Brain-Wide Functional Dysconnectivity in Schizophrenia: Parsing Diathesis,Resilience,and the Effects of Clinical Expression

Abstract

French

Background:

The functional dysconnectivity observed from functional magnetic resonance imaging (fMRI) studies in schizophrenia is also seen in unaffected siblings indicating its association with the genetic diathesis. We intended to apportion resting-state dysconnectivity into components that represent genetic diathesis, clinical expression or treatment effect, and resilience.

Methods:

fMRI data were acquired from 28 schizophrenia patients, 28 unaffected siblings, and 60 healthy controls. Based on Dosenbach’s atlas, we extracted time series of 160 regions of interest. After constructing functional network, we investigated between-group differences in strength and diversity of functional connectivity and topological properties of undirected graphs.

Results:

Using analysis of variance, we found 88 dysconnectivities. Post hoc t tests revealed that 62.5% were associated with genetic diathesis and 21.6% were associated with clinical expression. Topologically, we observed increased degree, clustering coefficient, and global efficiency in the sibling group compared to both patients and controls.

Conclusion:

A large portion of the resting-state functional dysconnectivity seen in patients represents a genetic diathesis effect. The most prominent network-level disruption is the dysconnectivity among nodes of the default mode and salience networks. Despite their predisposition, unaffected siblings show a pattern of resilience in the emergent connectomic topology. Our findings could potentially help refine imaging genetics approaches currently used in the pursuit of the pathophysiology of schizophrenia.

Keywords

functional connectivity schizophrenia sibling genetic risk graph theory

Introduction

Statistical dependencies among various brain regions in blood oxygen level–dependent (BOLD) signal are disrupted in patients with schizophrenia. Such disruptions observed using resting-state functional magnetic resonance imaging (fMRI) studies are taken to indicate macroscale functional dysconnectivity that is often invoked as a mechanistic explanation for the manifest-level disconnection in perception, action, emotion, and speech in patients with schizophrenia. Functional dysconnectivity is not localized to a single region or lobes but affects multiple information-processing subsystems including the default mode,^1,2 salience,^3,4 frontoparietal,^5,6 subcortical,⁷ and sensorimotor⁸ networks. When graph analytical approaches are employed to study the systemic effects of dysconnectivity at a whole-brain level, a pattern of reduced local clustering despite preserved global efficiency is noted,^9
–11 resulting in a disruption of overall topological organization of brain networks in schizophrenia.

Distributed functional dysconnectivity does not appear to be exclusive to patients but is also seen in their unaffected siblings indicating its association with the genetic diathesis of the illness.^12

–17 Nevertheless, when compared to patients, the extent of dysconnectivity appears to be limited both in spatial distribution and magnitude in siblings,^18

–21 suggesting that some of the abnormalities could be exclusively linked to the clinical expression or treatment effect rather than genetic diathesis. Our prior work indicates that siblings may be able to withstand some of the impairments in functional connectivity; for example, while the numerical proportion of distributed connections are reduced in both patients and siblings, siblings demonstrate a degree of strengthening of existing links, which is not seen in patients.¹⁶

We investigated brain-wide functional connectivity across six well-characterized functional networks (Dosenbach’s atlas of 160 regions²²) using the graph theory approach to apportion resting-state dysconnectivity into components that represent genetic diathesis, clinical expression or treatment effect, and resilience. We used a specific framework of hypothesis testing to parse diathesis, expression, and resilience effects. In this framework (Table 1; see also Supplemental Figure S1), features of dysconnectivity seen both in unaffected siblings and patients (compared to healthy controls) are considered as markers of genetic diathesis, while the changes seen in siblings when compared to patients, but not in patients compared to healthy controls, may represent protective features that mark resilience. On the other hand, neuroimaging abnormalities seen in patients versus healthy controls but not in siblings versus healthy controls contrast may represent the pathways proximal to clinical expression of a psychotic state (expression markers), as long as secondary effects of interventions can be ruled out.^23,24 Importantly, we did not have a priori expectations on regional localizations of dysconnectivity pertaining to diathesis, resilience, or expression, but based on prior works,^18

–21 we expected a large proportion of dysconnectivity to be attributable to genetic diathesis.

Table 1.

Statistical Approaches and Interpretation.

Contrasts in which Neuroimaging Differences Are Observed*			Phenotype Interpretation**
HC vs. SCZ	HC vs. SIB	SIB vs. SCZ	Phenotype Interpretation**
YES	YES	NO	Genetic diathesis effect
YES	NO	YES	Clinical expression or treatment effect
NO	YES	YES	Resilience effect
Other combinations of results			Inconclusive attribute

Note. HC = healthy controls; SCZ = schizophrenia patients; SIB = schizophrenia siblings.

