Altered cerebral activities and functional connectivity in depression: a systematic review of fMRI studies

Abstract

Background: As one of the leading causes of global disability, major depressive disorder (MDD) places a noticeable burden on individuals and society. Despite the great expectation on finding accurate biomarkers and effective treatment targets of MDD, studies in applying functional magnetic resonance imaging (fMRI) are still faced with challenges, including the representational ambiguity, small sample size, low statistical power, relatively high false positive rates, etc. Thus, reviewing studies with solid methodology may help achieve a consensus on the pathology of MDD.

Methods: In this systematic review, we screened fMRI studies on MDD through strict criteria to focus on reliable studies with sufficient sample size, adequate control of head motion, and a proper multiple comparison control strategy.

Results: We found consistent evidence regarding the dysfunction within and among the default mode network (DMN), the frontoparietal network (FPN), and other brain regions. However, controversy remains, probably due to the heterogeneity of participants and data processing strategies.

Conclusion: Future studies are recommended to apply a comprehensive set of neuro-behavioral measurements, consider the heterogeneity of MDD patients and other potentially confounding factors, apply surface-based neuroscientific network fMRI approaches, and advance research transparency and open science by applying state-of-the-art pipelines along with open data sharing.

Graphical abstract

Keywords

depression / resting-state fMRI / task-based fMRI / default mode network / frontoparietal network

1 INTRODUCTION

Major depressive disorder (MDD) is a prevalent disorder and places noticeable societal burdens [1]. For the recent 20 years, it has been stably listed in the leading 20 causes of global disability across ages and genders, which account for 37% of the total disability caused by mental disorders [2]. Despite the burgeoning methodological advances of neuroimaging, the current diagnosis approach of MDD is mainly based on the clinical interview or patient rating scales, which is associated with high rates of misdiagnosis [3]. Recent advances in neuroimaging techniques have made it possible to leverage the brain’s functional architecture towards the objective biomarker of MDD. With advantages of simpleness, non-invasiveness, safety, and relatively high spatial and temporal resolutions [4], functional magnetic resonance imaging (fMRI) method, especially resting-state fMRI (R-fMRI) method, has enabled to depict of dynamic maps of the brain and carry great expectations on finding biomarkers and effective treatment targets of MDD.

Unfortunately, though a large number of studies have been conducted, few consensuses have been reached on the neural mechanism of depression. Several challenges for fMRI studies have been noted, including between-subject and within-subject variability, the representational ambiguity, small sample size, low statistical power, relatively high false positive rates, and lack of consistency in data preprocessing procedures and statistical analysis [5–8]. For instance, the challenge of small sample size has been proposed to dampen the reliability of the fMRI results directly. Moreover, the between-participant variability that significantly impacts the reliability of fMRI studies is limited by the measurement and the experimental design, especially in task-based fMRI studies [8].

To deal with the current challenges in fMRI studies, researchers have highlighted a couple of recommendations on the designs of study, preprocessing pipelines, and strategies to correct multiple comparisons. These recommendations included performing adequate confound regression strategies [9–11] and multiple comparison correction, enlarging the sample size, and openly data sharing [12–14]. Chen and his colleagues assessed the test-retest reliability of fMRI studies and found studies with small sample sizes (< 40 per group) are not well reliable [14]. Head motions of subjects in an MRI scanner can produce artifacts, leading to results not caused by the “real” intrinsic brain functions [15,16]. Thus, adequate head motion corrections should be included to get reliable results. Eklund et al. reported that the liberal thresholding strategies commonly used in the field of neuroimaging could cause high false positive rates [17], which emphasized the significance of the effective multiple comparison correction, such as permutation test with threshold-free cluster enhancement (TFCE) and false discovery rate (FDR) based correction [14,18,19]. Moreover, within-subjects designs, paired repetitive measurements, and meta-analyses have also been noted as effective ways to reduce the false positivity in brain imaging studies [12,20].

Therefore, it is clear that adequate sample size, confound regression strategies, and strict enough threshold for the multiple comparison correction are important and necessary. In the present systematic review, we screened studies according to the criteria such as sufficient sample size (i.e., more than 40 participants per group according to the findings of our previous study [14]), proper head motion correction, appropriate correction strategies for multiple comparisons (e.g., FWE correction with voxel-wise P < 0.001 and cluster-wise P < 0.05 or permutation test with TFCE, FDR correction with q < 0.05). Through this strict screening for studies, we intend to focus on more reliable studies in this review and to get more reproducible and reliable results with less false positivity to some extent. We aim to reach a consensus on the key brain circuits in MDD. Finally, we also intended to raise some suggestions and directions for further fMRI studies on MDD.

2 RESULTS

2.1 General information

In the first step for article identification in PubMed, 1,012 papers were obtained and fed into the next screening procedures. After the first screening of titles and abstracts, most review and comment articles, the structural MRI studies, and the studies not focusing on MDD were excluded, with 462 lefts. Then, the full texts of these fMRI studies on depression were read through and screened by the relatively strict criteria (see materials and methods) for their qualities, leaving a final sample of 39 studies (Tab.1), of which 25 were R-fMRI studies, 17 were task-based fMRI (included 9 meta-analysis studies [49–51,53,55–59] and 3 multi-modality studies with both resting-state and task-based fMRI [36,46,56]). Note that some multi-modality studies only reported significant results regarding one modality (i.e., resting-state or task-based fMRI), so they were listed in only one result table.

Most of the early MRI studies on depression were structural MRI studies, in which the earliest one found in this searching was published in 1993 [60]. The recent 10 years have witnessed the rise of fMRI studies and multi-modality studies (the earliest fMRI study included in this review was published in 2011 [22], see Tab.1). Among studies reviewed here, a meta-analysis examined multiple neuroimaging methods and reported the lower sensitivity and specificity of structural MRI and task-based fMRI methods than R-fMRI in the differentiation of MDD patients from healthy control subjects (HCs) [56], which may occur due to the introduction of the potential complexity from task design and manipulation in measuring dynamic brain functions and which implies the advantages of R-fMRI in identifying neuroimaging markers for MDD [61].

