اتصال عملکردی تغییر یافته مغز در رابطه با ادراک از موشکافی در اختلال اضطراب اجتماعی
کد مقاله | سال انتشار | تعداد صفحات مقاله انگلیسی |
---|---|---|
39194 | 2012 | 10 صفحه PDF |
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Psychiatry Research: Neuroimaging, Volume 202, Issue 3, 30 June 2012, Pages 214–223
چکیده انگلیسی
Abstract Although the fear of being scrutinized by others in a social context is a key symptom in social anxiety disorder (SAD), the neural processes underlying the perception of scrutiny have not previously been studied by functional magnetic resonance imaging (fMRI). We used fMRI to map brain activation during a perception-of-scrutiny task in 20 SAD patients and 20 controls. A multi-dimensional analytic approach was used. Scrutiny perception was mediated by activation of the medial frontal cortex, insula–operculum region and cerebellum, and the additional recruitment of visual areas and the thalamus in patients. Between-group comparison demonstrated significantly enhanced brain activation in patients in the primary visual cortex and cerebellum. Functional connectivity mapping demonstrated an abnormal connectivity between regions underlying general arousal and attention. SAD patients showed significantly greater task-induced functional connectivity in the thalamo-cortical and the fronto-striatal circuits. A statistically significant increase in task-induced functional connectivity between the anterior cingulate cortex and scrutiny-perception-related regions was observed in the SAD patients, suggesting the existence of enhanced behavior-inhibitory control. The presented data indicate that scrutiny perception in SAD enhances brain activity in arousal–attention systems, suggesting that fMRI may be a useful tool to explore such a behavioral dimension.
مقدمه انگلیسی
1. Introduction Social anxiety disorder (SAD) is a prevalent and disabling psychiatric illness (Stein, 2006) characterized by heightened fear and avoidance of social situations, such as speaking in public, meeting people or attending gatherings. Typically, SAD patients demonstrate anxious behavior when faced with the prospect of becoming the focus of attention and being evaluated by other people. Specifically, the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) (American Psychiatric Association, 2000) defines SAD as a “marked and persistent fear of one or more social or performance situations in which the person is exposed to unfamiliar people or to possible scrutiny by others”. However, SAD patients often report a sense of uneasiness and anxiety arising from their concern as to how others will perceive them, even in non-performance situations, and merely on the basis of having their physical appearance judged by other individuals (Leary and Kowalski, 1995). Such concern, dubbed as public self-consciousness (Theron et al., 1991), may be generated in certain situations, such as being in a waiting room with people, where no performance is required and it is merely the person's physical self that is exposed to judgment by others. In this sense, the SAD population reports lower self-physical attractiveness than in the ratings made by observers (Rapee and Abbott, 2006). Other studies in SAD (beyond functional magnetic resonance imaging (fMRI)) involving scrutiny perception situations with no required performance demonstrate high levels of fear and avoidance behavior merely upon eye contact (Schneier et al., 2011) or even in virtual situations where there is no real, direct contact with another person (Wieser et al., 2010). Virtual reality exposure has been widely used in the treatment of social phobia, to reproduce such situations (for a recent review, see de Carvalho et al., 2010). Consistent with the clinical definition, activity in fear-related brain circuits has been assessed by functional imaging during performance or exposure to socially relevant stimuli, such as public speaking (Tillfors et al., 2001), speech anticipatory anxiety (Tillfors et al., 2001 and Lorberbaum et al., 2004), language-based criticism stimuli (Blair et al., 2008a), and responses to facial expressions, as relevant elements of social interaction (Straube et al., 2004a, Amir et al., 2005, Phan et al., 2006, Campbell et al., 2007, Stein et al., 