* Significant for individual minimum t statistics; results in the same direction in both contrasts.

** If all three contrasts are positive for the individual minimum t statistic, we may infer that the observed differences to be shared features of expression and diathesis (e.g., genetic dose effect).

Methods

Participants

Twenty-eight patients with schizophrenia, 28 sex-matched siblings free of psychiatric disorders (1 per patient belong to the same family), and 60 healthy controls recruited from Changsha, China, were included in this study. The detailed description about sibling characteristics is included in Supplemental Materials.

Seven of the patients were drug naive, while the rest were receiving antipsychotic medications at the time of image acquisition. The clinical and demographic details of the samples were presented in Table 2 and in our previous works where we reported altered anatomical distance function and dynamic variability of functional connectivity in patients and siblings.^16,23 There was no significant difference between the three groups regarding sex, age, and education. All participants gave their written informed consent to participate in the study after the risks and benefits were discussed in detail. The study was approved by the ethics committee of the Second Xiangya Hospital, Central South University.

Table 2.

Demographic and Clinical Characteristics.

Variables	Schizophrenia Patients (n = 28)	Healthy Sibling (n = 28)	Healthy Control (n = 60)	P Value
Age (year)	25.3571 ± 5.8323; range (18 to 41)	25.7857 ± 6.4369; range (17 to 39)	27.1667 ± 6.6388; range (18 to 46)	0.3978
Sex (M/F)	15/13	15/13	35/25	0.8752
Education (year)	12.3929 ± 2.4846	12.5714 ± 2.7679	13.5167 ± 3.1596	0.1626
Illness duration (month)	18.32 ± 15.84	—	—	—
PANSS
Total score	85.9259 ± 12.584	—	—	—
Positive scale score	21.667 ± 4.795	—	—	—
Negative scale score	23.407 ± 5.759	—	—	—
General psychopathology scale score	40.880 ± 7.051	—	—	—

Note. PANSS = Positive and Negative Syndrome Scale.

Imaging Acquisition

Before scanning, the participants were explicitly instructed to lie supine, stay relaxed with their eyes closed, and move as little as possible. Images were acquired on a 1.5-Tesla GE SignaTwinspeed scanner (General Electric Medical System, Milwaukee, WI). A standard head coil was used for radio frequency transmission and reception of magnetic resonance signal. Foam pads and earplugs were used to minimize head motion and scanner noise, respectively. Functional images were acquired by using a gradient-echo echo-planar imaging sequence sensitive to BOLD signal (Repetition Time (TR)/Echo Time (TE) = 2,000/40 ms, flip angle = 90°, Field of View (FOV) = 240 mm × 240 mm). Whole-brain volumes were acquired with 20 contiguous 5-mm thick transverse slices with a 1-mm gap and 3.75 mm × 3.75 mm in-plane resolution. For each participant, the fMRI scanning lasted for 6 min.

Data Preprocessing

fMRI data preprocessing was then conducted by SPM8 and a Data Processing Assistant for Resting-State fMRI.²⁵ Briefly, the first 10 volumes were dropped to allow scanner stabilization;²⁶ the remaining functional scans were first corrected for within-scan acquisition time differences between slices and then realigned to the middle volume to correct for interscan head motions. Subsequently, the functional scans were spatially normalized to a standard template (Montreal Neurological Institute) and resampled to 3 mm × 3 mm × 3 mm, smoothed using a 4-mm Gaussian kernel. After that, BOLD signal of each voxel was firstly detrended to abandon linear trend and then passed through a band-pass filter (0.01 to 0.08 Hz) to reduce low-frequency drift and high-frequency physiological noise. Analysis of variance (ANOVA) results revealed a significant difference for framewise displacement of head motion parameter²⁷ among the three groups (Supplemental Materials for detail). Head motion parameters, global mean signals, white matter signals, and cerebrospinal signals were regressed out from the BOLD signals (rationale provided in the Supplemental Material). Dosenbach’s atlas²² was employed to parcellate the brain into 160 regions of interest (ROIs; 6 mm spheres for each ROI, six networks including cingulo-opercular, frontal-parietal, default, sensorimotor, occipital, cerebellum were provided), and the mean time courses were extracted from each ROI. After data preprocessing, the time series was extracted in each ROI by averaging the signals of all voxels within that region. Pearson correlation coefficients were computed between all pairs of ROIs, to obtain 160 × 160 correlation matrices r_ij , i, j = 1, 2,…, 160, indicating the functional connectivity strength for each pair of regions for each subject.

