Identifying neuroanatomical subtypes of psychosis

Background: The quantity and heterogeneity of structural magnetic resonance imaging(sMRI) studies in schizophrenia present a challenge to their meaningful interpretation.This thesis sought to synthesise the available literature on neuroanatomical alterations inpeople with schizophrenia, and explore brain volume differences among subgroups ofschizophrenia defined on the basis of cognitive ability. The thesis also investigatedwhether analogous cognitive subgroups extending across the psychosis-mood spectrum(including bipolar-I disorder) may be differentiated by brain volume.Methods: Meta-review of the existing sMRI literature was conducted using electronicdatabases. Random-effects meta-analysis was used to statistically collate studiesreporting insular cortex volume in schizophrenia. Subsequent analysis of structuralimaging data from two separate samples explored differences in grey and white mattervolume among cognitive subtypes of psychotic disorder relative to controls. Subgroupswithin a large schizophrenia sample were allocated to cognitively spared (CS;; N=157)or cognitively deficit (CD;; N=106), compared to healthy controls (N=181).Subsequently, a mixed sample of schizophrenia and bipolar-I disorder was stratifiedinto executively spared (ES;; N=38) and executively deficit subgroups (ED;; N=32) toassess differences in grey and white matter volume.Results: Critical synthesis of the sMRI literature identified limited high qualityevidence for widespread regional grey and white matter aberration in schizophrenia.Specific assessment of insular cortex identified moderate volume reductions, with largeheterogeneity not explained by age, sex, illness duration, medication, whole brainvolume, or hemisphere. Assessment of cognitive subtypes of schizophrenia identifieddistinct neuroanatomical profiles of each group relative to controls, with the cognitively impaired group showing more diffuse brain volume reductions. Assessment of thediagnostic specificity of these findings identified no differences between groups definedby bipolar-I disorder and schizophrenia diagnoses, but similarly diffuse volume deficitsacross the psychosis-mood spectrum in the executively impaired subgroup.Conclusions: Stratification according to intermediate cognitive phenotypes identifiedtrans-diagnostic subtypes that exhibited markedly discrepant brain aberrations andclinical features. These findings provide insight into potential neuropathologicalregions, and suggest the existence of discrete illness types who may be associated withdivergent neurodevelopmental or genetic vulnerabilities