Disrupted dynamic local brain functional connectivity patterns in generalized anxiety disorder

Disrupted dynamic local brain functional connectivity patterns in generalized anxiety disorder

Progress in Neuropsychopharmacology & Biological Psychiatry 99 (2020) 109833 Contents lists available at ScienceDirect Progress in Neuropsychopharma...

1MB Sizes 0 Downloads 10 Views

Progress in Neuropsychopharmacology & Biological Psychiatry 99 (2020) 109833

Contents lists available at ScienceDirect

Progress in Neuropsychopharmacology & Biological Psychiatry journal homepage: www.elsevier.com/locate/pnp

Disrupted dynamic local brain functional connectivity patterns in generalized anxiety disorder


Qian Cuia, , Yuyan Chenb, Qin Tangb, Shaoqiang Hanb, Shan Hua, Yajing Pangb, Fengmei Lub, ⁎ Xiaoyu Nana, Wei Shengb, Qian Shenc, Yifeng Wangb, Zongling Heb, Huafu Chenb, ⁎


School of Public Affairs and Administration, University of Electronic Science and Technology of China, Chengdu, China The Clinical Hospital of Chengdu Brain Science Institute, MOE Key Lab for Neuroinformation, School of Life Science and Technology, University of Electronic Science and Technology of China, Chengdu, China c Education Center for Students Cultural Qualities, University of Electronic Science and Technology of China, Chengdu, China b



Keywords: Generalized anxiety disorder Resting-state functional magnetic resonance imaging Dynamic regional phase synchrony

Previous studies have reported abnormalities in static brain activity and connectivity in patients with generalized anxiety disorder (GAD). However, the dynamic patterns of brain connectivity in patients with GAD have not been fully explored. In this study, we aimed to investigate the dynamic local brain functional connectivity in patients with GAD using dynamic regional phase synchrony (DRePS), a newly developed method for assessing intrinsic dynamic local functional connectivity. Seventy-four patients with GAD and 74 healthy controls (HCs) were enrolled and underwent resting-state functional magnetic resonance imaging. Compared to the HCs, patients with GAD exhibited decreased DRePS values in the bilateral caudate, left hippocampus, left anterior insula, left inferior frontal gyrus, and right fusiform gyrus extending to inferior temporal gyrus. The DRePS value of the left hippocampus was negatively correlated with the Hamilton Anxiety Rating Scale scores. Moreover, these abnormal DRePS patterns could be used to distinguish patients with GAD from HCs in an independent sample (18 patients with GAD and 21 HCs). Our findings provide further evidence on brain dysfunction in GAD from the perspective of the dynamic behaviour of local connections, suggesting that patients with GAD may have an insufficient brain adaptation. This study provides new insights into the neurocognitive mechanism of GAD and could potentially inform the diagnosis and treatment of this disease. Future studies on GAD could benefit from combining the DRePS method with task-related functional magnetic resonance imaging and non-invasive brain stimulation.

1. Introduction Generalized anxiety disorder (GAD) is one of the most common mental disorders. It is characterized by excessive and persistent worry and anxiety (Goossen et al., 2019; Mochcoyitch et al., 2014) and is highly prevalent worldwide (Wittchen and Jacobi, 2005; Wittchen et al., 2011). Patients with GAD often suffer from a variety of anxietyrelated physical symptoms, such as concentration difficulties, irritability, muscle tension, and disturbed sleep, which cause them to have a poor quality of life and diminished social function (Cuijpers et al., 2014). Although many studies have investigated functional and structural abnormalities in GAD, its neuropathological mechanisms remain unclear (Hilbert et al., 2014). Elucidating these mechanisms and exploring potential neuro-markers could improve the understanding of GAD and may contribute to the enhancement of clinical treatment for

GAD. Various neuroimaging techniques have been employed to explore GAD pathogenesis. Patients with GAD have been reported to have brain dysfunction in widely distributed brain regions, including the prefrontal cortex, parietal cortex, temporal gyrus, and the limbic system (Chen and Etkin, 2013; Etkin et al., 2009; Li et al., 2016; Roy et al., 2013). Consistent with these findings, structural alterations have also been repeatedly reported in patients with GAD, such as grey matter volume alterations in the amygdala (De Bellis et al., 2000; Makova et al., 2016; Schienle et al., 2011), reduced cortical thickness in the frontal gyrus, and hypo-gyrification in the inferior temporal gyrus (ITG) (Molent et al., 2018). At the network level, GAD shows widespread abnormal functional connectivity of regions in the limbic system (Chen and Etkin, 2013; Roy et al., 2013), executive system, and default mode network (Andreescu et al., 2014) with the whole brain. Furthermore,

Corresponding authors. E-mail addresses: [email protected] (Q. Cui), [email protected] (H. Chen).

https://doi.org/10.1016/j.pnpbp.2019.109833 Received 21 March 2019; Received in revised form 1 November 2019; Accepted 3 December 2019 Available online 05 December 2019 0278-5846/ © 2019 Published by Elsevier Inc.