Functional Connectivity: Strength and Diversity

Connectivity strength is a global measure of connectivity. The regional strength of connectivity $\bar{R} (i)$ was likewise defined for the ith region as the average of the correlations between it and all other regions in the brain:

\bar{R} (i) = \frac{\sum_{j = 1}^{N} r_{i j}}{N - 1}, i = 1, 2, \dots, N,

where N is the number of region. The regional diversity of connections, Var(R(i)), was defined as the variance of the correlations between the ith index region and all other regions:

V a r (R (i)) = \frac{\sum_{j = 1}^{N} {(r_{i j} - \bar{R} (i))}^{2}}{N - 1}, i = 1, 2, \dots, N

Globally, connectivity diversity was defined as the average regional diversity across the N brain regions.

Graph Theoretic Analyses

Undirected graphs were constructed from the functional connectivity matrices. Any correlation r_ij , i, j = 1, 2,…, N in the functional connectivity matrix greater than a given threshold, τ, was retained as an edge connecting regions i and j in the adjacency matrix. That is, if r_ij < τ, no edge connects regions i and j. Graphs of different connection densities are produced by thresholding at different values of τ. In this study, the following topological properties were estimated for graphs that were binarized with respect to a series of connection strength thresholds ranging between 0.1 and 0.6 (in line with study by Zeng et al.²⁸ and Ganella et al.²⁹) applied to functional connectivity matrices of each individual. Note that there were no significant differences across the groups in terms of global functional connectivity strength, thus making it unlikely that one group would have had weaker edges than the other for a given density threshold.^30
–32 See Supplementary Materials for further consideration of this issue.

The patterns of relationship among brain regions can be described using three topological properties (importance of a node [Degree], degree of overall integration [Global Efficiency], and regional segregation [Clustering Coefficient]).³³

Degree, K(i), is simply equal to the number of edges connecting the ith region to the rest of the network:

K (i) = \sum_{j} A_{i j}, i = 1, 2, \dots, N,

where A = (A_ij ), i, j = 1, 2,…, N is the binary adjacency matrix obtained by thresholding the functional connectivity matrix. If r_ij < τ, A_ij = 0; if r_ij ≥ τ, A_ij = 1. Global degree of the graph is the mean degree over all nodes.

Global efficiency, Eg(i), is computed for each node in a graph, as follows:

E g (i) = \frac{1}{N - 1} \sum_{j} \frac{1}{L_{i j}}

Here L_ij is the minimum path length between regions i and j. Global efficiency of the graph is the mean regional efficiency over all of the nodes.

Clustering coefficient, C(i) of a node i (with degree > 2) is the ratio of connected triangles, δ _i , to connected triples, τ _i , in the subgraph of node i. The clustering coefficient of a graph is as follows:

C (G) = \frac{1}{| V |} \sum_{i \in V} \frac{δ_{i}}{τ_{i}}

where V is the set of nodes with degree > 2.

Statistical Analysis

Correlation coefficients were first normalized with the application of the Fisher’s z transform before statistical analysis. Two kinds of statistical analyses were performed in this study. ANOVA was first performed to test the null hypothesis of equality among the three groups: schizophrenia patients, siblings, and healthy controls, with false discovery rate (FDR) correction. Post hoc t tests were then performed to test the null hypothesis of between-groups equality. Specifically, two-sample t test was performed between (1) schizophrenia patients and matched healthy controls (SCZ vs. HC), (2) schizophrenia siblings and matched healthy controls (SIB vs. HC), and (3) paired t test between schizophrenia patients and matched schizophrenia siblings (SCZ vs. SIB) with age, sex, and root mean square displacements of head movement being regressed out. We used a conjunction analysis to infer the genetic, diathesis, and resilience effects as outlined in Table 1. Support vector machine (SVM) was used to supplement the above analysis. SVM methods are presented in the Supplemental Materials.

Results

Pair-Wise Connections: Diathesis, Resilience, and the Effects of Clinical Expression

Using ANOVA (FDR corrected P < 0.05), we found 88 of 12,720 pairs of functional links to be significantly different among the three groups which are listed in Supplemental Table S2. 48.8% of these 88 links included nodes from the default mode network (DMN), as shown in Figure 1A. Figure 1B is the proportion of the intra- and interconnectivity for each network, 38.6% of the 88 significant links is interconnectivity from DMN, with the largest portion involving salience network/DMN connectivity (14.8%). The detailed results are listed in Supplemental Table S2.

