Clinical neuroscience 567
Emotional intelligence is associated with reduced insula responses to masked angry faces Anna Alkozeia and William D.S. Killgorea,b High levels of emotional intelligence (EI) have been associated with increased success in the workplace, greater quality of personal relationships, and enhanced wellbeing. Evidence suggests that EI is mediated extensively by the interplay of key emotion regions including the amygdala, insula, and ventromedial prefrontal cortex, among others. The insula, in particular, is important for processing interoceptive and somatic cues that are interpreted as emotional responses. We investigated the association between EI and functional brain responses within the aforementioned neurocircuitry in response to subliminal presentations of social threat. Fifty-four healthy adults completed the Mayer–Salovey–Caruso Emotional Intelligence Test (MSCEIT) and underwent functional magnetic brain imaging while viewing subliminal presentations of faces displaying anger, using a backward masked facial affect paradigm to minimize conscious awareness of the expressed emotion. In response to masked angry faces, the total MSCEIT scores correlated negatively with a cluster of activation located within the left
Introduction Individuals vary widely in their ability to accurately perceive, understand, and manage emotions in themselves and in others. These capacities are crucial for engaging in successful social interactions and have been termed ‘emotional intelligence’ (EI) [1]. Individuals with high levels of EI are less likely to report mental health problems, have fewer comorbid psychosomatic health difficulties such as chronic fatigue [2], and report greater satisfaction in close relationships [3] and success in the workplace [4]. Despite the importance of EI to so many areas of human functioning, very little is known about the neurocircuitry that contributes to these capacities. Successful functioning in contemporary social contexts requires the ability to detect and understand subtle emotional cues in others and the ability to engage in constructive prosocial actions in response to such cues. Whereas primitive biological survival mechanisms are generally primed to engage immediate defensive reactions to danger or threatening stimuli [5], survival in complex social situations often requires more refined responses that minimize or prevent further escalation of the social threat (i.e. in social contexts, an angry, defensive, or physically confrontational response may be counterproductive). Thus, an individual with greater EI would, by definition, be better able to regulate or minimize these socially inappropriate emotional responses
insula, but not with activation in any other region of interest. Considering the insula’s role in the processing of interoceptive emotional cues, the results suggest that greater EI is associated with reduced emotional visceral reactivity and/or more accurate interoceptive prediction when confronted with stimuli indicative of social threat. NeuroReport 26:567–571 Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved. NeuroReport 2015, 26:567–571 Keywords: amygdala, emotional intelligence, insula, ventromedial prefrontal cortex a Social, Cognitive, and Affective Neuroscience Lab, Department of Psychiatry, University of Arizona, Tucson, Arizona and bMcLean Hospital, Harvard Medical School, Boston, Massachusetts, USA
Correspondence to William D.S. Killgore, PhD, Social, Cognitive and Affective Neuroscience Lab, Department of Psychiatry, University of Arizona, PO Box 245002, Tucson, AZ 85724-5002, USA Tel: + 1 520 621 0605; fax: + 1 520 626 6050; e-mail: [email protected] Received 14 April 2015 accepted 6 May 2015
according to the contextual requirements and their longterm goals. The neurocircuitry that underlies these EI skills is not fully understood, but key brain regions that contribute to emotional responses and their regulation include the amygdala, the insula, and the ventromedial prefrontal cortex (vmPFC), among others [6]. Recent work has suggested that higher EI is correlated with greater gray matter volume within the vmPFC and insula [7,8]. Functional MRI studies have demonstrated that greater EI may be associated with reduced responsiveness of the amygdala, insula, and vmPFC among children and adolescents when viewing facial expressions of fear, suggesting greater neural efficiency of select affect processing systems in those with higher EI (i.e. requiring less neural effort during emotional processing) [9]. A recent study in adults has suggested that higher EI is associated with greater activation of the vmPFC and the rostral anterior cingulate cortex in response to dynamic facial expressions that changed in terms of perceived trustworthiness [10]. Thus, EI is associated with the structure and function of a core neurocircuitry involved in affective perception, evaluation, and regulation, particularly with regard to overtly perceived stimuli. Social interactions often involve subtle communications that occur rapidly and often outside the focus of conscious awareness. The responsiveness of this core affective neurocircuitry to subtle affective stimuli presented
0959-4965 Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
DOI: 10.1097/WNR.0000000000000389
568
NeuroReport 2015, Vol 26 No 10
below the level of normal conscious perception has been well studied [11,12] but not in the context of EI. In the present study, therefore, we used functional MRI to examine the association between EI scores and the responsiveness of this core affective neurocircuitry in healthy individuals during a subliminal angry face perception task. We hypothesized that greater EI would correlate negatively with activation within the amygdala and insula, reflecting reduced spontaneous emotional reactivity to social threat cues and increased activation within the vmPFC, reflecting increased regulation of these responses.
