. 2018 Sep 12;45(4):911–923. doi: 10.1093/schbul/sby121

Linking Cortical and Connectional Pathology in Schizophrenia

Maria Angelique Di Biase ^1,^2,^3,^✉, Vanessa L Cropley ^1,³, Luca Cocchi ⁴, Alexander Fornito ⁵, Fernando Calamante ⁶, Eleni P Ganella ^1,^3,⁷, Christos Pantelis ^1,^3,^7,^8,⁹, Andrew Zalesky ^1,^3,¹⁰

PMCID: PMC6581130 PMID: 30215783

Abstract

Schizophrenia is associated with cortical thickness (CT) deficits and breakdown in white matter microstructure. Whether these pathological processes are related remains unclear. We used multimodal neuroimaging to investigate the relationship between regional cortical thinning and breakdown in adjacent infracortical white matter as a function of age and illness duration. Structural magnetic resonance and diffusion images were acquired in 218 schizophrenia patients and 167 age-matched healthy controls to map CT and fractional anisotropy in regionally adjacent infracortical white matter at various cortical depths. We found a robust and reproducible relationship between thickness and anisotropy deficits, which were inversely correlated across cortical regions (r = −.5, P < .0001): the most anisotropic infracortical white matter was found adjacent to regions with extensive cortical thinning. This pattern was evident in early (20 y: r = −.3, P = .005) and middle life (30 y: r = −.4, P = .004, 40 y: r = −.3, P = .04), but not beyond 50 years (P > .05). Frontal pathology contributed most to this pattern, with cortical thinning in patients compared to controls at all ages (P < .05); in contrast to initially elevated frontal white matter anisotropy in patients at 30 years, followed by rapid white matter decline with age (rate of annual decline; patients: 0.0012, controls 0.0006, P < .001). Our findings point to pathological dependencies between gray and white matter in a large sample of schizophrenia patients. We argue that elevated frontal anisotropy reflects regionally-specific, compensatory responses to cortical thinning, which are eventually overwhelmed with increasing illness duration.

Keywords: neuropsychiatry, multimodal imaging, computational psychiatry, cortical thickness, diffusion, tractography

Introduction

Schizophrenia is characterized by cortical thickness (CT) deficits,^1,2 as measured with structural magnetic resonance imaging (MRI) and widespread abnormalities in white matter fractional anisotropy (FA),³ as measured with diffusion-weighted MRI. However, it remains unclear whether CT and white matter alterations are independent pathological processes. Histological evidence implicates smaller neuronal cell bodies⁴ and fewer dendritic spines⁵ in the hippocampus and cortex of schizophrenia patients, which may underlie reduced CT. In addition, ultrastructural findings report abnormalities in the morphology and number of oligodendrocytes.⁶ As oligodendrocytes are critically involved in the formation and maintenance of myelin, neuronal function, and neurotransmission,⁷ it is likely that alterations in neurons and components of white matter are intimately linked.⁸

The interplay between cortical thinning and changes in white matter has been investigated macroscopically with multimodal MRI acquisitions in healthy human development and aging.^9–12 Of note, Whitaker and colleagues recently found brain regions that underwent greater cortical thinning during normal adolescent development were associated with increased infracortical myelination.¹² This normal gray-white matter consolidation and maturation was posited to facilitate healthy cognitive and behavioral development and in contrast, abnormal gray-white matter associations were proposed to characterize psychiatric conditions, such as schizophrenia. In line with this notion, CT and white matter FA have previously revealed differential associations in schizophrenia patients compared to healthy controls,¹³ potentially indicating poor neurodevelopmental consolidation between gray and white matter.

Given that microstructural features of gray and white matter remain linked throughout the lifespan in health,^11,12 it is plausible that disease-related reductions in CT and white matter microstructure in schizophrenia would characterize related processes that share a common aetiology. To this end, a previous study by Douaud and colleagues¹⁴ observed spatial overlap between gray matter volume and diffusion-derived white matter estimates in adolescent-onset schizophrenia patients, suggesting that gray and white matter deficits co-localize in the disorder. However, it would be instructive to explicitly quantify the relationship between the degree of gray and white pathology. In doing so, abnormal associations between these biological properties may be attributed to disease-related processes.

To address this multimodal line of inquiry, we examined the relationship between pathological gray and white matter in a sample of 218 schizophrenia patients and 167 healthy comparison individuals. Diffusion-weighted MRI was used to measure microstructural tissue organization. Structural MRI was used to estimate regional CT. In one of the largest multimodal schizophrenia samples analyzed to date, we tested whether between-group differences in white matter anisotropy and diffusivity measured at 4 cortical depths (cortical ribbons) were correlated across regions with between-group differences in CT. We hypothesized that regional heterogeneity in the severity of abnormalities in CT and white matter anisotropy would covary across cortical regions, such that the most extensive cortical thinning in schizophrenia would be found adjacent to the most extensively disrupted infracortical white matter. The impact of age and illness duration on this relationship was also characterized to explore whether one process represents an epiphenomenon or delayed sequela of the other.

Methods

Participants

Data were collected across 5 Australian states and territories. Full details related to recruitment procedures have been published elsewhere.¹⁵ Exclusion criteria included any neurological disorder of MR-visible pathology, history of brain trauma followed by a long period of amnesia (>24 h), mental retardation (full-scale IQ < 70), current drug or alcohol dependence, as well as electroconvulsive therapy in the past 6 months. Healthy controls with a personal or family history of psychosis or bipolar I disorder were also excluded. Patients had a confirmed diagnosis of schizophrenia or schizoaffective disorder according to DSM-IV and ICD-10 diagnostic criteria. Participants were English speaking, predominately of European ancestry and aged between 18 and 65 years.

For this study, participants with both structural and diffusion-weighted imaging (DWI) scans passing quality control procedures were selected (described below). This resulted in 385 participants: 218 individuals (156 male) diagnosed with schizophrenia (n = 182) or schizoaffective disorder (n = 36) and 167 healthy controls (81 male). Approval for data analysis was provided by the Melbourne Health Human Research Ethics Committee (MHREC: 2010.250) and written informed consent was obtained from all participants, according to the Declaration of Helsinki. Procedures for assessing clinical status and symptoms are described in supplementary material.

