Neuroscience Letters 590 (2015) 106–110

Contents lists available at ScienceDirect

Neuroscience Letters journal homepage: www.elsevier.com/locate/neulet

Research article

Enhanced amygdala–cortical functional connectivity in meditators Mei-Kei Leung a,b,1 , Chetwyn C.H. Chan c,1 , Jing Yin d , Chack-Fan Lee d , Kwok-Fai So e,f,g , Tatia M.C. Lee a,b,g,h,∗ a

Laboratory of Neuropsychology, The University of Hong Kong, Hong Kong Laboratory of Cognitive Affective Neuroscience, The University of Hong Kong, Hong Kong Applied Cognitive Neuroscience Laboratory, Department of Rehabilitation Sciences, The Hong Kong Polytechnic University, Hong Kong d Centre of Buddhist Studies, The University of Hong Kong, Hong Kong e Department of Ophthalmology, The University of Hong Kong, Hong Kong f GMH Institute of CNS Regeneration, and Guangdong Medical Key Laboratory of Brain Function and Diseases, Jinan University, Guangzhou, China g The State Key Laboratory of Brain and Cognitive Sciences, The University of Hong Kong, Hong Kong h Institute of Clinical Neuropsychology, The University of Hong Kong, Hong Kong b c

h i g h l i g h t s • • • •

Functional connectivity in meditation experts during emotion processing was studied. Enhanced neural connectivity in meditators during affective processing. The enhanced neural connectivity may relate to the mental training of meditation. Amygdala involves in meditation-related affective neuroplasticity.

a r t i c l e

i n f o

Article history: Received 8 October 2014 Received in revised form 9 January 2015 Accepted 20 January 2015 Available online 23 January 2015 Keywords: Meditation Amygdala Anterior cingulate cortex Emotions Mirror neuron Functional connectivity

a b s t r a c t Previous studies have demonstrated that meditation is associated with neuroplastic changes in the brain regions including amygdala, anterior cingulate cortex (ACC), and temporal–parietal junction. Extended from these previous works, this study examined the functional connectivity of the amygdala in meditation experts during affective processing and observed that these experts had significantly stronger left amygdala (LA) connectivity with the dorsal ACC (dACC), premotor, and primary somatosensory cortices (PSC) while viewing affectively positive stimuli when compared to the novices. The current findings have implications for further understanding of affective neuroplastic changes associated with meditation in the amygdala. © 2015 Elsevier Ireland Ltd. All rights reserved.

1. Introduction Loving-kindness meditation is a mental practice of the cultivation of unconditionally positive feelings toward the self and others [1]. Meditation practitioners usually commence their practice in forms that train attention or mindfulness and then move on to other forms of practice such as loving-kindness meditation. The

∗ Corresponding author at: Room 656, Laboratory of Neuropsychology, The Jockey Club Tower, The University of Hong Kong, Pokfulam Road, Hong Kong. Tel.: +852 3917 8394; fax: +852 2819 0978. E-mail address: [email protected] (T.M.C. Lee). 1 These authors contributed equally to this work. http://dx.doi.org/10.1016/j.neulet.2015.01.052 0304-3940/© 2015 Elsevier Ireland Ltd. All rights reserved.

initial attention practice is thought to be complementary to the subsequent compassion practice, which in turn provides a calm and peaceful state for one to enter an attentive state. The practice of meditation is associated with a reduction in anxiety and negative affect [2–4]. Lutz et al. [5] observed that practice of compassion meditation is associated with a significantly higher level of neural activity in the amygdala, right temporo–parietal junction, and right posterior superior temporal sulcus. Lee et al. [6] have demonstrated that experts of loving-kindness meditation showed significant activity in the ventral anterior cingulate cortex (ACC) and inferior frontal gyrus while viewing happy pictures and that in the caudate and middle frontal gyrus while viewing sad pictures. These regions are