Progress in Neuropsychopharmacology & Biological Psychiatry 99 (2020) 109833

Q. Cui, et al.

connectivity disruptions have been reported between the default mode, salience, and central executive networks in patients with GAD (Chen and Etkin, 2013; Li et al., 2016). These brain network abnormalities are correlated with GAD symptoms (Andreescu et al., 2014; Roy et al., 2013). Moreover, patients with GAD exhibit atypical brain activation in the anterior cingulate cortex, prefrontal cortex, and limbic regions when exposed to emotional stimuli, especially threatening ones that could induce anxiety (McClure et al., 2007; Monk et al., 2008; Via et al., 2018). These findings have been associated with maladaptive responses to the external environment, which is a typical clinical characteristic of GAD (Grupe and Nitschke, 2013; McClure et al., 2007; Shechner and Bar-Haim, 2016; Via et al., 2018). Most previous functional magnetic resonance imaging (fMRI) studies on GAD have assessed static brain activity. However, there have been increasing reports of the dynamic nature of brain activity and connections (Hutchison et al., 2013; Liu and Duyn, 2013; Zalesky et al., 2014). As an emerging field in computational neuroscience (Omidvarnia et al., 2016), dynamic functional connectivity (dFC) analysis allows flexibility to be assessed in the functional coordination between different neural systems (Allen et al., 2014); this strategy has been applied to detect variations in the temporal FC patterns in patients with GAD. Moreover, dFC can be used as a feature to distinguish patients with GAD subtypes from healthy controls (HCs) (Li et al., 2019; Yao et al., 2017). Nevertheless, these very preliminary dFC studies on GAD only focused on inter-regional functional connectivity. As a modular and hierarchical organization, the brain is comprised of several local functional units that are clustered with adjacent “neighbourhoods”, in which the local functional units are further integrated at the global level (Park and Friston, 2013). This “neighbourhood” configuration is a key aspect of brain functionality and its changes are related to cognitive state shifts (Gu et al., 2015). The dynamic organization of “local neighbourhoods” is important for brain function and can be evaluated using a newly proposed method known as dynamic regional phase synchrony (DRePS), a “time-varying version of regional homogeneity (ReHo)” (Omidvarnia et al., 2016). DRePS evaluates dynamic local functional connectivity at a high temporal resolution and reflects how the spatial distributions of local functional units fluctuate across time (Omidvarnia et al., 2016). A DRePS study on epilepsy reported increased DRePS in patients with focal epilepsy, which is associated with an adaptive brain reaction to epileptic activity (Pedersen et al., 2017a). Since GAD has been shown to be a brain network disease (Andreescu et al., 2014; Chen and Etkin, 2013; Pedersen et al., 2017a; Roy et al., 2013) and involves maladaptive responses to the external environment (Grupe and Nitschke, 2013; McClure et al., 2007; Shechner and Bar-Haim, 2016; Via et al., 2018), DRePS is a potentially valuable tool for investigating brain dysfunction in patients with GAD from the novel perspective of dynamic local functional connectivity. This study investigated the properties of dynamic local connectivity of the brain in patients with GAD using the DRePS method. Furthermore, the altered DRePS patterns observed in patients with GAD were evaluated to assess whether they could be used as features to distinguish patients with GAD from HCs in an independent sample at an individual level. We hypothesized that DRePS in patients with GAD would indicate abnormal patterns. This is because GAD is a brain network disease and DRePS can delineate the dynamic organization of local networks, comprised of local neighbourhoods, which supports brain functionality. The findings of this study may provide novel imaging evidence that can help explain the clinical manifestations of GAD.

Table 1 Demographic and clinical characteristics of the main sample. Variables

HCs (n = 74)

GAD (n = 74)


Age (years) Gender (male / female) Handedness (left / right) Education (years) Mean FD Duration of illness (months) Age of first onset (years) No. of anxiety episodes Duration of single anxiety episode HAMA scores Medical Medication load index

38.58 ± 14.62 33/41 3/71 12.98 ± 3.49 0.09 ± 0.03 – – – – –

38.00 ± 10.66 31/43 2/72 12.46 ± 3.21 0.09 ± 0.06 47.24 ± 58.32 34.20 ± 11.29 2.01 ± 1.18 5.65 ± 5.98 23.81 ± 5.78

0.87a 0.74b 0.65b 0.10a 0.41a – – – – –

Medications Antianxiety Fluoxetine Sertraline Paroxetine citalopram Escitalopram Venlafaxine Duloxetine Mirtazapine Antipsychotics Olanzapine Quetiapine Aripiprazole Antianxiety + Antipsychotics Duloxetine + Quetiapine Paroxetine + Aripiprazole Mirtazapine + Quetiapine Paroxetine + Olanzapine Venlafaxine + Quetiapine Duloxetine + Olanzapine

1.88 ± 0.45 No. of patients – – – – – – – –

2 8 23 1 10 4 14 2

3 4 1 2 1 1 1 1 2

Abbreviations: HCs, healthy controls; GAD, generalized anxiety disorder; FD, frame-wise displacement; HAMA, 14-item Hamilton Anxiety Rating Scale. Values are mean ± standard deviation. a Mann-Whitney U-test. b Chi-square test.

by two experienced psychiatrists using the Structured Clinical Interview for the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV)Patient Edition (SCID-P) (First et al., 1997). These patients were then diagnosed according to the DSM-IV criteria for GAD. Exclusion criteria included schizophrenia, mental retardation, personality disorder, a history of any loss of consciousness, substance abuse, and severe medical or neurological disease. Personality disorders were diagnosed using the Chinese version of the Personality Disorder subscale (Pfohl et al., 1997) of the SCID-II instrument. The Hamilton Anxiety Rating Scale (HAMA), which consists of 14 items, was used to assess the clinical state of each patient. All patients in this study were prescribed antianxiety and antipsychotic drugs; the patients did not consume any other drugs during the course of this study. Medication information is shown in Table 1. Next, HCs were screened using the non-patient version of the SCID to confirm the lifetime absence of psychiatric illnesses. Age, gender, handedness, years of education, and mean frame-wise displacement (FD) (see section 2.3 for information on FD calculations) were matched between patients and HCs. Written informed consent was obtained from each participant before the experiments. The study was approved by the research ethical committee of the University of Electronic Science and Technology of China. Group comparison in age, years of education and mean FD were analysed using the MannWhitney U‐test, and differences in gender and handedness were analysed using the Chi-square test. The detailed demographic information is shown in Table 1. An independent sample comprising 18 patients with GAD and 21 HCs was used to assess whether abnormal DRePS patterns found in patients with GAD (main sample) could be used as features to

2. Materials and methods 2.1. Participants Seventy-four patients with GAD and 74 HCs were enrolled in the current study. The patients with GAD were recruited from the Clinical Hospital of Chengdu Brain Science Institute. Patients were interviewed 2

Progress in Neuropsychopharmacology & Biological Psychiatry 99 (2020) 109833

Q. Cui, et al.

distinguish between patients with GAD and HCs. Therefore, statistical analyses were conducted based on the data obtained from the main sample to assess between-group differences in DRePS and the correlation between DRePS and symptomatology. The independent sample was used only for classification analysis. The enrolment criteria for participants and the methodology for the acquisition and analysis of fMRI data from the independent sample were similar to that used for the main sample. Demographic and clinical information for the independent sample is shown in Table S1.

all the subjects; 4) spatially smoothing with a Gaussian kernel (fullwidth at half maximum = 8 mm) was further performed to improve the signal-to-noise ratio. The DRePS time series at each voxel represents time-varying phase coherence relationships between a central voxel and its immediate 26 neighbouring voxels over the course of the scan. 2.5. Statistical analyses A two-sample t-test was performed to compare the DRePS maps between the GAD and HC groups, with age, gender, years of education, handedness and mean FD being regressed out as these variables may confound the results (Zuo et al., 2012). Results were corrected using the false discovery rate (FDR) (p < .01, and minimum cluster size of 50 voxels). Brain areas surviving the multiple comparison correction were selected as regions of interest (ROIs) for a subsequent post-hoc analysis.

