Reduced Cortical Thickness in Schizophrenia and Schizotypal Disorder

Abstract

Schizotypal disorder is characterized by odd behavior and attenuated forms of schizophrenic features without the manifestation of overt and sustained psychoses. Past studies suggest that schizotypal disorder shares biological and psychological commonalties with schizophrenia. Structural magnetic resonance imaging (MRI) studies have demonstrated both common and distinct regional gray matter changes between schizophrenia and schizotypal disorder. However, no study has compared cortical thickness, which is thought to be a specific indicator of cortical atrophy, between schizophrenia and schizotypal disorder. The subjects consisted of 102 schizophrenia and 46 schizotypal disorder patients who met the International Classification of Diseases, 10th edition criteria and 79 gender- and age-matched healthy controls. Each participant underwent a T1-weighted 3-D MRI scan using a 1.5-Tesla scanner. Cortical thickness was estimated using FreeSurfer. Consistent with previous studies, schizophrenia patients exhibited wide-spread cortical thinning predominantly in the frontal and temporal regions as compared with healthy subjects. Patients with schizotypal disorder had a significantly reduced cortical thickness in the left fusiform and parahippocampal gyri, right medial superior frontal gyrus, right inferior frontal gyrus, and right medial orbitofrontal cortex as compared with healthy controls. Schizophrenia patients had thinner cortices in the left precentral and paracentral gyri than those with schizotypal disorder. Common cortical thinning patterns observed in schizophrenia and schizotypal disorder patients may be associated with vulnerability to psychosis. Our results also suggest that distinct cortical changes in schizophrenia and schizotypal disorder may be associated with the differences in the manifestation of clinical symptoms among these disorders.

Introduction

Schizotypal (personality) disorder is characterized by odd behavior and attenuated forms of schizophrenic features without the manifestation of overt and sustained psychoses.^1,2 According to the criteria of the International Classification of Diseases, 10th edition (ICD-10),¹ individuals with schizotypal disorder reveal an enduring pattern of unusual speech (eg, odd speech), perception (eg, unusual perceptual experiences), beliefs (odd beliefs or magical thinking), and behaviors (behavior or appearance that is odd, eccentric, or peculiar), which do not exceed the requirements of schizophrenia. Schizotypal disorder is thought to be a prototypic disorder within the schizophrenia spectrum.^3,4 Indeed, individuals with schizotypal disorder share genetic and neurocognitive abnormalities with patients with schizophrenia.⁴

Although previous structural magnetic resonance imaging (MRI) studies on schizotypal disorder have repeatedly demonstrated volume reductions of temporal lobe regions, especially medial temporal structures,^5–7 frontal lobe regions were spared⁵ or even exhibited larger-than-controls patterns.^6,7 However, the majority of these previous studies measured limited numbers of regions of interest (ROIs) and there were considerable inconsistencies regarding regional volume changes in schizotypal disorder compared with the reference group. Several studies that employed a whole-brain voxel-based morphometry (VBM) reported gray matter reductions in frontal and temporal regions in patients with schizotypal (personality) disorder compared with healthy controls.^8,9

Cortical thickness measurement using a surface-based method is an accurate and powerful tool that can detect focal cortical atrophy¹⁰ and has widely been used to investigate the brain structural characteristics of neurological and psychiatric disorders, including schizophrenia.^11–18 Using the surface-based approach, the components of cortical volume (ie, cortical thickness and surface area) can be measured separately. In addition, the subvoxel precision is achieved by the thickness measurement because its value is assigned to individual vertex instead of voxel.¹⁰ Furthermore, it has been shown that the surface-based method is less susceptible to partial volume effects than VBM.¹⁹ Pereira and colleagues²⁰ used 3 different imaging measures (ie, cortical thickness, cortical folding, and VBM) in patients with Parkinson’s disease (PD) and concluded that surface-based measurement of cortical thickness may be more sensitive than VBM for detecting focal gray matter changes in PD. However, to our knowledge, no study has used a surface-based and whole-brain analysis to compare cortical thickness between individuals with schizotypal disorder and healthy subjects or patients with other mental disorders such as schizophrenia.