M.-K. Leung et al. / Neuroscience Letters 590 (2015) 106–110

important for identifying the emotional value of stimuli, generation of affective states, and regulation of emotional responses [7]. Amygdala is a prime neural correlate of the limbic system essential for emotion processing [8]. Altered amygdala activity toward emotional stimuli has been consistently observed in patients with depression, anxiety, bipolar disorder, or post-traumatic stress disorder [9]. Stein et al. [10] reported that the amygdala is functionally connected with an extended cortical–subcortical network consisting of the anterior and posterior cingulate, insular, prefrontal, and parahippocampal cortices during the perception of angry and fearful faces. Specifically, the amygdala is strongly inhibited by the supragenual ACC when processing fearful/angry stimuli. Such inhibition of the amygdala is important for the down-regulation of the fear response [11]. Failure to engage the pregenual ACC to downregulate the amygdalar response relates to difficulty in resolving emotional conflict in patients with generalized anxiety disorder [12]. Activity of the amygdala is also strongly inhibited by the activity of the posterior cingulate when processing fearful/angry stimuli [10]. A recent study which found that eight weeks of compassion training increased the activity of the right amygdala (RA) in response to negative images at trend-level in a group of meditation novices, and such increase correlates with a decrease in depression scores [13]. Change of functional synchrony of the amygdala with other brain regions during affective processing after long-term meditation practice remains unanswered. Taken together, because of the importance of the neural synchrony between the amygdala and its connected regions in emotion processing and initial evidence of meditation-related affective plasticity in the amygdala, this study explored, for the first time, the impact of long-term meditation practice on the amygdalar functional connectivity during emotion processing. A passive viewing paradigm was employed to probe the basic process of emotion perception in the absence of other cognitive demands. The hypothesis that meditation experts would exhibit a different pattern of amygdalar connectivity in regions involved in affective and reward processing such as the nucleus accumbens, ACC, and OFC compared to novices during emotion processing was verified in this study. 2. Materials and methods 2.1. Participants This study was approved by the Institutional Review Board of the University of Hong Kong and the Hospital Authority (Hong Kong West Cluster). There were altogether 24 right-handed Chinese men participated in this study. The right-handedness was confirmed with the Edinburgh Handedness Inventory [14]. They were free of any medical or psychiatric conditions that could confound the results at the time of recruitment. There were 10 meditation experts who have practiced meditation following the Theravada tradition for at least five years, which include the practice of attention and loving-kindness meditation. The 14 matched novices did not have long-term meditation experience but were interested in meditation and had undergone seven hours of home-based basic meditation practice (to control for motivation of meditation practice between the two groups). Both groups were matched in ages [t(22) = 0.651, p > 0.5] and years of education [t(22) = −1.683, p > 0.1]. All participants gave their written informed consent before the study began. 2.2. Self-report measures The 20-item Chinese Affect Scale [15] was used to assess positive affect (PA) and negative affect (NA). It is culturally adapted to be