Methods Participants
Fifty-four healthy adults (50% women) participated in a large investigation of the neural correlates of EI. Participants ranged in age from 18 to 45 years (M = 30.37, SD = 8.43), and completed an average of 14.74 years (SD = 1.98) of education. Participants had no history of psychiatric, neurological, or substance-use disorders. All participants provided written informed consent. Partial data from this sample have been reported elsewhere [7,10], but the associations between EI and maskedanger responses are novel and have not previously been published. This research protocol was reviewed and approved by the Institutional Review Board of McLean Hospital and the US Army Human Research Protections Office.
Each block consisted of 20 1.5-s trials. The control condition was identical to the anger condition, but it consisted of neutral faces masked by neutral faces from the same poser. Under both conditions, 10 posers were used. Each trial was separated by a 3-s interstimulus interval. A 15-s fixation cross was presented at the beginning and the end of the run. Each block started with a neutral and ended with a neutral condition to ensure that the participants returned to a baseline state before the masked angry faces were presented. Because neutral affect was the baseline condition to which the affective condition was compared, we believed it was important to ensure adequate sampling of stabilized neutral conditions before and after masked affect presentations. The order of the stimuli, therefore, was as follows: + , N, A, N, A, N, + . Neuroimaging methods
Neuroimaging was conducted at 3 T (Siemens Tim Trio scanner, Erlangen, Germany) with a 12-channel head coil. T1-weighted structural 3D MPRAGE images were acquired (TR/TE/flip angle = 2.1 s/2.25 ms/12°) covering 128 sagittal slices (256 × 256), with a slice thickness of 1.33 mm (voxel size = 1.33 × 1 × 1). Functional T2-weighted scans were acquired over 43 transverse slices (3.5 mm thickness). Sixty images were collected per slice with an interleaved sequence (TR/TE/flip angle = 3.0 s/30 ms/90°). The field of view was 22.4 cm, with a 64 × 64 acquisition matrix. Image processing
Materials The Mayer–Salovey–Caruso Emotional Intelligence Test
The Mayer–Salovey–Caruso Emotional Intelligence Test (MSCEIT) was used to assess participants’ performance when having to reason about and solve emotional problems. The MSCEIT is a 141-item performancebased measure of EI with good psychometric properties [13]. Wechsler Abbreviated Scale of Intelligence
The Wechsler Abbreviated Scale of Intelligence (WASI) [14] was used as a measure of intellectual ability or the intelligence quotient (IQ) and demonstrates strong psychometric properties. Backward masked affect task
During functional MRI, participants completed a backward masked affect task for angry faces, lasting 180 s, which has been used in previous studies [15]. Participants were presented with five alternating 30-s epochs consisting of either masked angry or masked neutral faces. Each masked stimulus trial consisted of an emotional angry target face presented for 20 ms (generally below the threshold of conscious visual perception), followed immediately by a neutral face for 100 ms from the same poser taken from the Ekman and Friesen [16] series.
Processing and analysis of neuroimaging scans were carried out in SPM8 (Wellcome Department of Cognitive Neurology, London, UK; http://www.fil.ion.ucl.ac.uk/spm). Raw functional images were realigned, unwarped, and coregistered to each participant’s MRPAGE image in accordance with standard algorithms. Images were then normalized to Montreal Neurological Institute coordinate space, spatially smoothed (6 mm full-width at half maximum), and resliced to 2 × 2 × 2 mm voxels. The standard canonical hemodynamic response function in SPM was used, serial autocorrelation was corrected with the AR(1) function, and low-frequency confounds were minimized with a 128-s high-pass filter. The Artifact Detection Tool (http://www.nitrc.org/projects/artifact_detect/) was used to regress out scans as nuisance covariates in the first-level analysis, exceeding 3 SD in the mean global intensity and scan-to-scan motion that exceeded 1.0 mm. Absolute head movement did not exceed 2.0 mm for any participant. Statistical analysis
On an individual basis, a general linear model was specified to contrast activation under the masked-anger condition versus activation under the masked neutral condition. These contrast images were entered into a second-level random effects regression analysis with MSCEIT scores as the predictor variable and full-scale
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Emotional intelligence and masked angry faces Alkozei and Killgore 569
WASI scores entered as a nuisance covariate. On the basis of our a-priori hypotheses, bilateral search territories were created using the Wake Forest University PickAtlas Utility [17] and the boundaries were defined by the Automated Anatomical Labeling Atlas [18], focusing on the vmPFC, amygdala, and insula bilaterally. Analyses were thresholded at P less than 0.001 (uncorrected) and subjected to small-volume correction for multiple comparisons within each search territory, false discovery rate (FDR) corrected at P less than 0.10, and k (extent) of at least 10 contiguous voxels. In addition, whole-brain analyses were carried out to explore any other brain region associated with EI in response to masked anger. Analyses were thresholded at P less than 0.001 (uncorrected), FDR corrected at P less than 0.10 and k (extent) of at least 10 contiguous voxels.
