Skip to main content
NCBI home page
Human Brain Mapping logoLink to Human Brain Mapping
. 2017 Nov 6;39(2):758–771. doi: 10.1002/hbm.23880

A dimensional approach to determine common and specific neurofunctional markers for depression and social anxiety during emotional face processing

Lizhu Luo 1,†, Benjamin Becker 1,†, Xiaoxiao Zheng 1, Zhiying Zhao 1, Xiaolei Xu 1, Feng Zhou 1, Jiaojian Wang 1, Juan Kou 1, Jing Dai 1, Keith M Kendrick 1,✉
PMCID: PMC6866417  PMID: 29105895

Abstract

Major depression disorder (MDD) and anxiety disorder are both prevalent and debilitating. High rates of comorbidity between MDD and social anxiety disorder (SAD) suggest common pathological pathways, including aberrant neural processing of interpersonal signals. In patient populations, the determination of common and distinct neurofunctional markers of MDD and SAD is often hampered by confounding factors, such as generally elevated anxiety levels and disorder‐specific brain structural alterations. This study employed a dimensional disorder approach to map neurofunctional markers associated with levels of depression and social anxiety symptoms in a cohort of 91 healthy subjects using an emotional face processing paradigm. Examining linear associations between levels of depression and social anxiety, while controlling for trait anxiety revealed that both were associated with exaggerated dorsal striatal reactivity to fearful and sad expression faces respectively. Exploratory analysis revealed that depression scores were positively correlated with dorsal striatal functional connectivity during processing of fearful faces, whereas those of social anxiety showed a negative association during processing of sad faces. No linear relationships between levels of depression and social anxiety were observed during a facial‐identity matching task or with brain structure. Together, the present findings indicate that dorsal striatal neurofunctional alterations might underlie aberrant interpersonal processing associated with both increased levels of depression and social anxiety.

Keywords: biomarker, depression, face emotion, putamen, social anxiety, trait

1. INTRODUCTION

Major depression disorder (MDD) and anxiety disorders have high estimated lifetime prevalence rates of 16.6% and 12.1% respectively, and are a major cause of distress both for patients and their families (Adams, Balbuena, Meng, & Asmundson, 2016; Kessler et al., 2003, 2005). Epidemiological data indicates high rates of comorbidity between MDD and anxiety disorders (Adams et al., 2016; Kessler, Stang, Wittchen, Stein, & Walters, 1999), particularly social anxiety disorder (SAD) (Adams et al., 2016; Chavira, Stein, Bailey, & Stein, 2004; Ohayon and Schatzberg, 2010). Such comorbidities suggest that they share common vulnerabilities and pathways (Langer and Rodebaugh, 2014), including genetic influences (Kendler et al., 2011), personality traits (Bienvenu, Hettema, Neale, Prescott, & Kendler, 2007), and affective‐cognitive symptoms, that is, biased emotional processing and poor social functioning (Ladouceur et al., 2005; Wells and Carter, 2002).

On the symptomatic level, both MDD and SAD are characterized by impaired decoding of interpersonal affective‐cognitive signals, which has been associated with poor social functioning, and might constitute an etiological factor contributing to the development and maintenance of both disorders (Demenescu, Kortekaas, den Boer, & Aleman, 2010). Across both disorders, aberrant, often negatively biased, processing of facial expressions has been consistently reported (Bourke, Douglas, & Porter, 2010; Gilboa‐Schechtman, Foa, Vaknin, Marom, & Hermesh, 2008; Gilboa‐Schechtman, Foa, & Amir, 1999; Pishyar, Harris, & Menzies, 2004).

Notably, previous findings on emotional face processing in MDD and SAD suggest an emotion‐specific negative processing bias in both disorders. Whereas both are characterized by increased interpersonal sensitivity for, and a response bias toward, angry facial expressions, MDD has been associated with an enhanced vigilance and selective attention for sad facial expressions (Bourke et al., 2010; Stuhrmann, Suslow, & Dannlowski, 2011). However, the few studies that directly compared facial emotion processing in MDD and SAD patients suggest a more complex picture, with MDD patients displaying a higher sensitivity for sad expressions and a higher identification bias for both angry and happy expressions, while SAD patients display a stronger response bias for sad faces (Gilboa‐Schechtman et al., 2008; Joormann and Gotlib, 2006).

In accordance with the common and distinct facial emotion processing alterations on the behavioral level, recent meta‐analyses of functional imaging studies on altered neural emotional face processing comparing separate samples of MDD and SAD with healthy controls have revealed both overlapping and disorder‐specific networks. Meta‐analytic data in SAD has revealed increased neural activity in the amygdala, globus pallidus, hippocampal and temporal regions, and the prefrontal and anterior cingulate cortex (Binelli et al., 2014; Gentili et al., 2016; Hattingh et al., 2013). Similar analyses in MDD have most consistently reported increased activity in striatal, including globus pallidus, and limbic, particularly amygdala and hippocampal, regions and decreased frontal and anterior cingulate activity (Delvecchio et al., 2012; Fitzgerald, Laird, Maller, & Daskalakis, 2008; Groenewold, Opmeer, de Jonge, Aleman, & Costafreda, 2013; Lai, 2014; Palmer, Crewther, & Carey, 2015). Together, the available neuroimaging data suggest common alterations in striatal and limbic regions as well as distinct, disorder‐specific, alterations in prefrontal regions during emotional processing. Notably, emotional valence including face emotions has been shown to play an important role in modulating these neural abnormalities in both disorders (Groenewold et al., 2013; Hattingh et al., 2013).

Some initial studies have directly compared SAD and MDD patients to determine common and disorder‐specific neural markers. Given the high comorbidity between the disorders, a common methodological approach in this context is the comparison of patients with either MDD or SAD with a comorbid group exhibiting both disorders. For example, one study employing this approach reported that MDD plus SAD patients could be specifically characterized by alterations in the middle cingulate cortex and precentral gyrus (less activity in MDD alone and more in SAD alone) and the posterior cingulate cortex (more activity in MDD alone and less in SAD alone) (Waugh, Hamilton, Chen, Joormann, & Gotlib, 2012).

However, the identification of common and distinct neurofunctional markers for the two disorders is often hampered by serious confounding factors inherent in patient‐based studies such as medication, complex interactions between co‐morbid disorders, and disorder‐specific brain structural alterations (van Tol et al., 2010). These issues have led researchers to argue that traditional patient‐based research strategies alone might not reveal valid biomarkers and encouraged the development of dimensional disorder conceptualizations such as the Research Domain Criteria (RDoC) framework (Cuthbert, 2014; Insel et al., 2010; Sanislow et al., 2010; Wakschlag et al., 2015).

Against this background, the present study employed a dimensional approach assessing individual variations in the levels of depression and social anxiety in healthy subjects to determine common and specific functional MRI‐based neural markers for MDD and SAD while controlling for a range of important confounders inherent in patient studies, such as medication. To this end, associations between levels of depression and social anxiety with fMRI‐based neural activity and connectivity were investigated during explicit emotional face processing in a large sample (n = 91) of healthy subjects. To facilitate the determination of distinct neurofunctional markers, trait anxiety was controlled for as it represents a stable general anxiety marker across different domains with a strong involvement in both depression and social anxiety (Anagnostou et al., 2012). Moreover, recent studies even found that depression biomarkers overlapped with those for generalized anxiety disorder (Drysdale et al., 2017; Wager and Woo, 2017). To further control for potential confounding effects of associations between levels of depression or social anxiety with simple face recognition, participants additionally underwent a face and house matching fMRI paradigm. Finally, to account for potential associations between the disorder‐relevant dimensions and brain structure a voxel‐based morphometry (VBM) analysis was conducted. Based on previous meta‐analytic findings in MDD and SAD patients, we expected that higher levels of depression and social anxiety would be associated with emotion processing‐related increases of neural reactivity in limbic and striatal regions and differential alterations in prefrontal responses.

2. MATERIALS AND METHODS

2.1. Participants

A total of 92 healthy, young right‐handed Chinese (Han) students (47 males; age range = 18–27 years; M ± SD = 21.68 ± 2.22 years) participated in the study after providing written informed consent. All volunteers reported no history of medical, neurological, or psychiatric disorders, and no history of head injury as well as frequent drug, cigarette or alcohol use and were free of MRI contraindications. This study was approved by the local Ethics Committee at the University of Electronic Science and Technology of China (UESTC), and conformed to the latest revision of the Declaration of Helsinki.

