Abstract
BACKGROUND: Although cognitive impairment is a recognized feature of schizophrenia (SZ) and bipolar disorder (BD), there is increasing evidence to suggest substantial patient-to-patient variation in cognitive ability that may reflect differences in underlying neurobiology. An expanding body of work has focused on probing this hypothesis by examining brain morphological patterns within subgroups of patients with SZ or BD classified using cognitive clustering approaches. This review aimed to summarise this work, with a view to providing insight into the extent to which brain morphological differences map on to cognitive subgroups within these disorders. METHODS: Science Direct, PubMed, NCBI, and Web of Science databases were searched for papers investigating cognitive subgroups and brain morphology (cortical thickness, surface area and volume) in SZ and BD. RESULTS: Twenty-one studies have investigated the presence of cognitive subgroups and associated brain morphology, 16 focusing on SZ, one on BD, and four examining SZ and BD cross-diagnostically. Within these, the subgroup structure of cognitive heterogeneity was generally similar – anchored by a cognitively near-normal subgroup and a severely impaired subgroup. Generally speaking, brain morphological differences were evident between cognitive subgroups and healthy controls, but not between subgroups themselves. When between-subgroup differences were evident, the magnitude of morphological alterations was typically the distinguishing factor rather than the pattern of brain abnormality. The case - control deficits in patients classified as cognitively impaired were more severe and widespread than those classified as cognitively spared/near-normal. Most studies analyzed brain volume, although the general pattern of findings was similar in the few studies that examined cortical thickness and surface area. DISCUSSION: Relative to controls, brain morphological differences between cognitive subgroups generally appear to reflect variations in the magnitude of brain abnormality rather than unique patterns of abnormality. This coupled with the presence of morphological alterations in cognitively spared/near-normal patients, suggests that brain abnormalities may map more closely to the presence of psychopathology rather than cognition itself.