Past studies have shown broad cortical thinning patterns prominent in the prefrontal and temporal regions in schizophrenia patients relative to healthy controls.^11–18 The cortical thinning in the prefrontal and temporal regions is thought to underlie the symptoms of schizophrenia based on the functions of these brain regions.

In this study, we evaluated cortical thickness using a whole-brain surface-based method in individuals with schizotypal disorder, patients with schizophrenia, and healthy controls. Based on previous studies, we hypothesized that schizotypal and schizophrenia patients exhibit common and distinct cortical thickness changes. We expect cortical thinning in the prefrontal and temporal regions among schizophrenia patients^11–18 and medial temporal regions^5–7 in individuals with schizotypal disorder as compared with healthy controls based upon previous literature.

Methods

Participants

The participants consisted of 46 schizotypal disorder patients (mean age = 25.0 ± 5.5 years), 102 patients with schizophrenia (mean age = 25.5 ± 5.5 years), and 79 healthy subjects (mean age = 24.3 ± 5.8 years) who were recruited from the clinics of the Department of Neuropsychiatry, Toyama University Hospital. All subjects were included in our past study, which evaluated the sulco-gyral patterns of the orbitofrontal cortex.²¹ Of 227 participants, 89 (43 controls, 24 individuals with schizotypal disorder, and 22 schizophrenia patients) had been included in our previous study that used a whole-brain analysis (ie, VBM).⁸ The schizotypal patients visited our hospital due to the distress and problems (eg, anxiety, suicidal ideation, irritability, or insomnia) stemming from their schizotypal features for which they received clinical care, including antipsychotic medications. They met the criteria for schizotypal disorder of ICD-10¹ and the criteria for schizotypal personality disorder of the Diagnostic and Statistical Manual of Mental Disorders, fourth edition (DSM- IV).² These subjects were diagnosed by the consensus of 2 experienced psychiatrists using the Comprehensive Assessment of Symptoms and History (CASH)²² and the Structured Clinical Interview for DSM- IV axis II disorders.²³ The mental condition of each subject was regularly assessed by experienced psychiatrists to check for the emergence of frank psychotic symptoms, and none of the individuals with schizotypal disorder have developed full-blown schizophrenia (mean follow-up period after MRI scanning = 3.0 ± 2.6 years). At the time of scanning, 14 were taking typical neuroleptics and 25 were taking atypical neuroleptics. The remaining seven were antipsychotic naïve. The individuals with schizophrenia fulfilled both the ICD-10 and DSM- IV criteria for schizophrenia. All schizophrenia patients but 2 were taking neuroleptic medications, 42 were taking typical neuroleptics, and 58 were taking atypical neuroleptics. The clinical symptoms of the schizotypal and schizophrenia patients were evaluated using the Scale for the Assessment of Negative and Positive Symptoms (SANS/SAPS)^24,25 within 1 month of scanning. The healthy controls (HC) were recruited from the community, hospital staff, and university students. They completed a questionnaire with 15 items regarding their present/past illness and family histories of illness. None of the control subjects had any personal and family history of psychiatric illness in their first-degree relatives. In addition, all control subjects underwent an interview using the Minnesota Multiphasic Personality Inventory, and subjects who had abnormal profiles (ie, who had any T-score exceeding 70) were excluded from the study.

All subjects were right-handed and physically healthy when they were enrolled in the study, and their history was free from serious head trauma, neurological illness, and substance abuse disorder. Written informed consent was obtained from all subjects. This study was approved by the Committee on Medical Ethics of Toyama University.

Imaging Procedures

The subjects underwent MRI scanning using a 1.5-Tesla scanner (Magnetom Vision, Siemens Medical System, Inc) with a 3-D gradient-echo sequence FLASH (fast low-angle shots) yielding 160–180 contiguous T1-weighted slices of 1.0-mm thickness in the sagittal plane. All subjects were scanned using the same MRI scanner. The imaging parameters were as follows: repetition time = 24 ms; echo time = 5 ms; flip angle = 40°; field of view = 256 mm; and matrix size = 256 × 256 pixels. The voxel size was 1.0 × 1.0 × 1.0 mm³.