2.2. Measurements

To test associations between brain structure, neural activity and pathology relevant symptom dimensions, levels of depression, social anxiety, and trait anxiety were assessed using validated Chinese versions of the Beck Depression Inventory, BDI‐II (Beck, Steer, Ball, & Ranieri, 1996; Wang et al., 2011) which assess the level of depressive symptoms during the last two weeks; Liebowitz Social Anxiety Scale, LSAS (He and Zhang, 2004; Liebowitz, 1987), and the State‐Trait Anxiety Inventory, STAI (Li and Qian, 1995; Speilberger, Gorsuch, Lushene, Vagg, & Jacobs, 1983) (see Table 1 for details). The covariance matrix and correlation coefficients of the scales used in the regression models are presented in Supporting Information, Table S1 (BDI, LSAS, STAI‐Trait, for completeness also STAI‐State is presented). Given that collinear regressors in fMRI models might lead to unreliable estimations (Andrade, Paradis, Rouquette, & Poline, 1999; Mumford, Poline, & Poldrack, 2015) collinearity between the regressors was initially assessed using the variance inflation factor (VIF) (for a similar approach see also (Chau et al., 2017; Ohashi et al., 2017)). A VIF >5 is typically considered to indicate problematic collinearity (Mumford et al., 2015; O'Brien, 2007). VIFs in this study were <1.6 (Supporting Information, Table S2), arguing against problematic collinearity.

Table 1.

Group characteristics of subject age and questionnaire scores in this study (N = 91) and from the lab database for structural data analysis (N = 234)

Variables N = 91 N = 234
Mean SD Mean SD
Age (years) 21.69 2.24 21.65 2.13
State‐Trait Anxiety Inventory (STAI)
Trait 40.76 7.86 39.93 7.54
State 38.87 8.46 \ \
Liebowitz Social Anxiety Scale (LSAS) 45.25 20.52 51.18 23.53
Fear/anxiety 24.03 11.15 26.97 12.58
Avoidance 21.22 10.95 24.28 12.13
Beck Depression Inventory (BDI‐II) 7.03 6.43 6.36 5.42
Open in a new tab

Note. SD, standard deviation.

2.3. Stimuli

For the emotional face processing task, a total of 150 faces were selected from two standardized Asian facial expression databases: Chinese Facial Affective Picture System (Gong, Huang, Wang, & Luo, 2011) and Taiwanese Facial Expression Image Database (TFEID) (Chen and Yen, 2007), including happy, angry, fearful, sad, and neutral faces each from 30 different individual actors (15 males). Faces were rescaled and covered with an oval frame to mask individual characteristics (e.g., hair) (Figure 1a).

Figure 1.

Figure 1

Open in a new tab

The procedure for (a) the emotional face processing task and (b) the face matching task

For the face‐matching task, a total of 60 neutral faces (30 pairs, 15 male pairs) and 60 houses (30 pairs) were selected from our own database (faces were previously used in (Gao et al., 2016)) (Figure 1b). Given that subjects had to match a previously shown stimulus with two simultaneously presented stimuli from the same condition (house or face) we assessed the similarity between stimulus pairs in a pre‐study. An independent sample (n = 20, 10 males) rated the similarity between each pair of faces or houses on a 9‐point Likert scale (1 = not similar at all, 9 = very similar). No significant differences between the face and house stimuli set were found (M face ± SD = 4.25 ± 1.19; Mhouse ± SD = 4.27 ± 1.28; p = .903), indicating that differences in cognitive load between conditions were controlled for.

2.4. fMRI paradigms

The emotional face processing paradigm used an event‐related design. Trials were distributed over 3 subsequent runs of 50 trials each, and balanced for face emotion and gender (duration = 570 s per run). For each trial, a face was shown for 7000 ms, with the first 1500 ms for passive viewing, and then 2500 ms for subjects to identify the emotion shown (select from the five alternatives neutral, sad, happy, angry, fearful presented below the facial picture) and finally 3000 ms to rate the emotional intensity of the emotion displayed (1 = very weak to 9 = very strong). After each trial a jittered fixation‐cross was presented for 3600–4400 ms (mean ITI = 4000 ms, see Figure 1a) that served as low‐level baseline.

The face‐matching task used an ABBA‐block design incorporating 10 blocks (5 for faces) in a single run of 460 s duration. There were 6 trials in each block. In each trial, a fixation was shown first for 1500 ms, and then a face or house was presented for 1500 ms, followed by the paired faces or houses for 2500 ms during which subjects had to choose which face or house had been presented previously. Blocks were separated by a low‐level baseline (fixation cross) presented for 12 s. The order of blocks was balanced across the gender of the participants, with half of the males and half of the females starting with a facial stimuli block (Figure 1 b).

Stimuli were presented using E‐Prime 2.0 software (Psychological Software Tools, Pittsburgh, Pennsylvania), and presented with a cloned projection display in the scanner gantry.

2.5. Image acquisitions

MRI data were obtained on a 3 T GE Discovery MR750 system (General Electric, Milwaukee, WI, USA) located in the neuroimaging center of the UESTC. High‐resolution whole‐brain T1‐weighted images were acquired using a spoiled gradient echo pulse sequence (TR = 6 ms; TE = 1.964 ms; flip angle = 9°; number of slices = 156; slice thickness, 1 mm; FOV = 256 × 256 mm2; matrix = 256 × 256). A total of 230 volumes of T2*‐weighted echo planar images were acquired during the face matching task and 285 volumes per run of the emotional face processing task with the following acquisition parameters: repetition time (TR) = 2000 ms; echo time (TE) = 30 ms; number of slices = 39; slice orientation = axial; slice thickness = 3.4 mm; spacing = 0.6 mm; field of view (FOV) = 240 × 240 mm2; matrix = 64 × 64; flip angle = 90°.

2.6. Data analysis

2.6.1. MRI data preprocessing

Structural MRI data processing was performed in Matlab (R2014b, MathWorks, Inc., USA) using the Computational Anatomy Toolbox (CAT12; http://dbm.neuro.uni-jena.de/cat) based on the Statistical Parametric Mapping (SPM12; http://www.fil.ion.ucl.ac.uk/spm) software. During preprocessing, default settings were used as described in the toolbox manual (http://dbm.neuro.uni-jena.de/cat12/CAT12-Manual.pdf). The SPM12 tissue probability maps (TPMs) were used for the initial spatial registration and segmentation. The images were segmented into three tissue types: gray matter (GM), white matter (WM), and cerebrospinal fluid (CSF). A group‐specific template based on all participants was created using the DARTEL algorithm. Next, participants' scans were warped onto it through a flow field which stored the deformation information. Finally, gray matter images were affine spatially normalized to Montreal Neurological Institute (MNI) space and smoothed with an 8 mm FWHM Gaussian kernel. Data quality was assessed and total intracranial volume (TIV) was estimated to be used as covariate on the second level.

Functional MRI data were preprocessed using SPM12. For each functional time series, the first five volumes were discarded to allow for MR equilibration, and the remaining functional images were initially realigned to the first image. The T1 images of each subject were segmented into GM, WM, and CSF and a skull‐stripped bias corrected brain image was created using the ImCalc function. Next, the functional images and the structural images were coregistered. Finally, the images were normalized to the standard MNI template, resampled at 3 × 3 × 3 mm3 voxel size, and spatially smoothed using an 8 mm FWHM Gaussian kernel.

After preprocessing, the first‐level design matrix was built for each paradigm including the six movement parameters as covariates. The face emotion task included separate regressors for the five face viewing phases (happy, angry, fearful, sad, and neutral) as contrasts of interest, as well as separate regressors for the responses for emotional identity and emotional intensity (across all conditions) as regressors of no interest. The face matching task incorporated separate regressors for face and house matching, generating face > house and house > face as contrasts of interest.

2.6.2. Statistical analysis

Statistical analyses of the behavioral indices (accuracy and emotional intensity for the emotional face processing task; accuracy for the face matching task) were conducted using SPSS Statistics 22 (Armonk, NY: IBM Corp.).

Before assessing associations between levels of depression and social anxiety with neural indices the normal distribution of all questionnaire scales was assessed. Given that the BDI (Shapiro‐Wilk, p < .05) did not show a normal distribution in the present sample nonparametric tests were employed to determine associations between levels of depression and social anxiety with neural indices. To evaluate associations between depression and social anxiety with brain structure, we used simple regression analyses as implemented in the Statistical nonParametric Mapping (SnPM13) toolbox (http://warwick.ac.uk/snpm), for BDI or LSAS with STAI‐Trait, gender and TIV as covariates, and a voxel‐wise significance threshold of p < .05 FDR‐corrected. Only clusters of k ≥ 10 contiguous voxels that survived the voxel‐wise FDR‐correction are reported. Concordant analyses were run for the two neural activity tasks (without including TIV as covariate). Brain regions were identified using MNI coordinates and the Automated Anatomic Labelling (AAL) atlas (Tzourio‐Mazoyer et al., 2002) as implemented in the WFU PickAtlas (School of Medicine, Winston‐Salem, NC). To further explore common and distinct brain regions associated with levels of depression and social anxiety, overlap maps between results from neural associations of both symptom dimensions were calculated. To this end, the thresholded activity maps for the significant associations with levels of depression or social anxiety (p < .05, FDR‐corrected) were exported and the overlap was examined using SPM's Image Calculator (ImCalc) and the function i1.*i2.

To identify associations between the symptom dimensions and neural processing on the network level, an exploratory functional connectivity analysis was conducted using a seed‐to‐whole brain approach and a generalized form of context‐dependent psychophysiological interactions (gPPI) (McLaren, Ries, Xu, & Johnson, 2012), with a threshold of p < .001 uncorrected and cluster extent k ≥ 10. For this analysis seed regions were constructed by placing 6‐mm‐radius sphere centered at peak coordinates of each cluster within the common region using MarsBaR (Brett, Anton, Valabregue, & Poline, 2002).