Figure 1.

(A) Proportion of links for each network of all 88 dysconnectivities. (B) Proportion of the intra- and interconnectivity for each network. (C) Dysconnectivities show genetic diathesis effect, clinical expression or treatment effect, resilience effect, and inconclusive effect.

Post hoc t tests were performed on each significant connection to classify them into three mutually exclusive categories: (i) genetic diathesis effect, (ii) clinical expression or treatment effect, and (iii) resilience effect, as shown in Table 1. The majority of connections (n = 55, 62.5% of significant connections) show genetic diathesis effect, which represents abnormal functional connectivities in both schizophrenia patients and siblings. A smaller number of connections (n = 19, 21.6%) show clinical expression or treatment effect, which represents abnormal functional connectivities in schizophrenia patients, but not in siblings. In contrast, far fewer connections (n = 6, 7%) show resilience effect, which represents abnormal functional connectivity in siblings when compared to the other two groups, but not apparent in patients versus control comparisons. Eight connections (9%) showed inconclusive effect, as shown in Figure 1C and Supplemental Table S2.

When exploring the correlation between the strength of the 19 links related to clinical expression and Positive and Negative Syndrome Scale positive, negative, and general scores, only one link (inf cerebellum.L–mid insula.L) significantly correlated with positive and general score, but this did not survive FDR correction (more details in the Supplemental Material).

Brain-Wide Connectivity: Strength and Diversity

For each of 160 anatomically defined brain regions, we estimated the strength and diversity of its functional connectivity to the rest of the brain. The overall strength (average of all ROIs: Supplemental Figure S3) was not significantly different among the three groups, mean (SD) strength: 0.0189 (0.0329) for siblings, 0.0151 (0.0283) for patients, and 0.0158 (0.0312) for healthy controls. As shown in Supplemental Figure S4, for individual nodes, functional connectivity strength was numerically smaller in patients while greater in siblings over different brain regions than in healthy controls. With FDR correction, only one node (inf cerebellum.L, P = 4.8492 × 10⁻⁴) was significantly different among the three groups. Post hoc t test of this node indicates it is an inconclusive attribute as per Table 1 (P = 3.4513 × 10⁻⁴ for HC vs. SCZ, P = 6.4514 × 10⁻⁴ for SCZ vs. SIB, P = 0.5377 for HC vs. SIB; see Figure 2 for a bar plot).

Figure 2.

Upper panels display brain regions with significant group differences among patients, siblings, and healthy control subjects. Group differences in functional connectivity strength (A) and functional diversity (B). Post hoc differences that survived false discovery rate correction (P < 0.05) are shown with an asterisk (*). Lower panels are plotted as a function of the graph binarized threshold for healthy controls (blue), schizophrenia patients (red), and siblings (green) groups for (C) Degree, (D) global efficiency, (E) clustering coefficient. The inner subplot is the error bar for each group.

Functional connectivity diversity was generally greater in siblings while was generally smaller in schizophrenia patients over different brain regions than in healthy controls, mean (SD) diversity: 0.2408 (0.0501) for sibling, 0.2247 (0.0469) for patients, 0.2288 (0.0489) for healthy controls, as shown in Supplemental Figure S4. Four ROIs were significantly different among three groups under FDR correction (occipital.L, P = 0.0008; lat cerebellum.L, P = 0.0015; vlPFC.R, P = 0.0016; fusiform.R, P = 0.0025). Post hoc t test of these ROIs indicates that vlPFC.R and fusiform.R are genetic diathesis effects while occipital.L and lat cerebellum.L are inconclusive attributes. The results are shown in Figure 2 and Supplemental Table S4.

Global Network Topology

We measured the topological properties of binary graphs derived by thresholding with respect to a series of connection strength thresholds ranging between 0.1 and 0.6 applied to individual functional connectivity matrices. The threshold is commonly used in prior studies.^28,29 The area under curve (AUC) was used as a summary statistic of global degree, global efficiency, and clustering coefficient.

The null hypothesis of equality in AUC between HC, SCZ, and SIB was rejected for degree (F = 3.56, P = 0.032), global efficiency (F = 5.11, P = 0.008), and clustering coefficient (F = 4.36, P = 0.015). Topologically, we observed a significant increase in global degree, clustering coefficient, and global efficiency in the sibling group compared to both patients and controls, indicating a resilience (or endurance) effect for these measures, as shown in Table 3. SVM results are presented in Supplemental Materials.

Table 3.

Mean (SD) of Area under the Curve of the Topological Measures and P Value of Post Hoc t Tests between Different Groups.