107

equivalent to the Positive and Negative Affect Schedule [16]. Participants rated the frequency of 10 positive and 10 negative affective states in the previous two weeks on a 5-point scale. Separate summation scores were calculated for PA and NA. 2.3. Emotion processing task The task included 20 affectively positive (happy), negative (sad), and neutral (neutral) pictures each from the International Affective Picture System (IAPS) with the highest valence and arousal ratings [17]. Each emotion valence had equal proportions of pictures with human and nonhuman images. All stimuli appeared once on a dark background randomly in two 30-trial runs. Each picture was displayed for 3000 milliseconds (ms), separated by a white central fixation cross with varying ISI: 500 ms, 1000 ms, 1500 ms, 2000 ms, and 2500 ms. The experimental conditions were the trials viewing happy and sad pictures whereas the trials viewing neutral pictures represented the control condition (see the Supplementary Result and Fig. S1 in the Supplemental Material for further details of validation of the task using the novices’ data). The duration of each run was about 155 s, and there were two runs. So the total duration of the fMRI scan was about 310 s. Participants also rated the valence from 1 (very negative) to 9 (very positive) and arousal from 1 (not arousing) to 9 (very arousing) for each happy and sad picture outside the scanner after scanning. 2.4. Image acquisition Whole-brain axial scanning was performed with a 3.0 Tesla Philips Medical Systems Achieva scanner equipped with an 8-channel SENSE head coil. Thirty-two functional slices were acquired using a T2*-weighted gradient echo planar imaging sequence [slice thickness = 4 mm, time to repetition (TR) = 1800 ms, time to echo (TE) = 30 ms, flip angle = 90◦ , matrix = 64 × 64, field of view (FOV) = 230 × 230 × 128 mm, voxel size = 3.59 × 3.59 × 4 mm3 ]. The duration of each fMRI run was about 3 min, and there were two runs. So the total duration of the fMRI scan was about 6 min. The axial slices were adjusted to be parallel to the anterior commissure–posterior commissure (AC–PC) plane. The first six volumes were discarded to allow for T1 equilibration effects. A three-dimensional, T1-weighted, magnetization-prepared rapid-acquisition gradient-echo (MPRAGE) sequence was used to acquire high-resolution anatomical images (164 contiguous sagittal slices, 1 mm thick, TR = 7 ms, TE = 3.2 ms, flip angle = 8◦ , FOV = 164 mm, matrix = 256 × 240 mm, voxel size = 1 mm3 ). 2.5. fMRI data analysis The fMRI data were preprocessed using SPM8 (Wellcome Department of Cognitive Neurology, London, UK) in MATLAB 7.7 (Mathworks Inc. Natick, MA, USA). The preprocessing steps included slice-timing correction, realignment, coregistration, segmentation, normalization, smoothing with an 8 mm full width half maximum (FWHM) kernel, and band-pass filtering (0.009 < f < 0.08 Hz) to reduce the effect of low-frequency drift and high-frequency noise. Functional connectivity analysis, using a seed-driven approach, was performed with the ‘conn’ toolbox v13.1 [18] [http://www.nitrc.org/projects/conn]. The toolbox computes the correlation coefficients between the fMRI signal in a seed region and each voxel in the brain separately to generate the parametric seed-voxel correlation map, which is one of the most common techniques for studying functional connectivity [19]. The left amygdala (LA) and RA were defined using the MarsBar toolbox (v0.42, http://marsbar.sourceforge.net/) [20] as the seed region-of-interest (ROI), based on the anatomical masks provided by the anatomical

108

M.-K. Leung et al. / Neuroscience Letters 590 (2015) 106–110

Table 1 Stronger functional connectivity with the left amygdala while viewing happy (versus neutral) pictures in meditation experts compared with novices. Contrast

Brain regions

(A) Experts > novices

Dorsal anterior cingulate cortex (dACC) Premotor cortex (PMC) Primary somatosensory cortex (PSC) No suprathreshold voxels

(B) Experts < novices

Peak coordinates

Cluster FWE- corrected

x

Y

z

p-Value

Cluster size (Voxels)

2 52 32

−6 −4 −38

44 44 74

0.002 0.009 0.042

234 181 129

Coordinates are in Montreal Neurological Institute (MNI) space.

automatic labeling package [21] defined in Montreal Neurological Institute (MNI) space. Due to the threat of introducing spurious anticorrelations into the data via global signal regression during the removal of noise signals [22], the ‘conn’ toolbox instead used a component-based noise correction method (CompCor) to identify and remove the principle components of physiological and other spurious sources of noises from white matter and cerebral spinal fluid (five dimensions with their zeroth order derivative). This method enhanced the sensitivity and specificity of positive correlations and produced comparable anticorrelations compared with the standard approach with mean global signal removal (subtracting global signals from noise regions of interest) [23]. Additionally, the confounding effect of the movement-related parameters (six dimensions with their first order derivative) and that of the sessionspecific time series (one dimension with their first order derivative) were also removed. The onset and duration of each experimental (happy and sad) and control trial was supplied to the ‘conn’ toolbox so as to distinguish the connectivity maps generated for the happy, sad and neutral conditions. In the first-level analysis, a seed-to-voxel correlation map was produced for each subject per each condition. This was done by extracting the corresponding residual blood oxygenation leveldependent (BOLD) time course from the seed ROI and computing Pearson’s correlation coefficients between that time course and the time course of all other voxels. Correlation coefficients were converted to normally distributed z-scores using Fisher’s transform to allow for subsequent general linear model (GLM) analyses. The connectivity difference between happy/sad and neutral conditions was obtained by subtracting the amygdalar connectivity obtained during happy/sad condition from that during neutral condition (i.e., happy/sad versus neutral). In the second-level random-effects analysis of variance (ANOVA), connectivity maps of the left and right amygdala seeds from all participants were analyzed separately. Regions that showed a group difference (experts > novices, or novices > experts) in the amygdalar connectivity during positive emotion processing (happy versus neutral conditions) or negative emotion processing (sad versus neutral conditions) were reported if they survived a stringent cluster-extent threshold at family-wise error (FWE)corrected p < 0.05 and voxel-height threshold at uncorrected p < 0.001. The connectivity value (Fisher’s z-score) between the amygdala seed and each resultant cluster obtained from the ANOVA test was extracted from the corresponding amygdalar connectivity map per each condition for each participant. The Fisher’s z-scores were converted back to correlation coefficients (r-values) for reporting purposes. A positive r-value indicates a correlation between the amygdala and the resultant cluster whereas a negative r-value indicates an anticorrelation between the amygdala and the resultant cluster. Paired t-tests were performed to verify the direction of a within-group difference. For region that showed significant between-group difference in connectivity with the amygdala, separate correlational analyses were performed to examine the relationship between the average connectivity measures extracted from that resultant cluster and