3. RESULTS

One participant with clinically relevant symptom loads (BDI = 41, STAI‐Trait = 72) was excluded, leaving a total of 91 subjects (47 males; age range = 18–27 years; M ± SD = 21.69 ± 2.24 years) for the final analyses (Table 1). Four subjects had head motions >2.5 mm within a run of the emotional face processing task (two for Run 1, one for Run 3, and one for Runs 2 and 3), consequently these runs were excluded from further analyses. The quality of the scales was checked through reliability analysis with all three scales yielding excellent internal consistencies (Supporting Information).

3.1. Task and gender effects

During the face emotion recognition paradigm, subjects showed high recognition accuracy (happy, M ± SD = 97.87% ± 3.56%; angry, 91.12% ± 8.58%; fearful, 88.56% ± 9.87%; sad, 93.64% ± 6.45%; neutral, 91.07% ± 11.43%), confirming that they attended reliably to the facial stimuli. In addition, emotional intensity ratings given indicated that the emotional expression faces were, as expected, rated higher than the neutral ones (happy, M ± SD = 5.01 ± 0.88; angry, 5.19 ± 0.79; fearful, 5.21 ± 0.69; sad, 4.81 ± 0.82; neutral, 2.95 ± 1.77). Additional repeated measures ANOVAs for both, the emotional face processing task and the face‐matching task yielded main effects of stimulus type; however, no main or interaction effects involving participants gender (details see Supporting Information). Together, the behavioral findings confirmed first that subjects attentively processed the stimuli and second that there were no effects of subject gender.

3.2. Associations between levels of depression and social anxiety with behavior

To match the analysis of the brain imaging data, associations between levels of social anxiety and depression with emotion recognition accuracy were explored using nonparametric correlation analyses with BDI and LSAS scores as separate predictors and emotion recognition accuracy as dependent variable (computed with the Permutation Analysis of Linear Models toolbox ‐PALM, version alpha 109, https://fsl.fmrib.ox.ac.uk/fsl/fslwiki/PALM (Winkler, Ridgway, Webster, Smith, & Nichols, 2014); permutation test, number of permutations 10,000). The analyses revealed a significant negative association between BDI scores and emotion recognition accuracy for happy facial expressions (r = −0.205, p = .046; controlling for STAI‐Trait and gender as covariates). However, the very high accuracy for happy facial expressions (mean recognition accuracy 97.87%, 58 of 91 subjects reached 100% accuracy for happy faces) suggests a ceiling effect thus inhibiting an interpretation of this finding. No significant associations were observed between the BDI and accuracy for the other face conditions or the LSAS and emotion recognition accuracy (all p > .383, controlling for STAI‐Trait and gender as covariates).

3.3. Voxel‐based morphometry (VBM)

There were no significant linear relationships between levels of depression or social anxiety and brain structure at p < .05 (FDR‐corrected). To further evaluate the lack of associations, we included participants from a larger database of brain structural data from our lab (total n = 234, age range = 17–27 years, M ± SD = 21.65 ± 2.13) (Table 1), again there were no significant linear relationships with brain structure. The analyses were additionally repeated using non‐DARTEL based preprocessing and quality assessments as implemented in CAT12. Again, no significant linear relationships were observed.

3.4. Face matching task: Neural activity

There were no significant linear relationships between depression or social anxiety with face recognition‐related neural activity under both face and face > house conditions at p < .05 FDR‐corrected threshold, controlling for both gender and trait anxiety.

3.5. Emotional face processing task: Neural activity

Results from the regression analysis revealed an association between levels of depression and neural activity during processing fearful faces. The BDI scores, controlling for trait anxiety and gender as covariates, were positively correlated with activity during processing of fearful faces in the left caudate (k = 443, t 87 = 4.69, x/y/z: −18/23/−7), right globus pallidus (k = 168, t 87 = 4.59, x/y/z: 15/5/2), left inferior parietal lobule (k = 260, t 87 = 4.41, x/y/z: −33/−55/44), bilateral anterior cingulate cortex (right, k = 13, t 87 = 3.08, x/y/z: 15/38/−1; left, k = 68, t 87 = 4.19, x/y/z: −9/35/5), bilateral posterior cingulate cortex (right, k = 14, t 87 = 3.17, x/y/z: 6/−43/5; left, k = 105, t 87 = 4.21, x/y/z: −6/37/17), bilateral inferior frontal gyrus (right, k = 22, t 87 = 4.60, x/y/z: 54/29/5; left, k = 52, t 87 = 3.95, x/y/z: −45/38/−10), left medial prefrontal cortex (k = 112, t 87 = 3.78, x/y/z: −3/35/50), right hippocampus (k = 20, t 87 = 3.52, x/y/z: 33/−13/−16), bilateral fusiform gyrus (right, k = 58, t 87 = 4.14, x/y/z: 39/−67/−16; left, k = 10, t 87 = 3.29, x/y/z: −30/−58/−19), and other regions including right precentral gyrus, left postcentral gyrus, right precuneus, bilateral middle temporal gyrus, left lingual gyrus, right superior frontal gyrus, left middle frontal gyrus, left supplemental motor area, and bilateral cerebellum (Table 2 and Figures 2a and 3). No significant negative linear relationship was observed, and none during processing of other emotional or neutral faces. Levels of social anxiety (LSAS scores) were positively associated with activity during processing of sad faces in the left globus pallidus (k = 95, t 87 = 5.93, x/y/z: −21/−4/−7), and left superior temporal gyrus (k = 11, t 87 = 4.90, x/y/z: −48/2/−4) (Figure 2b), controlling for trait anxiety and gender as covariates. Findings remained stable after additionally controlling for state anxiety levels as assessed by the state anxiety subscale of the STAI, as covariate. To further evaluate interaction effects of levels of depression and social anxiety on the observed associations, follow‐up analyses specifically examined the interaction term for the BDI and LSAS using SnPM simple regression models. These analyses did not reveal significant interaction effects on processing of fearful or sad faces. No significant linear relationships were found for the other face emotion conditions. Finally, no significant positive or negative linear relationships were observed between levels of trait anxiety (trait subscales of STAI) and neural activity with gender and BDI or LSAS as covariates.

Table 2.

Simple regression for BDI and LSAS, with gender and STAI‐Trait as nuisance covariates, using SnPM on whole‐brain level and showing positive effects in response to fearful and sad faces

Coordinates
Regions Cluster k x y z t
BDI: fearful faces
Positive correlation
Calcarine sulcus, R 373 24 −91 −1 4.99
Inferior occipital gyrus, R 33 −82 −4 4.16
Cerebellum, R 9 −76 −16 4.01
Caudate, L 443 −18 23 −7 4.69
Putamen, L −21 −10 11 4.55
Midbrain 9 −28 −22 3.98
Triangular inferior frontal gyrus, R 22 54 29 5 4.60
Globus pallidus, R 168 15 5 2 4.59
Globus pallidus, R 12 −4 −7 3.67
Caudate, R 9 8 14 3.60
Inferior parietal lobule, L 260 −33 −55 44 4.41
Inferior parietal lobule, L −36 −52 35 4.04
Middle occipital gyrus, L −30 −64 29 3.97
Middle temporal gyrus, R 28 39 −64 11 4.26
Posterior cingulate cortex, L 105 −6 −37 17 4.21
Precuneus, L −9 −43 5 3.46
Lingual gyrus, L 169 −24 −91 −16 4.19
Inferior occipital gyrus, L −27 −82 −4 4.13
Inferior occipital gyrus, L −15 −94 −10 3.80
Anterior cingulate cortex, L 68 −9 35 5 4.19
Anterior cingulate cortex, L −18 50 −1 3.19
Precentral gyrus, R 57 21 −19 59 4.17
Superior frontal gyrus, R 27 −13 65 3.02
Fusiform gyrus, R 58 39 −67 −16 4.14
Inferior temporal gyrus, R 48 −58 −13 4.13
Fusiform gyrus, R 27 −61 −16 3.14
Precentral gyrus, R 43 54 5 38 4.11
Cerebellum, R 124 12 −73 −40 4.03
Cerebellum, R 33 −58 −40 3.86
Cerebellum, R 3 −58 −43 3.76
Orbital inferior frontal gyrus, L 52 −45 38 −10 3.95
Triangular inferior frontal gyrus, L −39 38 8 3.70
Postcentral gyrus, L 56 −51 −7 38 3.94
Medial prefrontal cortex, L 112 −3 35 50 3.78
Medial prefrontal cortex, L −15 35 26 3.63
Medial prefrontal cortex, L −6 32 35 3.60
Middle temporal gyrus, L 73 −63 −22 −10 3.68
Hippocampus, L −42 −31 −7 3.41
Middle temporal gyrus, L −54 −34 −7 3.33
Superior frontal gyrus, R 23 21 56 11 3.64
Middle frontal gyrus, L 22 −33 11 50 3.53
Superior frontal gyrus, R 27 −13 65 3.02
Hippocampus, R 20 33 −13 −16 3.52
Medial prefrontal cortex, L 10 −12 65 8 3.42
Cerebellum, L 11 −33 −64 −43 3.38
Fusiform gyrus, L 10 −30 −58 −19 3.29
Precuneus, R 18 3 −64 35 3.22
Precuneus, L −3 −55 41 3.04
Cerebellum, L 17 −12 −73 −40 3.20
Posterior cingulate cortex, R 14 6 −43 5 3.17
Parahippocampal gyrus, R 12 −37 2 2.95
Cerebellum, R 10 18 −37 −22 3.10
Anterior cingulate cortex, R 13 15 38 −1 3.08
Supplemental motor area, L 10 0 5 44 3.07
LSAS: Sad faces
Positive correlation
Globus pallidus, L 95 −21 −4 −7 5.93
Putamen, L −18 8 −1 4.79
Superior temporal gyrus, L 11 −48 2 −4 4.90
Open in a new tab

The results are reported threshold at p < .05 FDR‐corrected on the voxel level, only cluster with k ≥10 on whole‐brain level are reported. Coordinates of peak voxels (x/y/z) are given in Montreal Neurological Institute space. Abbreviations: LSAS = Liebowitz Social Anxiety Scale; BDI = Beck Depression Inventory; STAI = State‐Trait Anxiety Inventory.