The MR images were preprocessed using FreeSurfer software suit (version 5.3; http://surfer.nmr.harvard.edu/). The cortical surface was reconstructed using the FreeSurfer’s standard auto-reconstruction algorithm, including tissue intensity inhomogeneity normalization, nonbrain tissue removal, transformation to Talairach-like space, and segmentation of gray/white matter tissue. Cortical thickness measurements were obtained by calculating the shortest distance from the gray/white boundary to the gray/cerebrospinal fluid boundary at each vertex on the tessellated surface.²⁶ Each image was carefully inspected and any segmentation errors were manually corrected by a trained investigator (D.S.) who was blinded to the subjects’ identities. Each image was resampled onto the average subject (fsaverage) and smoothed with a 10-mm, full-width, half-maximum (FWHM) Gaussian kernel.

When a significant cortical thickness change was found between patient groups (ie, schizophrenia vs schizotypal disorder), the ROI was extracted and mapped onto each subject and the mean thickness of the ROI was calculated for each subject.

Statistical Analysis

Demographic and clinical variables were compared using one-way ANOVA, independent 2-sample t-test, or chi-square test among diagnostic groups. We used FreeSurfer’s general linear model with a 10-mm FWHM Gaussian kernel to compare cortical thickness across groups (schizophrenia vs HC, schizotypal vs HC, and schizotypal vs schizophrenia) and to correlate cortical thickness with clinical variables (ie, onset age, illness duration, daily antipsychotics dosage, and SANS/SAPS subscores), while age, gender, and intracranial volume were treated as nuisance variables. Correlation analyses were performed separately for schizotypal and schizophrenia groups. Monte-Carlo simulation implemented in the AlphaSim program of the Analysis of Functional NeuroImages (AFNI) was used to correct for multiple comparisons.²⁷ To define significant clusters, 10 000 simulations were performed for each comparison in total. To see the association between symptom measures (ie, SANS/SAPS subscores) and thickness of ROI where group difference was found among patient groups, we calculated Spearman’s rho. The significance level was set at P < .005 (2-tailed and corrected for multiple testing), given that we conducted 3 group comparisons for each hemisphere (.005 < .05/6). We also employed this P < .005 threshold for the correlational analyses regarding type II errors for these analyses (12 tests in 2 groups for each hemisphere).

Results

Demographic and Clinical Measures

Demographic and clinical characteristics of the participants are summarized in table 1. Groups did not differ in age, gender, or parental education, whereas control subjects had a higher self-education attainment than patients. SAPS hallucination/delusion and SANS blunted affect/attention deficit subscores were significantly higher in schizophrenia patients than in schizotypal patients. In addition, the daily antipsychotic dosage was significantly higher for patients with schizophrenia than for those with schizotypal disorder.

Table 1.

Demographic and clinical characteristics

Variables	Group
	HC (n = 79)	SZ (n = 102)	ST (n = 46)	Statistics
Age (years)	24.3 ± 5.8	25.5 ± 5.5	25.0 ± 5.5	F (2, 224) = 1.00, P = .37
Age range (years)	18.0–45.9	16.7–45.6	16.0–37.0
Gender (female/male)	35/44	47/55	17/29	χ 2 = 1.01, P = .58
Education years	16.0 ± 2.5	13.4 ± 1.9	13.0 ± 2.0	F (2, 224) = 39.42, P < .001; HC > SZ, ST
Parental education years	12.9 ± 2.3	12.4 ± 2.1	12.4 ± 1.7	F (2, 215) = 1.34, P = .27
Onset age (years)		22.0 ± 4.5
Illness duration (years)		3.6 ± 4.6
Medication dosage (haloperidol equivalent, mg/day)		10.5 ± 8.9	5.9 ± 5.9	t (1, 136) = 3.01, P = .003
Duration of antipsychotic medication (years)		2.6 ± 3.9	1.8 ± 3.2	t (1, 138) = 1.09, P = .28
SAPS
Hallucinations		7.5 ± 7.9	1.9 ± 2.4	t (1, 138) = 4.65, P < .001
Delusions		12.4 ± 9.5	7.6 ± 5.5	t (1, 138) = 3.11, P = .002
Bizarre behavior		4.1 ± 4.1	4.1 ± 2.6	t (1, 138) = 0.06, P = .95
Positive formal thought disorder		3.9 ± 6.4	2.5 ± 3.7	t (1, 138) = 1.37, P = .17
SANS
Blunted affect		13.7 ± 9.0	10.2 ± 7.5	t (1, 138) = 2.28, P = .024
Alogia		6.8 ± 4.7	5.3 ± 3.8	t (1, 138) = 1.84, P = .07
Avolition-apathy		9.9 ± 4.8	9.2 ± 4.5	t (1, 138) = 0.86, P = .39
Anhedonia-asociality		10.6 ± 6.7	10.2 ± 6.3	t (1, 138) = 0.28, P = .78
Attention deficit		8.4 ± 4.6	6.8 ± 4.2	t (1, 138) = 1.99, P = .049