2.2. Data acquisition Participants' imaging data were collected using a 3 T GE DISCOVERY MR750 scanner (General Electric, Fairfield Connecticut, USA) equipped with a high-speed gradient and an 8-channel prototype quadrature birdcage head coil. Head movement was minimized using foam pads. The participants were instructed to simply rest with their eyes closed, relax their minds, avoid falling asleep, and remain motionless. Functional images were collected through an echo-planar imaging sequence with the following parameters: repetition time (TR)/ echo time (TE) = 2000/30 ms; slices = 43; matrix size = 64 × 64; flip angle = 90°; field of view = 240 mm × 240 mm; voxel size = 3.75 mm × 3.75 mm × 3.2 mm; thickness = 3.2 mm, no gap; and a total of 255 volumes.

2.6. Associations with symptom severity Pearson's correlation coefficients were calculated between the averaged DRePS values of each ROI and the HAMA scores of patients with GAD. The statistical level at p < .05/N was considered significant, where N is the number of ROIs (Bonferroni corrected). 2.7. Classification analysis The support vector machine implemented in the LIBSVM library (Chang and Lin, 2011) was used to assess the potential of using DRePS as a diagnostic biomarker for distinguishing patients with GAD from HCs in an independent sample. Specifically, the ROIs were defined from the group analysis based on the main sample data. Next, the averaged DRePS values of each ROI were extracted for each participant in the main sample and were used as features to train the classifier. The averaged DRePS values of the same ROIs were also extracted for each participant in the independent sample, entered into the classifier, and used as features to distinguish patients with GAD from HCs in the independent sample. A non-parametric permutation test was employed to calculate the statistical significance of the classification performance (the area under the curve (AUC) of the receiver-operating characteristic curves, ROC). In each permutation test trial, the actual group labels (GAD and HC) were shuffled randomly, and the same classification procedure was conducted to obtain a classification accuracy value based on the shuffled dataset. This procedure was repeated 5000 times and the level of statistical significance, p, was determined by computing the proportion of AUC that exceeded the original AUC value.

2.3. Data preprocessing The DPABI toolbox (DPABI V4.0, http://rfmri.org/dpabi) (Yan et al., 2016) was used to preprocess resting-state blood oxygenation level-dependent (BOLD) images. The first five volumes were discarded due to the initial MRI signal instability and participant inadaptation to the circumstances. Next, the slice timing was performed on the remaining time points to correct for their acquisition time, and the realignment was conducted to correct for head motion. Participant data were discarded if the translational and rotational displacement exceeded 3.0 mm or 3.0°. Thereafter, the resting-state BOLD images were normalized to the standard echo-planar imaging template and resampled into a voxel size of 3 mm × 3 mm × 3 mm. Subsequently, the linear trend removal was applied to the data. Nuisance covariates (Friston-24 parameters of head motion, white matter signal, cerebrospinal fluid signal, and global signal) were regressed out from the data. Finally, the effects of high-frequency physiological noise were reduced through filtering (0.01–0.08 Hz). In addition, the mean FD of each participant was calculated and between-group difference in head movement was assessed (Li et al., 2018b; Lu et al., 2017). There was no significant between-group difference in the mean FD.

2.8. Clinical analysis The total medication load of each patient was measured based on a previously described method (Almeida et al., 2009; Hassel et al., 2008; Phillips et al., 2008; Versace et al., 2008). For each patient, each medication was converted to a level from 1 to 4 based on previously reported criteria (Sackeim, 2001) according to the medication dosage and the duration of its course. Medications on levels 1 and 2 were classified as low-dose (1) and those on levels 3 and 4 as high-dose (2). A no-dose subtype was added for patients who did not take any medication. The sum of the codes of all medications taken by each patient was used as the total medication load index of that patient.

2.4. Dynamic regional phase synchrony calculation The computational process for DRePS was proposed by Omidvarnia et al. (2016). The DRePS measures regional similarity of 27 immediate voxels within a 3-dimensional (3D) cube (3 × 3 × 3 mm3). The whole brain DRePS maps were obtained by following steps: 1) the instantaneous phase information of voxel-wise fMRI time series was obtained by Hilbert transform; 2) the instantaneous mean phase coherence between a given voxel x and its adjacent neighbouring voxels yi (i = 1, ⋯, M − 1), where M is the entire number of non-zero voxels within a 3D cube centred at x, was computed by: Dx [n] =

1 M


M 1


cos( i =1

x [n]

yi [n])


3. Results


M 1


x [n]

yi [n])

3.1. Dynamic regional phase synchrony alterations in patients with generalized anxiety disorder


where Φx[n] and Φyi[n] were the instantaneous phase functions of time series with T time-points (n = 1, …, T) at voxel x and its adjacent neighbouring voxels yi; 3) the standard deviation of the function Dx[n] of each voxel was calculated to obtain the whole brain DRePS maps of

Compared with HCs, patients with GAD showed significantly decreased DRePS in widespread brain areas (FDR-corrected, p < .01, and minimum cluster size of 50 voxels), including the bilateral caudate, left 3

Progress in Neuropsychopharmacology & Biological Psychiatry 99 (2020) 109833

Q. Cui, et al.

Table 2 Regions with decreased DRePS in the GAD group compared with the HC group. Brain region


Peak (MNI)

Brodmann area

Cluster Size

t value

20/37 47 54 48 54 48 47

208 106 79 61 52 52 51

−6.02 −4.88 −5.20 −5.39 −5.50 −4.62 −5.29

(x, y, z) Fusiform extending to ITG AI Hippocampus Caudate Hippocampus Caudate IFG

Right Left Left Left Left Right Left

42, −15, −24 −36, 15, −6 −33, −27, −9 −18, 0, 21 −15, −36, 9 12, 6, 21 −51, 24, −9

Abbreviations: DRePS, dynamic regional phase synchrony; GAD, generalized anxiety disorder; HC, healthy control; ITG, inferior temporal gyrus; AI, anterior insula; IFG, inferior frontal gyrus.

hippocampus, left anterior insula (AI), left inferior frontal gyrus (IFG), and right fusiform gyrus extending to right inferior temporal gyrus (ITG). No regions had increased DRePS in the patients with GAD. Detailed information on between-group differences in DRePS is presented in Table 2 and Fig. 1.

the classifier and obtained an AUC of 0.80 (p < .001), indicating good classification power. 3.4. Correlations with medication Correlations between the DRePS values of each ROI and the total medication load of patients were assessed to evaluate the possible influence of medicine on the DRePS-related results. No significant correlation was observed between the DRePS values of each ROI and the medication load (all p > .1).