ratings of IAPS pictures in the meditator and novice groups respectively. If there were significant between-group differences when processing positive emotion, the connectivity difference between the happy and neutral conditions would be tested with the ratings of happy pictures. On the same token, if significant group differences when processing negative emotion were identified, the connectivity difference between the sad and neutral conditions would be tested with the ratings of sad pictures. 3. Results 3.1. Behavioral result The two groups did not differ in their level of positive affect [t(22) = 0.432, p > 0.5], and their valence and arousal ratings for the happy [valence: t(22) = 0.760, p > 0.1; arousal: t(22) = 0.724, p > 0.1] and sad [valence: t(22) = −0.120, p > 0.5; arousal: t(22) = 1.358, p > 0.1] pictures. The meditation experts had significantly lower negative affect [t(22) = −2.844, p < 0.01]. In general, all participants gave higher arousal ratings to the sad pictures compared to the happy pictures [t(23) = 2.220, p < 0.05]. 3.2. Functional connectivity result For processing positive emotion, relative to the novices, the LA of the meditation experts had significantly greater functional connectivity with the dorsal anterior cingulate cortex (dACC), right premotor cortex (PMC), and right primary somatosensory cortex (PSC) while viewing happy compared to neutral pictures (Table 1,

Fig. 1. The left amygdala (seed ROI) of the meditation experts (experts) had significantly stronger connectivity with the dorsal anterior cingulate cortex (dACC), right premotor cortex (PMC), and right primary somatosensory cortex (PSC) than novices while viewing happy compared to neutral pictures (surviving a combination of voxel threshold at uncorrected p < 0.001 and cluster threshold at FWE-corrected p < 0.05). The dashed-line box showed that the LA-dACC co-activation cluster extended to the anterior region of ACC in the experts when they were viewing happy pictures.

M.-K. Leung et al. / Neuroscience Letters 590 (2015) 106–110

Fig. 1 and Supplementary Fig. S2). Although the LA-dACC cluster located at a rather caudal part of the ACC in the between-group result, within-group analysis (cluster-level FWE-corrected p < 0.05) showed that the effect also extended to the anterior region in the experts in the happy condition (referenced by the yellow dashed arrow in Fig. 1). No significantly greater RA connectivity was found in the experts compared to the novices. There was no significantly greater LA or RA connectivity observed in the novices compared to the experts. For processing affectively negative stimuli, there was no significant between-group difference for either LA or RA connectivity while viewing sad versus neutral pictures was observed. The connectivity difference of LA-dACC, LA-PMC, and LA-PSC between happy and neutral conditions did not correlate with the valence and arousal ratings of happy pictures for both meditator and novice groups (p > 0.05).