Figure 2.

Figure 2

Open in a new tab

BOLD level analysis showed positive associations between (a) BDI and neural responses to fearful faces and (b) LSAS and neural responses to sad faces, with gender and trait anxiety as covariates. (c) Both BDI and LSAS scores were associated with responses in the left globus pallidus (GP, −21/−7/−7) and left putamen (PUT, −18/8/−1). All voxels are significant at p < .05 (FDR corrected) and cluster size k ≥10 voxels. Coordinates (x/y/z) are given in standard Montreal Neurological Institute space. Abbreviations: LSAS = Liebowitz Social Anxiety Scale; BDI = Beck Depression Inventory; L = left; R = right [Color figure can be viewed at http://wileyonlinelibrary.com]

Figure 3.

Figure 3

Open in a new tab

BOLD level analysis showed specific positive associations between BDI scores and neural responses to fearful faces in (a) left inferior parietal lobule (IPL, −33/−55/44), (b) bilateral anterior cingulate cortex (ACC; right, 15/38/−1; left, −9/35/5), (c) bilateral inferior frontal gyrus (IFG; right, 54/29/5; left, −45/38/−10), (d) left medial prefrontal cortex (mPFC, −3/35/50), (e) right hippocampus (HIPP, 33/−13/−16), and (f) bilateral fusiform gyrus (FFG; right, 39/−67/−16; left, −30/−58/−19). All voxels are significant at p < .05 (FDR corrected) and cluster size k ≥10 voxels. Coordinates (x/y/z) are given in standard Montreal Neurological Institute space. Abbreviations: BDI = Beck Depression Inventory; L = left; R = right [Color figure can be viewed at http://wileyonlinelibrary.com]

Examining the overlap of the neural activity maps with significant associations for the symptom dimensions revealed that both were positively associated with increased reactivity in two clusters: the left globus pallidus (k = 10, t 87 = 4.26, x/y/z: −21/−7/−7) and left putamen (k = 27, t 87 = 4.26, x/y/z: −18/8/−1) (Figure 2c and 4a).

Figure 4.

Figure 4

Open in a new tab

(a) 6‐mm‐radius spheres centered at the peak MNI coordinates (putamen, PUT; globus pallidus, GP) involved in both depression and social anxiety symptoms were used to explore functional connectivity. (b) The functional connectivity of left PUT–right superior temporal gyrus (60/−4/−10) and left PUT–right amygdala (33/−1/−16) to fearful faces were negatively correlated with BDI, (c) whereas that of left PUT–right middle cingulate cortex (12/−10/47) to sad faces was positively associated with LSAS. All voxels are significant at p < .001 (uncorrected) and cluster size k ≥10 voxels. Coordinates (x/y/z) are given in standard Montreal Neurological Institute space. Abbreviations: LSAS = Liebowitz Social Anxiety Scale; BDI = Beck Depression Inventory; L = left; R = right [Color figure can be viewed at http://wileyonlinelibrary.com]

3.6. Emotional face processing task: Functional connectivity

The exploratory analysis of associations between the symptom levels and functional connectivity revealed negative associations between BDI‐scores and functional connectivity strengths between left putamen and right superior temporal gyrus (k = 23, p = .0004, t 87 = 3.49, x/y/z: 60/−4/−10), and right amygdala (k = 32, t 87 = 2.94, p = .0007, x/y/z: 33/−1/−16; including right insula, t 87 = 1.89, p = .0009, x/y/z: 33/14/−16) during processing of fearful faces (Figure 4b). On the other hand, positive associations were observed between LSAS scores and connectivity of the left putamen and right middle cingulate cortex (k = 32, p = .0001, t 87 = 5.07, x/y/z: 12/−10/47) during processing of sad faces (Table 3 and Figure 4c). No significant linear relationships were observed between the BDI or LSAS and functional connectivity of the left globus pallidus.

Table 3.

Significant associations between BDI and LSAS, with gender and STAI‐Trait as nuisance covariates, and functional connectivity with the left putamen during processing negative faces

Coordinates
Regions Cluster k x y z T
BDI: fearful faces
Negative correlation
Superior temporal gyrus, R 23 60 −4 −10 3.49
Amygdala, R 32 33 −1 −16 2.94
Insula, R 33 14 −16 1.89
LSAS: sad faces
Positive correlation
Middle cingulate cortex, R 32 12 −10 47 5.07
Open in a new tab

The results are reported threshold at p < .001 uncorrected on the voxel level, only cluster with k ≥10 on whole‐brain level are reported. Coordinates of peak voxels (x/y/z) are given in Montreal Neurological Institute space. Abbreviations: LSAS = Liebowitz Social Anxiety Scale; BDI = Beck Depression Inventory; STAI = State‐Trait Anxiety Inventory.

4. DISCUSSION

This study employed a dimensional approach in healthy subjects to determine common and distinct neurofunctional markers of depression and social anxiety while controlling for confounders such as medication and generally increased levels of trait anxiety inherent in traditional patient‐based studies. The implementation of this approach demonstrated that both symptom dimensions were positively associated with increased reactivity in left globus pallidus (GP) and putamen (PUT), although during processing of different negative emotional face stimuli. While higher levels of depression symptoms were associated with increased reactivity to fearful faces, higher levels of social anxiety symptoms were associated with increased reactivity to sad faces. Moreover, depressive symptom load was associated with increased reactivity to fearful faces in a network comprising core regions of the fear reactivity and emotion regulation network, including bilateral inferior frontal gyrus, anterior cingulate cortex, fusiform gyrus, right hippocampus, and left inferior parietal lobule and medial prefrontal cortex. Further exploring associations between levels of depression and social anxiety with functional connectivity of the globus pallidus and putamen revealed that higher levels of social anxiety were associated with increased left putamen—right middle cingulate cortex interplay during processing of sad faces, while higher levels of depression were associated with decreased left putamen—right amygdala functional connectivity during processing of fearful faces. Analysis of linear relationships of the symptom dimensions with neural activity during a facial identity matching task and with grey matter volumes did not reveal significant linear relationships, arguing against strong confounding effects of individual variations in face perception or brain structure on the present findings.

Only few previous studies applied a dimensional approach to examine the associations between individual variations in levels of depression and social anxiety with neurofunctional markers. In line with the present observation of increased reactivity to threatening fearful faces and levels of depression, a previous study which also used a dimensional approach found that levels of depressive symptoms were positively associated with increased stress‐reactivity during a public speaking task (Benson, Arck, Blois, Schedlowski, & Elsenbruch, 2011). Moreover, a study examining associations between anxiety, depression and fear learning, reported that high‐anxiety participants demonstrated less discriminative skin conductance responses during acquisition of fear, whereas high‐depression was associated with higher skin conductance reactivity during extinction of fear (Dibbets, van den Broek, & Evers, 2015). Together with the present data, the previous dimensional approach studies might indicate an association between levels of depression and increased reactivity to threatening stimuli, and additionally suggest that dimensional approaches might be an important complementary research strategy in addition to traditional patient‐based research strategies.

In the present sample of healthy subjects, increased GP and PUT reactivity towards negative faces was found to be positively associated with both depression and social anxiety symptoms. This is consistent with previous meta‐analyses on depression and social anxiety disorders showing increased reactivity in these regions during emotional face processing (Binelli et al., 2014; Delvecchio et al., 2012; Gee et al., 2013; Kanske and Kotz, 2011a, 2011b; MacQueen, 2012; Wu et al., 2016). As core regions of the dorsal striatum and the basal ganglia, both GP and PUT, are not only involved in motor functions, but also play important roles in other domains, such as emotional regulation (Frank et al., 2014; Sztainberg, Kuperman, Justice, & Chen, 2011), salience processing (Gentili et al., 2016; Menon, 2015; Smith, Berridge, & Aldridge, 2011) and social‐cognitive functions (MacQueen, 2012). Previous clinical studies demonstrated that both depression and social anxiety patients demonstrate marked impairments in emotion regulation (Jazaieri, Morrison, Goldin, & Gross, 2015; Joormann, 2010; Joormann and Stanton, 2016), aberrant salience processing (Hamilton et al., 2016; Pannekoek et al., 2013; Yuen et al., 2014), and deficient social cognition (Lavoie, Battaglia, & Achim, 2014; Weightman, Air, & Baune, 2014). Together with the present data from a dimensional approach, these findings suggest an important role of GP and PUT hyper‐responsivity to negative social emotional stimuli as a core neuropathological feature in both depression and social anxiety disorders.