3.2. Associations with symptom severity The correlation results showed a significantly negative correlation between the DRePS values of the left hippocampus and the HAMA scores of patients with GAD (r = −0.3175, p = .0058, Bonferroni corrected, Fig. 2A). In addition, we also found the negative correlation between HAMA scores and DRePS values of the left AI (r = −0.2681, p = .0209, Fig. 2B) and the right caudate (r = −0.2484, p = .0328, Fig. 2C).

4. Discussion This is the first study to investigate the dynamic local functional connectivity of the brain in patients with GAD using the DRePS method. Patients with GAD exhibited decreased DRePS in the bilateral caudate, left hippocampus, left AI, left IFG, and right fusiform gyrus extending to right ITG. The altered DRePS in the left hippocampus was significantly correlated with HAMA scores in patients with GAD. Furthermore, classification analysis on an independent sample using the DRePS values of these regions as features showed an acceptable accuracy of 79.49% in distinguishing patients with GAD from HCs. This study provides new insights into brain dysfunction in GAD and indicates the possibility of using DRePS as a potential biomarker of GAD. The hippocampus is a pivotal region of the limbic system and plays a critical role in emotion processing (Moon and Jeong, 2015), mediation of anxiety states (Caliskan and Stork, 2019), and manifestation of anxiety reactions (Engin and Treit, 2007). Hippocampal anomalies are critical for the development and maintenance of pathological anxiety (Bannerman et al., 2004; Cha et al., 2016; Kheirbek et al., 2012). Patients with GAD are reported to show hippocampal hyperactivation during the processing of anxiety inducing pictures (Moon and Jeong, 2015; Park et al., 2016). This hippocampal hyperactivation was found correlate with pathological anxiety response of GAD (Moon and Jeong, 2015). Additionally, altered hippocampal gray matter integrity has also been reported in patients with GAD (Abdallah et al., 2013; Hettema et al., 2012; Moon et al., 2014), and such structural abnormality may consequently increase anxiety symptom because of its impact on functional abnormality (Cha et al., 2016). Consistent with these findings, the current study found hippocampal dysfunction in GAD, which manifested as decreased dynamic local functional connectivity. Intriguingly, the decreased DRePS in the left hippocampus was negatively correlated with the anxiety symptom severity, which suggested that the hypo-variability of brain instantaneous local connectivity in this region might be associated with excessive worry state in patients with GAD and may reflect the potential pathophysiological mechanisms underlying GAD. The fronto-insula cortex was also found to show abnormal DRePS in patients with GAD, including decreased DRePS in the insula and IFG. Insula is a known critical component of the salience network (SN)

3.3. Classification analysis using an independent sample Fig. 3 presents the classification analysis results. The classifier, trained by the main sample, was tested on an independent sample for its ability to distinguish patients with GAD from HCs and obtained an accuracy of 79.49%. A ROC was utilized to evaluate the performance of

Fig. 1. (A) The results of one-sample t-tests of DRePS in the GAD and HC groups. (B) Group difference in DRePS between patients with GAD and HCs was detected using a two-sample t-test (FDR-corrected, p < .01 and minimum cluster size of 50 voxels). The brain regions with cold colours (blue and green) have decreased DRePS values in patients with GAD compared with HCs, including the bilateral caudate, left hippocampus, left AI, left IFG, and right fusiform gyrus extending to right ITG. Abbreviations: DRePS, dynamic regional phase synchrony; GAD, generalized anxiety disorder; HC, healthy control; FDR, false discovery rate; AI, anterior insula; IFG, inferior frontal gyrus; ITG, inferior temporal gyrus; L, left; R, right.; FDR, false discovery rate. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) 4

Progress in Neuropsychopharmacology & Biological Psychiatry 99 (2020) 109833

Q. Cui, et al.

Fig. 2. Correlation between DRePS and HAMA scores of patients with GAD. (A) Significantly negative correlation between DRePS values in the left hippocampus and HAMA scores (r = −0.3175; p = .0058, Bonferroni corrected). (B) Negative correlation between DRePS values in the left AI and HAMA scores (r = −0.2681; p = .0209). (C) Negative correlation between DRePS values in the right caudate and HAMA scores (r = −0.2484; p = .0328). First column depicts the schematic diagrams of the different brain regions; the second column represents the box diagrams which show the DRePS values between the GAD and HC groups; the third column shows the plots of Pearson correlation between the DRePS values and HAMA scores. Red colour represents patients with GAD, whereas blue colour represents the HCs. Abbreviations: DRePS, dynamic regional phase synchrony; HAMA, Hamilton Anxiety Rating Scale; GAD, generalized anxiety disorder; HC, healthy control; AI, anterior insula; L, left; R, right. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

(Moran et al., 2013; Supekar et al., 2019), which is responsible for monitoring individual internal and external conflicts and transmitting them to the frontoparietal control network to solve the conflict (Menon and Uddin, 2010; Paulus and Stein, 2010; Seeley et al., 2007). The fronto-insula region is also a key node of the SN and plays a critical role in switching between exogenous and endogenous processes (Sridharan et al., 2008). The fronto-insula dysfunction may lead to aberrant salience processing in patients with GAD (McClure et al., 2007). DRePS is a method of detecting dynamic local functional connectivity patterns in the brain with a high temporal resolution (Omidvarnia et al., 2016; Pedersen et al., 2018). The hyper-variability (increased temporal variance) or hypo-variability (decreased temporal variance) of brain dynamic connectivity (Christoff et al., 2016; Pedersen et al., 2017b) may contribute to altered cognitive functions and particular pathological states (Preti et al., 2017). Given the high temporal variability of brain functional networks in the insula in healthy individuals (Zhang et al., 2016), we speculated that the decreased DRePS in the fronto-insula observed in this study might be associated with the inability of patients with GAD to integrate external sensory information with internal