Tab.1 shows that SPM [62], DPABI [63], FSL [64], and AFNI [65], realignment tools are most commonly used in the literature. All the included studies have applied head motion corrections, but some omitted key details regarding the nuisance covariates regression they applied. For example, simply “realignment” or “applying head motion correction” was declared in the methods section without further description. Moreover, a large proportion of studies were excluded due to the small sample size and inadequate multiple comparison strategies.

2.2 Results from R-fMRI studies

2.2.1 Altered spontaneous functional activities in MDD

Nine R-fMRI studies reported the abnormal functional metric values in MDD patients vs. HCs, including the regional homogeneity (ReHo), the amplitude of low-frequency fluctuations (ALFF), and the fractional amplitude of low-frequency fluctuations (fALFF) (Tab.2).

Four studies reported abnormally increased ReHo values in the left precuneus, the inferior frontal gyrus (IFG), and the medial prefrontal cortex (MPFC), as well as abnormally decreased ReHo values in the left putamen, the right postcentral gyrus (poCG), the right poCG and the lingual gyrus (LG) in patients with MDD [24,41,48,52]. Three studies reported abnormally increased ALFF in the IFG, the supplementary motor area (SMA), the left parahippocampal gyrus (PHG), the left anterior cingulate cortex (ACC), and the left superior temporal gyrus (STG) and abnormally decreased activation in the orbitofrontal cortices (OFC), the left cerebellum and the left middle temporal gyrus (MTG) [41,55,57]. Three studies reported abnormally increased fALFF in the visual cortex (VC), as well as abnormally decreased fALFF in the cuneus, the thalamus, MTG, the hippocampus, PHG, the amygdala, the dorsolateral prefrontal cortex (dlPFC), the insula, ACC, the superior frontal gyrus (SFG) and the inferior parietal lobule (IPL) [36,44,57].

Among these findings, abnormally decreased ALFF and fALFF in MTG [44,57] and reduced ReHo in poCG [41,48] were reported convergently. Furthermore, altered spontaneous activities in PHG and ACC were reported, albeit in opposite directions [36,55,57].

2.2.2 Altered resting-state functional connectivity in MDD

Fourteen studies reported abnormal functional connectivity in the resting state (Tab.3), in which most of them performed the seed-based analysis except one applied the voxel-mirrored homotopic connectivity (VMHC) [28], and two performed the independent component analysis (ICA) [26,38].

Decreased VMHC was found in STG, the insula, and the precuneus [28]. In ICA studies, abnormally increased functional connectivity was found between the precuneus/posterior cingulate cortex (PCC) and the left lateral parietal cortex (LPC), and the abnormally decreased coupling was found between the bilateral anterior portion of dlPFC [26,38]. In the seed-based analysis, the most common region of interest (ROI) is the amygdala. The abnormally increased connectivity was found between the amygdala and the prefrontal cortex, the precentral gyrus (prCG), poCG, PCC, the uncus, STG, the superior occipital gyrus (SOG), and the insula [31,34,43]. Moreover, abnormally decreased connectivity was found between the amygdala and IPL, the middle frontal gyrus (MFG), the insula, the cerebellum posterior lobe (CBPL), the cerebellar tonsil, IFG, the temporal pole, and the ventromedial prefrontal cortex (vmPFC) [31,35,43]. As for studies with multiple ROIs across brain networks, abnormally increased connectivity within the reward network and the default mode network (DMN) as well as between DMN and the executive control network was reported [32,43,48]. Meanwhile, abnormally decreased within DMN connectivity, within-network superior occipital and superior temporal connectivity, and between-network hippocampal connectivity was reported [37,42,43,59].

Among these findings, though the increased and decreased connectivity within or between networks were reported, DMN was the most involved brain network in MDD, suggesting not only the limited statical power but the complex neuropathobiology underlying the interactions of DMN and other confounding factors. One recent study investigated the DMN functional connectivity in a large sample of 1,300 depressed patients and 1,128 HCs and then found a significantly decreased functional connectivity within DMN in recurrent MDD vs. HCs as well as recurrent MDD vs. first-episode drug-naïve MDD patients. Furthermore, this effect was associated with medication usage rather than MDD duration [66]. Moreover, the abnormal connectivity between the amygdala and the insula, the precuneus, PCC, IPL, and CBPL was reported in more than one study.

2.3 Results from task-based fMRI studies

2.3.1 Altered activations in local brain regions of MDD

Fourteen task-based fMRI studies reported abnormal activations in local brain regions in patients with MDD (Tab.4). Tasks in these studies can be roughly classified into three categories: the emotional processing tasks (including the angry faces processing task and the emotion regulation task), the reward learning and valuation tasks (including the reward guessing task, the monetary reward tasks, probabilistic selection task, the probabilistic reward task, and other reward-related tasks) and the cognitive tasks (including the working memory tasks and the Tower of London paradigm).

In the studies performing the emotional processing tasks [21,25,40,45,50], abnormally increased activations were found in the right ventrolateral prefrontal cortex (vlPFC), OFC, the bilateral MTG, STG, and MFG in dlPFC, the amygdala, the striatum, the parahippocampal, the cerebellar, the fusiform and ACC, and abnormally decreased activation was found in the dACC. In the studies performing the reward learning and valuation tasks [27,33,47,51,58], abnormally increased activation was found in the dorsal medial prefrontal cortex (dmPFC) and MFG. Abnormally decreased activations were found in the striatum and the nucleus accumbens (NAcc), the caudate, the putamen, and the globus pallidus. In the studies performing the cognitive tasks [22,30,53], abnormally increased activation was found in the left dlPFC, and abnormally decreased activations were found in the middle occipital gyri (MOG), MFG, the right prCG, the cerebellum, and the left IPL. Moreover, a meta-analysis [53] reviewed the altered brain responses to working memory loads in MDD patients vs. HCs and reported the increased activations in the left IFG and MFG, the left prCG, the left insula, the right STG, and the right supramarginal gyrus (SG) and the decreased activations in the right prCG, the right precuneus, and the right insula.