2007, Blair et al., 2008b, Evans et al., 2008 and Pujol et al., 2009a). The research has revealed enhanced neural activity in neuronal systems related to the amygdala, whose responsivity suggests a biased processing of social information possibly related to the disorder (Cooney et al., 2006 and Kober et al., 2008). Recent advances in neuroimaging coincide in their reports of alterations in limbic and paralimbic regions in SAD, as well as in prefrontal and striatal regions, thus highlighting the complexity of the pathology, and the cortico-subcortical pathways still constitute a wide field to be researched within the framework of SAD (for a review, see Freitas-Ferrari et al., 2010). Although static images of negative facial expressions are indeed a relevant social cue to reveal a person's disapproval of others, they capture a partial aspect of the fear-related systems in SAD samples in a social context. Despite some shared key features, the perception of being publicly scrutinized triggers an exaggerated fear when the individual is confronted with an external audience, and constitutes a common day-to-day situation that SAD patients attempt to avoid. In such a context, the goal of the present study was to directly examine brain activity by fMRI under real exposure to scrutiny by other people in SAD. When confronting unfamiliar and/or feared settings, the socially phobic individual demonstrates increased reactivity, self-reporting an enhanced physical arousal (Edelmann and Baker, 2002); however, the essential feature of SAD (the fear of being scrutinized) has not been directly investigated using fMRI. Therefore, we developed a novel fMRI paradigm primarily to examine whether the mere perception of being scrutinized is associated with alterations in the attention and arousal thalamo-related networks in SAD patients. In a very recent study of general emotion processing in SAD, the perceptual-attention domain comprising the thalamus was found to be altered (Brühl et al., 2011). Because standard statistical parametric mapping (SPM) analysis of task-related activation only partially evaluates possible brain fMRI abnormal responses, a whole-brain functional connectivity analysis was also performed. Three specific objectives were considered: (a) to evaluate possible brain activation enhancement in the SAD group while under scrutiny; (b) to assess between-group differences in task-induced functional connectivity by means of psychophysiological interaction (PPI) analyses; and (c) to test for possible between-group differences in functional connectivity considering the whole fMRI run as a global “modulated resting-state” context. We hypothesize that the perception of being under scrutiny will affect brain activation and functional connectivity in neural systems associated with the management of emotionally aversive situations, including arousal systems, fear brain circuits and related cortico-subcortical neuronal networks engaged in the top-down regulation of emotional response (Goldin et al., 2008, Kober et al., 2008 and Brühl et al., 2011). On the basis of the context-dependent change in our current fMRI task (scrutiny on/off), a PPI analysis was included to ascertain whether the effect of a specific influencing region upon another depends on the experimental context. Interestingly, the PPI approach has proved to be very useful in assessing for the modulation of thalamo-cortical systems in response to fear signals (Das et al., 2005). Finally, in keeping with previous evidence that abnormal sustained emotion-related responses are longer than the stimulus duration in certain pathologies (Pujol et al., 2009b), we suspect that the sensation of being scrutinized may evoke brain responses extending beyond the scrutiny blocks, inducing both an abnormal, sustained anxious and arousal-related state for the whole assessment period. We therefore performed a seed-based correlation mapping analysis in order to capture the functional connectivity between different brain regions across the task as a whole “modulated resting-state” (interpreting the scanning as “steady-state” time series data). Thus, functional connectivity assessments may be complementary to conventional neuroimaging characterizing brain dysfunctions in SAD.