emotion. Decreased DRePS in GAD was also observed in the caudate. This region is part of the frontal-striatal-thalamic-cortical (FSTC) circuitry (Roth et al., 2007) and has been associated with motor control (Sowman et al., 2017), response inhibition (Roth et al., 2007), and cognitive and emotional processing (Monk, 2008). Previous studies have shown that the caudate plays a major role in cognitive processing and influences cognitive reappraisal through the FSTC loop (Opialla et al., 2015). Cognitive reappraisal deficits have been implicated in the pathology of GAD/worry (Karim et al., 2017), especially in failing to implement adaptive reappraisal strategies, which may result in neutral or non-conflict stimuli being treated as threatening stimuli and eventually causes an abnormal activation of anxiety circuitry (Choi et al., 2012). Early studies have reported a caudate size reduction in patients with anxiety (Herringa et al., 2012) and decreased ReHo in the caudate, which indicates a decline in responses to external stimuli in patients with GAD (Li et al., 2018a). Thus, the present findings suggest that decreased DRePS in the caudate might signify disturbed FSTC integrity, which could explain the abnormal cognitive reappraisal in patients with 5

Progress in Neuropsychopharmacology & Biological Psychiatry 99 (2020) 109833

Q. Cui, et al.

eliminate the potential effects of medications. Third, there is currently no clear conclusion on the biological and cognitive significance of DRePS; therefore, future studies are needed to further substantiate the clinical applications. Nevertheless, clinical practice and future research in this area may benefit from the following perspectives. First, verifying the relationship between DRePS and brain metabolism is advantageous since previous studies have shown a link between ReHo and brain energy consumption; however, there is currently no direct evidence regarding the relationship between DRePS (a time-varying version of ReHo) and brain metabolism. Further research in this area could reveal more exact information on the biological significance of DRePS, thus expanding the clinical implications and applications of the present results on GAD. Second, future task-related studies may apply DRePS to assess dynamic temporal changes in local brain networks during specific cognitive processing, especially in those that show typical deficits in GAD such as vigilance, attention, and emotion processing and regulation. These studies could provide important evidence that deepens our understanding of the neurocognitive mechanisms of GAD. Third, future taskrelated studies may combine DRePS with brain stimulation, which could reveal a causal relationship between DRePS and cognitive function; moreover, this approach may guide the application of non-invasive brain stimulation therapy by identifying the important networks or regions that can be stimulated to alleviate and improve the anxiety symptoms in patients with GAD.

Fig. 3. Receiver operating characteristic curve analysis. The ROC curve was used to assess the discrimination between patients with GAD and HCs in an independent sample. The observed AUC value signifies good classification power (AUC = 0.80; p < .001). Abbreviations: ROC, receiver operating characteristic; AUC, area under the ROC curve; GAD, generalized anxiety disorder; HCs, healthy controls.

5. Conclusion

GAD. In addition, patients with GAD showed decreased DRePS in the right fusiform gyrus, extending to the right ITG. The fusiform gyrus has been implicated in emotion processing, especially the perception of facial emotions (Rossion et al., 2000). Patients with GAD showed hypoactivation in the fusiform gyrus during processing pictures with threating stimuli (Moon and Jeong, 2015). Hypoactivation of the left fusiform gyrus in response to facial expression of fear and anger was also found in individuals with social phobia, another subtype of anxiety disorders, and was interpreted as the result of patients' excessive wariness to threats (Gentili et al., 2008). ITG is a region reported to be involved in multiple cognitive functions, including language, visual perception, and memory (Noppeney and Price, 2002; Ojemann et al., 2002; Wu et al., 2017). The ITG has been demonstrated to be a major association area that has widespread connections with multiple brain regions, including the amygdala and orbital frontal cortex (Tang et al., 2018), which are involved in emotion processing and emotion regulation. Connections with these regions allow the ITG to be involved in emotion regulation (Pannekoek et al., 2014). The reduced grey matter volume in the ITG was also found in patients with social anxiety disorder, which was considered as a possible anatomical substrate underlying negative emotion regulation (Liao et al., 2011). Our findings of decreased DRePS in the fusiform gyrus and ITG in GAD may indicate insufficient local network adaptation in these regions, which might be associated with emotional dysregulation in response to negative and threatening stimuli, a clinical characteristic of GAD (Monk et al., 2008; Palm et al., 2011).

This study applied the DRePS method to investigate dynamic local functional connectivity in patients with GAD. Decreased DRePS may be associated with insufficient brain adaptation to external stimuli in patients with GAD. The acceptable classification accuracy of an independent sample suggests that DRePS may be a potential neuromarker of GAD. This study provides further evidence of brain dysfunction in GAD from the perspective of dynamic local network behaviour and provides novel insights that help explain cognitive deficits in GAD; this is essential because DRePS delineates the dynamic organization of “local neighbourhoods” that have been shown to support cognitive function (Zang et al., 2004). Thus, these findings reveal some of the underpinnings of the clinical GAD manifestation and provide avenues for future studies on its diagnosis and treatment. Ethical statement The study was approved by the research ethical committee of University of Electronic Science and Technology of China. Declaration of Competing Interest All authors in this study declare that they have no conflicts of interest. Acknowledgements We thank all subjects participating in this study. This study was supported by the Key Project of Research and Development of Ministry of Science and Technology (2018AAA0100705), the Natural Science Foundation of China (61533006, 81771919, U1808204, and 31600930), the Science Foundation of Ministry of Education of China (14XJC190003), the Scientific research project of Sichuan Medical Association (S15012), the Youth Innovation Project of Sichuan Provincial Medical Association (Q14014), the Fundamental Research Funds for the Central Universities (ZYGX2013Z004), Sichuan Science and Technology Program (2018TJPT0016).