Topological measures	HC	SCZ	SIB	P Value (t Statistics)
Degree	166.67 (28.85)	162.06 (26.63)	181.89 (34.43)	SCZ vs. HC: 0.4954 HC vs. SIB: 0.032* SCZ vs. SIB: 0.0317*
Global efficiency	3.49 (0.36)	3.47 (0.37)	3.74 (0.37)	SCZ vs. HC: 07366 HC vs. SIB: 0.0046* SCZ vs. SIB: 0.0108*
Clustering coefficient	3.8 (0.75)	3.56 (0.75)	4.12 (0.64)	SCZ vs. HC: 0.1682 HC vs. SIB: 0.0481* SCZ vs SIB: 0.0052*

Note. HC = healthy controls; SCZ = schizophrenia patients; SIB = schizophrenia siblings.

*contrasts that were significant at the threshold of p < 0.05.

Discussion

Using resting fMRI to study the effects of diathesis, resilience, and clinical expression on brain-wide functional dysconnectivity in Schizophrenia, we report three major findings: (1) A major proportion (62.5%) of pairwise functional dysconnectivity seen in patients from schizophrenia can be attributed to the genetic diathesis to develop this illness. (2) DMN is the most affected subsystem with respect to functional connectivity in patients and siblings, with DMN–salience network interactions being the most prominent cross-network abnormality. (3) Changes that mark resilience in siblings are most prominent when examining the topological architecture of brain-wide connectivity. We also report that using SVM approach to supplement these findings, the discriminatory ability to separate HC from patients, siblings, and the combined patients and siblings group is excellent, while it is not possible to separate SIB from SCZ using the same set of connectivity features (see Supplemental Materials for details). Taken together, these results support a genetic basis for the origin of specific functional network disruptions seen in schizophrenia, the effects of which appear to be modulated by mounting global topological reorganization in unaffected siblings.

In line with our observation that genetic diathesis is the major contributor of functional dysconnectivity, quantitative studies of heritability and environmental effects suggest that functional dysconnectivity observed in schizophrenia has a significant genetic basis.¹³ In addition, task-based fMRI studies^34
–36 and Electroencephalographic (EEG) studies³⁷ corroborate the notable similarities between patients and siblings in the degree of aberrant functional connectivity. While these prior studies focused on investigating the shared “trait-like” endophenotypes in connectivity, we separated the specific features of genetic diathesis from that of clinical expression using a stringent statistical framework. Nevertheless, it is important to note that our statistical partitioning does not imply that diathesis and expression are discrete phenomena. A continuous metric of genetic change (often quantified as polygenic risk score in recent studies) may underlie both diathesis and expression. Similarly, resolving the compensatory effects from the effects of symptoms requires multiple observations of both illness severity and brain features across time, which we lacked in this study (see Supplemental Figure S1).

Several prior studies report shared abnormalities between patients and siblings in the DMN.^1,20 We report that of all brain networks, the DMN is the most influenced by genetic diathesis for schizophrenia. In particular, its interaction with the cingulo-opercular system (salience network, comprising of insula, anterior cingulate, and thalamic nodes) seems to be the most genetically constrained abnormality relevant to this illness. The connectivity of DMN itself is deemed to be highly heritable in the healthy population.³⁸ Our observation that the most prominent effect of genetic diathesis converge on DMN suggests possible overlap between the polygenic determinants of DMN connectivity and schizophrenia. These results are also consistent with previous observations in the same data set, where we reported high degree of dynamic instability of the precuneus node in the DMN²³ as well as the disproportionate reduction in long-distance connections¹⁶ in patients with schizophrenia and siblings.

To our knowledge, we report for the first time that topological features of brain network organization are heavily influenced by resilience or endurance against the expression of schizophrenia. This observation also indicates that neurobiological resilience is best considered as a systemic (“gestalt”) phenomenon, rather than being a product of focused regional adaptations. In unaffected siblings, the presence of higher clustering coefficient may represent superior local specialization whereas higher global efficiency may enable an economical distributed processing of information represented by BOLD signal dependencies. In terms of topology of complex systems, the combination of higher clustering and higher global efficiency represents a robust protection against random failures that could occur in the system.^39,40