MRI Acquisition

T1-weighted and DWI scans were acquired in each participant with a Siemens Avanto 1.5-Tesla system (Siemens) across 5 different sites in Australia. The same model of MRI scanner, acquisition sequences, and exclusion/inclusion criteria were used at each site. No scanner upgrades were performed during the study lifetime. Calibration scans were conducted at each site using an identical Siemens phantom and collection of MRI scans from the same individual traveling to each site using the same acquisition protocol. Quality control procedures did not detect any site variation and have been detailed elsewhere.¹⁵

For T1-weighted images, an optimized magnetization-prepared rapid acquisition gradient echo (MP-RAGE) sequence was acquired with the following parameters: 176 sagittal slices of 1 mm thickness without gap; field of view = 250 × 250 mm²; repetition time/echo time = 1980/4.3 ms; data matrix size = 256 × 256; voxel dimensions = 0.98 × 0.98 × 1.0 mm³. For DWI, 64 gradient-weighted volumes were acquired using a spin-echo EPI sequence with the following parameters: b-value, 1000 s/mm²; 65 consecutive axial slices of thickness 2.4 mm; 104 × 104 image matrix with an in-plane voxel resolution of 2.4 × 2.4 mm; field of view, 25 × 25 cm; repetition time, 8.4 s; echo time, 88 ms; flip angle, 90 degrees. The images were visually inspected for artifacts, during which 4 scans were removed from the analysis due to poor image quality (gross artifact, cerebellar cropping, and movement). This resulted in the sample population shown in table 1.

Table 1.

Sample demographics and clinical characteristics

	Schizophrenia (n = 218)		Healthy controls (n = 167)
Variables	Mean	SD	Mean	SD	t or χ²	P
Age	37.7	10	39.9	13.9	1.70	>.05
Wechsler Abbreviated Scale of Intelligence (current IQ)	104	14.8	117.3	11.1	10.11	<.001
Age of symptom onset	23.3	6.1
Duration of illness, y	14.4	9.5
Scale for the Assessment of Negative Symptoms (total score)	26.6	18.3
Diagnostic Interview for Psychosis (lifetime positive symptom score)	7.6	3.5
	n	%	N	%
Gender, M/F	156/62		81/86		21.55	<.001
Site
Melbourne	62	28.30	64	38.30	4.32	.04
Sydney	51	23.30	34	20.40	0.47	.49
Brisbane	74	34.00	31	18.60	11.70	<.001
Perth	21	9.60	24	14.40	2.10	.14
Newcastle	10	4.60	14	8.40	2.37	.12
Medication
Currently taking antipsychotics	188	86.24	0	0.00
Taking atypical antipsychotics	179	82.11	0	0.00
Taking typical antipsychotics	23	10.55	0	0.00
Taking antidepressants	72	33.03	0	0.00

Open in a new tab

CT Estimates

Structural MRI images were processed using the Freesurfer (version 5.1.0)¹⁶ automated neuroanatomical segmentation software to generate a 3-dimensional model of the cortical surface for each participant, which was spherically inflated and registered to the manually delineated Desikan-Killiany (DK-68)¹⁷ and Destrieux (DS-148)¹⁸ brain atlases. Two atlases were employed to evaluate robustness to the choice of atlas and variation in region size.¹⁹ All gray/white and gray/cerebral spinal fluid boundaries were manually examined and corrected by trained raters to increase the accuracy of CT estimates. Images were then reprocessed to produce final CT estimates.²⁰ Further details about thickness estimation are provided in supplementary material.

Intra- and Infracortical Tissue Structure

Regional FA was measured at 4 distinct cortical depths to enable comprehensive characterization of tissue microstructure across the cortex. Diffusion-weighted MRI were pre-processed with the FMRIB Software Library (FSL).²¹ In brief, pre-processing involved visual inspection for artifacts; eddy correction for motion and gradient-induced eddy current distortions using affine registration to the first volume; computation of FA and volumes and rigid-body registration of these volumes to each participant’s respective T1 structural space. Regional radial diffusivity (RD) estimates were computed as a supplementary measure to aid interpretation of the primary analysis performed with FA estimates (see supplementary material for results pertaining to RD).

Cortical ribbons were delineated in Matlab (version 2014b), using a fully automated algorithm based on the tessellated cortical meshes estimated by Freesurfer. The cortical ribbon produced by Freesurfer was segmented into 4 mutually exclusive ribbons: (1) intracortical gray matter between the pial surface and the gray/white matter interface (ribbon 1), (2) gray/white matter interface (ribbon 2), (3) infracortical white matter within ~2–4 mm from the gray/white matter interface (ribbon 3), (4) relatively deep infracortical white matter within a band between ~4–6 mm from the gray/white matter interface (ribbon 4; figure 2A). The thickness of ribbon 1 was equal to the thickness of the cortical surface. The remaining 3 ribbons had a thickness that was approximately equal to the native voxel dimension (2.4 mm isotropic). Ribbons 3 and 4 were segmented by identifying the normal (perpendicular vector) to the gray/white matter interface delineated with Freesurfer and displacing this interface in the normal direction. The interface was displaced along the normal direction by 2.4 mm (native voxel dimension) to define ribbon 3 and 4.8 mm (×2 native voxel dimension) to define ribbon 4. Voxels intersected by the displaced interface were assigned to the relevant ribbon. Each voxel was assigned to at most one ribbon. Voxels intersected by multiple ribbons were assigned to the ribbon with which they shared the greatest interface. It is important to note that the gray/white matter interface is not a discrete “black-and-white” boundary,²² and thus ribbon 2 is likely to have included partial volume effects between gray and white matter, whereas these effects were diminished at ribbons 3 and 4. For each region of the DK-68 and DS-148 atlas, mean FA was computed at each of the 4 ribbons by averaging over the appropriate set of voxels delineating each ribbon. Thus, 4 mean FA measurements were determined for each region. Regions delineating the frontal pole were excluded from ribbons 3 and 4 because the gray/white matter interface could not be displaced sufficiently deep before intersecting with itself. Therefore, a total of 66 regions were examined in ribbons 3 and 4.

Fig. 2. — Cortical distribution of CT and FA deficits as a function of age. Polynomial regression (see text) was used to map between-group differences in cortical thickness (left column) and infracortical white matter anisotropy (right; ribbon 4) at ages 20, 30, 40, 50, and 60 y (top to bottom). Between-group differences were expressed as the percentage, Δ = (patients – controls) / controls × 100%, and thus warmer colors denote an increase in patients relative to controls. In early adulthood (20–30 y), frontal regions exhibit up to a 4% increase in infracortical white matter anisotropy in patients relative to controls. With increasing age and illness duration, this increase is supplanted by global reductions in white matter anisotropy. Frontotemporal cortical thickness reductions are evident in patients at all ages. L: left, R: right, CT: cortical thickness, FA: fractional anisotropy.