In addition, the associations between the two dimensions were found for neural reactivity during negative emotional face processing, with levels of depression being associated with higher neural reactivity in regions involved in emotion regulation and emotion reactivity during processing of fearful faces. Interestingly, and in contrast with our initial hypothesis and a number of previous studies reporting decreased neural activity in frontal and anterior cingulate regions in depressive patients (Delvecchio et al., 2012; Fitzgerald et al., 2008; Groenewold et al., 2013; Lai, 2014; Palmer et al., 2015), the present dimensional approach revealed an association between higher reactivity to fearful faces in these regions, including the anterior cingulate and the inferior frontal cortex, and levels of depression. However, one study has reported increased left amygdala responses to masked fearful faces compared to other emotional faces in MDD patients, which were normalized with anti‐depressant treatment (Sheline et al., 2001) and another has similarly reported increased medial prefrontal activity to fearful faces (Homan, Drevets, & Hasler, 2014). Furthermore, another study on individuals at risk of developing depression has reported increased amygdala and nucleus accumbens responses to fearful compared to happy faces (Monk et al., 2008). Inconsistent findings might be possibly explained in terms of frontal compensation for increased emotional reactivity in depression, as previously reported in a study comparing anxiety and depression patients (Etkin and Schatzberg, 2011).

The only region demonstrating positive associations with social anxiety symptoms was the superior temporal gyrus (STG). Increased reactivity of this region is in line with a recent meta‐analysis in SAD reporting increased STG reactivity during face processing (Gentili et al., 2016), and original studies in patients with SAD (Binelli et al., 2014; Marazziti et al., 2014). Findings on increased STG activity in depression are less consistent with some meta‐analyses reporting no consistent evidence for altered STG activity in MDD during emotion recognition (Dalili, Penton‐Voak, Harmer, & Munafò, 2015) or emotional face processing (Delvecchio et al., 2012), whereas others have reported both decreased STG activity during positive (Groenewold et al., 2013) and negative stimuli (Fitzgerald et al., 2008), and hyperactivity in STG to negative versus neutral valence in depressive patients (Miller, Hamilton, Sacchet, & Gotlib, 2015).

An additional exploratory analysis examined differential connectivity patterns of the dorsal striatal regions showing overlapping increased neural reactivity in the context of both, increased levels of depression and social anxiety. This approach revealed that higher levels of depression were associated with reduced coupling of the left putamen with the contralateral STG, amygdala and insula whereas higher levels of social anxiety were specifically associated with increased coupling of the left putamen with the contralateral middle cingulate cortex. Given the exploratory nature of the present connectivity findings the corresponding results need to be interpreted cautiously. The negative associations between levels of depression and reduced putamen connectivity is in accordance with previous studies reporting decreased putamen–insula interaction during rest in MDD patients (Guo et al., 2015), and decreased functional coupling between the putamen, amygdala, insula, and STG in MDD patients before treatment, which normalized during the course of successful antidepressive treatment (Chen et al., 2008). Together with the putamen, the amygdala, insula and STG are core nodes of the emotional face processing networks (Fusar‐Poli et al., 2009), with successful encoding of emotional faces being dependent on the integration of information across nodes primarily involved in visual processing, including the STG (Stuhrmann et al., 2011), as well as limbic and striatal emotion processing regions, including the striatum and the amygdala (Haxby, Hoffman, & Gobbini, 2000). Overall, this suggests that the higher reactivity in emotional processing and emotion regulation nodes during presentation of fearful faces might interfere with the integrative processing of these faces on the network level. The specific importance of disruptions in the integrative fronto‐striatal‐limbic interplay in MDD is further emphasized by recent studies reporting that the intrinsic connectivity of this circuitry appears to reliably discriminate depression sub‐types (Drysdale et al., 2017; Wager and Woo, 2017). With regard to social anxiety only the pathway between the putamen and middle cingulate cortex demonstrated increased connectivity. Both regions share functional connections (Hoffstaedter et al., 2012), thought to underlie the integration of emotion and motor‐responses (Shackman et al., 2011), with stronger connectivity in this pathway during negative face possibly reflecting increased tendency to avoid or withdrawal negative social signals which constitutes a core feature of social anxiety disorders (Mkrtchian, Aylward, Dayan, Roiser, & Robinson, 2017).

In contrast to levels of depression where neural effects were associated with fearful faces those for social anxiety were associated with sad ones. Most previous patient‐based studies showed an opposite pattern, suggesting that MDD patients are more sensitive to sad faces (Arnone et al., 2012; Surguladze et al., 2005) while SAD patients are more sensitive to fearful faces (Frick, Howner, Fischer, Kristiansson, & Furmark, 2013; Killgore and Yurgelun‐Todd, 2005). However, as discussed above a number of studies on MDD patients or individuals at risk of MDD have also reported enhanced neural responses to fearful faces. Similarly, other studies have reported increased sensitivity to sad as well as fearful faces (Arrais et al., 2010) and hyperactivity to sad faces in prefrontal regions (Labuschagne et al., 2012) in SAD patients.

There are some limitations in this study. First, the BDI values were not normally distributed in our subject cohort, although nonparametric analyses were used to account for the non‐normal distribution. Second, the implementation of the nonparametric approach limited the flexibility of the second level regression models leading to separate regression models for the BDI and LSAS as well as for each emotion, limiting the statistical strengths and the interpretation of emotion‐specific associations. Similarly, the limitations did not allow direct implementation of a conjunction model to determine overlapping regions between the symptom scales while controlling for multiple comparisons on the voxel level. The computed overlap thus refers to a regional overlay of the activity clusters and needs to be interpreted with caution. Third, in line with previous studies in healthy subjects that examined associations between levels of depression with behavioral and neural indices (Benson et al., 2011; Dibbets et al., 2015; Levita et al., 2014), the current study employed the BDI‐II to assess levels of depression. However, the BDI‐II has been developed in a clinical context and might have a limited sensitivity to determine individual variations in a healthy sample. Fourth, participants in this study showed moderate levels of depression and social anxiety symptoms, and thus the neural markers might additionally represent compensatory mechanisms that prevent the transition to a clinically relevant disorder.

5. CONCLUSION

The current findings add to the growing literature on the neural reactivity toward emotional face processing associated with depression and social anxiety, and additionally suggest a key role of altered putamen reactivity as core pathological substrate across both disorders, whereas connectivity of this region might reflect specific alterations.

FUNDING

This work was supported by the National Natural Science Foundation of China (NSFC) grant (grant numbers: 31530032; 91632117), the Fundamental Research Funds for the Central Universities (ZYGX2015Z002), and Fundamental Research Funds from Science & Technology Department of Sichuan Province of China (2017JY0031, JD).

CONFLICT OF INTEREST

The authors declare no financial interests or potential conflict of interest.

AUTHOR CONTRIBUTIONS

LL, BB, and KMK designed the experiments. LL and XZ prepared all the materials and programed the procedures. LL, XZ, ZZ, XX, JK, and JD acquired the data. LL, BB, XZ, FZ, JW, and KMK analyzed the data. LL, BB, and KMK interpreted the data and drafted the manuscript. All the authors revised it critically for intellectual content, especially LL, BB, and KMK. All the authors gave final approval to the version to be published, and agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Supporting information

Additional Supporting Information may be found online in the supporting information tab for this article.

Supporting Information

Click here for additional data file. (29.7KB, docx)

ACKNOWLEDGMENTS

Lizhu Luo and Benjamin Becker contributed equally to this work.