Results Overall, participants showed average levels of EI, as measured with the MSCEIT (M = 103.06, SD = 11.78), and high average levels of IQ, as measured with the WASI (M = 111.35, SD = 15.34). Main effect of task activation
A one-sample t-test for the masked angry versus masked neutral contrast was not significant within the regions of interest (ROIs), suggesting that, on the whole, the mean activation did not differ between conditions. Correlations Region of interest analysis
To determine whether activation in response to the masked angry versus masked neutral faces differed according to the level of EI, the data were analyzed with a multiple regression in SPM8, focusing on the three primary ROIs. In response to masked angry faces, the total MSCEIT scores correlated negatively with a cluster of activation located within the left posterior insula (47 voxels; T = 3.91; Montreal Neurological Institute coordinates: x = − 38, y = 2, z = 12; P = 0.06, FDR corrected; Fig. 1). No significant correlations between EI and activation within the vmPFC or the amygdala were observed. Whole-brain analysis
To identify potentially relevant regions that were not initially hypothesized, we conducted a whole-brain regression analysis, applying the same correction thresholds across the entire brain. There were no brain regions outside of the predicted ROIs showing any significant correlation with MSCEIT scores.
Discussion Greater EI was associated with reduced activation of a cluster located within the left posterior insula during subliminal presentations of masked angry faces,
consistent with our expectation that individuals with greater EI capacities would show less interoceptive reactivity to such social displays. However, contrary to expectations, EI was not correlated with activation within the vmPFC or amygdala. These findings suggest that although greater EI was not directly associated with modulation of primary emotional attention processes, it was associated with generally lower visceral interoceptive reactivity to social threat cues. The insula cortex is involved in the interoceptive awareness of the visceral sensations of emotion, making sense of such responses and integrating them with ongoing cognition [19]. Individuals with high interoceptive sensitivity are often prone to anxiety disorders and exaggerated responses to emotional stimuli [20]. The presently observed finding of reduced activation within the insula in response to subtle cues of social threat in high EI individuals might therefore suggest a reduced visceral emotional response, which in turn might contribute to greater composure and less emotional lability within the first moments of a socially threatening encounter. Although this interpretation remains speculative, consistent with this interpretation, Elite Warfighters show increased insula responses to angry faces in comparison with healthy male controls, suggesting an exaggerated visceral reaction that facilitates quick activation of primitive survival responses to a potential source of threat [21]. It is possible that highly emotionally intelligent individuals are able to remain calm (through reduced insula reactivity) when first detecting a socially threatening stimulus, which might explain their particular success in navigating socially complex environments, such as professional work settings and interpersonal relationships [3,4]. However, given that the present findings are correlational in nature, it remains uncertain whether it is greater EI that precedes and leads to lower reactivity while encountering angry expressions, or whether it is the reduced reactivity that contributes to higher EI scores. Future research will be necessary to investigate the directionality between insula activation and behavioral responses in greater detail. In contrast to our previous neuroimaging studies on EI [9,10], no associations between EI and activation within the vmPFC and amygdala were found. This discrepancy is most plausibly accounted for by the present use of backward masking (vs. overt perception), which has been shown to bypass higher cortical processing. This technique may reveal an important aspect of EI, suggesting that EI may be negligibly associated with differences in early amygdalar threat responses or later emotion-regulating processes by the prefrontal cortex, but instead may be characterized by a relatively imperturbable visceral interoception system. Some theoretical models propose the insula as an interoceptive prediction system that compares observed and expected visceral somatic states, leading to increased emotional responses when there is
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
570 NeuroReport 2015, Vol 26 No 10
Fig. 1
2.0 R2 = 0.23
Left insula parameter estimate
1.5
1.0
0.5
0.0
−0.5
−1.0 −30
−20
−10
0
10
20
Residualized MSCEIT score Brain responses to masked angry faces within the left insula (MNI: x = − 38, y = 2, z = 12; k = 47) were negatively correlated with total Mayer–Salovey–Caruso Intelligence Test (MSCEIT) scores (P < 0.10, false discovery rate corrected). The scatter plot illustrates the partial regression plot between MSCEIT scores and the peak voxel value parameter estimate, adjusting for full-scale intelligence (IQ). IQ, intelligence quotient; MNI, Montreal Neurological Institute.
an error in prediction [22]. One possibility is that high EI individuals are more accurate in predicting future interoceptive states within a particular social context; hence, there is less error in the interoceptive prediction signal. Thus, the role of EI in emotional processing of subliminal social cues may be most evident in the
intermediate steps between the immediate amygdalar responses and later regulation by the higher-order cortex (i.e. interoceptive error prediction within the insula). Additional research will be necessary to determine the relationship between interoceptive prediction signals and EI.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Emotional intelligence and masked angry faces Alkozei and Killgore 571
Conclusion
7
Greater EI was associated with reduced activation within the insula cortex in response to subliminal presentations of anger. This suggests an association between higher EI and lower interoceptive reactivity in response to subtle threat cues, perhaps as a result of an interoceptive prediction system that is more accurate in predicting future visceral somatic states within a particular social context. The higher levels of professional success and satisfaction in personal and professional settings often reported for highly emotionally intelligent individuals might be partially explained by their generally lower emotional reactivity during complex situations involving subtle cues of social threat.
