotic medication affects the brain in schizophrenia, but not in the heterogeneous group of other psychoses. In schizophrenia, brain changes associated to antipsychotic medication cannot be explained by illness duration or symptom severity.Disclosure of interestThe authors have not supplied their declaration of competing interest.