In sum, studies on working memory generally reported the decreased activation in the right prCG. With both cognitive tasks and emotional processing tasks, researchers generally found altered activations in dlPFC and STG.

2.3.2 Altered functional connectivity during tasks in MDD

Five task-based fMRI studies reported the altered functional connectivity in patients with MDD (Tab.5). Tasks performed in the five studies can be roughly classified into four categories, three of which are the same as the above-mentioned tasks: the emotion processing tasks (including the emotion regulation task, conscious and non-conscious emotional faces processing tasks), the reward learning and valuation tasks (including the monetary incentive delay task, probabilistic selection task, and the probabilistic reward learning task), and the cognitive tasks (including the external attention task, the auditory oddball task, the continuous performance task, and the Go-No Go task). Moreover, the self-appraisal task is classified into the fourth category as self-perception and self-understanding.

In studies performing the emotion processing tasks, decreased functional connectivity between the amygdala and vlPFC in patients with MDD was reported [40]. In studies with the reward learning and valuation tasks, decreased connectivity between the prefrontal cortex and the ventral striatum (VS) as well as between NAcc and the midcingulate cortex (MCC) was reported [33,47]. In the study of Korgaonkar et al. [46], cognitive and emotion processing abilities were assessed through 5 fMRI tasks, and decreased DMN–FPN connectivity was observed in MDD non-remitters. And in the study with self-appraisal task, the negative modulatory effect of the medial prefrontal cortex (mPFC) on IPL was reported [29].

3 DISCUSSION

Due to the lack of consensus and reproducibility in fMRI studies on MDD, a growing body of literature has highlighted the methodological issues in neuroimaging research. According to these studies, we screened previous fMRI studies on MDD using criteria including sample size, preprocessing pipelines, and multiple comparison correction strategies. Contrary to our assumptions, after screening the previous studies with the above-mentioned criteria, both convergent and contradicted results were reported. Here, we focused on those convergent findings and discussed some implications accordingly.

3.1 Dysfunctions of brain regions and networks in MDD

Evidence from the R-fMRI studies reviewed here suggests that the alterations in DMN (especially in ACC, PCC, and the precuneus) and its couplings with other brain networks may play an important role in the pathology of MDD. As a key node of DMN, convergent dysfunction of the precuneus regarding abnormal regional activities and functional connectivity [26] in MDD patients versus HCs in resting-state were reported. The precuneus dysfunction has been found in several different mental disorders, even in migraines [67,68]. Thus the dysfunction of the precuneus may be a generalized functional marker across mental disorders. It has been proposed that DMN underlies the self-referential process and the negative rumination in MDD [69–72]. It is also found that the hyper-connectivity within DMN may predict the better outcomes of sertraline treatment [43], which was consistent with the recent study reporting the decreased functional connectivity within DMN in recurrent MDD patients [66]. To sum up, these studies implicated that DMN may be a potential biomarker of MDD and a possible target for future MDD treatment.

We found that task-based fMRI studies get some convergent results about cognitive and emotional processing. Specifically, the hyperactivity in dlPFC and altered functional connectivity between DMN and FPN, especially dlPFC, were reported in studies applying paradigms with both cognitive or emotional stimuli. These findings align with previous studies indicating the recruitment of dlPFC in emotion regulation [73]. Specifically, dlPFC has been repetitively reported for blunted activity in R-fMRI studies and proved as an effective target for the TMS treatment of MDD [74]. Of note, dlPFC was a key node of the FPN, which underlies executive and control functions [75]. FPN does not generate emotions directly but may underlie the reappraisal and reactive processes regarding emotions. Emotion regulation is identified as a process with conscious or non-conscious strategies to change the initial emotional reaction, especially negative emotions [76]. Regulating strategies can impact the generation and reactivity to emotions at different time points [77]. Thus, the causal relationship between cognition and emotion yields a confounding pattern asking for further examinations.

3.2 Caveats for future fMRI studies on MDD

We also observed inconsistency among these results. For example, both decreased and increased spontaneous activities in the PHG and ACC were reported [36,55,57]. And both increased and decreased functional connectivity between the amygdala and the insula, as well as within- and between-network connectivity regarding DMN was reported [28,31,43,78]. Issues including the lack of integrative assessments and interpretations of MDD, differences in experimental design, data acquisition and preprocessing procedures, the mixture of the heterogeneous participant populations, and inappropriate methodologies to depict the anatomical and functional brain may lead to such inconsistency in R-fMRI studies.

It is not surprising that task-based fMRI studies yielded such inconsistent results, given the differences in the psychological processes they investigated. In previous studies, these processes were commonly investigated by three sets of paradigms, which correspond to three functional systems according to the research domain criteria (RDoC) [79,80] framework: the emotion processing tasks, the reward learning and valuation tasks and the cognitive tasks. The cognitive tasks mainly assess the working memory and attention, which corresponds to the cognitive system. The reward learning and valuation tasks mainly assess the learning and valuating ability to rewards, which corresponds to the positive valence system. The emotion processing tasks mainly assessed the responses to threats and loss, which are closely related to the negative valence system [81]. Note that these psychological functions were investigated separately in most previous MDD studies lacking integrative assessments and interpretations. However, as a disorde characterized by a constellation of behavioral, emotional, and cognitive symptoms [82], MDD is posited to involve dysfunctions in many aspects of cognitive and emotional processes including inhibitory processes, deficits in working memory, rumination and reappraisal [83]. This suggests that applying a carefully selected and comprehensive set of neuro-behavioral measurements that covers multiple psychological functions may shed new light on depression research.