Note: Values represent means ± SD unless otherwise stated.

HC, healthy controls; SANS, Scale for the Assessment of Negative Symptoms; SAPS, Scale for the Assessment of Positive Symptoms; ST, schizotypal disorder; SZ, schizophrenia.

Comparison of Cortical Thickness

Schizophrenia patients exhibited wide-spread cortical thinning predominantly in frontal and temporal regions compared with healthy subjects (table 2; figure 1). Patients with schizotypal disorder had a significantly reduced cortical thickness in the left fusiform and parahippocampal gyri, right medial superior frontal gyrus, right inferior frontal gyrus, and right medial orbitofrontal cortex compared with healthy controls (table 3; figure 2). The regions where cortical thinning was observed in schizotypal patients as compared with controls mostly overlapped with the areas found to be thinner in schizophrenia patients as compared with controls. Furthermore, schizophrenia patients had thinner cortices in the left precentral and paracentral gyri than those with schizotypal disorder (table 4; figure 3).

Table 2.

Reduced cortical thickness in schizophrenia patients as compared with healthy controls

			MNI coordinates (maximum vertex)
Cluster no.	Cluster size (mm2)	Cluster-wise P	x	y	z	Annotation
1	19 636	.0001	−22.1	41.5	24.1	Left caudal anterior cingulate, superior frontal, medial/lateral orbitofrontal, caudal/rostral middle frontal, and precentral gyri, pars opecularis/triangularis/orbitalis, and insula
2	2238	.0001	−35.5	−44.1	−21.1	Left fusiform and lingual gyri
3	671	.0001	−41.3	−1.3	−37.6	Left middle temporal and inferior temporal gyri
4	20 371	.0001	38.7	41	24	Right caudal/rostral anterior cingulate, medial/lateral orbitofrontal, superior frontal, caudal/rostral middle frontal, superior temporal and precentral gyri, pars opecularis/ triangularis/orbitalis, and insula
5	6296	.0001	44	−67.7	7	Right middle/inferior temporal, entorhinal, parahippocampal, fusiform, lateral occipital, and lingual gyri

MNI, The Montreal Neurological Institute.

Fig. 1.

Cortical statistical maps comparing the cortical thickness between schizophrenia (SZ) patients and healthy controls (HC). The maps are shown for the right and left hemispheres in lateral (upper) and medial (bottom) views. The horizontal bar indicates the P-value (<.005, corrected). LH, left hemisphere; RH, right hemisphere. The color maps are available in the online version of the manuscript.

Table 3.

Reduced cortical thickness in schizotypal patients as compared with healthy controls

			MNI coordinates (maximum vertex)
Cluster no.	Cluster size (mm2)	Cluster-wise P	x	y	z	Annotation
1	554	.0007	−35.1	−23.3	−22.8	Left parahippocampal and fusiform gyri
2	1167	.0001	48.4	25.9	5.2	Right pars opercularis and pars triangularis
3	1033	.0001	21.1	60.3	6.1	Right superior frontal and medial orbitofrontal gyri

MNI, The Montreal Neurological Institute.

Fig. 2.