Tractography

To provide a complementary characterization of white matter that was less susceptible to partial volume effects and potential inaccuracies in the delineation of the gray/white matter interface, probabilistic tractography was used to map all white fiber tracts connected to each cortical region, thus allowing tract-averaged FA to be investigated in addition to ribbon-averaged infracortical FA. While FA can also be measured in deep white matter structures delineated by a regional parcellation atlas, there is no explicit one-to-one correspondence between regions comprising white and gray matter parcellations, and thus associations cannot be tested across regions. Tractography procedures are detailed in supplementary material.

Statistical Modeling

All statistical modeling was performed in Matlab (version 2014b). Two-sample t-tests were used to independently test the null hypothesis of equality in group means for age, current IQ (Wechsler Abbreviated Scale of Intelligence) and acquisition site. Sex was compared between groups using Fisher’s exact test (2-tailed).

Case-Control Differences in CT and FA.

Regional CT and FA were normally distributed within each group, as confirmed by Shapiro-Wilk’s test (>.05) and homogeneity of variances was checked by Levine’s test. Between-group differences in thickness and FA were expressed as the percentages,

Δ_{i}^{CT} = \frac{μ_{patients}^{CT} (i) - μ_{controls}^{CT} (i)}{μ_{controls}^{CT} (i)} \times 100 %, Δ_{i j}^{FA} = \frac{μ_{patients}^{FA} (i, j) - μ_{controls}^{FA} (i, j)}{μ_{controls}^{FA} (i, j)} \times 100 %,

where $μ^{CT} (i)$ denotes group-averaged CT in the cortical region with index i and $μ^{FA} (i, j)$ denotes group-averaged FA in region i at the cortical depth corresponding to ribbon j. Cortical regions were delineated using 2 alternative atlases: DK-66 ( $i = 1, \dots, 66)$ and DS-148 ( $i = 1, \dots, 148)$ . Between-group differences in thickness were therefore quantified for multiple cortical regions and between-group differences in FA were quantified for the same regions at 4 distinct cortical depths.

Relation Between CT and FA.

Two-tailed Pearson correlation coefficient was used to test for a linear association between $Δ_{i}^{CT}$ and $Δ_{i j}^{FA}$ for each cortical ribbon $j = 1, \dots, 4$ . This correlation analysis tested whether between-group differences in CT were correlated across cortical regions with between-group differences in FA, thereby evaluating our hypothesis that the most extensive CT deficits would be found adjacent to the most extensively disrupted infracortical white matter.

Error control across the 8 correlations tested was achieved with the Bonferroni correction (4 FA ribbons × 2 cortical atlases, adjusted α = .006). It is important to note that all correlations were computed across cortical regions, not across subjects, and are thus particularly robust to demographic and sample confounds (eg, gender and acquisition site) because these variables remain fixed across regions of an individual’s cortex. Complementary analyses examined correlations between $Δ_{i}^{CT}$ and $Δ_{i j}^{RD}$ for each cortical ribbon $j = 1, \dots, 4$ , as well as between $Δ_{i}^{CT}$ and $Δ_{i}^{Tract - averaged FA}$ , instead of ribbon-averaged anisotropy (supplementary material).

Statistical Modeling of Age.

To characterize the impact of age on CT and FA in a cross-sectional manner, the following polynomial regression model was fitted to each cortical region,

Y_{n} = β_{1} + β_{2} \times {diagnosis}_{n} + β_{3} ({age}_{n} - A) + β_{4} \times ({age}_{n} - A) \times {diagnosis}_{n} + β_{5} \times {sex}_{n} + C_{n} + ϵ_{n}

where $Y_{n}$ denotes either regional FA or CT for the participant with index n, ${diagnosis}_{n}$ is a binary indicator of the diagnostic status of participant n (patient: 1, control: 0), ${age}_{n}$ is the age of participant n and $C_{n}$ represents the nuisance confounds, including sex, scanner site, and intracranial volume. The explanatory variables can be understood as follows: β₁ is the intercept term and β₂ models the between-group difference in the dependent variable (CT or FA); β₃ models the rate of loss at age A, while β₄ models the between-group difference in this rate; and, β₅ models sex differences. The regression model defined by the above equation was independently fitted with age centered between 20 and 65 years in yearly increments; ie, A = 20, 21,…,65. When the regression was estimated with age centered at A, the amount of loss (β₁ and β₂) in the dependent variable specifically pertained to an age of A years, therefore, yielding annual snapshots of CT or FA across the lifespan of each participant, within the constraints of a cross-sectional design. Explanatory variables involving the square of age did not significantly improve the model fit and were thus excluded.

Results

Demographics and Symptoms

Diffusion-weighted and structural MRI were acquired in schizophrenia patients (n = 218) and healthy comparison individuals (n = 167) to address the aims of this study. Sample demographics and patient symptom ratings are shown in table 1. The patient and control samples were well matched on demographic variables, with the exception of sex and IQ. The patient sample comprised a significantly greater proportion of males. Sex was therefore included as a nuisance covariate when appropriate. Although IQ was significantly lower in patients compared to controls, disentangling this effect from schizophrenia is generally difficult, given that the disorder is characterized by pervasive cognitive impairments²³ and in turn, controlling for IQ in statistical analyses may erode statistical power to detect genuine changes related to the disorder.^24,25

CT and White Matter Anisotropy

Case-Control Differences in CT and FA.

Frontotemporal cortical regions were significantly thinner in patients, consistent with previous reports,^1,2 and widespread differences were found in infracortical FA between patients and controls (supplementary figure 1).

Relation Between CT and FA.

For each ribbon, we tested whether the between-group differences in FA and CT were linearly associated across regions. Ribbon 1 delineated intracortical gray matter between the pial surface and gray/white matter interface (see Methods). Regions with lower ribbon 1 FA in patients were significantly positively correlated with the extent of regional CT deficits in patients (r = .3, P < .006; figure 1B). This positive correlation was previously observed²⁶ and most likely due to cellular and synaptic breakdown in gray matter, resulting in fewer obstacles to diffusion.²⁷ Ribbon 1 FA was given no further consideration here.