Furthermore, the lack of convergent results across previous fMRI studies investigating the same cognitive or emotional function indicates that differences in experimental design and data acquisition procedures may also contribute to inconsistency [84]. For example, in task-based studies included in the current review, working memory was tested by different versions of n-back tasks (e.g., continuous performance task, or n-back tasks that consist of conditions of 0-back, 2-back, and 3-back), mental arithmetic tasks, Tower of London tasks and other paradigms. In these experiments, participants were instructed to respond to different stimuli, such as letters or numbers [30,46,53]. Even in R-fMRI studies, the MRI data were acquired on different scanners, with different parameters and prompted for subjects (eyes open or closed), and through different scan durations, followed by various pre-processing workflows [28,32,36,44]. The absence of a “gold standard” for the data acquisition in R-fMRI studies may lead to the recent replicability and reproducibility crises as well as difficulties in interpreting this inconsistency [85,86]. Some researchers have raised the discussion of replicability issues and called for disciplines to advance research transparency and open science [87,88]. In practice, we have initiated the REST-meta-MDD Project with a standardized MRI data sharing and preprocessing protocol based on data processing assistant for resting-state fMRI (DPARSF) [89] and achieved preliminary success in open data sharing and collaborative research [66].

One other possible contribution to the inconsistency of previous fMRI studies on MDD is the heterogeneity of the investigated populations [84] regarding medications, treatment outcomes, onset ages, and subtypes (e.g., melancholic vs. atypical MDD) [90]. Therefore, future studies on MDD should carefully divide the MDD samples into subgroups according to these confounding factors so that a clearer understanding of the relationships between representations of fMRI brain alternations and MDD can be obtained.

Finally, the functional systems of the human brain have features of an intricate network with multiple temporal and spatial levels, which are largely distributed/embedded on the intrinsic two-dimensional structure of the cortical surface [91,92]. The network neuroscientific approach provided efficient new ways to map, analyze and model the elements and interactions of neurobiological systems as a graph [93]. There have been fMRI studies [94,95] using the network and graph theory analysis that shows some altered functional locations in line with the voxel-wise metrics, including Reho, ALFF, and fALFF and functional connectivity strength findings. For example, Long and his colleagues find local changes in the default-mode, sensorimotor and subcortical areas using a novel dynamic network-based approach [94]. What’s more, in order to properly understand the brain functional systems, it is necessary to obtain an accurate and explicit representation of the cortical surface considering its topology of a 2-D sheet and a highly folded geometry. The surface-based fMRI approach is a principled way to achieve this goal, which was reported to be nearly three times better than the traditional volume-based approaches in special localization of cortical areas [96]. However, most previous studies are still based on traditional voxel-wise metrics (e.g., Reho, ALFF, and fALFF) and functional connectivity. Future studies could leverage a surface-based network neuroscientific approach to advance understanding the brain dysfunctions in depressions from a more integrative perspective.

3.3 Limitations

Despite the strict criteria we applied, controversy still existed in the remaining studies. Due to the limited number of papers included in the present review, we did not further group studies according to factors such as age, race, severity of disease, or medication usage. Moreover, we note that graph theory metrics can depict the brain as a complex networked system and be worth considering. However, findings from graph theory studies may need to be interpreted in a more sophisticated framework and thus are out of the scope of the current review.

4 CONCLUSION

In this systematic review, we found that MDD was consistently characterized by abnormalities within the DMN and the FPN, as well as altered connectivity between them and other brain networks. However, highly inconsistent results remained, probably due to issues including the lack of integrative assessments and interpretations of MDD, differences in experimental design and data acquisition procedures, the mixture of the heterogeneous participant populations, and the relatively inappropriate methodologies to depict the anatomical and functional brain. Apart from a sufficient sample size, adequate head motion artifact correction, and multiple comparison correction strategies, future studies are recommended to perform a comprehensive set of neuro-behavioral measurements, consider the heterogeneity of MDD patients and other potentially confounding factors, apply surface-based neuroscientific network fMRI approaches and advance research transparency and open science by movements including developing state-of-the-art pipeline with open data sharing.

5 MATERIALS AND METHODS

5.1 Literature search strategy

Studies that are electronically published until June 10^th, 2020, were searched in PubMed (Fig.1). The search terms are “((English [Language]) AND ((MRI [Title/Abstract]) OR (fMRI [Title/Abstract]))) AND ((depression [Title/Abstract]) OR (depressive disorder [Title/Abstract]))”. The filter of PubMed was used to constrain the article types of the searching results, which could exclude the nonscientific papers, such as news, books, and documents. Moreover, the reference lists of the included articles were also screened.

5.2 Inclusion and exclusion criteria

We reviewed the titles and abstracts to exclude studies that are not fMRI studies on MDD, systematic reviews, and commentary articles. Then the following exclusion criteria were applied: (i) publications that have not been peer-reviewed; (ii) studies regarding other mental diseases, such as post-traumatic stress disorder (PTSD) or bipolar disorders. (iii) studies based solely on the structural MRI or diffusion MRI. (iv) studies reporting no fMRI findings; (v) studies focusing on topics other than the human brain, such as genetics or gut-brain axis.

To ensure the quality of the included studies, we further applied additional methodological criteria: (i) at least 40 participants per group; (ii) performed proper head motion artifact correction; (iii) with proper correction for multiple comparisons, e.g., FWE-based or FDR-based. Except for the permutation test with TFCE, the accepted thresholds for other FWE-based correction are voxel-wise P < 0.001 with cluster-wise P < 0.05, and for FDR-based correction q < 0.05.

