Cortical thickness reduction in individuals at ultra-high-risk for psychosis

Abstract

Although schizophrenia is characterized by gray matter (GM) abnormalities, particularly in the prefrontal and temporal cortices, it is unclear whether cerebral cortical GM is abnormal in individuals at ultra-high-risk (UHR) for psychosis. We addressed this issue by studying cortical thickness in this group with magnetic resonance imaging (MRI). We measured cortical thickness of 29 individuals with no family history of psychosis at UHR, 31 patients with schizophrenia, and 29 healthy matched control subjects using automated surface-based analysis of structural MRI data. Hemispheric mean and regional cortical thickness were significantly different according to the stage of the disease. Significant cortical differences across these 3 groups were found in the distributed area of cerebral cortices. UHR group showed significant cortical thinning in the prefrontal cortex, anterior cingulate cortex, inferior parietal cortex, parahippocampal cortex, and superior temporal gyrus compared with healthy control subjects. Significant cortical thinning in schizophrenia group relative to UHR group was found in all the regions described above in addition with posterior cingulate cortex, insular cortex, and precentral cortex. These changes were more pronounced in the schizophrenia group compared with the control subjects. These findings suggest that UHR is associated with cortical thinning in regions that correspond to the structural abnormalities found in schizophrenia. These structural abnormalities might reflect functional decline at the prodromal stage of schizophrenia, and there may be progressive thinning of GM cortex over time.

Fig. 1.

Mean Cortical Thickness in 3 Groups for Each Hemisphere. The plots show gradual decreases in the mean cortical thickness according to the psychotic stages. UHR, ultra-high-risk (n = 29); schizophrenia (n = 31); HC subjects, healthy control subjects (n = 29).

Fig. 2.

Regional Maps of the Differences in the Mean Cortical Thickness (A) and Statistical Maps Between Groups: HC Subjects vs UHR Subjects, HC Subjects vs Schizophrenia Subjects, and UHR Subjects vs Schizophrenia Subjects (B). UHR, ultra-high-risk (n = 29); schizophrenia (n = 31); HC subjects, healthy control subjects (n = 29).

Fig. 3.

Brain Regions with Significant Reduced Cortical Thickness in UHR Subjects Compared with HC Subjects. In the left hemisphere, the cortical thickness was decreased in superior temporal gyrus (L1), anterior cingulate cortex (ACC, L2), parahippocampal cortex (L3), and medial superior frontal cortex (L4). In the right hemisphere, the cortical thickness was reduced in dorsal ACC (R1), rostral ACC (R2), inferior frontal cortex (R3), and inferior parietal cortex (R4). All these regions showed statistically significant differences among 3 groups. In the boxplot, the x-axis shows brain regions with significant reduction of cortical thickness, and the y-axis indicates cortical thickness (millimeters). The horizontal thick line inside each box indicates the median value. The upper and lower boundaries of each box mean lower quartile and upper quartile values, respectively. The whiskers represent smallest and largest nonoutlier observations. The circle depicts a mild outlier, and asterisk indicates an extreme outlier. Scatterplots and age correlation slopes for mean cortical thickness (millimeters) with increasing age within regions for UHR and HC subjects. The x-axis indicates age (y), and the y-axis indicates cortical thickness (millimeters). Significant correlations between age and mean cortical thickness within each region are marked (asterisk). UHR, ultra-high-risk (n = 29); schizophrenia (n = 31); HC subjects, healthy control subjects (n = 29); ACC, anterior cingulate cortex.