8
9 10
11
12
13
Acknowledgements This study was supported by a USAMRAA grant (W81XWH-09-1-0730).
14 15
Conflicts of interest
There are no conflicts of interest. 16
References 1 Mayer JD, Salovey P, Caruso DR. Mayer–Salovey–Caruso Emotional Intelligence Test (MSCEIT) user’s manual. North Tonawanda, NY: MHS; 2002. 2 Schutte NS, Malouff JM, Thorsteinsson EB, Bhullar N, Rooke SE. A metaanalytic investigation of the relationship between emotional intelligence and health. Pers Individ Dif 2007; 42:921–933. 3 Lopes PN, Brackett MA, Nezlek JB, Schütz A, Sellin I, Salovey P. Emotional intelligence and social interaction. Pers Soc Psychol Bull 2004; 30:1018–1034. 4 Van Rooy DL, Viswesvaran C. Emotional intelligence: a meta-analytic investigation of predictive validity and nomological net. J Vocat Behav 2004; 65:71–95. 5 Baldwin DV. Primitive mechanisms of trauma response: an evolutionary perspective on trauma-related disorders. Neurosci Biobehav Rev 2013; 37:1549–1566. 6 Bar-On R, Tranel D, Denburg NL, Bechara A. Exploring the neurological substrate of emotional and social intelligence. Brain 2003; 126 (Pt 8):1790–1800.
17
18
19 20
21
22
Killgore WD, Weber M, Schwab ZJ, Deldonno SR, Kipman M, Weiner MR, Rauch SL. Gray matter correlates of Trait and Ability models of emotional intelligence. Neuroreport 2012; 23:551–555. Takeuchi H, Taki Y, Sassa Y, Hashizume H, Sekiguchi A, Fukushima A, Kawashima R. Regional gray matter density associated with emotional intelligence: evidence from voxel-based morphometry. Hum Brain Mapp 2011; 32:1497–1510. Killgore WD, Yurgelun-Todd DA. Neural correlates of emotional intelligence in adolescent children. Cogn Affect Behav Neurosci 2007; 7:140–151. Killgore WD, Schwab ZJ, Tkachenko O, Webb CA, DelDonno SR, Kipman M, et al. Emotional intelligence correlates with functional responses to dynamic changes in facial trustworthiness. Soc Neurosci 2013; 8:334–346. Almeida J, Pajtas PE, Mahon BZ, Nakayama K, Caramazza A. Affect of the unconscious: visually suppressed angry faces modulate our decisions. Cogn Affect Behav Neurosci 2013; 13:94–101. Whalen PJ, Rauch SL, Etcoff NL, McInerney SC, Lee MB, Jenike MA. Masked presentations of emotional facial expressions modulate amygdala activity without explicit knowledge. J Neurosci 1998; 18:411–418. Brackett MA, Mayer JD. Convergent, discriminant, and incremental validity of competing measures of emotional intelligence. Pers Soc Psychol Bull 2003; 29:1147–1158. Wechsler D. Wechsler Abbreviated Scale of Intelligence. San Antonio, Texas: Psychological Corporation; 1999. Cui J, Olson EA, Weber M, Schwab ZJ, Rosso IM, Rauch SL, Killgore WD. Trait emotional suppression is associated with increased activation of the rostral anterior cingulate cortex in response to masked angry faces. Neuroreport 2014; 25:771–776. Ekman P, Friesen WV. Pictures of facial affect. Paolo Alto, CA: Consulting Psychologists Press; 1976. Maldjian JA, Laurienti PJ, Kraft RA, Burdette JH. An automated method for neuroanatomic and cytoarchitectonic atlas-based interrogation of fMRI data sets. Neuroimage 2003; 19:1233–1239. Tzourio-Mazoyer N, Landeau B, Papathanassiou D, Crivello F, Etard O, Delcroix N, et al. Automated anatomical labeling of activations in SPM using a macroscopic anatomical parcellation of the MNI MRI single-subject brain. Neuroimage 2002; 15:273–289. Singer T, Critchley HD, Preuschoff K. A common role of insula in feelings, empathy and uncertainty. Trends Cogn Sci 2009; 13:334–340. Stein MB, Simmons AN, Feinstein JS, Paulus MP. Increased amygdala and insula activation during emotion processing in anxiety-prone subjects. Am J Psychiatry 2007; 164:318–327. Paulus MP, Simmons AN, Fitzpatrick SN, Potterat EG, Van Orden KF, Bauman J, Swain JL. Differential brain activation to angry faces by elite warfighters: neural processing evidence for enhanced threat detection. PLoS One 2010; 5:e10096. Paulus MP, Stein MB. An insular view of anxiety. Biol Psychiatry 2006; 60:383–387.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Emotional intelligence is associated with reduced insula responses to masked angry faces Anna Alkozeia and William D.S. Killgorea,b High levels of emotional intelligence (EI) have been associated with increased success in the workplace, greater quality of personal relationships, and enhanced wellbeing. Evidence suggests that EI is mediated extensively by the interplay of key emotion regions including the amygdala, insula, and ventromedial prefrontal cortex, among others. The insula, in particular, is important for processing interoceptive and somatic cues that are interpreted as emotional responses. We investigated the association between EI and functional brain responses within the aforementioned neurocircuitry in response to subliminal presentations of social threat. Fifty-four healthy adults completed the Mayer–Salovey–Caruso Emotional Intelligence Test (MSCEIT) and underwent functional magnetic brain imaging while viewing subliminal presentations of faces displaying anger, using a backward masked facial affect paradigm to minimize conscious awareness of the expressed emotion. In response to masked angry faces, the total MSCEIT scores correlated negatively with a cluster of activation located within the left