References

[1] Friedrich,M. (2017). Depression is the leading cause of disability around the world. JAMA, 317: 1517

[2] Vos,T., Lim,S. S., Abbafati,C., Abbas,K. M., Abbasi,M., Abbasifard,M., Abbasi-Kangevari,M., Abbastabar,H., Abd-Allah,F., Abdelalim,A. . (2020). Global burden of 369 diseases and injuries in 204 countries and territories, 1990‒2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet, 396: 1204–1222

[3] rez-Stable,E. J., Miranda,J., oz,R. F. Ying,Y. (1990). Depression in medical outpatients. Underrecognition and misdiagnosis. Arch. Intern. Med., 150: 1083–1088

[4] Bandettini,P. (2012). Twenty years of functional MRI: the science and the stories. Neuroimage, 62: 575–588

[5] Bijsterbosch,J., Harrison,S. J., Jbabdi,S., Woolrich,M., Beckmann,C., Smith,S. Duff,E. (2020). Challenges and future directions for representations of functional brain organization. Nat. Neurosci., 23: 1484–1495

[6] Ioannidis,J. (2005). Why most published research findings are false. PLoS Med., 2: e124

[7] Button,K. S., Ioannidis,J. P., Mokrysz,C., Nosek,B. A., Flint,J., Robinson,E. S. (2013). Power failure: why small sample size undermines the reliability of neuroscience. Nat. Rev. Neurosci., 14: 365–376

[8] Zuo,X. N., Xu,T. Milham,M. (2019). Harnessing reliability for neuroscience research. Nat. Hum. Behav., 3: 768–771

[9] Ciric,R., Wolf,D. H., Power,J. D., Roalf,D. R., Baum,G. L., Ruparel,K., Shinohara,R. T., Elliott,M. A., Eickhoff,S. B., Davatzikos,C. . (2017). Benchmarking of participant-level confound regression strategies for the control of motion artifact in studies of functional connectivity. Neuroimage, 154: 174–187

[10] Yan,C. Cheung,B., Kelly,C., Colcombe,S., Craddock,R. C., Di Martino,A., Li,Q., Zuo,X. N., Castellanos,F. X. Milham,M. (2013). A comprehensive assessment of regional variation in the impact of head micromovements on functional connectomics. Neuroimage, 76: 183–201

[11] Ciric,R., Rosen,A. F. G., Erus,G., Cieslak,M., Adebimpe,A., Cook,P. A., Bassett,D. S., Davatzikos,C., Wolf,D. H. Satterthwaite,T. (2018). Mitigating head motion artifact in functional connectivity MRI. Nat. Protoc., 13: 2801–2826

[12] Eklund,A., Nichols,T. E. (2016). Cluster failure: Why fMRI inferences for spatial extent have inflated false-positive rates. Proc. Natl. Acad. Sci. USA, 113: 7900–7905

[13] Algermissen,J. Mehler,D. M. (2018). May the power be with you: are there highly powered studies in neuroscience, and how can we get more of them? J. Neurophysiol., 119: 2114–2117

[14] Chen,X., Lu,B. Yan,C. (2018). Reproducibility of R-fMRI metrics on the impact of different strategies for multiple comparison correction and sample sizes. Hum. Brain Mapp., 39: 300–318

[15] Van Dijk,K. R., Sabuncu,M. R. Buckner,R. (2012). The influence of head motion on intrinsic functional connectivity MRI. Neuroimage, 59: 431–438

[16] Power,J. D., Barnes,K. A., Snyder,A. Z., Schlaggar,B. L. Petersen,S. (2012). Spurious but systematic correlations in functional connectivity MRI networks arise from subject motion. Neuroimage, 59: 2142–2154

[17] Eklund,A., Nichols,T. E. (2016). Cluster failure: Why fMRI inferences for spatial extent have inflated false-positive rates. Proc. Natl. Acad. Sci. USA, 113: 7900–7905

[18] Nichols,T. (2003). Controlling the familywise error rate in functional neuroimaging: a comparative review. Stat. Methods Med. Res., 12: 419–446

[19] Genovese,C. R., Lazar,N. A. (2002). Thresholding of statistical maps in functional neuroimaging using the false discovery rate. Neuroimage, 15: 870–878

[20] Grimes,D. A. Schulz,K. (1996). Determining sample size and power in clinical trials: the forgotten essential. Semin. Reprod. Endocrinol., 14: 125–131

[21] Demenescu,L. R., Renken,R., Kortekaas,R., van Tol,M. J., Marsman,J. B., van Buchem,M. A., van der Wee,N. J., Veltman,D. J., den Boer,J. A. (2011). Neural correlates of perception of emotional facial expressions in out-patients with mild-to-moderate depression and anxiety. A multicenter fMRI study. Psychol. Med., 41: 2253–2264

[22] van Tol,M. J., van der Wee,N. J., Demenescu,L. R., Nielen,M. M., Aleman,A., Renken,R., van Buchem,M. A., Zitman,F. G. Veltman,D. (2011). Functional MRI correlates of visuospatial planning in out-patient depression and anxiety. Acta Psychiatr. Scand., 124: 273–284

[23] Bermingham,R., Carballedo,A., Lisiecka,D., Fagan,A., Morris,D., Fahey,C., Meaney,J., Gill,M. (2012). Effect of genetic variant in BICC1 on functional and structural brain changes in depression. Neuropsychopharmacology, 37: 285–2862

[24] Yang,X., Ma,X., Li,M., Liu,Y., Zhang,J., Huang,B., Zhao,L., Deng,W., Li,T. (2015). Anatomical and functional brain abnormalities in unmedicated major depressive disorder. Neuropsychiatr. Dis. Treat., 11: 2415–2423

[25] Gollier-Briant,F., re-Martinot,M. L., Lemaitre,H., Miranda,R., Vulser,H., Goodman,R., Struve,M., Fadai,T., Kappel,V. . (2016). Neural correlates of three types of negative life events during angry face processing in adolescents. Soc. Cogn. Affect. Neurosci., 11: 1961–1969

[26] Posner,J., Cha,J., Wang,Z., Talati,A., Warner,V., Gerber,A., Peterson,B. (2016). Increased default mode network connectivity in individuals at high familial risk for depression. Neuropsychopharmacology, 41: 1759–1767

[27] Casement,M. D., Keenan,K. E., Hipwell,A. E., Guyer,A. E. Forbes,E. (2016). Neural reward processing mediates the relationship between insomnia symptoms and depression in adolescence. Sleep, 39: 439–447

[28] Hermesdorf,M., Sundermann,B., Feder,S., Schwindt,W., Minnerup,J., Arolt,V., Berger,K., Pfleiderer,B. (2016). Major depressive disorder: Findings of reduced homotopic connectivity and investigation of underlying structural mechanisms. Hum. Brain Mapp., 37: 1209–1217