There are several limitations in the presented work. Firstly, the data were collected using a 1.5-T MRI scanner; in general, higher field strength can improve anatomical resolution of the connectivity estimates, but there is a notable concordance among the estimated connectivity metrics across scanners of various field strength, especially when spatial smoothing is employed. Second, there are many parcellation approaches that can be employed to derive nodes of interest for graph analysis. We used an established anatomical atlas based on prior knowledge of functional network localization. Third, the age of siblings matched with the age of patients and healthy subjects at a group level; nevertheless, as some cases of schizophrenia can have onset after the mean age of siblings in this sample, we cannot consider that they were absolutely discordant from patients in terms of clinical expression. In the presence of such unobserved clinical concordance, the direction of patients versus siblings difference would be biased toward the null, with eventual interpretations biased toward the genetic diathesis rather than resilience model. Nevertheless, as we purposefully recruited the non-help-seeking siblings with no current or prior contact with psychiatric services, and interviewed using the Structured Clinical Interview Nonpatient Version for DSM-IV, to exclude any Axis I/II features. Thus, at the time of obtaining the imaging data, none of the siblings had any clinical features overlapping with patients. Fourth, we lacked nonpsychotic symptom data for siblings, precluding the study of nonpsychotic clinical expression phenotype. Most of our patients were treated with antipsychotics. While the dose prescribed was not related to almost all dysconnectivities, given the lack of lifetime exposure data, plasma drug levels or compliance metrics, we cannot fully rule out a treatment effect on our results. A sample of treatment-naive subjects and matched siblings will be required to confidently rule out this confound. When testing our primary hypothesis, the smallest degree of freedom (df = 54) pertains to the comparison between 28 SCZ versus 28 SIB. To make inferences at the significance level of α = 0.05, with 1:1 group allocation for a two-tailed test, with 80% power, our sample had an ability to detect an effect size d = 0.76 of difference between the two groups. In contrast, for the two comparisons involving the healthy controls (df = 86 each, with 2:1 allocation ratio), our sample had an ability to detect an effect size d = 0.65 of difference between the groups. Thus, all of our comparisons had the ability to detect medium to large effect size differences, corresponding to reporting >33% nonoverlap between groups.⁴¹ Nevertheless, given that our sample had 15.6% more sensitivity to report differences of SCZ and SIB groups from HC, than between SCZ and SIB groups, we urge caution and further replication before generalizing these results.

In summary, our observations support the notion that genetic diathesis is the most critical factor influencing network-level functional dysconnectivity in schizophrenia, whereas resilience against symptom expression may be conferred by topological reorganization in the genetically primed brain. On a speculative note, this also raises the question of whether putative interventions to delay or prevent onset of psychosis are likely to reorganize the functional topology toward higher resilience, as seen in the siblings sample in our study.

Supplemental Material

Supplemental Material, 19037-c-CJP-2019-178-OR.R2-Abstract__for_translation - Brain-Wide Functional Dysconnectivity in Schizophrenia: Parsing Diathesis, Resilience, and the Effects of Clinical Expression

Supplemental Material, 19037-c-CJP-2019-178-OR.R2-Abstract__for_translation for Brain-Wide Functional Dysconnectivity in Schizophrenia: Parsing Diathesis, Resilience, and the Effects of Clinical Expression by Shuixia Guo, Ningning He, Zhening Liu, Zeqiang Linli, Haojuan Tao and Lena Palaniyappan in The Canadian Journal of Psychiatry

Supplemental Material

Supplemental Material, supplement20190916 - Brain-Wide Functional Dysconnectivity in Schizophrenia: Parsing Diathesis, Resilience, and the Effects of Clinical Expression

Supplemental Material, supplement20190916 for Brain-Wide Functional Dysconnectivity in Schizophrenia: Parsing Diathesis, Resilience, and the Effects of Clinical Expression by Shuixia Guo, Ningning He, Zhening Liu, Zeqiang Linli, Haojuan Tao and Lena Palaniyappan in The Canadian Journal of Psychiatry

Footnotes

Authors’ Note

The study was designed by Lena Palaniyappan, Shuixia Guo, and Zhening Liu. Haojuan Tao and Zhening Liu collected the data from Second Xiangya Hospital, Central South University, had full access to all of the data in the study, and takes responsibility for the integrity of the data. Lena Palaniyappan developed the analytical plan. Shuixia Guo, Ningning He, and Zeqiang Linli undertook the statistical analyses and take responsibility for its accuracy. Shuixia Guo and Lena Palaniyappan wrote the first draft of the manuscript. All authors contributed to writing the manuscript. Reasonable requests for access to the data should be directed to the corresponding authors. The data reported in this study are archived as per institutional requirements.