Introduction Individuals vary widely in their ability to accurately perceive, understand, and manage emotions in themselves and in others. These capacities are crucial for engaging in successful social interactions and have been termed ‘emotional intelligence’ (EI) [1]. Individuals with high levels of EI are less likely to report mental health problems, have fewer comorbid psychosomatic health difficulties such as chronic fatigue [2], and report greater satisfaction in close relationships [3] and success in the workplace [4]. Despite the importance of EI to so many areas of human functioning, very little is known about the neurocircuitry that contributes to these capacities. Successful functioning in contemporary social contexts requires the ability to detect and understand subtle emotional cues in others and the ability to engage in constructive prosocial actions in response to such cues. Whereas primitive biological survival mechanisms are generally primed to engage immediate defensive reactions to danger or threatening stimuli [5], survival in complex social situations often requires more refined responses that minimize or prevent further escalation of the social threat (i.e. in social contexts, an angry, defensive, or physically confrontational response may be counterproductive). Thus, an individual with greater EI would, by definition, be better able to regulate or minimize these socially inappropriate emotional responses
insula, but not with activation in any other region of interest. Considering the insula’s role in the processing of interoceptive emotional cues, the results suggest that greater EI is associated with reduced emotional visceral reactivity and/or more accurate interoceptive prediction when confronted with stimuli indicative of social threat. NeuroReport 26:567–571 Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved. NeuroReport 2015, 26:567–571 Keywords: amygdala, emotional intelligence, insula, ventromedial prefrontal cortex a Social, Cognitive, and Affective Neuroscience Lab, Department of Psychiatry, University of Arizona, Tucson, Arizona and bMcLean Hospital, Harvard Medical School, Boston, Massachusetts, USA
Correspondence to William D.S. Killgore, PhD, Social, Cognitive and Affective Neuroscience Lab, Department of Psychiatry, University of Arizona, PO Box 245002, Tucson, AZ 85724-5002, USA Tel: + 1 520 621 0605; fax: + 1 520 626 6050; e-mail: [email protected] Received 14 April 2015 accepted 6 May 2015
according to the contextual requirements and their longterm goals. The neurocircuitry that underlies these EI skills is not fully understood, but key brain regions that contribute to emotional responses and their regulation include the amygdala, the insula, and the ventromedial prefrontal cortex (vmPFC), among others [6]. Recent work has suggested that higher EI is correlated with greater gray matter volume within the vmPFC and insula [7,8]. Functional MRI studies have demonstrated that greater EI may be associated with reduced responsiveness of the amygdala, insula, and vmPFC among children and adolescents when viewing facial expressions of fear, suggesting greater neural efficiency of select affect processing systems in those with higher EI (i.e. requiring less neural effort during emotional processing) [9]. A recent study in adults has suggested that higher EI is associated with greater activation of the vmPFC and the rostral anterior cingulate cortex in response to dynamic facial expressions that changed in terms of perceived trustworthiness [10]. Thus, EI is associated with the structure and function of a core neurocircuitry involved in affective perception, evaluation, and regulation, particularly with regard to overtly perceived stimuli. Social interactions often involve subtle communications that occur rapidly and often outside the focus of conscious awareness. The responsiveness of this core affective neurocircuitry to subtle affective stimuli presented
0959-4965 Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
DOI: 10.1097/WNR.0000000000000389
568
NeuroReport 2015, Vol 26 No 10
below the level of normal conscious perception has been well studied [11,12] but not in the context of EI. In the present study, therefore, we used functional MRI to examine the association between EI scores and the responsiveness of this core affective neurocircuitry in healthy individuals during a subliminal angry face perception task. We hypothesized that greater EI would correlate negatively with activation within the amygdala and insula, reflecting reduced spontaneous emotional reactivity to social threat cues and increased activation within the vmPFC, reflecting increased regulation of these responses.