Cortical statistical maps comparing the cortical thickness between patients with schizotypal disorder (ST) and healthy controls (HC). The maps are shown for the right and left hemispheres in lateral (upper) and medial (bottom) views. The horizontal bar indicates the P-value (<.005, corrected). LH, left hemisphere; RH, right hemisphere. The color maps are available in the online version of the manuscript.

Table 4.

Reduced cortical thickness in schizophrenia patients as compared with schizotypal disorder patients

			MNI coordinates (maximum vertex)
Cluster no.	Cluster size (mm2)	Cluster-wise P	x	y	z	Annotation
1	546	.0009	−12.8	−11.1	67.7	Left precentral and paracentral gyri

MNI, The Montreal Neurological Institute.

Fig. 3.

Cortical statistical maps comparing the cortical thickness between patients with schizotypal disorder (ST) and schizophrenia (SZ) patients. The maps are shown for the right and left hemispheres in lateral (upper) and medial (bottom) views. The horizontal bar indicates the P-value (<.005, corrected). LH, left hemisphere; RH, right hemisphere. The color maps are available in the online version of the manuscript.

Correlation between Cortical Thickness and Clinical Measures

In the whole-brain correlational analyses, there were no significant correlations between cortical thickness and clinical variables (ie, onset age, illness duration, daily antipsychotics dosage, and SANS/SAPS subscores) in any patient group. The mean thickness of the ROI where schizophrenia group showed cortical thinning relative to individuals with schizotypal disorder (figure 3) was calculated for each participant. We found a trend-level negative correlation between mean thickness of ROI and SANS anhedonia-asociality subscore (rho = −0.25, P = 0.01) in schizophrenia patients, while there was no significant correlation of symptoms with the mean thickness of ROI in schizotypal disorder group.

Discussion

In this study, we found cortical thinning in medial temporal and frontal regions (ie, fusiform gyrus, parahippocampal gyrus, medial superior frontal gyrus, inferior frontal gyrus, and medial orbitofrontal cortex) in both schizotypal and schizophrenia patients as compared with healthy controls. Such common cortical changes may be associated with the vulnerability to psychosis. Schizophrenia patients exhibited further wide-spread cortical thinning patterns, predominantly in frontal and temporal regions as compared with healthy subjects. In addition, a reduced cortical thickness in the left precentral gyrus differentiated schizophrenia patients from individuals with schizotypal disorder. These distinct cortical changes in schizophrenia and schizotypal disorder may be related to the manifestation of psychotic symptoms.

Our sample overlaps substantially with that in our previous studies,^28–31 which evaluated regional gray matter volume using manual parcellation of ROIs in schizotypal disorder patients. However, the cortical regions found to be thinner in patients with schizotypal disorder than in controls in this study were not fully consistent with the areas where schizotypal patients exhibited gray matter volume reductions. Although we demonstrated reduced cortical thickness in fusiform and orbitofrontal gyri,^28,30 we failed to detect cortical atrophy in the superior temporal and postcentral gyri.^29,31 Similarly, our results are partly inconsistent with those of our previous VBM study that demonstrated gray matter reductions in medial temporal and inferior frontal regions, as well as the superior temporal gyrus among schizotypal patients who were also included in the present study.⁸ In order to see if the difference in image processing methodology account for these regional discrepancies, we applied the same methodology we used in this study to the participants who were also included in the study by Kawasaki et al⁸ (43 controls, 24 schizotypal and 22 schizophrenia patients) that used VBM. The clinical characteristics of per groups were similar between current study and Kawasaki et al⁸ (supplementary table 1). However, we only found cortical thinning in the left inferior frontal gyrus in schizophrenia patients as compared with controls using a more moderate threshold (P ≤ .05). Therefore, this additional analysis failed to replicate the group differences found in the analyses with all 227 subjects. The negative finding in the additional analysis may be due to its small sample size.

Our results are partly consistent with those of a recent VBM study by Asami and colleagues,⁹ but they demonstrated gray matter reductions in broader regions, including the superior temporal gyrus, prefrontal, frontolimbic, and parietal regions in male schizotypal patients as compared with healthy controls. In addition to the difference in methodology (ie, voxel-based gray matter density measurement or surface-based cortical thickness measurement), the disparities in the gender distribution and the severity of symptoms (ie, help-seeking patients taking antipsychotics vs antipsychotic-naïve participants recruited through advertisements) between our study and that by Asami et al may be relevant to the somewhat different results.