4. Discussion We found that the LA functional connectivity with the dACC, PMC, and PSC while viewing happy (versus neutral) pictures among meditation experts was greater than that among novices, which coupled with a non-significant group difference in the levels of positive affect. On the other hand, there was no group difference in LA connectivity during negative emotion processing, but the experts did show significantly lower levels of negative emotions than novices. This overall discrepancy may reflect differential amygdala-related processes for evaluating the level of affect of the self and processing emotions of others. In sum, the LA of the experts had a strong positive connectivity with the dACC, PMC, and PSC in the happy condition. It was found that the LA is important for differentiating emotional and neutral stimuli during encoding and responding in the recency-probes task that examined the effect of emotions in facilitating interference resolution [24]. In particular, patients with lesion in the LA, but not the RA, did not exhibit the emotional facilitation effect on resolving interference [25]. Thus, the authors speculated that the LA may be crucial in providing salience and arousal signals to other brain regions for distinguishing emotional and neutral information, and integrating affect into working memory. Together with the findings on positive emotion processing in the current study, this function of the LA may have implication for generating or maintaining positive emotion, which is often impaired in people with depression [26]. The dACC is one of the components in the dorsal neural system for emotion cognition and regulation. The role of the dorsal neural system in emotion processing is to regulate the activity of the ventral neural system such as the amygdala, insula, or OFC during emotion processing [7]. Therefore, the functional coupling between the dorsal and ventral neural system (e.g. dACC and amygdala) during emotion processing has important clinical implications for understanding emotion dysregulation disorders [11,27,28]. Here, we observed a stronger positive coupling between the amygdala and dACC activity while viewing happy (versus neutral) pictures in the meditation experts compared to novices. Although the amygdala–ACC coupling is typically linked with down-regulation of negative emotion (e.g. fear) [10–12], a metaanalysis revealed that the effect size of amygdalar activation is actually stronger for positive than negative stimuli [29]. Because the activity of the ACC is related to the coding of subjective pleasantness [30], the increased amygdala–ACC coupling may reflect enhancement of incentive salience of a rewarding stimulus [31,32]. Although both meditation experts and novices rated the valence and arousal of the happy pictures similarly, and there was no correlation between the connectivity measures and ratings of happy

109

pictures, the experts may differ from the novices in terms of other properties of a felt emotion such as the duration of the positive feeling that they could maintain. It has been suggested that long-term meditation practice may be associated with bottom-up emotion regulation strategy [33]; and the application of some emotion regulation strategy may affect the duration but not the intensity of a perceived emotion [34]. Future study should test if the maintenance of positive emotions differs in meditators compared to that in novices using, for example, mood induction task. The PMC and PSC may be involved in emotion processing via their mirror property that allows us to transform the actions and states of others into a vicarious representation of corresponding actions and states in ourselves, so that we can empathetically share the states of others [35–37]. The somatosensory cortex, especially the right PSC, is important for recognizing emotions from facial expressions. Lesions in this area impair the ability to feel the emotions in others [8]. The somatosensory cortices may mediate the recognition of others’ emotional states by generating internal somatosensory representations that mimic others’ feelings [38]. In line with this thought, a recent effective connectivity study showed that the amygdala has rich connections with different higher-order cortical regions, including the PSC, to modulate various processes relating to cognitive, sensorimotor, and somatosensory aspects of basic emotions [39]. We hence speculate that the enhanced coupling of LA-PMC and LA-PSC while viewing happy (versus neutral) pictures in the experts may facilitate their understanding and sharing of positive emotions. The sample size of this study was limited by the availability of meditation experts. However, riding on the same limitation, the findings on processing of affectively positive stimuli in meditation experts could be considered very robust. Future longitudinal studies employing male and female meditation experts would help address the above unresolved questions.

5. Conclusions Our findings add to the literature of neuroplastic changes associated with meditation practice, specifically in the amygdalar functional connectivity during emotion processing. The positive LA-dACC coupling may be important for the cultivation of positive emotion, in contrast to the negative coupling of amygdala–ACC for the extinction of negative emotion. Future longitudinal studies examining the relationship between changes in behavior, brain structures, and effects of meditation on emotion regulation are worthwhile.

Funding This work was supported by the Research Grants Council General Research Fund [HKU747612H to T. L.] and K.K. Ho International Charitable Foundation.

Conflicts of interest There are no conflicts of interest including any financial, personal, or other relationships with persons or organizations for any author related to the work described in this article.

Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.neulet. 2015.01.052.