4.1. Limitations and future work This study has some limitations. First, the statistical results may be underpowered by the relatively small sample size; therefore, future studies with larger samples should retest the present findings. Second, most of the patients with GAD have received medication treatment. Although there were no significant correlations between medication and the DRePS-related findings in this study, future studies should verify these findings using medication-free and first-episode patients to 6

Progress in Neuropsychopharmacology & Biological Psychiatry 99 (2020) 109833

Q. Cui, et al.

Appendix A. Supplementary data

generalization: a new approach to stratify and treat anxiety disorders. Nat. Neurosci. 15, 1613. Li, W., Cui, H., Zhu, Z., Kong, L., Guo, Q., Zhu, Y., et al., 2016. Aberrant functional connectivity between the amygdala and the temporal pole in drug-free generalized anxiety disorder. Front. Hum. Neurosci. 10. Li, C., Su, S., Wu, H., Zhu, Y., 2018a. Abnormal spontaneous brain activity in patients with generalized anxiety disorder revealed by resting-state functional MRI. Neuroreport. 29, 397–401. Li, R., Liao, W., Yu, Y., Chen, H., Guo, X., Tang, Y.L., et al., 2018b. Differential patterns of dynamic functional connectivity variability of striato–cortical circuitry in children with benign epilepsy with centrotemporal spikes. Hum. Brain Mapp. 39, 1207–1217. Li, C., Xia, L., Ma, J., Li, S., Liang, S., Ma, X., et al., 2019. Dynamic functional abnormalities in generalized anxiety disorders and their increased network segregation of a hyperarousal brain state modulated by insomnia. J. Affect. Disord. 246, 338–345. Liao, W., Xu, Q., Mantini, D., Ding, J., Machado-de-Sousa, J.P., Hallak, J.E.C., et al., 2011. Altered gray matter morphometry and resting-state functional and structural connectivity in social anxiety disorder. Brain Res. 1388, 167–177. Liu, X., Duyn, J.H., 2013. Time-varying functional network information extracted from brief instances of spontaneous brain activity. Proc. Natl. Acad. Sci. U. S. A. 110, 4392–4397. Lu, F.-M., Zhou, J.-S., Wang, X.-P., Xiang, Y.-T., Yuan, Z., 2017. Short-and long-range functional connectivity density alterations in adolescents with pure conduct disorder at resting-state. Neuroscience. 351, 96–107. Makova, E., Meeten, F., Watson, D.R., Garfinkel, S.N., Critchley, H.D., Ottaviani, C., 2016. Neurostructural abnormalities associated with axes of emotion dysregulation in generalized anxiety. Neuroimage-Clinical. 10, 172–181. McClure, E.B., Monk, C.S., Nelson, E.E., Parrish, J.M., Adler, A., Blair, R.J.R., et al., 2007. Abnormal attention modulation of fear circuit function in pediatric generalized anxiety disorder. Arch. Gen. Psychiatry 64, 97–106. Menon, V., Uddin, L.Q., 2010. Saliency, switching, attention and control: a network model of insula function. Brain Struct. Funct. 214, 655–667. Mochcoyitch, M.D., da Rocha Freire, R.C., Garcia, R.F., Nardi, A.E., 2014. A systematic review of fMRI studies in generalized anxiety disorder: evaluating its neural and cognitive basis. J. Affect. Disord. 167, 336–342. Molent, C., Maggioni, E., Cecchetto, F., Garzitto, M., Piccin, S., Bonivento, C., et al., 2018. Reduced cortical thickness and increased gyrification in generalized anxiety disorder: a 3 T MRI study. Psychol. Med. 48, 2001–2010. Monk, C.S., 2008. The development of emotion-related neural circuitry in health and psychopathology. Dev. Psychopathol. 20, 1231–1250. Monk, C.S., Telzer, E.H., Mogg, K., Bradley, B.P., Mai, X., Louro, H.M.C., et al., 2008. Amygdala and ventrolateral prefrontal cortex activation to masked angry faces in children and adolescents with generalized anxiety disorder. Arch. Gen. Psychiatry 65, 568–576. Moon, C.-M., Jeong, G.-W., 2015. Functional neuroanatomy on the working memory under emotional distraction in patients with generalized anxiety disorder. Psychiatry Clin. Neurosci. 69, 609–619. Moon, C.-M., Kim, G.-W., Jeong, G.-W., 2014. Whole- brain gray matter volume abnormalities in patients with generalized anxiety disorder: voxel- based morphometry. Neuroreport. 25, 184–189. Moran, L.V., Tagamets, M.A., Sampath, H., O’Donnell, A., Stein, E.A., Kochunov, P., et al., 2013. Disruption of anterior insula modulation of large-scale brain networks in schizophrenia. Biol. Psychiatry 74, 467–474. Noppeney, U., Price, C.J., 2002. Retrieval of visual, auditory, and abstract semantics. Neuroimage. 15, 917–926. Ojemann, G.A., Schoenfield-McNeill, J., Corina, D.P., 2002. Anatomic subdivisions in human temporal cortical neuronal activity related to recent verbal memory. Nat. Neurosci. 5, 64–71. Omidvarnia, A., Pedersen, M., Walz, J.M., Vaughan, D.N., Abbott, D.F., Jackson, G.D., 2016. Dynamic regional phase synchrony (DRePS): an instantaneous measure of local fMRI connectivity within spatially clustered brain areas. Hum. Brain Mapp. 37, 1970–1985. Opialla, S., Lutz, J., Scherpiet, S., Hittmeyer, A., Jaencke, L., Rufer, M., et al., 2015. Neural circuits of emotion regulation: a comparison of mindfulness-based and cognitive reappraisal strategies. Eur. Arch. Psychiatry Clin. Neurosci. 265, 45–55. Palm, M.E., Elliott, R., McKie, S., Deakin, J.F.W., Anderson, I.M., 2011. Attenuated responses to emotional expressions in women with generalized anxiety disorder. Psychol. Med. 41, 1009–1018. Pannekoek, J.N., van der Werff, S.J.A., Meens, P.H.F., van den Bulk, B.G., Jolles, D.D., Veer, I.M., et al., 2014. Aberrant resting-state functional connectivity in limbic and salience networks in treatment-naive clinically depressed adolescents. J. Child Psychol. Psychiatry 55, 1317–1327. Park, H.-J., Friston, K.J., 2013. Structural and functional brain networks: from connections to cognition. Science 342 579−+. Park, J.-I., Kim, G.-W., Jeong, G.-W., Chung, G.H., Yang, J.-C., 2016. Brain activation patterns associated with the effects of emotional distracters during working memory maintenance in patients with generalized anxiety disorder. Psychiatry Investig. 13, 152–156. Paulus, M.P., Stein, M.B., 2010. Interoception in anxiety and depression. Brain Struct. Funct. 214, 451–463. Pedersen, M., Omidvarnia, A., Curwood, E.K., Walz, J.M., Rayner, G., Jackson, G.D., 2017a. The dynamics of functional connectivity in neocortical focal epilepsy. Neuroimage-Clinical. 15, 209–214. Pedersen, M., Omidvarnia, A., Walz, J.M., Zalesky, A., Jackson, G.D., 2017b. Spontaneous brain network activity: analysis of its temporal complexity. Network neuroscience (Cambridge, Mass) 1, 100–115. Pedersen, M., Omidvarnia, A., Zalesky, A., Jackson, G.D., 2018. On the relationship