[29] Davey,C. G., Breakspear,M., Pujol,J. Harrison,B. (2017). A brain model of disturbed self-appraisal in depression. Am. J. Psychiatry, 174: 895–903

[30] ksel,D., Dietsche,B., Forstner,A. J., Witt,S. H., Maier,R., Rietschel,M., Konrad,C., then,M. M., Dannlowski,U., Baune,B. T. . (2017). Polygenic risk for depression and the neural correlates of working memory in healthy subjects. Prog. Neuropsychopharmacol. Biol. Psychiatry, 79: 67–76

[31] Ye,J., Shen,Z., Xu,X., Yang,S., Chen,W., Liu,X., Lu,Y., Liu,F., Lu,J., Li,N. . (2017). Abnormal functional connectivity of the amygdala in first-episode and untreated adult major depressive disorder patients with different ages of onset. Neuroreport, 28: 214–221

[32] Pan,P. M., Sato,J. R., Salum,G. A., Rohde,L. A., Gadelha,A., Zugman,A., Mari,J., Jackowski,A., Picon,F., Miguel,E. C. . (2017). Ventral striatum functional connectivity as a predictor of adolescent depressive disorder in a longitudinal community-based sample. Am. J. Psychiatry, 174: 1112–1119

[33] Admon,R., Kaiser,R. H., Dillon,D. G., Beltzer,M., Goer,F., Olson,D. P., Vitaliano,G. Pizzagalli,D. (2017). Dopaminergic enhancement of striatal response to reward in major depression. Am. J. Psychiatry, 174: 378–386

[34] Lopez,K. C., Luby,J. L., Belden,A. C. Barch,D. (2018). Emotion dysregulation and functional connectivity in children with and without a history of major depressive disorder. Cogn. Affect. Behav. Neurosci., 18: 232–248

[35] Mehta,N. D., Haroon,E., Xu,X., Woolwine,B. J., Li,Z. Felger,J. (2018). Inflammation negatively correlates with amygdala-ventromedial prefrontal functional connectivity in association with anxiety in patients with depression: Preliminary results. Brain Behav. Immun., 73: 725–730

[36] Qi,S., Yang,X., Zhao,L., Calhoun,V. D., Perrone-Bizzozero,N., Liu,S., Jiang,R., Jiang,T., Sui,J. (2018). MicroRNA132 associated multimodal neuroimaging patterns in unmedicated major depressive disorder. Brain, 141: 916–926

[37] Tokuda,T., Yoshimoto,J., Shimizu,Y., Okada,G., Takamura,M., Okamoto,Y., Yamawaki,S. (2018). Identification of depression subtypes and relevant brain regions using a data-driven approach. Sci. Rep., 8: 14082

[38] Tu,Z., Jia,Y. Y., Wang,T., Qu,H., Pan,J. X., Jie,J., Xu,X. Y., Wang,H. Y. (2018). Modulatory interactions of resting-state brain functional connectivity in major depressive disorder. Neuropsychiatr. Dis. Treat., 14: 2461–2472

[39] Wang,D., Xue,S. W., Tan,Z., Wang,Y., Lian,Z. (2019). Altered hypothalamic functional connectivity patterns in major depressive disorder. Neuroreport, 30: 1115–1120

[40] Fitzgerald,J. M., Klumpp,H., Langenecker,S. Phan,K. (2019). Transdiagnostic neural correlates of volitional emotion regulation in anxiety and depression. Depress. Anxiety, 36: 453–464

[41] Xia,M., Si,T., Sun,X., Ma,Q., Liu,B., Wang,L., Meng,J., Chang,M., Huang,X., Chen,Z. . (2019). Reproducibility of functional brain alterations in major depressive disorder: Evidence from a multisite resting-state functional MRI study with 1,434 individuals. Neuroimage, 189: 700–714

[42] Yao,Z., Zou,Y., Zheng,W., Zhang,Z., Li,Y., Yu,Y., Zhang,Z., Fu,Y., Shi,J., Zhang,W. . (2019). Structural alterations of the brain preceded functional alterations in major depressive disorder patients: Evidence from multimodal connectivity. J. Affect. Disord., 253: 107–117

[43] Chin Fatt,C. R., Jha,M. K., Cooper,C. M., Fonzo,G., South,C., Grannemann,B., Carmody,T., Greer,T. L., Kurian,B., Fava,M. . (2020). Effect of intrinsic patterns of functional brain connectivity in moderating antidepressant treatment response in major depression. Am. J. Psychiatry, 177: 143–154

[44] Zhu,D. M., Zhang,C., Yang,Y., Zhang,Y., Zhao,W., Zhang,B., Zhu,J. (2020). The relationship between sleep efficiency and clinical symptoms is mediated by brain function in major depressive disorder. J. Affect. Disord., 266: 327–337

[45] Hilland,E., Harmer,C. J., Browning,M., Maglanoc,L. A. (2020). Attentional bias modification is associated with fMRI response toward negative stimuli in individuals with residual depression: a randomized controlled trial. J. Psychiatry Neurosci., 45: 23–33

[46] Korgaonkar,M. S., Goldstein-Piekarski,A. N., Fornito,A. Williams,L. (2020). Intrinsic connectomes are a predictive biomarker of remission in major depressive disorder. Mol. Psychiatry, 25: 1537–1549

[47] Rupprechter,S., Romaniuk,L., Series,P., Hirose,Y., Hawkins,E., Sandu,A. L., Waiter,G. D., McNeil,C. J., Shen,X., Harris,M. A. . (2020). Blunted medial prefrontal cortico-limbic reward-related effective connectivity and depression. Brain, 143: 1946–1956

[48] Yang,Y., Zhu,D. M., Zhang,C., Zhang,Y., Wang,C., Zhang,B., Zhao,W., Zhu,J. (2020). Brain structural and functional alterations specific to low sleep efficiency in major depressive disorder. Front. Neurosci., 14: 50