Fig. 1. — Correlation across cortical regions in CT and white matter FA deficits. Between-group differences in cortical thickness (Δ CT) and fractional anisotropy (Δ FA) were measured in 4 cortical ribbons (A) for each cortical region comprising the Desikan-Killiany (DK-66) and Destrieux (DS-148) parcellation atlases. Between-group differences in CT and FA were expressed as the percentage, Δ = (mean patients – mean controls) / mean controls × 100%. (B) Scatter plots and lines of best fit show the association between Δ CT (horizontal axis) and Δ FA (vertical) for each cortical ribbon. Each data point denotes a cortical region. The Pearson correlation coefficient (r) and corresponding P-value are indicated above each scatter plot. For ribbon 4, both the DK-66 (left) and DS-148 (right) cortical parcellations are shown. The DK-66 parcellation is shown for ribbons 1–3. The relation between Δ CT and Δ FA differed as a function of cortical depth: positive correlation in gray matter (ribbon 1); no correlation at the gray/white matter interface (ribbon 2); negative correlations in superficial (ribbon 3) and deep white matter (ribbon 4). (C) Probabilistic tractography was initiated from each cortical region to map white matter fiber tracts. Fiber tracts originating from a frontal region (red) are shown as an example. The negative correlation between Δ CT and Δ FA remained significant when using tract-averaged FA instead of ribbon-averaged FA for both the DK-66 (left) and DS-148 (right) cortical parcellations.

Regional between-group differences in CT and FA were not significantly correlated at ribbon 2, possibly due to partial volume effects at this transitionary ribbon and the limited resolution of diffusion MRI to resolve the gray/white matter interface. Ribbons 3 and 4 comprised infracortical and relatively deep white matter, respectively. Between-group differences in CT and FA were significantly inversely correlated across the 66 cortical regions at both ribbons 3 (r = −.4, P = .001) and 4 (r = −.5, P < .0001; figure 1B) and remained significant when ribbon 3 and 4 were combined (supplementary figure 2). This negative correlation was replicated using the 148 cortical regions comprising the DS atlas,¹⁸ suggesting robustness to the choice of regional parcellation (ribbon 3: r = −.4, P < .0001; ribbon 4: r = −.5, P < .0001; figure 1B).

Although whole-brain FA and CT measurements were significantly and positively correlated (ribbons 1–4: r = .4, P < .0001), regional FA and CT measurements were not correlated (P > .05) in ribbons 3 and 4 in: (1) patients, (2) controls, or (3) the combined sample of patients and controls (supplementary figure 3). Only between-group differences in regional FA and CT were significantly correlated, indicating that disease processes were the most likely explanation for this relation. The relations found across each ribbon remained significant after using partial correlation to control for variation in volume across regions (supplementary figure 4) and after regressing sex, acquisition site and intracranial volume from FA and CT measurements (supplementary figures 5 and 6). Furthermore, the correlations remained significant when FA was substituted for RD, a measure that is particularly sensitive to myelination (ribbon 3: r = .4, P = .0001; ribbon 4: r = .6, P < .0001; supplementary figure 7).

A potential explanation for these intriguing correlations was that brain registration to standard anatomical space was inaccurate in regions with extensive cortical thinning, resulting in partial volume effects and systematic displacement of the gray/white matter interface towards the pial surface in thinner regions. To rule out this confound, probabilistic tractography was initiated from ribbon 4 of each region to map regionally-specific white matter fiber tracts. For each region, tract-averaged FA was then measured to characterize FA along the entire length of fiber tracts, thereby minimizing the influence of FA measurements in the vicinity of a potentially inaccurately delineated gray/white matter interface. Consistent with infracortical FA measured in white matter (ribbons 3 and 4), between-group differences in CT and tract-averaged FA were also inversely correlated across the 66 regions comprising the DK atlas (r = −.4, P = .002) and 148 regions of the DS atlas (r = −.3, P < .001; figure 1B).

Influence of Age and Illness Duration

Relation Between CT and White Matter Anisotropy as a Function of Age.

Having found an inverse relation between group-changes in CT and infracortical FA, we next sought to characterize this relation in terms of age. Age also served as a surrogate measure of illness duration, since both variables were strongly correlated (r = .8, P < .0001). Polynomial regression was used to model the impact of diagnosis (controls: β₁, patients: β₂), age (β₃) and the interaction between age and diagnosis (β₄) on CT and ribbon 4 FA measurements (see Methods). A separate regression was fitted for CT and FA measured at each of the 66 regions comprising the DK atlas. Figure 2 shows cortical renderings of regional between-group differences in CT (left column) and ribbon 4 FA (right) for each decade of age between 20 and 60 years. Between-group differences for each age band were expressed at the percentage β₂⁄β₁ × 100%, which is equivalent to (patients – controls) / controls × 100%, controlling for sex and other nuisance confounds (see Methods). In early adulthood (20–30 y), temporal and frontal regions (bilaterally) were found to show the most marked CT reductions in patients relative to healthy controls of the same age, but these regions also exhibited abnormally increased infracortical white matter. Moreover, for some frontal regions (medial prefrontal cortex), in early adulthood, ribbon 4 FA was increased by up to 4% in patients relative to controls. With increasing age, the inverse relation diminished, with ribbon 4 FA deficits becoming widespread and encompassing frontal and temporal regions after middle age (40–50 y; figure 2).

To quantify these observations, between-group differences in CT and ribbon 4 FA from the regression model (β₂) were converted to effect sizes (Cohen’s d) and correlated across the 66 DK regions for each decade of age between 20 and 60 years. Effects sizes for between-group differences in CT and ribbon 4 FA were inversely correlated at age 20 (r = −.3, P = .005), 30 (r = −.4, P = .004) and 40 (r = −.3, P = .04; figure 3A). However, significant associations were not found at 50 and 60 years (P > .05). Figure 3A shows that the inverse relation evident in early adulthood and middle life (20–40 y) is due to the contribution of frontal regions (blue data points), where cortical thinning in patients is extensive, but ribbon 4 FA is preserved, and even elevated by 3%–4% for some frontal regions in patients compared to controls.

Fig. 3. — Age-resolved correlation across cortical regions in CT and white matter FA deficits. Effect sizes (Cohen’s d) of between-group differences in cortical thickness (CT) and infracortical white matter fractional anisotropy (FA; ribbon 4) were correlated at different ages. (A) Scatter plots and lines of best fit show the association between effect sizes in CT (vertical axis) and FA (horizontal) at ages 20, 30, 40, 50, and 60 y (left to right). Each data point denotes a cortical region of the DK-66 brain parcellation atlas. The Pearson correlation coefficient (r) and corresponding P-value are indicated above each scatter plot. (B) Pearson correlation coefficient shown as a function of age for the association between effect sizes in CT and FA. Between-group differences in FA and CT were significantly and negatively correlated at ages 20, 30, and 40 y (P < .05). Significant correlations were not evident at ages 50 and 60 y (P > .05). Effect sizes were computed based on the β₂ regression coefficients.