Supplementary data to this article can be found online at https:// doi.org/10.1016/j.pnpbp.2019.109833. References Abdallah, C.G., Coplan, J.D., Jackowski, A., Sato, J.R., Mao, X., Shungu, D.C., et al., 2013. A pilot study of hippocampal volume and N-acetylaspartate (NAA) as response biomarkers in riluzole-treated patients with GAD. Eur. Neuropsychopharmacol. 23, 276–284. Allen, E.A., Damaraju, E., Plis, S.M., Erhardt, E.B., Eichele, T., Calhoun, V.D., 2014. Tracking whole-brain connectivity dynamics in the resting state. Cereb. Cortex 24, 663–676. Almeida, J.R.C., Akkal, D., Hassel, S., Travis, M.J., Banihashemi, L., Kerr, N., et al., 2009. Reduced gray matter volume in ventral prefrontal cortex but not amygdala in bipolar disorder: significant effects of gender and trait anxiety. Psychiatry ResearchNeuroimaging. 171, 54–68. Andreescu, C., Sheu, L.K., Tudorascu, D., Walker, S., Aizenstein, H., 2014. The ages of anxietydifferences across the lifespan in the default mode network functional connectivity in generalized anxiety disorder. International Journal of Geriatric Psychiatry. 29, 704–712. Bannerman, D., Rawlins, J., McHugh, S., Deacon, R., Yee, B., Bast, T., et al., 2004. Regional dissociations within the hippocampus—memory and anxiety. Neurosci. Biobehav. Rev. 28, 273–283. Caliskan, G., Stork, O., 2019. Hippocampal network oscillations at the interplay between innate anxiety and learned fear. Psychopharmacology. 236, 321–338. Cha, J., Greenberg, T., Song, I., Simpson, H.B., Posner, J., Mujica-Parodi, L.R., 2016. Abnormal hippocampal structure and function in clinical anxiety and comorbid depression. Hippocampus. 26, 545–553. Chang, C.-C., Lin, C.-J., 2011. LIBSVM: a library for support vector machines. ACM Trans. Intell. Syst. Technol. 2. Chen, A.C., Etkin, A., 2013. Hippocampal network connectivity and activation differentiates post-traumatic stress disorder from generalized anxiety disorder. Neuropsychopharmacology. 38, 1889–1898. Choi, J.M., Padmala, S., Pessoa, L., 2012. Impact of state anxiety on the interaction between threat monitoring and cognition. Neuroimage. 59, 1912–1923. Christoff, K., Irving, Z.C., Fox, K.C.R., Spreng, R.N., Andrews-Hanna, J.R., 2016. Mindwandering as spontaneous thought: a dynamic framework. Nat. Rev. Neurosci. 17, 718–731. Cuijpers, P., Sijbrandij, M., Koole, S., Huibers, M., Berking, M., Andersson, G., 2014. Psychological treatment of generalized anxiety disorder: a meta-analysis. Clin. Psychol. Rev. 34, 130–140. De Bellis, M.D., Casey, B.J., Dahl, R.E., Birmaher, B., Williamson, D.E., Thomas, K.M., et al., 2000. A pilot study of amygdala volumes in pediatric generalized anxiety disorder. Biol. Psychiatry 48, 51–57. Engin, E., Treit, D., 2007. The role of hippocampus in anxiety: intracerebral infusion studies. Behav. Pharmacol. 18, 365–374. Etkin, A., Prater, K.E., Schatzberg, A.F., Menon, V., Greicius, M.D., 2009. Disrupted Amygdalar subregion functional connectivity and evidence of a compensatory network in generalized anxiety disorder. Arch. Gen. Psychiatry 66, 1361–1372. First, M., Spitzer, R., Williams, J., 1997. Structured Clinical Interview for DSM-IV Axis I Disorders SCID-I. American Psychiatric Press, Washington, DC. Gentili, C., Gobbini, M.I., Ricciardi, E., Vanello, N., Pietrini, P., Haxby, J.V., et al., 2008. Differential modulation of neural activity throughout the distributed neural system for face perception in patients with social phobia and healthy subjects. Brain Res. Bull. 77, 286–292. Goossen, B., van der Starre, J., van der Heiden, C., 2019. A review of neuroimaging studies in generalized anxiety disorder: “so where do we stand?”. J. Neural Transm. 126, 1203–1216. Grupe, D.W., Nitschke, J.B., 2013. Uncertainty and anticipation in anxiety: an integrated neurobiological and psychological perspective. Nat. Rev. Neurosci. 14, 488–501. Gu, S., Pasqualetti, F., Cieslak, M., Telesford, Q.K., Yu, A.B., Kahn, A.E., et al., 2015. Controllability of structural brain networks. Nat. Commun. 6. Hassel, S., Almeida, J.R.C., Kerr, N., Nau, S., ladouceur, C.D., Fissell, K., et al., 2008. Elevated striatal and decreased dorsolateral prefrontal cortical activity in response to emotional stimuli in euthymic bipolar disorder: no associations with psychotropic medication load. Bipolar Disord. 10, 916–927. Herringa, R., Phillips, M., Almeida, J., Insana, S., Germain, A., 2012. Post-traumatic stress symptoms correlate with smaller subgenual cingulate, caudate, and insula volumes in unmedicated combat veterans. Psychiatry Research-Neuroimaging. 203, 139–145. Hettema, J.M., Kettenmann, B., Ahluwalia, V., McCarthy, C., Kates, W.R., Schmitt, J.E., et al., 2012. Pilot multimodal twin imaging study of generalized anxiety disorder. Depression and Anxiety. 29, 202–209. Hilbert, K., Lueken, U., Beesdo-Baum, K., 2014. Neural structures, functioning and connectivity in generalized anxiety disorder and interaction with neuroendocrine systems: a systematic review. J. Affect. Disord. 158, 114–126. Hutchison, R.M., Womelsdorf, T., Allen, E.A., Bandettini, P.A., Calhoun, V.D., Corbetta, M., et al., 2013. Dynamic functional connectivity: promise, issues, and interpretations. Neuroimage. 80, 360–378. Karim, H.T., Tudorascu, D.L., Butters, M.A., Walker, S., Aizenstein, H.J., Andreescu, C., 2017. In the grip of worry: cerebral blood flow changes during worry induction and reappraisal in late-life generalized anxiety disorder. Transl. Psychiatry 7. Kheirbek, M.A., Klemenhagen, K.C., Sahay, A., Hen, R., 2012. Neurogenesis and