Methods Participants
Fifty-four healthy adults (50% women) participated in a large investigation of the neural correlates of EI. Participants ranged in age from 18 to 45 years (M = 30.37, SD = 8.43), and completed an average of 14.74 years (SD = 1.98) of education. Participants had no history of psychiatric, neurological, or substance-use disorders. All participants provided written informed consent. Partial data from this sample have been reported elsewhere [7,10], but the associations between EI and maskedanger responses are novel and have not previously been published. This research protocol was reviewed and approved by the Institutional Review Board of McLean Hospital and the US Army Human Research Protections Office.
Each block consisted of 20 1.5-s trials. The control condition was identical to the anger condition, but it consisted of neutral faces masked by neutral faces from the same poser. Under both conditions, 10 posers were used. Each trial was separated by a 3-s interstimulus interval. A 15-s fixation cross was presented at the beginning and the end of the run. Each block started with a neutral and ended with a neutral condition to ensure that the participants returned to a baseline state before the masked angry faces were presented. Because neutral affect was the baseline condition to which the affective condition was compared, we believed it was important to ensure adequate sampling of stabilized neutral conditions before and after masked affect presentations. The order of the stimuli, therefore, was as follows: + , N, A, N, A, N, + . Neuroimaging methods
Neuroimaging was conducted at 3 T (Siemens Tim Trio scanner, Erlangen, Germany) with a 12-channel head coil. T1-weighted structural 3D MPRAGE images were acquired (TR/TE/flip angle = 2.1 s/2.25 ms/12°) covering 128 sagittal slices (256 × 256), with a slice thickness of 1.33 mm (voxel size = 1.33 × 1 × 1). Functional T2-weighted scans were acquired over 43 transverse slices (3.5 mm thickness). Sixty images were collected per slice with an interleaved sequence (TR/TE/flip angle = 3.0 s/30 ms/90°). The field of view was 22.4 cm, with a 64 × 64 acquisition matrix. Image processing
Materials The Mayer–Salovey–Caruso Emotional Intelligence Test
The Mayer–Salovey–Caruso Emotional Intelligence Test (MSCEIT) was used to assess participants’ performance when having to reason about and solve emotional problems. The MSCEIT is a 141-item performancebased measure of EI with good psychometric properties [13]. Wechsler Abbreviated Scale of Intelligence
The Wechsler Abbreviated Scale of Intelligence (WASI) [14] was used as a measure of intellectual ability or the intelligence quotient (IQ) and demonstrates strong psychometric properties. Backward masked affect task
During functional MRI, participants completed a backward masked affect task for angry faces, lasting 180 s, which has been used in previous studies [15]. Participants were presented with five alternating 30-s epochs consisting of either masked angry or masked neutral faces. Each masked stimulus trial consisted of an emotional angry target face presented for 20 ms (generally below the threshold of conscious visual perception), followed immediately by a neutral face for 100 ms from the same poser taken from the Ekman and Friesen [16] series.
Processing and analysis of neuroimaging scans were carried out in SPM8 (Wellcome Department of Cognitive Neurology, London, UK; http://www.fil.ion.ucl.ac.uk/spm). Raw functional images were realigned, unwarped, and coregistered to each participant’s MRPAGE image in accordance with standard algorithms. Images were then normalized to Montreal Neurological Institute coordinate space, spatially smoothed (6 mm full-width at half maximum), and resliced to 2 × 2 × 2 mm voxels. The standard canonical hemodynamic response function in SPM was used, serial autocorrelation was corrected with the AR(1) function, and low-frequency confounds were minimized with a 128-s high-pass filter. The Artifact Detection Tool (http://www.nitrc.org/projects/artifact_detect/) was used to regress out scans as nuisance covariates in the first-level analysis, exceeding 3 SD in the mean global intensity and scan-to-scan motion that exceeded 1.0 mm. Absolute head movement did not exceed 2.0 mm for any participant. Statistical analysis
On an individual basis, a general linear model was specified to contrast activation under the masked-anger condition versus activation under the masked neutral condition. These contrast images were entered into a second-level random effects regression analysis with MSCEIT scores as the predictor variable and full-scale
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Emotional intelligence and masked angry faces Alkozei and Killgore 569
WASI scores entered as a nuisance covariate. On the basis of our a-priori hypotheses, bilateral search territories were created using the Wake Forest University PickAtlas Utility [17] and the boundaries were defined by the Automated Anatomical Labeling Atlas [18], focusing on the vmPFC, amygdala, and insula bilaterally. Analyses were thresholded at P less than 0.001 (uncorrected) and subjected to small-volume correction for multiple comparisons within each search territory, false discovery rate (FDR) corrected at P less than 0.10, and k (extent) of at least 10 contiguous voxels. In addition, whole-brain analyses were carried out to explore any other brain region associated with EI in response to masked anger. Analyses were thresholded at P less than 0.001 (uncorrected), FDR corrected at P less than 0.10 and k (extent) of at least 10 contiguous voxels.