In this study, we found wide-spread cortical thinning patterns predominantly in the prefrontal and temporal regions in patients with schizophrenia as compared with healthy controls. These results are consistent with previous studies.^11–18 The prefrontal and temporal regions are thought to be involved in cognitive function, auditory/visual processing, speech, emotional processing, executive function, and decision-making, and these features are impaired in individuals with schizophrenia.^32–34 Thus, the broad cortical thinning in the prefrontal and temporal regions is thought to underlie the manifestation of the symptoms of schizophrenia.

We noted greater cortical thinning of the left precentral and paracentral gyri in schizophrenia patients than in those with schizotypal disorder. This cluster corresponds to the primary motor cortex (M1; Brodmann area 4). Schizophrenia patients often exhibit motor symptoms such as catatonia, neurological soft sings (NSS), extrapyramidal symptoms (EPS), and psychomotor slowing.³⁵ Functional imaging studies have revealed aberrant M1 activity during finger sequence learning and simple sensorimotor detection tasks (reviewed by Abboud and colleagues).³⁶ Structural MRI studies have demonstrated negative correlations of NSS with the gray matter volume,³⁷ cortical thickness,³⁸ and local gyrification index³⁹ of M1. Thus, a reduced thickness of the precentral and paracentral gyri may underlie the aberrant movements exhibited by schizophrenia patients. However, we were unable to examine the difference in the correlation of the precentral gyrus thickness with movement measures between schizophrenia and schizotypal patients because we did not have the movement data such as NSS and EPS. The results of the correlational analysis suggest that thinner cortex of this region might be related to worse negative symptoms (ie, anhedonia-asociality). The relationships between motor and negative symptoms have been shown in the literature.^40–42 Therefore, the manifestation of negative symptoms in schizophrenia patients might partly be associated with motor symptoms stemming from altered motor cortex. As noted above, schizotypal disorder is thought to be a prototypic disorder within the schizophrenia spectrum without the manifestation of frank and sustained psychoses. Given the significant but subtle cortical difference between schizophrenia and schizotypal disorder found in this study, even such small difference might underlie the difference in the manifestation of clinical symptoms (eg, attenuated form of schizophrenia futures vs full-blown psychotic symptoms) between schizotypal disorder and schizophrenia.

A few limitations of this study should be noted. First, we lacked the data for the evaluation of movement, as mentioned above. Second, as most of the patients in this study had taken antipsychotics, we were unable to exclude the influence of antipsychotics on brain morphology. For example, a recent meta-analysis revealed complex associations between antipsychotic medications and cortical gray matter changes.⁴³ In order to see the potential confounding by the use of antipsychotics, we performed an additional analysis including antipsychotic use (ie, duration of antipsychotic medication) as well as age, gender, and intracranial volume as nuisance covariates for the group comparisons of cortical thickness. The results of this additional analysis are generally consistent with those we reported above (supplementary figures 1–3), suggesting that our finding were not merely consequences of using antipsychotic medications.

Although we found common and distinct cortical thickness reductions in schizotypal disorder and schizophrenia patients, we could not answer the question that such cortical thinning would change over time among these mental disorders. Thus, longitudinal evaluations should be warranted in future studies.

In conclusion, shared cortical thinning patterns among schizophrenia and schizotypal disorder patients may be associated with the common clinical features within the schizophrenia spectrum. Distinct cortical changes in schizophrenia and schizotypal disorder may underlie the differences in the manifestation of clinical symptoms between these disorders.

Funding

This work was supported by the Japan Society for the Promotion of Science (18K07549 to Dr Takayanagi), (18K15509 to Dr Sasabayashi), and (18K07550 to Dr Takahashi); Health and Labour Sciences Research Grants for Comprehensive Research on Persons with Disabilities from Japan Agency for Medical Research and Development (AMED) (16dk0307029h0003 to Dr Suzuki). Dr Sasabayashi was also supported by grants from SENSHIN Medical Research Foundation.