CT and White Matter Anisotropy as a Function of Age and Cortical Lobe.

Given that frontal infracortical FA was abnormally high in younger patients, but abnormally low in middle life and beyond, we aimed to quantify the rate of anisotropy breakdown and cortical thinning (mm/y) in the frontal lobes, subject to the limitations of a cross-sectional design. The polynomial regression model described above was fitted to lobe-averaged CT and ribbon 4 FA estimates (frontal, temporal, parietal, occipital). Figure 4A shows ribbon 4 FA (upper panel) and CT (lower) as a function of age for patients (red line) and controls (blue). Significant cortical thinning in frontal and temporal lobes was evident in patients at all ages (P < .05; figure 4B) and further encompassed parietal and occipital lobes in individuals older than 30 years of age. However, the rate of thinning did not significantly differ between patients and controls (mm/y patients: 0.00460; controls: 0.00457; P > .05, figure 5), suggesting that reduced CT may occur early in the disorder and remain a relatively stable deficit thereafter. In contrast, frontal infracortical white matter was found to show increased ribbon 4 FA in patients very early in the disorder (ie, early adulthood), but declined at a significantly faster rate in patients compared to controls (loss/y patients: 0.0012; controls: 0.0006; P < .001; figure 5), with significant FA reductions emerging in patients older than 40 years within frontal and temporal regions (P < .05; figure 4B). Thus, early in the disorder, frontotemporal pathology was characterized by reduced CT but relatively normal infracortical white matter. In fact, early in the disorder, anisotropy in white matter underlying the medial prefrontal cortex (bilaterally) was increased in patients relative to controls. With increasing age and illness duration, frontal infracortical white matter rapidly reduced in patients at a rate that was significantly faster than attributable to healthy aging. This rapid decline in frontal white matter resulted in the inverse relation diminishing by middle life (40–50 y).

Fig. 4. — Trajectories of CT and white matter FA as a function of age and cortical lobe. (A) Lines of best fit representing trajectories of cortical thickness (CT) and infracortical fractional anisotropy (FA; ribbon 4) in schizophrenia patients and healthy controls for each of 4 cortical lobes. Each data point denotes an individual (N = 385). Dashed lines denote 95% confidence intervals estimated from bootstrapping. (B) Between-group differences in CT and FA shown as a function of age using the percentage, Δ = (mean patients – mean controls) / mean controls × 100%. The open circles denote ages at which the between-group difference is significant (P < .05). Reduced CT occurred early in the disorder and remained a relatively stable deficit thereafter. In contrast, frontal infracortical anisotropy was increased in patients early in the disorder (ie, early adulthood), but deteriorated thereafter.

Fig. 5. — Annual rate of change in CT and white matter FA as a function of cortical lobe. Boxplots show the rate of decrease in cortical thickness (mm per y) and fractional anisotropy (FA; arbitrary units a.u. per y) in patients (red) and controls (blue). The P-values test the null hypothesis of equality in the rates of decline between patients and controls. While cortical thinning was a relatively stable deficit across each lobe, frontal and parietal infracortical white matter deteriorated at a significantly faster rate in patients compared to controls.

Discussion

We found that reductions in CT and infracortical white matter FA were inversely related across cortical regions in schizophrenia patients. Our cross-sectional analysis across ages spanning 20–65 years suggests that this association was due to the frontal cortex where extensive cortical thinning was coupled with an initial increase in adjacent white matter FA in young patients, followed by an accelerated decline resulting in global FA reductions by middle age.

According with prior work,²⁸ we observed a positive linear correlation between whole-brain gray and white matter measurements (mean CT and FA), however relationships were not found across regional CT and FA values (supplementary figure 3). A linear relationship between whole-brain measures is proposed to reflect dependence of CT measurements on the T1-weighted contrast, which is sensitive to cortical myelin levels.²⁸ Indeed, the relationship between CT and FA is heritable.²⁹ In contrast, age-related trajectories in white and gray matter show regionally-specific time-courses,¹¹ and this heterogeneity could diminish gray-white matter relationships measured across cortical regions.¹³ The absence of association between regional FA and CT measures suggests that our findings, involving pathological gray and white mater, were specifically linked with disease processes.

We observed increased FA in young patients, which may reflect increased infracortical myelination. This is supported by replication of our findings with RD (supplementary figure 7), a measure that provides greater specificity to myelin content.³⁰ While speculative, excess myelin may reflect a compensatory mechanism to aid cortical reorganization in response to initial pathology.^31,32 Furthermore, increased white matter anisotropy may hold functional significance, with a relationship found between increased white matter FA and cognitive task performance in normal aging,³³ particularly in older individuals displaying greater microstructural variability. Thus, increased myelin may support function under healthy or abnormal conditions involving white matter breakdown.³⁴

The most extensive cortical deficits were localized to frontal regions—a hotspot in the pathophysiology of schizophrenia—previously shown to emerge during pre-psychosis³⁵ and suggested to reflect disrupted brain maturation.³⁶ In late illness, we found global reductions in CT and white matter FA, suggesting that pathology across both tissue classes eventually affects all brain lobes. Notably, rapid FA decline was evident in white matter adjacent to regions with early CT deficits. Thus, initially spared or increased myelin may represent a short-term compensatory mechanism in order to aid function in neighboring regions with cortical shrinkage. However, such compensation may be unsustainable for long-term functioning. Persistent hyperactivity may place compensatory systems under metabolic stress and are prone to oxidative damage,³⁷ which may underlie the gradual decline of both anisotropy and thickness with age. Recent resting-state functional MRI (fMRI) findings may accord with compensation early in the illness that is eventually overwhelmed with illness progression. For example, Anticevic and colleagues³⁸ identified hyper-connectivity between prefrontal-thalamic regions that show reduced connectivity in individuals with chronic schizophrenia. These increases have been hypothesized to represent a compensatory mechanism for early insults.³⁹ However, it is important to remark that hyper-connectivity per se does not necessarily represent a compensatory response and may be due to disinhibition caused by reduced top-down modulation of subcortical systems.³⁷

Alternative to compensatory mechanisms, our findings may represent aberrant development of white matter connectivity or neuroinflammatory processes occurring in early stages of schizophrenia. Fronto-posterior connectivity undergoes marked development in late adolescence,⁴⁰ coinciding with the period of peak risk for illness onset. A developmental anomaly in this maturational process could result in some regions being hyper-connected and others being hypo-connected, altering signaling dynamics across the brain. Over time, this abnormal state of activity may lead to a more global pattern of deterioration. Neuroinflammation may represent a key component of such global deterioration, by contributing to axonal degeneration as the disease progresses. Furthermore, increased markers of neuroinflammation are evident at the outset of psychosis⁴¹ and have been related to increased positive symptom severity.^42,43 In line with these findings, supplemental analyses revealed a faster annual rate of white matter decline in patients with high positive symptoms (0.001079; P < .05) compared to patients with low positive symptoms (0.000425; supplementary figure 8).