110

M.-K. Leung et al. / Neuroscience Letters 590 (2015) 106–110

References [1] S. Salzberg, Loving-Kindness: The Revolutionary Art of Happiness, first ed., Shambhala, Boston, 1995. [2] D.P. Johnson, D.L. Penn, B.L. Fredrickson, P.S. Meyer, A.M. Kring, M. Brantley, Loving-kindness meditation to enhance recovery from negative symptoms of schizophrenia, J. Clin. Psychol. 65 (2009) 499–509. [3] S. Sears, S. Kraus, I think therefore I om: cognitive distortions and coping style as mediators for the effects of mindfulness meditation on anxiety, positive and negative affect, and hope, J. Clin. Psychol. 65 (2009) 561–573. [4] P.R. Goldin, J.J. Gross, Effects of mindfulness-based stress reduction (MBSR) on emotion regulation in social anxiety disorder, Emotion 10 (2010) 83–91. [5] A. Lutz, J. Brefczynski-Lewis, T. Johnstone, R.J. Davidson, Regulation of the neural circuitry of emotion by compassion meditation: effects of meditative expertise, PLoS One 3 (2008) e1897. [6] T.M.C. Lee, M.K. Leung, W.K. Hou, J.C. Tang, J. Yin, K.F. So, C.F. Lee, C.C.H. Chan, Distinct neural activity associated with focused-attention meditation and loving-kindness meditation, PLoS One 7 (2012) e40054. [7] M.L. Phillips, C.D. Ladouceur, W.C. Drevets, A neural model of voluntary and automatic emotion regulation: implications for understanding the pathophysiology and neurodevelopment of bipolar disorder, Mol. Psychiatry 13 (2008) 833–857. [8] R. Adolphs, H. Damasio, D. Tranel, G. Cooper, A.R. Damasio, A role for somatosensory cortices in the visual recognition of emotion as revealed by three-dimensional lesion mapping, J. Neurosci. 20 (2000) 2683–2690. [9] W.C. Drevets, J.L. Price, M.L. Furey, Brain structural and functional abnormalities in mood disorders: implications for neurocircuitry models of depression, Brain Struct. Funct. 213 (2008) 93–118. [10] J.L. Stein, L.M. Wiedholz, D.S. Bassett, D.R. Weinberger, C.F. Zink, V.S. Mattay, A. Meyer-Lindenberg, A validated network of effective amygdala connectivity, NeuroImage 36 (2007) 736–745. [11] G.J. Quirk, R. Garcia, F. Gonzalez-Lima, Prefrontal mechanisms in extinction of conditioned fear, Biol. Psychiatry. 60 (2006) 337–343. [12] A. Etkin, K.E. Prater, F. Hoeft, V. Menon, A.F. Schatzberg, Failure of anterior cingulate activation and connectivity with the amygdala during implicit regulation of emotional processing in generalized anxiety disorder, Am. J. Psychiatry 167 (2010) 545–554. [13] G. Desbordes, L.T. Negi, T.W. Pace, B.A. Wallace, C.L. Raison, E.L. Schwartz, Effects of mindful-attention and compassion meditation training on amygdala response to emotional stimuli in an ordinary, non-meditative state, Front. Hum. Neurosci. 6 (2012) 292. [14] R.C. Oldfield, The assessment and analysis of handedness: the Edinburgh inventory, Neuropsychologia 9 (1971) 97–113. [15] P.N. Hamid, S.T. Cheng, The development and validation of an index of emotional disposition and mood state: the Chinese Affect Scale, Educ. Psychol. Meas. 56 (1996) 995–1014. [16] D. Watson, L.A. Clark, A. Tellegen, Development and validation of brief measures of positive and negative affect: the PANAS scales, J. Pers. Soc. Psychol. 54 (1988) 1063–1070. [17] M.M. Bradley, P.J. Lang, The international affective picture system (IAPS) in the study of emotion and attention, in: J.A. Coan, J.J.B. Allen (Eds.), Handbook of Emotion Elicitation and Assessment, Oxford University Press, New York, 2007, pp. 29–46. [18] S. Whitfield-Gabrieli, A. Nieto-Castanon, Conn: a functional connectivity toolbox for correlated and anticorrelated brain networks, Brain Connect. 2 (2012) 125–141. [19] B.P. Rogers, V.L. Morgan, A.T. Newton, J.C. Gore, Assessing functional connectivity in the human brain by fMRI, Magn. Reson. Imaging 25 (2007) 1347–1357.