Luo L, Becker B, Zheng X, et al. A dimensional approach to determine common and specific neurofunctional markers for depression and social anxiety during emotional face processing. Hum Brain Mapp. 2018;39:758–771. 10.1002/hbm.23880

Funding information National Natural Science Foundation of China (NSFC), Grant/Award Numbers: 31530032, 91632117; Fundamental Research Funds for the Central Universities, Grant/Award Number: ZYGX2015Z002; Fundamental Research Funds from Science & Technology Department of Sichuan Province of China, Grant/Award Number: 2017JY0031, JD

Work performed at: The Clinical Hospital of Chengdu Brain Science Institute, MOE Key Laboratory for Neuroinformation, University of Electronic Science and Technology of China, Chengdu, China

REFERENCES

  1. Adams, G. C. , Balbuena, L. , Meng, X. , & Asmundson, G. J. (2016). When social anxiety and depression go together: A population study of comorbidity and associated consequences. Journal of Affective Disorders, 206, 48–54. [DOI] [PubMed] [Google Scholar]
  2. Anagnostou, E. , Soorya, L. , Chaplin, W. , Bartz, J. , Halpern, D. , Wasserman, S. , … Kushki, A. (2012). Intranasal oxytocin versus placebo in the treatment of adults with autism spectrum disorders: A randomized controlled trial. Molecular Autism, 3, 16. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Andrade, A. , Paradis, A.‐L. , Rouquette, S. , & Poline, J.‐B. (1999). Ambiguous results in functional neuroimaging data analysis due to covariate correlation. NeuroImage, 10, 483–486. [DOI] [PubMed] [Google Scholar]
  4. Arnone, D. , McKie, S. , Elliott, R. , Thomas, E. J. , Downey, D. , Juhasz, G. , … Anderson, I. M. (2012). Increased amygdala responses to sad but not fearful faces in major depression: Relation to mood state and pharmacological treatment. American Journal of Psychiatry, 169, 841–850. [DOI] [PubMed] [Google Scholar]
  5. Arrais, K. C. , Machado‐de‐Sousa, J. P. , Trzesniak, C. , Santos Filho, A. , Ferrari, M. C. F. , Osório, F. L. , … Zuardi, A. W. (2010). Social anxiety disorder women easily recognize fearfull, sad and happy faces: The influence of gender. Journal of Psychiatric Research, 44, 535–540. [DOI] [PubMed] [Google Scholar]
  6. Beck, A. T. , Steer, R. A. , Ball, R. , & Ranieri, W. F. (1996). Comparison of Beck Depression Inventories‐IA and‐II in psychiatric outpatients. Journal of Personality Assessment, 67, 588–597. [DOI] [PubMed] [Google Scholar]
  7. Benson, S. , Arck, P. , Blois, S. , Schedlowski, M. , & Elsenbruch, S. (2011). Subclinical depressive symptoms affect responses to acute psychosocial stress in healthy premenopausal women. Stress (Amsterdam, Netherlands), 14, 88–92. [DOI] [PubMed] [Google Scholar]
  8. Bienvenu, O. J. , Hettema, J. M. , Neale, M. C. , Prescott, C. A. , & Kendler, K. S. (2007). Low extraversion and high neuroticism as indices of genetic and environmental risk for social phobia, agoraphobia, and animal phobia. The American Journal of Psychiatry, 164, 1714–1721. [DOI] [PubMed] [Google Scholar]
  9. Binelli, C. , Subirà, S. , Batalla, A. , Muñiz, A. , Sugranyés, G. , Crippa, J. , … Martín‐Santos, R. (2014). Common and distinct neural correlates of facial emotion processing in social anxiety disorder and Williams syndrome: A systematic review and voxel‐based meta‐analysis of functional resonance imaging studies. Neuropsychologia, 64, 205–217. [DOI] [PubMed] [Google Scholar]
  10. Bourke, C. , Douglas, K. , & Porter, R. (2010). Processing of facial emotion expression in major depression: A review. The Australian and New Zealand Journal of Psychiatry, 44, 681–696. [DOI] [PubMed] [Google Scholar]
  11. Brett, M. , Anton, J.‐L. , Valabregue, R. , & Poline, J.‐B. (2002). Region of interest analysis using the MarsBar toolbox for SPM 99. NeuroImage, 16, S497. [Google Scholar]
  12. Chau, S. A. , Herrmann, N. , Sherman, C. , Chung, J. , Eizenman, M. , Kiss, A. , & Lanctôt, K. L. (2017). Visual selective attention toward novel stimuli predicts cognitive decline in Alzheimer's disease patients. Journal of Alzheimer's Disease: JAD, 55, 1339–1349. [DOI] [PubMed] [Google Scholar]
  13. Chavira, D. A. , Stein, M. B. , Bailey, K. , & Stein, M. T. (2004). Comorbidity of generalized social anxiety disorder and depression in a pediatric primary care sample. Journal of Affective Disorders, 80, 163–171. [DOI] [PubMed] [Google Scholar]
  14. Chen, C.‐H. , Suckling, J. , Ooi, C. , Fu, C. H. , Williams, S. C. , Walsh, N. D. , … Bullmore, E. (2008). Functional coupling of the amygdala in depressed patients treated with antidepressant medication. Neuropsychopharmacology: Official Publication of the American College of Neuropsychopharmacology, 33, 1909–1918. [DOI] [PubMed] [Google Scholar]
  15. Chen, L. F. , & Yen, Y. S. (2007). Taiwanese Facial Expression Image Database [http://bml.ym.edu.tw/~download/html]. Brain Mapping Laboratory, Institute of Brain Science, National Yang‐Ming University, Taipei, Taiwan.
  16. Cuthbert, B. N. (2014). The RDoC framework: Facilitating transition from ICD/DSM to dimensional approaches that integrate neuroscience and psychopathology. World Psychiatry: Official Journal of the World Psychiatric Association (WPA), 13, 28–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Dalili, M. N. , Penton‐Voak, I. S. , Harmer, C. J. , & Munafò, M. R. (2015). Meta‐analysis of emotion recognition deficits in major depressive disorder. Psychological Medicine, 45, 1135–1144. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Delvecchio, G. , Fossati, P. , Boyer, P. , Brambilla, P. , Falkai, P. , Gruber, O. , … McIntosh, A. M. (2012). Common and distinct neural correlates of emotional processing in bipolar disorder and major depressive disorder: A voxel‐based meta‐analysis of functional magnetic resonance imaging studies. European Neuropsychopharmacology: The Journal of the European College of Neuropsychopharmacology, 22, 100–113. [DOI] [PubMed] [Google Scholar]
  19. Demenescu, L. R. , Kortekaas, R. , den Boer, J. A. , & Aleman, A. (2010). Impaired attribution of emotion to facial expressions in anxiety and major depression. PLoS One, 5, e15058. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Dibbets, P. , van den Broek, A. , & Evers, E. A. (2015). Fear conditioning and extinction in anxiety‐and depression‐prone persons. Memory (Hove, England), 23, 350–364. [DOI] [PubMed] [Google Scholar]
  21. Drysdale, A. T. , Grosenick, L. , Downar, J. , Dunlop, K. , Mansouri, F. , Meng, Y. , … Etkin, A. (2017). Erratum: Resting‐state connectivity biomarkers define neurophysiological subtypes of depression. Nature Medicine, 23, 264. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Etkin, A. , & Schatzberg, A. F. (2011). Common abnormalities and disorder‐specific compensation during implicit regulation of emotional processing in generalized anxiety and major depressive disorders. The American Journal of Psychiatry, 168, 968–978. [DOI] [PubMed] [Google Scholar]
  23. Fitzgerald, P. B. , Laird, A. R. , Maller, J. , & Daskalakis, Z. J. (2008). A meta‐analytic study of changes in brain activation in depression. Human Brain Mapping, 29, 683–695. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Frank, D. , Dewitt, M. , Hudgens‐Haney, M. , Schaeffer, D. , Ball, B. , Schwarz, N. , … Sabatinelli, D. (2014). Emotion regulation: Quantitative meta‐analysis of functional activation and deactivation. Neuroscience and Biobehavioral Reviews, 45, 202–211. [DOI] [PubMed] [Google Scholar]
  25. Frick, A. , Howner, K. , Fischer, H. , Kristiansson, M. , & Furmark, T. (2013). Altered fusiform connectivity during processing of fearful faces in social anxiety disorder. Translational Psychiatry, 3, e312. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Fusar‐Poli, P. , Placentino, A. , Carletti, F. , Landi, P. , Allen, P. , Surguladze, S. , … Barale, F. (2009). Functional atlas of emotional faces processing: A voxel‐based meta‐analysis of 105 functional magnetic resonance imaging studies. Journal of Psychiatry &Amp; Neuroscience: Japan, 34, 418. [PMC free article] [PubMed] [Google Scholar]