Declaration of Conflicting Interests

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Lena Palaniyappan reports personal fees from Janssen Canada, Otsuka Canada, SPMM Course Limited, UK, and the Canadian Psychiatric Association; book royalties from Oxford University Press; investigator-initiated educational grants from Sunovion, Janssen Canada, and Otsuka Canada; and travel support from Boehringer Ingelheim and Magstim Limited, outside the submitted work. In the last 5 years, L. Palaniyappan and/or his spouse have held shares in Shire Pharmaceuticals and GlaxoSmithKline in their pension funds for values less than US$10,000.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Zhening Liu is supported by the China Precision Medicine initiative grant (2016YFC0906300) and the National Natural Science Foundation of China (NSFC) grant (81561168021). Shuixia Guo is supported by the NSFC grant (11671129 and 31671134). Haojuan Tao is supported by the NSFC grant (81301161). Lena Palaniyappan acknowledges salary support from the Tanna Schulich Chair of Neuroscience and Mental Health and the Opportunities Fund of the Academic Health Sciences Centre Alternative Funding Plan of the Academic Medical Organization of Southwestern Ontario (AMOSO).

ORCID iD

Shuixia Guo

Supplemental Material

Supplemental material for this article is available online.

References

Whitfield-Gabrieli

Ford

. Default mode network activity and connectivity in psychopathology. Annu Rev Clin Psychol. 2012;8:49–76.

Anticevic

Cole

Murray

Corlett

Wang

Krystal

. The role of default network deactivation in cognition and disease. Trends Cogn Sci. 2012;16(12):584–592.

Palaniyappan

Liddle

. Does the salience network play a cardinal role in psychosis? An emerging hypothesis of insular dysfunction. J Psychiatry Neurosci. 2012;37(1):17.

Phan

Taylor

Welsh

Britton

Liberzon

. Neural correlates of individual ratings of emotional salience: a trial-related fMRI study. Neuroimage. 2004;21(2):768–780.

Cole

Repovš

Anticevic

. The frontoparietal control system: a central role in mental health. Neuroscientist. 2014;20(6):652–664.

Baker

Holmes

Masters

, et al. Disruption of cortical association networks in schizophrenia and psychotic bipolar disorder. JAMA Psychiatry. 2014;71(2):109–118.

Woodward

Coull

. Localization and organization of geniculocortical and corticofugal fiber tracts within the subcortical white matter. Neuroscience. 1984;12(4):1089–1099.

Grodd

Hülsmann

Lotze

Wildgruber

Erb

. Sensorimotor mapping of the human cerebellum: fMRI evidence of somatotopic organization. Hum Brain Mapp. 2001;13(2):55–73.

Kambeitz

Kambeitz-Ilankovic

Cabral

, et al. Aberrant functional whole-brain network architecture in patients with schizophrenia: a meta-analysis. Schizophr Bull. 2016;42(Suppl 1):S13–S21.

10.

Lynall

Bassett

Kerwin

, et al. Functional connectivity and brain networks in schizophrenia. J Neurosci. 2010;30(28):9477–9487.

11.

CYZ

Huang

, et al. Randomization and resilience of brain functional networks as systems-level endophenotypes of schizophrenia. Proc Natl Acad Sci USA. 2015;112(29):9123–9128.

12.

Peeters

SCT

van Bronswijk

van de Ven

, et al. Cognitive correlates of frontoparietal network connectivity ‘at rest’ in individuals with differential risk for psychotic disorder. Eur Neuropsychopharmacol. 2015;25(11):1922–1932.

13.

Peeters

SCT

van de Ven

Gronenschild

EHBM

, et al. Default mode network connectivity as a function of familial and environmental risk for psychotic disorder. PLoS One. 2015;10(3):e0120030.

14.

Chang

Shen

Wang

, et al. Altered default mode and fronto-parietal network subsystems in patients with schizophrenia and their unaffected siblings. Brain Res. 2014;1562:87–99.

15.

Saoud

d’Amato

Gutknecht

, et al. Neuropsychological deficit in siblings discordant for schizophrenia. Schizophr Bull. 2000;26(4):893.

16.

Guo

Palaniyappan

Yang

Liu

Xue

Feng

. Anatomical distance affects functional connectivity in patients with schizophrenia and their siblings. Schizophr Bull. 2013;40(2):449–459.

17.

Danesh

Youngman

Clark

Parish

Peto

Collins

. Helicobacter pylori infection and early onset myocardial infarction: case-control and sibling pairs study. BMJ. 1999;319(7218):1157–1162.

18.

Bassett

Meyer-Lindenberg

Achard

Duke

Bullmore

. Adaptive reconfiguration of fractal small-world human brain functional networks. Proc Natl Acad Sci USA. 2006;103(51):19518–19523.