Results Overall, participants showed average levels of EI, as measured with the MSCEIT (M = 103.06, SD = 11.78), and high average levels of IQ, as measured with the WASI (M = 111.35, SD = 15.34). Main effect of task activation
A one-sample t-test for the masked angry versus masked neutral contrast was not significant within the regions of interest (ROIs), suggesting that, on the whole, the mean activation did not differ between conditions. Correlations Region of interest analysis
To determine whether activation in response to the masked angry versus masked neutral faces differed according to the level of EI, the data were analyzed with a multiple regression in SPM8, focusing on the three primary ROIs. In response to masked angry faces, the total MSCEIT scores correlated negatively with a cluster of activation located within the left posterior insula (47 voxels; T = 3.91; Montreal Neurological Institute coordinates: x = − 38, y = 2, z = 12; P = 0.06, FDR corrected; Fig. 1). No significant correlations between EI and activation within the vmPFC or the amygdala were observed. Whole-brain analysis
To identify potentially relevant regions that were not initially hypothesized, we conducted a whole-brain regression analysis, applying the same correction thresholds across the entire brain. There were no brain regions outside of the predicted ROIs showing any significant correlation with MSCEIT scores.
Discussion Greater EI was associated with reduced activation of a cluster located within the left posterior insula during subliminal presentations of masked angry faces,
consistent with our expectation that individuals with greater EI capacities would show less interoceptive reactivity to such social displays. However, contrary to expectations, EI was not correlated with activation within the vmPFC or amygdala. These findings suggest that although greater EI was not directly associated with modulation of primary emotional attention processes, it was associated with generally lower visceral interoceptive reactivity to social threat cues. The insula cortex is involved in the interoceptive awareness of the visceral sensations of emotion, making sense of such responses and integrating them with ongoing cognition [19]. Individuals with high interoceptive sensitivity are often prone to anxiety disorders and exaggerated responses to emotional stimuli [20]. The presently observed finding of reduced activation within the insula in response to subtle cues of social threat in high EI individuals might therefore suggest a reduced visceral emotional response, which in turn might contribute to greater composure and less emotional lability within the first moments of a socially threatening encounter. Although this interpretation remains speculative, consistent with this interpretation, Elite Warfighters show increased insula responses to angry faces in comparison with healthy male controls, suggesting an exaggerated visceral reaction that facilitates quick activation of primitive survival responses to a potential source of threat [21]. It is possible that highly emotionally intelligent individuals are able to remain calm (through reduced insula reactivity) when first detecting a socially threatening stimulus, which might explain their particular success in navigating socially complex environments, such as professional work settings and interpersonal relationships [3,4]. However, given that the present findings are correlational in nature, it remains uncertain whether it is greater EI that precedes and leads to lower reactivity while encountering angry expressions, or whether it is the reduced reactivity that contributes to higher EI scores. Future research will be necessary to investigate the directionality between insula activation and behavioral responses in greater detail. In contrast to our previous neuroimaging studies on EI [9,10], no associations between EI and activation within the vmPFC and amygdala were found. This discrepancy is most plausibly accounted for by the present use of backward masking (vs. overt perception), which has been shown to bypass higher cortical processing. This technique may reveal an important aspect of EI, suggesting that EI may be negligibly associated with differences in early amygdalar threat responses or later emotion-regulating processes by the prefrontal cortex, but instead may be characterized by a relatively imperturbable visceral interoception system. Some theoretical models propose the insula as an interoceptive prediction system that compares observed and expected visceral somatic states, leading to increased emotional responses when there is
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
570 NeuroReport 2015, Vol 26 No 10
Fig. 1
2.0 R2 = 0.23
Left insula parameter estimate
1.5
1.0
0.5
0.0
−0.5
−1.0 −30
−20
−10
0
10
20
Residualized MSCEIT score Brain responses to masked angry faces within the left insula (MNI: x = − 38, y = 2, z = 12; k = 47) were negatively correlated with total Mayer–Salovey–Caruso Intelligence Test (MSCEIT) scores (P < 0.10, false discovery rate corrected). The scatter plot illustrates the partial regression plot between MSCEIT scores and the peak voxel value parameter estimate, adjusting for full-scale intelligence (IQ). IQ, intelligence quotient; MNI, Montreal Neurological Institute.
an error in prediction [22]. One possibility is that high EI individuals are more accurate in predicting future interoceptive states within a particular social context; hence, there is less error in the interoceptive prediction signal. Thus, the role of EI in emotional processing of subliminal social cues may be most evident in the
intermediate steps between the immediate amygdalar responses and later regulation by the higher-order cortex (i.e. interoceptive error prediction within the insula). Additional research will be necessary to determine the relationship between interoceptive prediction signals and EI.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Emotional intelligence and masked angry faces Alkozei and Killgore 571
Conclusion
7
Greater EI was associated with reduced activation within the insula cortex in response to subliminal presentations of anger. This suggests an association between higher EI and lower interoceptive reactivity in response to subtle threat cues, perhaps as a result of an interoceptive prediction system that is more accurate in predicting future visceral somatic states within a particular social context. The higher levels of professional success and satisfaction in personal and professional settings often reported for highly emotionally intelligent individuals might be partially explained by their generally lower emotional reactivity during complex situations involving subtle cues of social threat.