Some limitations are worth consideration. First, we were unable to determine the influence of antipsychotic dose on the relationship between gray and white matter pathology. Therefore, it is possible that medication effects contributed to the observed association, particularly as some studies indicate a protective effect of atypical medication on white matter volume.^44,45 In addition, cumulative antipsychotic use has been linked to reduced gray matter volume and thickness,⁴⁶ but this relationship may vary between drug classes.⁴⁷ The relationship between gray and white matter found in our study remained significant after covarying for medication class (supplementary figure 4). In addition, the proportion of patients on atypical medication did not differ between younger (≤40 y = 85%) and older (≥40 y = 83%) patient groups and thus, atypical medication is unlikely to underlie the pattern of white matter decline observed as a function of age and illness duration. Second, while we replicated our findings using 3 alternative measures of white matter microstructure, 2 cortical parcellation schemes and controlled for regional volume, our MRI-derived indices are associated with several caveats, including difficulty in reconstructing complex fiber orientations with diffusion data,⁴⁸ eliminating the contribution of fiber dispersion to FA measured in infracortical white matter and head motion in CT estimation.⁴⁹ Therefore, topological analyses⁵⁰ and recently developed MRI sequences that estimate axonal diameter,⁵¹ neurite dispersion and density⁵² and fiber density⁵³ would strengthen future examinations. In addition, further work is needed to validate the accuracy with which cortical ribbons can be delineated, possibly with rodent studies. Third, the cross-sectional design employed by this study cannot directly predict specific trajectories of white or gray matter progression and therefore, a longitudinal design could offer superior characterization of gray-white matter interactions in aging or over the course of illness.

Conclusions

We found that regional variation in CT pathology was inversely related to the extent of disruption to regionally adjacent infracortical white matter, as measured with diffusion-weighted MRI. Specifically, young schizophrenia patients exhibited increased FA in infracortical white matter that was adjacent to cortical regions with the most extensive cortical thinning. This relation was posited to represent: (1) regionally-specific, short-term compensatory responses to cortical thinning, which are eventually overwhelmed with increasing illness duration, and/or (2) aberrant development of white matter connectivity in schizophrenia patients. This work underscores a need to examine mechanisms of spread and linkages between different cell populations in schizophrenia.

Funding

This research was supported by Melbourne Bioinformatics at the University of Melbourne, grant number UOM0014. Data for this study were provided by the Australian Schizophrenia Research Bank (ASRB), which is supported by the National Health and Medical Research Council of Australia (enabling grant 386500), the Pratt Foundation, Ramsay Health Care, the Viertel Charitable Foundation, and the Schizophrenia Research Institute. M.A.D. was supported by an Australian Rotary Health, Ian Scott PhD Scholarship in Mental Health. V.L.C. was supported by a National Health and Medical Research Council (NHMRC) Early Career Fellowship (628880), a Brain and Behavior Research Foundation (NARSAD) Young Investigator Award (21660), and a University of Melbourne Faculty of Medicine, Dentistry, and Health Sciences Research Fellowship. L.C. was supported by two NHMRC grants (APP1099082, APP1138711). A.F. was supported by an Australian Research Council Future Fellowship (FT130100589) and the Charles and Sylvia Viertel Foundation. F.C. was supported by an NHMRC Senior Research Fellowship. C.P. was supported by an NHMRC Senior Principal Research Fellowship (628386 and 1105825). A.Z. was supported by an NHMRC Research Fellowship (1047648). None of the funding sources played any role in the study design; in the collection, analysis, or interpretation of data; in the writing of the manuscript; or in the decision to submit the manuscript for publication.

Supplementary Material

sby121_suppl_Supplementary_Material

Click here for additional data file.^{(3MB, doc)}

Acknowledgments

The authors thank the chief investigators and manager of the Australian Schizophrenia Research Bank (ASRB) and acknowledge the help of Jason Bridge for ASRB database queries. The authors have declared that there are no conflicts of interest in relation to the subject of this study.