نتیجه گیری انگلیسی
3. Results 3.1. Behavioral data 3.1.1. Anxiety before fMRI scanning: STAI-S pre fMRI Prior to the fMRI examination, SAD patients showed significantly higher STAI-S scores than the controls (patients: mean = 30.8 points, S.D. = 8.7; controls: mean = 7.8 points, S.D. = 5.9; t = − 9.78, p < 0.0005). 3.1.2. Fear and/or avoidance of social situations From the LSAS scores, we obtained indices evaluating fear and avoidance experienced by SAD patients (N = 20) in a range of representative ‘social interaction’ and ‘performance’ situations. Analysis of LSAS scores showed a mean for fear and avoidance of 39.4 points (S.D. = 8.1, maximum score: 66 points) for ‘social interaction’ situations and a mean for fear and avoidance of 40.9 points (S.D. = 9.1, maximum score: 78 points) for ‘performance’ situations. 3.1.3. Behavioral assessment of the scrutiny-perception paradigm Subjective out-of-scanner ratings were collected in a different behavioral sample of 10 SAD and 10 control subjects to measure the under-scrutiny perception globally and two specific dimensions involving alertness/tension and anxiety/distress. SAD patients reported higher scores than control subjects for both the perception of scrutiny ‘on’ condition and the difference between ‘on’ and ‘off’ conditions. See Table 2 for results. Table 2. Behavioral samplesa: out-of-scanner assessment of the scrutiny-perception paradigm. Scrutiny ‘on’ condition M (S.D.) Scrutiny ‘off’ condition M (S.D.) ‘On–off’ difference M (S.D.) Controls (N = 10) SAD (N = 10) p b Controls (N = 10) SAD (N = 10) p b Controls (N = 10) SAD (N = 10) p b Scrutiny perception 21.30 (26.10) 58.50 (24.47) 0.005 14.50 (16.74) 32.10 (27.64) 0.211 6.80 (11.31) 26.40 (21.59) 0.036 Alertness/tension 22.30 (20.18) 64.90 (24.35) 0.002 19.30 (17.99) 37.40 (27.35) 0.171 3.00 (4.83) 27.50 (24.12) 0.021 Anxiety/distress 15.00 (21.73) 52.00 (26.88) 0.004 9.20 (16.41) 26.70 (29.94) 0.085 5.50 (7.25) 25.30 (23.78) 0.047 a This table refers to the behavioral samples different from those undergoing the fMRI experiment. b Mann–Whitney U, 2-tailed, between the SAD patients and healthy controls. Range for the behavioral variables' scores: 0–100. Table options 3.2. fMRI data 3.2.1. fMRI group activation patterns and main effects of the task For the main intra-task contrast, in both groups, the perception of scrutiny activated a brain region network including the medial frontal cortex, a bilateral insula–operculum region extending to the inferior frontal cortex, and the cerebellum. SAD patients showed additional activation in the thalami, a part of basal ganglia and a region within the primary visual area. See Table 3 and Fig. 2. We found no activated regions during the scrutiny-off blocks compared to the scrutiny-on blocks, for either group. Table 3. Perception of scrutiny task: brain activation by groups. Anatomical regions Cluster size (num. voxels) Talairach coordinatesa x y z T Voxel level FDR pcorr b Control pattern L inferior frontal cortex and anterior insula 863 − 51 34 − 15 5.12 0.004 R inferior frontal cortex and anterior insula 2050 42 21 − 1 6.57 0.003 R middle frontal cortex 2083 24 59 23 4.76 0.007 Medial frontal cortex 3624 4 14 51 4.72 0.007 R precentral gyrus 566 51 8 49 4.77 0.006 R supramarginal gyrus 618 51 − 39 33 4.64 0.007 L cerebellum 1598 − 32 − 75 − 23 5.74 0.004 SAD pattern L medial and superior frontal cortex 797 − 22 39 13 4.77 0.016 R inferior frontal cortex and anterior insula 1021 55 14 9 4.17 0.017 Medial frontal cortex 1423 12 20 60 4.56 0.017 Bil. thalamus and R globus pallidus 1293 12 − 4 − 5 4.72 0.017 L cerebellum 640 − 46 − 75 − 16 4.51 0.017 R cerebellum and primary visual area 2585 42 − 63 − 20 4.82 0.016 Cerebellar vermis 839 − 2 − 62 − 32 5.16 0.016 Abbreviations: FDR: false discovery rate; L: left; R: right; Bil. bilateral. a Coordinates are provided for the local maxima within the cluster. Imaging coordinates (x y z) were transformed from SPM-Montreal Neurological Institute (MNI) to Talairach space. b All results significant at cluster level pcorr < 0.005 (except the cluster of between-group differences, significant at cluster level pcorr = 0.007). Table options Main task effects. Left: separate whole brain