[20] M. Brett, I.S. Johnsrude, A.M. Owen, The problem of functional localization in the human brain, Nat. Rev. Neurosci. 3 (2002) 243–249. [21] N. Tzourio-Mazoyer, B. Landeau, D. Papathanassiou, F. Crivello, O. Etard, N. Delcroix, B. Mazoyer, M. Joliot, Automated anatomical labelling of activations in spm using a macroscopic anatomical parcellation of the MNI MRI single subject brain, Neuroimage 15 (2002) 273–289. [22] K. Murphy, R.M. Birn, D.A. Handwerker, T.B. Jones, P.A. Bandettini, The impact of global signal regression on resting state correlations: are anti-correlated networks introduced, NeuroImage 44 (2009) 893–905. ˜ [23] X.J. Chai, A.N. Castanón, D. Ongür, S. Whitfield-Gabrieli, Anticorrelations in resting state networks without global signal regression, NeuroImage 59 (2012) 1420–1428. [24] S.M. Levens, E.A. Phelps, Insula and orbital frontal cortex activity underlying emotion interference resolution in working memory, J. Cogn. Neurosci. 22 (2010) 2790–2803. [25] S. Levens, O. Devinsky, E. Phelps, Role of the left amygdala and right orbital frontal cortex in emotional interference resolution facilitation in working memory, Neuropsychologia 49 (2011) 3201–3212. [26] M. Horner, G. Siegle, R. Schwartz, R. Price, A. Haggerty, A. Collier, E. Friedman, C’MON get happy: reduced magnitude and duration of response during a positive-affect induction in depression, Depress. Anxiety 31 (2014) 952–960. [27] S. Maren, G.J. Quirk, Neuronal signalling of fear memory, Nat. Rev. Neurosci. 5 (2004) 844–852. -Lindenberg, E.M. Drabant, B.A. Verchinski, K.E. Munoz, B.S. Kolachana, M.F. Egan, V.S. Mattay, A.R. Hariri, D.R. Weinberger, 5-HTTLPR polymorphism impacts human cingulate–amygdala interactions: a genetic susceptibility mechanism for depression, Nat. Neurosci. 8 (2005) 828–834. [29] K. Sergerie, C. Chochol, J.L. Armony, The role of the amygdala in emotion processing: a quantitative meta-analysis of functional neuroimaging studies, Neurosci. Biobehav. Rev. 32 (2008) 811–830. [30] K.C. Berridge, M.L. Kringelbach, Affective neuroscience of pleasure: reward in humans and animals, Psychopharmacology (Berl.) 199 (2008) 457–480. [31] O.M. Klimecki, S. Leiberg, C. Lamm, T. Singer, Functional neural plasticity and associated changes in positive affect after compassion training, Cereb. Cortex 23 (2013) 1552–1561. [32] O.M. Klimecki, S. Leiberg, M. Ricard, T. Singer, Differential pattern of functional brain plasticity after compassion and empathy training, Soc. Cogn. Affect. Neurosci. 9 (2014) 873–879. [33] A. Chiesa, A. Serretti, J.C. Jakobsen, Mindfulness: top–down or bottom–up emotion regulation strategy, Clin. Psychol. Rev. 33 (2013) 82–96. [34] M. Lernaire, W. El-Hage, S. Frangou, Reappraising suppression: subjective and physiological correlates of experiential suppression in healthy adults, Front. Psychol. 5 (2014) 571. [35] B. Wicker, C. Keysers, J. Plailly, J.P. Royet, V. Gallese, G. Rizzolatti, Both of us disgusted in my insula: the common neural basis of seeing and feeling disgust, Neuron 40 (2003) 655–664. [36] C. Keysers, B. Wicker, V. Gazzola, J.L. Anton, L. Fogassi, V. Gallese, A touching sight: SII/PV activation during the observation and experience of touch, Neuron 42 (2004) 335–346. [37] T. Singer, B. Seymour, J. O’Doherty, H. Kaube, R.J. Dolan, C.D. Frith, Empathy for pain involves the affective but not sensory components of pain, Science 303 (2004) 1157–1162. [38] T. Suslow, H. Kugel, A.V. Rauch, U. Dannlowski, J. Bauer, C. Konrad, V. Arolt, W. Heindel, P. Ohrmann, Attachment avoidance modulates neural response to masked facial emotion, Hum. Brain Mapp. 30 (2009) 3553–3562. [39] M. Tettamanti, E. Rognoni, R. Cafiero, T. Costa, D. Galati, D. Perani, Distinct pathways of neural coupling for different basic emotions, NeuroImage 59 (2012) 1804–1817.