References

1. Kuperberg GR, Broome MR, McGuire PK, et al. Regionally localized thinning of the cerebral cortex in schizophrenia. Arch Gen Psychiatry. 2003;60:878–888. [DOI] [PubMed] [Google Scholar]
2. van Haren NE, Schnack HG, Cahn W, et al. Changes in cortical thickness during the course of illness in schizophrenia. Arch Gen Psychiatry. 2011;68:871–880. [DOI] [PubMed] [Google Scholar]
3. Kelly S, Jahanshad N, Zalesky A, et al. Widespread white matter microstructural differences in schizophrenia across 4322 individuals: results from the ENIGMA Schizophrenia DTI Working Group. Mol Psychiatry. 2018;23:1261–1269. [DOI] [PMC free article] [PubMed] [Google Scholar]
4. Arnold SE, Franz BR, Gur RC, et al. Smaller neuron size in schizophrenia in hippocampal subfields that mediate cortical-hippocampal interactions. Am J Psychiatry. 1995;152:738–748. [DOI] [PubMed] [Google Scholar]
5. Konopaske GT, Lange N, Coyle JT, Benes FM. Prefrontal cortical dendritic spine pathology in schizophrenia and bipolar disorder. JAMA Psychiatry. 2014;71:1323–1331. [DOI] [PMC free article] [PubMed] [Google Scholar]
6. Uranova NA, Vikhreva OV, Rachmanova VI, Orlovskaya DD. Ultrastructural alterations of myelinated fibers and oligodendrocytes in the prefrontal cortex in schizophrenia: a postmortem morphometric study. Schizophr Res Treatment. 2011;2011:325789. [DOI] [PMC free article] [PubMed] [Google Scholar]
7. Baumann N, Pham-Dinh D. Biology of oligodendrocyte and myelin in the mammalian central nervous system. Physiol Rev. 2001;81:871–927. [DOI] [PubMed] [Google Scholar]
8. Takahashi N, Sakurai T, Davis KL, Buxbaum JD. Linking oligodendrocyte and myelin dysfunction to neurocircuitry abnormalities in schizophrenia. Prog Neurobiol. 2011;93:13–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
9. Gogtay N, Giedd JN, Lusk L, et al. Dynamic mapping of human cortical development during childhood through early adulthood. Proc Natl Acad Sci USA. 2004;101:8174–8179. [DOI] [PMC free article] [PubMed] [Google Scholar]
10. Courchesne E, Chisum HJ, Townsend J, et al. Normal brain development and aging: quantitative analysis at in vivo MR imaging in healthy volunteers. Radiology. 2000;216:672–682. [DOI] [PubMed] [Google Scholar]
11. Kochunov P, Thompson PM, Lancaster JL, et al. Relationship between white matter fractional anisotropy and other indices of cerebral health in normal aging: tract-based spatial statistics study of aging. Neuroimage. 2007;35:478–487. [DOI] [PubMed] [Google Scholar]
12. Whitaker KJ, Vértes PE, Romero-Garcia R, et al. Adolescence is associated with genomically patterned consolidation of the hubs of the human brain connectome. Proc Natl Acad Sci USA. 2016;113:9105–9110. [DOI] [PMC free article] [PubMed] [Google Scholar]
13. Ehrlich S, Geisler D, Yendiki A, et al. Associations of white matter integrity and cortical thickness in patients with schizophrenia and healthy controls. Schizophr Bull. 2014;40:665–674. [DOI] [PMC free article] [PubMed] [Google Scholar]
14. Douaud G, Smith S, Jenkinson M, et al. Anatomically related grey and white matter abnormalities in adolescent-onset schizophrenia. Brain. 2007;130:2375–2386. [DOI] [PubMed] [Google Scholar]
15. Loughland C, Draganic D, McCabe K, et al. Australian Schizophrenia Research Bank: a database of comprehensive clinical, endophenotypic and genetic data for aetiological studies of schizophrenia. Aust N Z J Psychiatry. 2010;44:1029–1035. [DOI] [PubMed] [Google Scholar]
16. Fischl B, Dale AM. Measuring the thickness of the human cerebral cortex from magnetic resonance images. Proc Natl Acad Sci USA. 2000;97:11050–11055. [DOI] [PMC free article] [PubMed] [Google Scholar]
17. Desikan RS, Ségonne F, Fischl B, et al. An automated labeling system for subdividing the human cerebral cortex on MRI scans into gyral based regions of interest. Neuroimage. 2006;31:968–980. [DOI] [PubMed] [Google Scholar]
18. Destrieux C, Fischl B, Dale A, Halgren E. Automatic parcellation of human cortical gyri and sulci using standard anatomical nomenclature. Neuroimage. 2010;53:1–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
19. Zalesky A, Fornito A, Harding IH, et al. Whole-brain anatomical networks: does the choice of nodes matter?Neuroimage. 2010;50:970–983. [DOI] [PubMed] [Google Scholar]
20. Fischl B, Salat DH, Busa E, et al. Whole brain segmentation: automated labeling of neuroanatomical structures in the human brain. Neuron. 2002;33:341–355. [DOI] [PubMed] [Google Scholar]
21. Jenkinson M, Beckmann CF, Behrens TE, Woolrich MW, Smith SM. FSL. Neuroimage. 2012;62:782–790. [DOI] [PubMed] [Google Scholar]
22. Annese J, Pitiot A, Dinov ID, Toga AW. A myelo-architectonic method for the structural classification of cortical areas. Neuroimage. 2004;21:15–26. [DOI] [PubMed] [Google Scholar]
23. Kahn RS, Keefe RS. Schizophrenia is a cognitive illness: time for a change in focus. JAMA Psychiatry. 2013;70:1107–1112. [DOI] [PubMed] [Google Scholar]
24. Miller GA, Chapman JP. Misunderstanding analysis of covariance. J Abnorm Psychol. 2001;110:40–48. [DOI] [PubMed] [Google Scholar]
25. Di Biase MA, Cropley VL, Baune BT, et al. White matter connectivity disruptions in early and chronic schizophrenia. Psychol Med. 2017;47:2797–2810. [DOI] [PubMed] [Google Scholar]
26. Weston PS, Simpson IJ, Ryan NS, Ourselin S, Fox NC. Diffusion imaging changes in grey matter in Alzheimer’s disease: a potential marker of early neurodegeneration. Alzheimers Res Ther. 2015;7:47. [DOI] [PMC free article] [PubMed] [Google Scholar]
27. Le Bihan D. Looking into the functional architecture of the brain with diffusion MRI. Nat Rev Neurosci. 2003;4:469–480. [DOI] [PubMed] [Google Scholar]
28. Kochunov P, Glahn DC, Lancaster J, et al. Fractional anisotropy of cerebral white matter and thickness of cortical gray matter across the lifespan. NeuroImage. 2011;58:41–49. [DOI] [PMC free article] [PubMed] [Google Scholar]
29. Kochunov P, Glahn DC, Nichols TE, et al. Genetic analysis of cortical thickness and fractional anisotropy of water diffusion in the brain. Front Neurosci. 2011;5:120. [DOI] [PMC free article] [PubMed] [Google Scholar]
30. Alexander AL, Lee JE, Lazar M, Field AS. Diffusion tensor imaging of the brain. Neurotherapeutics. 2007;4:316–329. [DOI] [PMC free article] [PubMed] [Google Scholar]