activation patterns in the control ... Fig. 2. Main task effects. Left: separate whole brain activation patterns in the control group (top) and in the social anxiety disorder (SAD) group (middle); between-group comparison is below. Results are overlaid on a standard T1 render template. The hot color bar indicates the t-statistics. All voxels significant at level false discovery rate (FDR) pcorr < 0.05 for patterns and p ≤ 0.01 for between-group comparisons are shown (cluster size ≥ 500 voxels). Images are shown with the subject's left side on the left side of the figure (neurological convention). Right: anatomical localization of the three seeds, Talairach coordinates for the seeds: primary visual seed, 2 − 81 6; thalamus seed, 6 − 13 4; medial frontal seed, 2 16 53. Abbreviation: L: left. Figure options Between-group comparison demonstrated significantly higher brain activation in the SAD group in a cluster involving the primary visual area and the cerebellum (cluster size = 1530 voxels, t = 4.83, p < 0.0005). No suprathreshold differences were observed in the opposite contrast (‘control group > SAD group’). 3.2.2. PPI analyses Significant task-induced functional connectivity between the ROIs and different brain regions was found only in the SAD group. The most relevant finding was that all three PPI maps systematically involved an extensive part of the cingulate cortex in patients (illustrated in Fig. 3). Sagittal view of the psychophysiological interactions (PPIs) assessing ... Fig. 3. Sagittal view of the psychophysiological interactions (PPIs) assessing task-induced functional connectivity maps between primary visual seed (left), the thalamic seed (middle) and the medial frontal seed (right), and anterior–posterior cingulate cortex (ACC–PCC) regions (tables extend the results at a whole brain level). Top: group patterns summarizing main task-induced functional connectivity for the PPI analyses. The figure demonstrates that only the social anxiety disorder (SAD) group shows significant task-induced functional connectivity. Bottom: between-groups comparisons (no suprathreshold results survived for the ‘control group > SAD group’ contrast). Results are overlaid on a standard T1 render template. The hot and violet color bars indicate the t-statistics. All voxels significant at level p ≤ 0.01 are shown (cluster size ≥ 500 voxels). For display purposes, overlaps in violet represent group differences at level p ≤ 0.05 (cluster size ≥ 500 voxels). Images are shown with the subject's left side on the left side of the figure (neurological convention). Abbreviation: n.s.c.: no suprathreshold clusters. Figure options Between-group comparison showed that task-induced functional connectivity was greater in SAD patients between the primary visual ROI and a parietal region, the anterior cingulate cortex (ACC), the left thalamus and right basal ganglia (globus pallidus and caudate nucleus), right cerebellum, and right visual areas and cuneus. Also, we found greater task-induced functional connectivity in SAD patients between the thalamic ROI and the middle and superior frontal gyrus. Finally, the SAD group showed enhanced task-induced functional connectivity between the medial frontal ROI and posterior cingulate cortex (PCC), the left thalamus, the globus pallidus and the hippocampus region. See Table 4 and Fig. 3 (for a sagittal view). Table 4. Psychophysiological interaction effects: independent group maps and between-group comparisons. Anatomical regions Cluster size (num. voxels) Talairach coordinatesa T Voxel level punc x y z Controls' PPI effects patterns n.s.c. for any PPI SAD PPI effects patterns Primary visual PPI effects Bil. anterior cingulate cortex 3029 6 17 21 4.81 < 0.0005 Thalamic PPI effects R anterior insula–operculum 797 40 8 14 3.57 < 0.0005 Bil. anterior cingulate cortex 3748 − 4 34 28 4.47 < 0.0005 R middle temporal gyrus 523 53 − 24 − 4 3.95 < 0.0005 L inferior parietal lobe 514 − 31 − 51 33 3.50 0.001 Medial frontal PPI effects Bil. anterior and posterior cingulate cortex, cuneus and thalamus 7973 − 16 − 57 22 4.74 < 0.0005 Bil. cerebellum 703 − 16 − 59 − 31 3.79 < 0.0005 R hippocampus and parahippocampal gyrus region 654 23 − 16 0 3.77 < 0.0005 SAD group > control group Primary visual PPI effects R inferior parietal region 916 61 − 24 33 4.68 < 0.0005 Bil anterior cingulate region, L thalamus, R basal ganglia 2204 − 16 − 23 3 4.36 < 0.0005 R cerebellum, primary visual area and cuneus 1427 18 − 73 15 4.85 < 0.0005 Thalamic PPI effects R middle and superior frontal gyrus 945 59 − 20 − 2 3.80 < 0.0005 Medial frontal PPI effects Bil. posterior cingulate region and L thalamus 5126 − 16 − 7 6 4.13 < 0.0005 R globus pallidus and hippocampal region 584 24 − 21 − 1 3.80 < 0.0005 Bil. cerebellum 724 − 18 − 61 − 24 3.91 < 0.0005 Abbreviations: L: left; PPI: psychophysiological interaction; R: right; Bil. bilateral; n.s.c., no suprathreshold clusters. a Coordinates are provided for the local maxima within the cluster. Imaging coordinates (x, y, z) were transformed from SPM-Montreal Neurological Institute (MNI) to Talairach space. Table options 3.2.3. Seed-based correlation analyses Fig. 4 presents functional connectivity maps between the three ROIs and the rest of the brain during the entire fMRI acquisition sequence. Seed-based correlation analyses. A) Violet color: seed-based correlation maps ... Fig. 4. Seed-based correlation analyses. A) Violet color: seed-based correlation maps between the primary visual seed (left), the thalamic seed (middle) and the medial frontal seed (right), and the rest of the brain. Top: controls' patterns; bottom: social anxiety disorder (SAD) patterns. Group map results remained significant at a voxel false discovery rate (FDR) pcorr < 0.0005 and cluster-level pcorr ≤ 0.0005. B) Hot colors: seed-based correlation map between-group differences. The color bars indicate the t-statistics for the differences between groups (‘control group > < SAD group’ contrasts). All voxels significant at level p ≤ 0.01 are shown (cluster size ≥ 500 voxels). All between-group results remained significant at a cluster-level pcorr ≤ 0.005. For display purposes, ‘SAD group > control group’ comparison within the thalamic seed results shown at level p < 0.05 (cluster size = 297 voxels). All results are overlaid on a standard T1 render template. Images are shown with the subject's left side on the left side of the figure (neurological convention). Abbreviation: L: left. Figure options Both groups demonstrated a similar connectivity pattern within the visual cortex, with the primary visual seed functionally and positively connected to the whole primary visual area, and also to parts of the extrastriate visual cortex. The thalamic seed was seen to be positively connected to the rest of the thalamus in both groups, but, while controls showed enhanced connectivity between the thalamic seed and some cerebellar regions, the SAD group presented a significant functional connectivity between the thalamic seed and an extended cluster involving the ACC. The medial frontal seed showed predominantly cortical–cortical functional connectivity in controls (focused on the ACC/medial frontal cortex), while the SAD group pattern presented additional cortical–subcortical functional connectivity involving the thalamus and the basal ganglia (caudate nucleus). Group comparisons of the functional connectivity maps did not show significant results for the primary visual seed analysis. Interestingly, between-group functional connectivity differences in relation to the thalamic seed showed that the cerebellum, bilaterally, was less functionally connected to the thalamus in the SAD group (cluster size = 2131 voxels, t = 4.66, p < 0.0005). At a subthreshold level, the thalamic seed was seen to be more functionally connected in the SAD group to an ACC subregion (cluster size = 297 voxels, p ≤ 0.05). The SAD patients showed the medial frontal seed to be significantly more connected to the thalamus and basal ganglia (caudate nucleus), bilaterally (cluster size = 1077 voxels, t = 4.67, p < 0.0005), and also to the left middle frontal gyrus (cluster size = 786 voxels, t = 3.69, p < 0.0005) ( Fig. 4, hot color). 3.3. Relationship between clinical measures and main task effects The LSAS, disorder onset and anxiety scores prior to the scanning did not show significant correlations (threshold of punc ≤ 0.01, cluster size ≥ 500 voxels) with the fMRI main task effects