[49] GrahamJ.,Salimi-Khorshidi G.,HaganC.,WalshN.,GoodyerI., LennoxB.. (2013) Meta-analytic evidence for neuroimaging models of depression: state or trait? J. Affect. Disord., 151, 423–431

[50] Groenewold,N. A., Opmeer,E. M., de Jonge,P., Aleman,A. Costafreda,S. (2013). Emotional valence modulates brain functional abnormalities in depression: evidence from a meta-analysis of fMRI studies. Neurosci. Biobehav. Rev., 37: 152–163

[51] Zhang,W. N., Chang,S. H., Guo,L. Y., Zhang,K. L. (2013). The neural correlates of reward-related processing in major depressive disorder: a meta-analysis of functional magnetic resonance imaging studies. J. Affect. Disord., 151: 531–539

[52] Iwabuchi,S. J., Krishnadas,R., Li,C., Auer,D. P., Radua,J. (2015). Localized connectivity in depression: a meta-analysis of resting state functional imaging studies. Neurosci. Biobehav. Rev., 51: 77–86

[53] Wang,X. L., Du,M. Y., Chen,T. L., Chen,Z. Q., Huang,X. Q., Luo,Y., Zhao,Y. J., Kumar,P. Gong,Q. (2015). Neural correlates during working memory processing in major depressive disorder. Prog. Neuropsychopharmacol. Biol. Psychiatry, 56: 101–108

[54] Zhong,X., Pu,W. (2016). Functional alterations of fronto-limbic circuit and default mode network systems in first-episode, drug-naïve patients with major depressive disorder: A meta-analysis of resting-state fMRI data. J. Affect. Disord., 206: 280–286

[55] Wang,W., Zhao,Y., Hu,X., Huang,X., Kuang,W., Lui,S., Kemp,G. J. (2017). Conjoint and dissociated structural and functional abnormalities in first-episode drug-naive patients with major depressive disorder: a multimodal meta-analysis. Sci. Rep., 7: 10401

[56] Kambeitz,J., Cabral,C., Sacchet,M. D., Gotlib,I. H., Zahn,R., Serpa,M. H., Walter,M., Falkai,P. (2017). Detecting neuroimaging biomarkers for depression: A meta-analysis of multivariate pattern recognition studies. Biol. Psychiatry, 82: 330–338

[57] Zhou,M., Hu,X., Lu,L., Zhang,L., Chen,L., Gong,Q. (2017). Intrinsic cerebral activity at resting state in adults with major depressive disorder: A meta-analysis. Prog. Neuropsychopharmacol. Biol. Psychiatry, 75: 157–164

[58] Keren,H., Callaghan,G., Vidal-Ribas,P., Buzzell,G. A., Brotman,M. A., Leibenluft,E., Pan,P. M., Meffert,L., Kaiser,A., Wolke,S. . (2018). Reward processing in depression: A conceptual and meta-analytic review across fMRI and EEG Studies. Am. J. Psychiatry, 175: 1111–1120

[59] Sha,Z., Xia,M., Lin,Q., Cao,M., Tang,Y., Xu,K., Song,H., Wang,Z., Wang,F., Fox,P. . (2018). Meta-connectomic analysis reveals commonly disrupted functional architectures in network modules and connectors across brain disorders. Cereb. Cortex, 28: 4179–4194

[60] Wu,J. C., Buchsbaum,M. S., Johnson,J. C., Hershey,T. G., Wagner,E. A., Tung,C. (1993). Magnetic resonance and positron emission tomography imaging of the corpus callosum: size, shape and metabolic rate in unipolar depression. J. Affect. Disord., 28: 15–25

[61] Biswal,B., Yetkin,F. Z., Haughton,V. M. Hyde,J. (1995). Functional connectivity in the motor cortex of resting human brain using echo-planar MRI. Magn. Reson. Med., 34: 537–541

[62] Friston,K. J., Ashburner,J., Frith,C. D., Poline,J. Heather,J. D. Frackowiak,R. S. (1995). Spatial registration and normalization of images. Hum. Brain Mapp., 3: 165–189

[63] Yan,C. G., Wang,X. D., Zuo,X. N. Zang,Y. (2016). DPABI: data processing & analysis for (resting-state) brain imaging. Neuroinformatics, 14: 339–351

[64] Jenkinson,M., Bannister,P., Brady,M. (2002). Improved optimization for the robust and accurate linear registration and motion correction of brain images. Neuroimage, 17: 825–841

[65] Cox,R. (1996). AFNI: software for analysis and visualization of functional magnetic resonance neuroimages. Comput. Biomed. Res., 29: 162–173

[66] Yan,C. G., Chen,X., Li,L., Castellanos,F. X., Bai,T. J., Bo,Q. J., Cao,J., Chen,G. M., Chen,N. X., Chen,W. . (2019). Reduced default mode network functional connectivity in patients with recurrent major depressive disorder. Proc. Natl. Acad. Sci. USA, 116: 9078–9083

[67] Castellanos,F. X., Margulies,D. S., Kelly,C., Uddin,L. Q., Ghaffari,M., Kirsch,A., Shaw,D., Shehzad,Z., Di Martino,A., Biswal,B. . (2008). Cingulate-precuneus interactions: a new locus of dysfunction in adult attention-deficit/hyperactivity disorder. Biol. Psychiatry, 63: 332–337

[68] Burstein,R., Noseda,R. (2015). Migraine: multiple processes, complex pathophysiology. J. Neurosci., 35: 6619–6629

[69] Zhou,H. X., Chen,X., Shen,Y. Q., Li,L., Chen,N. X., Zhu,Z. C., Castellanos,F. X. Yan,C. (2020). Rumination and the default mode network: Meta-analysis of brain imaging studies and implications for depression. Neuroimage, 206: 116287

[70] Chen,X., Chen,N. Shen,Y. Li,H. Li,L., Lu,B., Zhu,Z. C., Fan,Z. Yan,C. (2020). The subsystem mechanism of default mode network underlying rumination: A reproducible neuroimaging study. Neuroimage, 221: 117185