Progress in Neuropsychopharmacology & Biological Psychiatry 99 (2020) 109833

Q. Cui, et al. between instantaneous phase synchrony and correlation-based sliding windows for time-resolved fMRI connectivity analysis. Neuroimage. 181, 85–94. Pfohl, B., Blum, N., Zimmerman, M., 1997. Structured Interview for DSM-IV Personality: Sidp-IV. American Psychiatric Pub. Phillips, M.L., Travis, M.J., Fagiolini, A., Kupfer, D.J., 2008. Medication effects in neuroimaging studies of bipolar disorder. Am. J. Psychiatr. 165, 313–320. Preti, M.G., Bolton, T.A.W., Van De Ville, D., 2017. The dynamic functional connectome: state-of-the-art and perspectives. Neuroimage. 160, 41–54. Rossion, B., Dricot, L., Devolder, A., Bodart, J.M., Crommelinck, M., de Gelder, B., et al., 2000. Hemispheric asymmetries for whole-based and part-based face processing in the human fusiform gyrus. J. Cogn. Neurosci. 12, 793–802. Roth, R.M., Saykin, A.J., Flashman, L.A., Pixley, H.S., West, J.D., Mamourian, A.C., 2007. Event-related functional magnetic resonance imaging of response inhibition in obsessive-compulsive disorder. Biol. Psychiatry 62, 901–909. Roy, A.K., Fudge, J.L., Kelly, C., Perry, J.S.A., Daniele, T., Carlisi, C., et al., 2013. Intrinsic functional connectivity of amygdala-based networks in adolescent generalized anxiety disorder. J. Am. Acad. Child Adolesc. Psychiatry 52, 290–299. Sackeim, H.A., 2001. The definition and meaning of treatment-resistant depression. J. Clin. Psychiatry 62 s. Schienle, A., Ebner, F., Schaefer, A., 2011. Localized gray matter volume abnormalities in generalized anxiety disorder. Eur. Arch. Psychiatry Clin. Neurosci. 261, 303–307. Seeley, W.W., Menon, V., Schatzberg, A.F., Keller, J., Glover, G.H., Kenna, H., et al., 2007. Dissociable intrinsic connectivity networks for salience processing and executive control. J. Neurosci. 27, 2349–2356. Shechner, T., Bar-Haim, Y., 2016. Threat monitoring and attention-bias modification in anxiety and stress-related disorders. Curr. Dir. Psychol. Sci. 25, 431–437. Sowman, P.F., Ryan, M., Johnson, B.W., Savage, G., Crain, S., Harrison, E., et al., 2017. Grey matter volume differences in the left caudate nucleus of people who stutter. Brain Lang. 164, 9–15. Sridharan, D., Levitin, D.J., Menon, V., 2008. A critical role for the right fronto-insular cortex in switching between central-executive and default-mode networks. Proc. Natl. Acad. Sci. U. S. A. 105, 12569–12574. Supekar, K., Cai, W., Krishnadas, R., Palaniyappan, L., Menon, V., 2019. Dysregulated brain dynamics in a triple-network saliency model of schizophrenia and its relation to

psychosis. Biol. Psychiatry 85, 60–69. Tang, S., Lu, L., Zhang, L., Hu, X., Bu, X., Li, H., et al., 2018. Abnormal amygdala restingstate functional connectivity in adults and adolescents with major depressive disorder: a comparative meta-analysis. Ebiomedicine. 36, 436–445. Versace, A., Almeida, J.R.C., Hassel, S., Walsh, N.D., Novelli, M., Klein, C.R., et al., 2008. Elevated left and reduced right orbitomedial prefrontal fractional anisotropy in adults with bipolar disorder revealed by tract-based spatial statistics. Arch. Gen. Psychiatry 65, 1041–1052. Via, E., Fullana, M.A., Goldberg, X., Tinoco-Gonzalez, D., Martinez-Zalacain, I., SorianoMas, C., et al., 2018. Ventromedial prefrontal cortex activity and pathological worry in generalised anxiety disorder. Br. J. Psychiatry 213, 437–443. Wittchen, H.U., Jacobi, F., 2005. Size and burden of mental disorders in Europe - a critical review and appraisal of 27 studies. Eur. Neuropsychopharmacol. 15, 357–376. Wittchen, H.U., Jacobi, F., Rehm, J., Gustavsson, A., Svensson, M., Jonsson, B., et al., 2011. The size and burden of mental disorders and other disorders of the brain in Europe 2010. Eur. Neuropsychopharmacol. 21, 655–679. Wu, C., Zheng, Y., Li, J., Zhang, B., Li, R., Wu, H., et al., 2017. Activation and functional connectivity of the left inferior temporal gyrus during visual speech priming in healthy listeners and listeners with schizophrenia. Front. Neurosci. 11. Yan, C.-G., Wang, X.-D., Zuo, X.-N., Zang, Y.-F., 2016. DPABI: Data Processing & Analysis for (resting-state) brain imaging. Neuroinformatics. 14, 339–351. Yao, Z., Liao, M., Hu, T., Zhang, Z., Zhao, Y., Zheng, F., et al., 2017. An effective method to identify adolescent generalized anxiety disorder by temporal features of dynamic functional connectivity. Front. Hum. Neurosci. 11. Zalesky, A., Fornito, A., Cocchi, L., Gollo, L.L., Breakspear, M., 2014. Time-resolved resting-state brain networks. Proc. Natl. Acad. Sci. U. S. A. 111, 10341–10346. Zang, Y., Jiang, T., Lu, Y., He, Y., Tian, L., 2004. Regional homogeneity approach to fMRI data analysis. Neuroimage. 22, 394–400. Zhang, J., Cheng, W., Liu, Z., Zhang, K., Lei, X., Yao, Y., et al., 2016. Neural, electrophysiological and anatomical basis of brain-network variability and its characteristic changes in mental disorders. Brain. 139, 2307–2321. Zuo, X.-N., Ehmke, R., Mennes, M., Imperati, D., Castellanos, F.X., Sporns, O., et al., 2012. Network centrality in the human functional connectome. Cereb. Cortex 22, 1862–1875.