  27. Gao, S. , Beckera, B. , Luo, L. , Geng, Y. , Zhao, W. , Yin, Y. , … Kendrick, K. M. (2016). Oxytocin, the peptide that bonds the sexes also divides them. Proceedings of the National Academy of Sciences of the United States of America, 113, 7650–7654. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Gee, D. G. , Humphreys, K. L. , Flannery, J. , Goff, B. , Telzer, E. H. , Shapiro, M. , … Tottenham, N. (2013). A developmental shift from positive to negative connectivity in human amygdala–prefrontal circuitry. The Journal of Neuroscience: The Official Journal of the Society for Neuroscience, 33, 4584–4593. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Gentili, C. , Cristea, I. A. , Angstadt, M. , Klumpp, H. , Tozzi, L. , Phan, K. L. , & Pietrini, P. (2016). Beyond emotions: A meta‐analysis of neural response within face processing system in social anxiety. Experimental Biology and Medicine (Maywood, N.J.), 241, 225–237. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Gilboa‐Schechtman, E. , Foa, E. , Vaknin, Y. , Marom, S. , & Hermesh, H. (2008). Interpersonal sensitivity and response bias in social phobia and depression: Labeling emotional expressions. Cognitive Therapy and Research, 32, 605–618. [Google Scholar]
  31. Gilboa‐Schechtman, E. , Foa, E. B. , & Amir, N. (1999). Attentional biases for facial expressions in social phobia: The face‐in‐the‐crowd paradigm. Cognition and Emotion, 13, 305–318. [Google Scholar]
  32. Gong, X. , Huang, Y. X. , Wang, Y. , & Luo, Y. J. (2011). Revision of the Chinese facial affective picture system. Chinese Mental Health Journal, 25, 40–46. [Google Scholar]
  33. Groenewold, N. A. , Opmeer, E. M. , de Jonge, P. , Aleman, A. , & Costafreda, S. G. (2013). Emotional valence modulates brain functional abnormalities in depression: Evidence from a meta‐analysis of fMRI studies. Neuroscience and Biobehavioral Reviews, 37, 152–163. [DOI] [PubMed] [Google Scholar]
  34. Guo, W. B. , Liu, F. , Xiao, C. Q. , Zhang, Z. K. , Liu, J. R. , Yu, M. Y. , … Zhao, J. P. (2015). Decreased insular connectivity in drug‐naive major depressive disorder at rest. Journal of Affective Disorders, 179, 31–37. [DOI] [PubMed] [Google Scholar]
  35. Hamilton, J. P. , Glover, G. H. , Bagarinao, E. , Chang, C. , Mackey, S. , Sacchet, M. D. , & Gotlib, I. H. (2016). Effects of salience‐network‐node neurofeedback training on affective biases in major depressive disorder. Psychiatry Research, 249, 91–96. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Hattingh, C. J. , Ipser, J. , Tromp, S. , Syal, S. , Lochner, C. , Brooks, S. J. B. , & Stein, D. J. (2013). Functional magnetic resonance imaging during emotion recognition in social anxiety disorder: An activation likelihood meta‐analysis. Frontiers in Human Neuroscience, 6, 347. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Haxby, J. V. , Hoffman, E. A. , & Gobbini, M. I. (2000). The distributed human neural system for face perception. Trends in Cognitive Sciences, 4, 223–233. [DOI] [PubMed] [Google Scholar]
  38. He, Y. L. , & Zhang, M. Y. (2004). Psychometric investigation of Liebowitz Social Anxiety Scale. Journal of Diagnostics Concepts and Practice, 3, 89–93. [Google Scholar]
  39. Hoffstaedter, F. , Grefkes, C. , Caspers, S. , Roski, C. , Fox, P. , Zilles, K. , & Eickhoff, S. (2012). Functional connectivity of the mid‐cingulate cortex. Klinische Neurophysiologie, 43, P128. [Google Scholar]
  40. Homan, P. , Drevets, W. C. , & Hasler, G. (2014). The effects of catecholamine depletion on the neural response to fearful faces in remitted depression. The International Journal of Neuropsychopharmacology, 17, 1419–1428. [DOI] [PubMed] [Google Scholar]
  41. Insel, T. , Cuthbert, B. , Garvey, M. , Heinssen, R. , Pine, D. S. , Quinn, K. , … Wang, P. (2010). Research domain criteria (RDoC): Toward a new classification framework for research on mental disorders. American Journal of Psychiatry, 167, 748–751. [DOI] [PubMed] [Google Scholar]
  42. Jazaieri, H. , Morrison, A. S. , Goldin, P. R. , & Gross, J. J. (2015). The role of emotion and emotion regulation in social anxiety disorder. Current Psychiatry Reports, 17, 1–9. [DOI] [PubMed] [Google Scholar]
  43. Joormann, J. (2010). Cognitive inhibition and emotion regulation in depression. Current Directions in Psychological Science, 19, 161–166. [Google Scholar]
  44. Joormann, J. , & Gotlib, I. H. (2006). Is this happiness I see? Biases in the identification of emotional facial expressions in depression and social phobia. Journal of Abnormal Psychology, 115, 705. [DOI] [PubMed] [Google Scholar]
  45. Joormann, J. , & Stanton, C. H. (2016). Examining emotion regulation in depression: A review and future directions. Behaviour Research and Therapy, 86, 35–49. [DOI] [PubMed] [Google Scholar]
  46. Kanske, P. , & Kotz, S. A. (2011a). Conflict processing is modulated by positive emotion: ERP data from a flanker task. Behavioural Brain Research, 219, 382–386. [DOI] [PubMed] [Google Scholar]
  47. Kanske, P. , & Kotz, S. A. (2011b). Positive emotion speeds up conflict processing: ERP responses in an auditory Simon task. Biological Psychology, 87, 122–127. [DOI] [PubMed] [Google Scholar]
  48. Kendler, K. S. , Aggen, S. H. , Knudsen, G. P. , Røysamb, E. , Neale, M. C. , & Reichborn‐Kjennerud, T. (2011). The structure of genetic and environmental risk factors for syndromal and subsyndromal common DSM‐IV axis I and all axis II disorders. The American Journal of Psychiatry, 168, 29–39. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Kessler, R. C. , Berglund, P. , Demler, O. , Jin, R. , Koretz, D. , Merikangas, K. R. , … Wang, P. S. (2003). The epidemiology of major depressive disorder: Results from the National Comorbidity Survey Replication (NCS‐R). JAMA, 289, 3095–3105. [DOI] [PubMed] [Google Scholar]
  50. Kessler, R. C. , Berglund, P. , Demler, O. , Jin, R. , Merikangas, K. R. , & Walters, E. E. (2005). Lifetime prevalence and age‐of‐onset distributions of DSM‐IV disorders in the National Comorbidity Survey Replication. Archives of General Psychiatry, 62, 593–602. [DOI] [PubMed] [Google Scholar]
  51. Kessler, R. C. , Stang, P. , Wittchen, H.‐U. , Stein, M. , & Walters, E. E. (1999). Lifetime co‐morbidities between social phobia and mood disorders in the US National Comorbidity Survey. Psychological Medicine, 29, 555–567. [DOI] [PubMed] [Google Scholar]
  52. Killgore, W. D. , & Yurgelun‐Todd, D. A. (2005). Social anxiety predicts amygdala activation in adolescents viewing fearful faces. NeuroReport, 16, 1671–1675. [DOI] [PubMed] [Google Scholar]
  53. Labuschagne, I. , Phan, K. L. , Wood, A. , Angstadt, M. , Chua, P. , Heinrichs, M. , … Nathan, P. J. (2012). Medial frontal hyperactivity to sad faces in generalized social anxiety disorder and modulation by oxytocin. The International Journal of Neuropsychopharmacology, 15, 883–896. [DOI] [PubMed] [Google Scholar]
  54. Ladouceur, C. D. , Dahl, R. E. , Williamson, D. E. , Birmaher, B. , Ryan, N. D. , & Casey, B. (2005). Altered emotional processing in pediatric anxiety, depression, and comorbid anxiety‐depression. Journal of Abnormal Child Psychology, 33, 165–177. [DOI] [PubMed] [Google Scholar]
  55. Lai, C.‐H. (2014). Patterns of cortico‐limbic activations during visual processing of sad faces in depression patients: A coordinate‐based meta‐analysis. The Journal of Neuropsychiatry and Clinical Neurosciences, 26, 34–43. [DOI] [PubMed] [Google Scholar]
  56. Langer, J. K. , & Rodebaugh, T. L. (2014). Comorbidity of social anxiety disorder and depression In Richards C.S., O'Hara M.W. (Ed.), The Oxford handbook of depression and comorbidity (p 111). Oxford University Press. [Google Scholar]
  57. Lavoie, M.‐A. , Battaglia, M. , & Achim, A. M. (2014). A meta‐analysis and scoping review of social cognition performance in social phobia, posttraumatic stress disorder and other anxiety disorders. Journal of Anxiety Disorders, 28, 169–177. [DOI] [PubMed] [Google Scholar]
  58. Levita, L. , Bois, C. , Healey, A. , Smyllie, E. , Papakonstantinou, E. , Hartley, T. , & Lever, C. (2014). The Behavioural Inhibition System, anxiety and hippocampal volume in a non‐clinical population. Biology of Mood & Anxiety Disorders, 4, 4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Li, W. L. , & Qian, M. Y. (1995). Revision of the State‐trait Anxiety Inventory with sample of Chinese college students. ACTA Scientiarum Naturalium Universitatis Pekinensis, 31, 108–114. [Google Scholar]
  60. Liebowitz, M. R. (1987). Social phobia. Karger Publishers. [Google Scholar]
  61. MacQueen, G. M. (2012). Systematic review of the neural basis of social cognition in patients with mood disorders. Journal of Psychiatry &Amp; Neuroscience: Japan, 37, 154. [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Marazziti, D. , Abelli, M. , Baroni, S. , Carpita, B. , Piccinni, A. , & Dell'osso, L. (2014). Recent findings on the pathophysiology of social anxiety disorder. CNS Spectrums, 11, 91–100. [DOI] [PubMed] [Google Scholar]