[71] Greicius,M. D., Flores,B. H., Menon,V., Glover,G. H., Solvason,H. B., Kenna,H., Reiss,A. L. Schatzberg,A. (2007). Resting-state functional connectivity in major depression: abnormally increased contributions from subgenual cingulate cortex and thalamus. Biol. Psychiatry, 62: 429–437

[72] Hamilton,J. P., Farmer,M., Fogelman,P. Gotlib,I. (2015). Depressive rumination, the default-mode network, and the dark matter of clinical neuroscience. Biol. Psychiatry, 78: 224–230

[73] Silvers,J. A., Weber,J., Wager,T. D. Ochsner,K. (2015). Bad and worse: neural systems underlying reappraisal of high- and low-intensity negative emotions. Soc. Cogn. Affect. Neurosci., 10: 172–179

[74] George,M. S., Taylor,J. J. Short,E. (2013). The expanding evidence base for rTMS treatment of depression. Curr. Opin. Psychiatry, 26: 13–18

[75] Ochsner,K. N., Silvers,J. A. Buhle,J. (2012). Functional imaging studies of emotion regulation: a synthetic review and evolving model of the cognitive control of emotion. Ann. N. Y. Acad. Sci., 1251: E1–E24

[76] EysenckM.. (2020) Cognitive Psychology: A Student’s Handbook. London: Taylor & Francis Group

[77] GrossJ. J.. (2007) Emotion regulation: conceptual foundations. In: Handbook of Emotion Regulation, Gross, J. J. (Ed.), pp. 3–24. New York: The Guilford Press

[78] Wang,L., Zhao,Y., Edmiston,E. K., Womer,F. Y., Zhang,R., Zhao,P., Jiang,X., Wu,F., Kong,L., Zhou,Y. . (2020). Structural and functional abnormities of amygdala and prefrontal cortex in major depressive disorder with suicide attempts. Front. Psychiatry, 10: 923

[79] Peterson,B. (2015). Editorial: Research Domain Criteria (RDoC): a new psychiatric nosology whose time has not yet come. J. Child Psychol. Psychiatry, 56: 719–722

[80] Sanislow,C. A., Ferrante,M., Pacheco,J., Rudorfer,M. V. Morris,S. (2019). Advancing translational research using NIMH research domain criteria and computational methods. Neuron, 101: 779–782

[81] Tozzi,L., Staveland,B., Holt-Gosselin,B., Chesnut,M., Chang,S. E., Choi,D., Shiner,M., Wu,H., Lerma-Usabiaga,G., Sporns,O. . (2020). The human connectome project for disordered emotional states: Protocol and rationale for a research domain criteria study of brain connectivity in young adult anxiety and depression. Neuroimage, 214: 116715

[82] Association,A. (2013). Diagnostic and Statistical Manual of Mental Disorders (DSM-5®). American Psychiatric Publishing, Inc

[83] Gotlib,I. H. (2010). Cognition and depression: current status and future directions. Annu. Rev. Clin. Psychol., 6: 285–312

[84] ller,V. I., Cieslik,E. C., Serbanescu,I., Laird,A. R., Fox,P. T. Eickhoff,S. (2017). Altered brain activity in unipolar depression revisited: Meta-analyses of neuroimaging studies. JAMA Psychiatry, 74: 47–55

[85] Cesario,J. (2014). Priming, replication, and the hardest science. Perspect. Psychol. Sci., 9: 40–48

[86] Maxwell,S. E., Lau,M. Y. Howard,G. (2015). Is psychology suffering from a replication crisis? What does “failure to replicate” really mean? Am. Psychol., 70: 487–498

[87] Tackett,J. L., Lilienfeld,S. O., Patrick,C. J., Johnson,S. L., Krueger,R. F., Miller,J. D., Oltmanns,T. F. Shrout,P. (2017). It’s time to broaden the replicability conversation: Thoughts for and from clinical psychological Science. Perspect. Psychol. Sci., 12: 742–756

[88] Tackett,J. L., Brandes,C. M., King,K. M. Markon,K. (2019). Psychology’s Replication Crisis and Clinical Psychological Science. Annu. Rev. Clin. Psychol., 15: 579–604

[89] Chao-Gan,Y. (2010). DPARSF: A MATLAB toolbox for “pipeline” data analysis of resting-state fMRI. Front. Syst. Neurosci., 4: 13

[90] Ahmed,A. T., Frye,M. A., Rush,A. J., Biernacka,J. M., Craighead,W. E., McDonald,W. M., Bobo,W. V., Riva-Posse,P., Tye,S. J., Mayberg,H. S. . (2018). Mapping depression rating scale phenotypes onto research domain criteria (RDoC) to inform biological research in mood disorders. J. Affect. Disord., 238: 1–7

[91] Bullmore,E. (2009). Complex brain networks: graph theoretical analysis of structural and functional systems. Nat. Rev. Neurosci., 10: 186–198

[92] Dale,A. M., Fischl,B. Sereno,M. (1999). Cortical surface-based analysis. I. Segmentation and surface reconstruction. Neuroimage, 9: 179–194

[93] Bassett,D. S. (2017). Network neuroscience. Nat. Neurosci., 20: 353–364

[94] Long,Y., Cao,H., Yan,C., Chen,X., Li,L., Castellanos,F. X., Bai,T., Bo,Q., Chen,G., Chen,N. . (2020). Altered resting-state dynamic functional brain networks in major depressive disorder: Findings from the REST-meta-MDD consortium. Neuroimage Clin., 26: 102163

[95] Ye,M., Yang,T., Qing,P., Lei,X., Qiu,J. (2015). Changes of functional brain networks in major depressive disorder: A graph theoretical analysis of resting-state fMRI. PLoS One, 10: e0133775

[96] Coalson,T. S., Van Essen,D. C. Glasser,M. (2018). The impact of traditional neuroimaging methods on the spatial localization of cortical areas. Proc. Natl. Acad. Sci. USA, 115: E6356–E6365

RIGHTS & PERMISSIONS

The Authors (2022). Published by Higher Education Press.