8
9 10
11
12
13
Acknowledgements This study was supported by a USAMRAA grant (W81XWH-09-1-0730).
14 15
Conflicts of interest
There are no conflicts of interest. 16
References 1 Mayer JD, Salovey P, Caruso DR. Mayer–Salovey–Caruso Emotional Intelligence Test (MSCEIT) user’s manual. North Tonawanda, NY: MHS; 2002. 2 Schutte NS, Malouff JM, Thorsteinsson EB, Bhullar N, Rooke SE. A metaanalytic investigation of the relationship between emotional intelligence and health. Pers Individ Dif 2007; 42:921–933. 3 Lopes PN, Brackett MA, Nezlek JB, Schütz A, Sellin I, Salovey P. Emotional intelligence and social interaction. Pers Soc Psychol Bull 2004; 30:1018–1034. 4 Van Rooy DL, Viswesvaran C. Emotional intelligence: a meta-analytic investigation of predictive validity and nomological net. J Vocat Behav 2004; 65:71–95. 5 Baldwin DV. Primitive mechanisms of trauma response: an evolutionary perspective on trauma-related disorders. Neurosci Biobehav Rev 2013; 37:1549–1566. 6 Bar-On R, Tranel D, Denburg NL, Bechara A. Exploring the neurological substrate of emotional and social intelligence. Brain 2003; 126 (Pt 8):1790–1800.
17
18
19 20
21
22
Killgore WD, Weber M, Schwab ZJ, Deldonno SR, Kipman M, Weiner MR, Rauch SL. Gray matter correlates of Trait and Ability models of emotional intelligence. Neuroreport 2012; 23:551–555. Takeuchi H, Taki Y, Sassa Y, Hashizume H, Sekiguchi A, Fukushima A, Kawashima R. Regional gray matter density associated with emotional intelligence: evidence from voxel-based morphometry. Hum Brain Mapp 2011; 32:1497–1510. Killgore WD, Yurgelun-Todd DA. Neural correlates of emotional intelligence in adolescent children. Cogn Affect Behav Neurosci 2007; 7:140–151. Killgore WD, Schwab ZJ, Tkachenko O, Webb CA, DelDonno SR, Kipman M, et al. Emotional intelligence correlates with functional responses to dynamic changes in facial trustworthiness. Soc Neurosci 2013; 8:334–346. Almeida J, Pajtas PE, Mahon BZ, Nakayama K, Caramazza A. Affect of the unconscious: visually suppressed angry faces modulate our decisions. Cogn Affect Behav Neurosci 2013; 13:94–101. Whalen PJ, Rauch SL, Etcoff NL, McInerney SC, Lee MB, Jenike MA. Masked presentations of emotional facial expressions modulate amygdala activity without explicit knowledge. J Neurosci 1998; 18:411–418. Brackett MA, Mayer JD. Convergent, discriminant, and incremental validity of competing measures of emotional intelligence. Pers Soc Psychol Bull 2003; 29:1147–1158. Wechsler D. Wechsler Abbreviated Scale of Intelligence. San Antonio, Texas: Psychological Corporation; 1999. Cui J, Olson EA, Weber M, Schwab ZJ, Rosso IM, Rauch SL, Killgore WD. Trait emotional suppression is associated with increased activation of the rostral anterior cingulate cortex in response to masked angry faces. Neuroreport 2014; 25:771–776. Ekman P, Friesen WV. Pictures of facial affect. Paolo Alto, CA: Consulting Psychologists Press; 1976. Maldjian JA, Laurienti PJ, Kraft RA, Burdette JH. An automated method for neuroanatomic and cytoarchitectonic atlas-based interrogation of fMRI data sets. Neuroimage 2003; 19:1233–1239. Tzourio-Mazoyer N, Landeau B, Papathanassiou D, Crivello F, Etard O, Delcroix N, et al. Automated anatomical labeling of activations in SPM using a macroscopic anatomical parcellation of the MNI MRI single-subject brain. Neuroimage 2002; 15:273–289. Singer T, Critchley HD, Preuschoff K. A common role of insula in feelings, empathy and uncertainty. Trends Cogn Sci 2009; 13:334–340. Stein MB, Simmons AN, Feinstein JS, Paulus MP. Increased amygdala and insula activation during emotion processing in anxiety-prone subjects. Am J Psychiatry 2007; 164:318–327. Paulus MP, Simmons AN, Fitzpatrick SN, Potterat EG, Van Orden KF, Bauman J, Swain JL. Differential brain activation to angry faces by elite warfighters: neural processing evidence for enhanced threat detection. PLoS One 2010; 5:e10096. Paulus MP, Stein MB. An insular view of anxiety. Biol Psychiatry 2006; 60:383–387.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