19.

Shen

Zhang

Zeng

Xue

. Functional connectivity-based signatures of schizophrenia revealed by multiclass pattern analysis of resting-state fMRI from schizophrenic patients and their healthy siblings. Biomed Eng Online. 2013;12(1):10.

20.

van Buuren

Vink

Kahn

. Default-mode network dysfunction and self-referential processing in healthy siblings of schizophrenia patients. Schizophr Res. 2012;142(1):237–243.

21.

Schimmelmann

Michel

Martz-Irngartinger

, et al. Age matters in the prevalence and clinical significance of ultra-high-risk for psychosis symptoms and criteria in the general population: findings from the BEAR and BEARS-kid studies. World Psychiatry. 2015;14(2):189–197.

22.

Dosenbach

NUF

Nardos

Cohen

, et al. Prediction of individual brain maturity using fMRI. Science. 2010;329(5997):1358–1361.

23.

Guo

Zhao

Tao

Liu

Palaniyappan

. The instability of functional connectivity in patients with schizophrenia and their siblings: a dynamic connectivity study. Schizophr Res. 2018;195:183–189.

24.

Palaniyappan

Balain

Liddle

. The neuroanatomy of psychotic diathesis: a meta-analytic review. J Psychiatr Res. 2012;46(10):1249–1256.

25.

Yan

Zang

. DPARSF: a MATLAB toolbox for “pipeline” data analysis of resting-state fMRI. Front Syst Neurosci. 2010;4:13.

26.

Bandettini

Jesmanowicz

Van Kylen

Birn

Hyde

. Functional MRI of brain activation induced by scanner acoustic noise. Magn Reson Med. 1998;39(3):410–416.

27.

Power

Barnes

Snyder

, et al. Spurious but systematic correlations in functional connectivity MRI networks arise from subject motion. Neuroimage. 2012;59(3):2142–2154.

28.

Zeng

Shen

Liu

, et al. Identifying major depression using whole-brain functional connectivity: a multivariate pattern analysis. Brain. 2012;135:1498–1507.

29.

Ganella

Seguin

Bartholomeusz

, et al. Risk and resilience brain networks in treatment-resistant schizophrenia. Schizophr Res. 2018;193:284–292.

30.

van den Heuvel

de Lange

Zalesky

Seguin

Yeo

BTT

Schmidt

. Proportional thresholding in resting-state fMRI functional connectivity networks and consequences for patient-control connectome studies: issues and recommendations. Neuroimage. 2017;152:437–449.

31.

Váša

Bullmore

Patel

. Probabilistic thresholding of functional connectomes: application to schizophrenia. Neuroimage. 2018;172:326–340.

32.

Váša

Seidlitz

Romero-Garcia

, et al. Adolescent tuning of association cortex in human structural brain networks. Cereb Cortex. 2017;28(1):281–294.

33.

Rubinov

Sporns

. Complex network measures of brain connectivity: uses and interpretations. Neuroimage. 2010;52(3):1059–1069.

34.

Whalley

Harris

Lawrie

. The neurobiological underpinnings of risk and conversion in relatives of patients with schizophrenia. Int Rev Psychiatry. 2007;19(4):383–397.

35.

Rasetti

Mattay

Wiedholz

, et al. Evidence that altered amygdala activity in schizophrenia is related to clinical state and not genetic risk. Am J Psychiatry. 2009;166(2):216–225.

36.

Repovs

Csernansky

Barch

. Brain network connectivity in individuals with schizophrenia and their siblings. Biol Psychiatry. 2011;69(10):967–973.

37.

Winterer

Egan

Rädler

Hyde

Coppola

Weinberger

. An association between reduced interhemispheric EEG coherence in the temporal lobe and genetic risk for schizophrenia. Schizophr Res. 2001;49(1):129–143.

38.

Meda

Ruaño

Windemuth

, et al. Multivariate analysis reveals genetic associations of the resting default mode network in psychotic bipolar disorder and schizophrenia. Proc Natl Acad Sci USA. 2015;111(19):E2066–E2075.

39.

Achard

Bullmore

. Efficiency and cost of economical brain functional networks. PLoS Comput Biol. 2007;3(2):e17.

40.

Bullmore

Sporns

. Complex brain networks: graph theoretical analysis of structural and functional systems. Nat Rev Neurosci. 2009;10(3):186.

41.

Sullivan

Feinn

. Using effect size—Or why the P value is not enough. J Grad Med Educ. 2012;4(3):279–282.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.01 MB

0.44 MB