  63. McLaren, D. G. , Ries, M. L. , Xu, G. , & Johnson, S. C. (2012). A generalized form of context‐dependent psychophysiological interactions (gPPI): A comparison to standard approaches. NeuroImage, 61, 1277–1286. [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Menon, V. (2015). Salience network. Brain Mapping: An Encyclopedic Reference, 2, 597–611. [Google Scholar]
  65. Miller, C. H. , Hamilton, J. P. , Sacchet, M. D. , & Gotlib, I. H. (2015). Meta‐analysis of functional neuroimaging of major depressive disorder in youth. JAMA Psychiatry, 72, 1045–1053. [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Mkrtchian, A. , Aylward, J. , Dayan, P. , Roiser, J. P. , & Robinson, O. J. (2017). Modelling avoidance in mood and anxiety disorders using reinforcement‐learning. Biological Psychiatry. [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Monk, C. S. , Klein, R. G. , Telzer, E. H. , Schroth, E. A. , Mannuzza, S. , Moulton, III P. D. , J. L., … Fromm, S. (2008). Amygdala and nucleus accumbens activation to emotional facial expressions in children and adolescents at risk for major depression. American Journal of Psychiatry, 165, 90–98. [DOI] [PubMed] [Google Scholar]
  68. Mumford, J. A. , Poline, J.‐B. , & Poldrack, R. A. (2015). Orthogonalization of regressors in fMRI models. PLoS One, 10, e0126255. [DOI] [PMC free article] [PubMed] [Google Scholar]
  69. O'Brien, R. M. (2007). A caution regarding rules of thumb for variance inflation factors. Quality & Quantity, 41, 673–690. [Google Scholar]
  70. Ohashi, K. , Anderson, C. M. , Bolger, E. A. , Khan, A. , McGreenery, C. E. , & Teicher, M. H. (2017). Childhood maltreatment is associated with alteration in global network fiber‐tract architecture independent of history of depression and anxiety. NeuroImage, 150, 50–59. [DOI] [PMC free article] [PubMed] [Google Scholar]
  71. Ohayon, M. M. , & Schatzberg, A. F. (2010). Social phobia and depression: Prevalence and comorbidity. Journal of Psychosomatic Research, 68, 235–243. [DOI] [PubMed] [Google Scholar]
  72. Palmer, S. M. , Crewther, S. G. , & Carey, L. M. (2015). A meta‐analysis of changes in brain activity in clinical depression. Frontiers in Human Neuroscience, 8, 1045. [DOI] [PMC free article] [PubMed] [Google Scholar]
  73. Pannekoek, J. N. , Veer, I. M. , van Tol, M.‐J. , van der Werff, S. J. , Demenescu, L. R. , Aleman, A. , … van der Wee, N. J. (2013). Resting‐state functional connectivity abnormalities in limbic and salience networks in social anxiety disorder without comorbidity. European Neuropsychopharmacology: The Journal of the European College of Neuropsychopharmacology, 23, 186–195. [DOI] [PubMed] [Google Scholar]
  74. Pishyar, R. , Harris, L. M. , & Menzies, R. G. (2004). Attentional bias for words and faces in social anxiety. Anxiety, Stress & Coping, 17, 23–36. [Google Scholar]
  75. Sanislow, C. A. , Pine, D. S. , Quinn, K. J. , Kozak, M. J. , Garvey, M. A. , Heinssen, R. K. , … Cuthbert, B. N. (2010). Developing constructs for psychopathology research: Research domain criteria. Journal of Abnormal Psychology, 119, 631. [DOI] [PubMed] [Google Scholar]
  76. Shackman, A. J. , Salomons, T. V. , Slagter, H. A. , Fox, A. S. , Winter, J. J. , & Davidson, R. J. (2011). The integration of negative affect, pain and cognitive control in the cingulate cortex. Nature Reviews. Neuroscience, 12, 154–167. [DOI] [PMC free article] [PubMed] [Google Scholar]
  77. Sheline, Y. I. , Barch, D. M. , Donnelly, J. M. , Ollinger, J. M. , Snyder, A. Z. , & Mintun, M. A. (2001). Increased amygdala response to masked emotional faces in depressed subjects resolves with antidepressant treatment: An fMRI study. Biological Psychiatry, 50, 651–658. [DOI] [PubMed] [Google Scholar]
  78. Smith, K. S. , Berridge, K. C. , & Aldridge, J. W. (2011). Disentangling pleasure from incentive salience and learning signals in brain reward circuitry. Proceedings of the National Academy of Sciences USA, 108, E255–E264. [DOI] [PMC free article] [PubMed] [Google Scholar]
  79. Speilberger, C. , Gorsuch, R. , Lushene, R. , Vagg, P. , & Jacobs, G. (1983). Manual for the state‐trait anxiety inventory (STAI) (p. 77). California: Mind Garden. [Google Scholar]
  80. Stuhrmann, A. , Suslow, T. , & Dannlowski, U. (2011). Facial emotion processing in major depression: A systematic review of neuroimaging findings. Biology of Mood & Anxiety Disorders, 1(1), [DOI] [PMC free article] [PubMed] [Google Scholar]
  81. Surguladze, S. , Brammer, M. J. , Keedwell, P. , Giampietro, V. , Young, A. W. , Travis, M. J. , … Phillips, M. L. (2005). A differential pattern of neural response toward sad versus happy facial expressions in major depressive disorder. Biological Psychiatry, 57, 201–209. [DOI] [PubMed] [Google Scholar]
  82. Sztainberg, Y. , Kuperman, Y. , Justice, N. , & Chen, A. (2011). An anxiolytic role for CRF receptor type 1 in the globus pallidus. The Journal of Neuroscience: The Official Journal of the Society for Neuroscience, 31, 17416–17424. [DOI] [PMC free article] [PubMed] [Google Scholar]
  83. Tzourio‐Mazoyer, N. , Landeau, B. , Papathanassiou, D. , Crivello, F. , Etard, O. , Delcroix, N. , … Joliot, M. (2002). Automated anatomical labeling of activations in SPM using a macroscopic anatomical parcellation of the MNI MRI single‐subject brain. NeuroImage, 15, 273–289. [DOI] [PubMed] [Google Scholar]
  84. van Tol, M.‐J. , van der Wee, N. J. , van den Heuvel, O. A. , Nielen, M. M. , Demenescu, L. R. , Aleman, A. , … Veltman, D. J. (2010). Regional brain volume in depression and anxiety disorders. Archives of General Psychiatry, 67, 1002–1011. [DOI] [PubMed] [Google Scholar]
  85. Wager, T. D. , & Woo, C.‐W. (2017). Imaging biomarkers and biotypes for depression. Nature Medicine, 23, 16–17. [DOI] [PubMed] [Google Scholar]
  86. Wakschlag, L. S. , Estabrook, R. , Petitclerc, A. , Henry, D. , Burns, J. L. , Perlman, S. B. , … Briggs‐Gowan, M. L. (2015). Clinical implications of a dimensional approach: The normal: Abnormal spectrum of early irritability. Journal of the American Academy of Child and Adolescent Psychiatry, 54, 626–634. [DOI] [PMC free article] [PubMed] [Google Scholar]
  87. Wang, Z. , Yuan, C. M. , Huang, J. , Li, Z. Z. , Chen, J. , Zhang, H. Y. , … Xiao, Z. P. (2011). Reliability and validity of the Chinese version of Beck Depression Inventory‐II among depression patient. Chinese Mental Health Journal, 25, 476–480. [Google Scholar]
  88. Waugh, C. E. , Hamilton, J. P. , Chen, M. C. , Joormann, J. , & Gotlib, I. H. (2012). Neural temporal dynamics of stress in comorbid major depressive disorder and social anxiety disorder. Biology of Mood &Amp; Anxiety Disorders, 2, 11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  89. Weightman, M. J. , Air, T. M. , & Baune, B. T. (2014). A review of the role of social cognition in major depressive disorder. Frontiers in Psychiatry, 5, 179. [DOI] [PMC free article] [PubMed] [Google Scholar]
  90. Wells, A. , & Carter, K. (2002). Further tests of a cognitive model of generalized anxiety disorder: Metacognitions and worry in GAD, panic disorder, social phobia, depression, and nonpatients. Behavior Therapy, 32, 85–102. [Google Scholar]
  91. Winkler, A. M. , Ridgway, G. R. , Webster, M. A. , Smith, S. M. , & Nichols, T. E. (2014). Permutation inference for the general linear model. NeuroImage, 92, 381–397. [DOI] [PMC free article] [PubMed] [Google Scholar]
  92. Wu, M. , Kujawa, A. , Lu, L. H. , Fitzgerald, D. A. , Klumpp, H. , Fitzgerald, K. D. , … Phan, K. L. (2016). Age‐related changes in amygdala–frontal connectivity during emotional face processing from childhood into young adulthood. Human Brain Mapping, 37, 1684–1695. [DOI] [PMC free article] [PubMed] [Google Scholar]
  93. Yuen, G. S. , Gunning‐Dixon, F. M. , Hoptman, M. J. , AbdelMalak, B. , McGovern, A. R. , … Alexopoulos, G. S. (2014). The salience network in the apathy of late‐life depression. International Journal of Geriatric Psychiatry, 29, 1116–1124. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

Additional Supporting Information may be found online in the supporting information tab for this article.

Supporting Information

Click here for additional data file. (29.7KB, docx)

Articles from Human Brain Mapping are provided here courtesy of Wiley

RESOURCES