31. Mierzwa AJ, Marion CM, Sullivan GM, McDaniel DP, Armstrong RC. Components of myelin damage and repair in the progression of white matter pathology after mild traumatic brain injury. J Neuropathol Exp Neurol. 2015;74:218–232. [DOI] [PMC free article] [PubMed] [Google Scholar]
32. Hoban AE, Stilling RM, Ryan FJ, et al. Regulation of prefrontal cortex myelination by the microbiota. Transl Psychiatry. 2016;6:e774. [DOI] [PMC free article] [PubMed] [Google Scholar]
33. Grydeland H, Walhovd KB, Tamnes CK, Westlye LT, Fjell AM. Intracortical myelin links with performance variability across the human lifespan: results from T1- and T2-weighted MRI myelin mapping and diffusion tensor imaging. J Neurosci. 2013;33:18618–18630. [DOI] [PMC free article] [PubMed] [Google Scholar]
34. Sugranyes G, Kyriakopoulos M, Dima D, et al. Multimodal analyses identify linked functional and white matter abnormalities within the working memory network in schizophrenia. Schizophr Res. 2012;138:136–142. [DOI] [PMC free article] [PubMed] [Google Scholar]
35. Bartholomeusz CF, Cropley VL, Wannan C, Di Biase M, McGorry PD, Pantelis C. Structural neuroimaging across early-stage psychosis: Aberrations in neurobiological trajectories and implications for the staging model. Aust N Z J Psychiatry. 2017;51:455–476. [DOI] [PubMed] [Google Scholar]
36. Pantelis C, Yücel M, Wood SJ, McGorry PD, Velakoulis D. Early and late neurodevelopmental disturbances in schizophrenia and their functional consequences. Aust N Z J Psychiatry. 2003;37:399–406. [DOI] [PubMed] [Google Scholar]
37. Fornito A, Zalesky A, Breakspear M. The connectomics of brain disorders. Nat Rev Neurosci. 2015;16:159–172. [DOI] [PubMed] [Google Scholar]
38. Anticevic A, Hu X, Xiao Y, et al. Early-course unmedicated schizophrenia patients exhibit elevated prefrontal connectivity associated with longitudinal change. J Neurosci. 2015;35:267–286. [DOI] [PMC free article] [PubMed] [Google Scholar]
39. Satterthwaite TD, Baker JT. How can studies of resting-state functional connectivity help us understand psychosis as a disorder of brain development?Curr Opin Neurobiol. 2015;30:85–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
40. Baker ST, Lubman DI, Yücel M, et al. Developmental changes in brain network hub connectivity in late adolescence. J Neurosci. 2015;35:9078–9087. [DOI] [PMC free article] [PubMed] [Google Scholar]
41. Pasternak O, Westin CF, Bouix S, et al. Excessive extracellular volume reveals a neurodegenerative pattern in schizophrenia onset. J Neurosci. 2012;32:17365–17372. [DOI] [PMC free article] [PubMed] [Google Scholar]
42. Takano A, Arakawa R, Ito H, et al. Peripheral benzodiazepine receptors in patients with chronic schizophrenia: a PET study with [11C]DAA1106. Int J Neuropsychopharmacol. 2010;13:943–950. [DOI] [PubMed] [Google Scholar]
43. Holmes SE, Hinz R, Drake RJ, et al. In vivo imaging of brain microglial activity in antipsychotic-free and medicated schizophrenia: a [11C]®-PK11195 positron emission tomography study. Mol Psychiatry. 2016;21:1672–1679. [DOI] [PubMed] [Google Scholar]
44. Bartzokis G, Lu PH, Nuechterlein KH, et al. Differential effects of typical and atypical antipsychotics on brain myelination in schizophrenia. Schizophr Res. 2007;93:13–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
45. Bartzokis G, Lu PH, Stewart SB, et al. In vivo evidence of differential impact of typical and atypical antipsychotics on intracortical myelin in adults with schizophrenia. Schizophr Res. 2009;113:322–331. [DOI] [PMC free article] [PubMed] [Google Scholar]
46. Fusar-Poli P, Smieskova R, Kempton MJ, Ho BC, Andreasen NC, Borgwardt S. Progressive brain changes in schizophrenia related to antipsychotic treatment? A meta-analysis of longitudinal MRI studies. Neurosci Biobehav Rev. 2013;37:1680–1691. [DOI] [PMC free article] [PubMed] [Google Scholar]
47. Krause D, Pogarell O. Shrinking brain and schizophrenia: a review of current studies on the effect of antipsychotic medication on gray matter volume. Psych Mental Disord. 2017;1:102. [Google Scholar]
48. Tuch DS, Reese TG, Wiegell MR, Wedeen VJ. Diffusion MRI of complex neural architecture. Neuron. 2003;40:885–895. [DOI] [PubMed] [Google Scholar]
49. Yao N, Winkler AM, Barrett J, et al. Inferring pathobiology from structural MRI in schizophrenia and bipolar disorder: modeling head motion and neuroanatomical specificity. Hum Brain Mapp. 2017;38:3757–3770. [DOI] [PMC free article] [PubMed] [Google Scholar]
50. Liu Y, Liang M, Zhou Y, et al. Disrupted small-world networks in schizophrenia. Brain. 2008;131:945–961. [DOI] [PubMed] [Google Scholar]
51. Assaf Y, Blumenfeld-Katzir T, Yovel Y, Basser PJ. AxCaliber: a method for measuring axon diameter distribution from diffusion MRI. Magn Reson Med. 2008;59:1347–1354. [DOI] [PMC free article] [PubMed] [Google Scholar]
52. Zhang H, Schneider T, Wheeler-Kingshott CA, Alexander DC. NODDI: practical in vivo neurite orientation dispersion and density imaging of the human brain. Neuroimage. 2012;61:1000–1016. [DOI] [PubMed] [Google Scholar]
53. Raffelt DA, Tournier JD, Smith RE, et al. Investigating white matter fibre density and morphology using fixel-based analysis. Neuroimage. 2017;144:58–73. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

sby121_suppl_Supplementary_Material

Click here for additional data file.^{(3MB, doc)}

PERMALINK

Linking Cortical and Connectional Pathology in Schizophrenia

Maria Angelique Di Biase

Vanessa L Cropley

Luca Cocchi

Alexander Fornito

Fernando Calamante

Eleni P Ganella

Christos Pantelis

Andrew Zalesky

Abstract

Introduction

Methods

Participants

MRI Acquisition

Table 1.

CT Estimates

Intra- and Infracortical Tissue Structure

Fig. 2.

Tractography

Statistical Modeling

Case-Control Differences in CT and FA.

Relation Between CT and FA.

Statistical Modeling of Age.

Results

Demographics and Symptoms

CT and White Matter Anisotropy

Case-Control Differences in CT and FA.

Relation Between CT and FA.

Fig. 1.

Influence of Age and Illness Duration

Relation Between CT and White Matter Anisotropy as a Function of Age.

Fig. 3.

CT and White Matter Anisotropy as a Function of Age and Cortical Lobe.

Fig. 4.

Fig. 5.

Discussion

Conclusions

Funding

Supplementary Material

Acknowledgments

References

Associated Data

Supplementary Materials

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases