Neural correlates of alexithymia: A meta-analysis of emotion processing studies

Page 1

R

Np

JRa

b

c

d

e

a

ARRA

KAABCEFIMMNP

C

f

(

0h

Neuroscience and Biobehavioral Reviews 37 (2013) 1774–1785

Contents lists available at ScienceDirect

Neuroscience and Biobehavioral Reviews

journa l h om epa ge: www.elsev ier .com/ locate /neubiorev

eview

eural correlates of alexithymia: A meta-analysis of emotionrocessing studies

orien van der Veldea,∗, Michelle N. Servaasa, Katharina S. Goerlichb,ichard Bruggemanc, Paul Hortond, Sergi G. Costafredad, André Alemana,e

Neuroimaging Center, University Medical Center Groningen, Antonius Deusinglaan 2, 9713 AW Groningen, The NetherlandsDepartment of Psychiatry, Psychotherapy, and Psychosomatics, Medical School, RWTH Aachen University, Pauwelstrasse 30, 52074 Aachen, GermanyDepartment of Neuroscience and Psychiatry, University Medical Center Groningen, Hanzeplein 1, 9713 GZ Groningen, The NetherlandsDepartment of Old Age Psychiatry, Institute of Psychiatry, King’s College London, PO Box 070, De Crespigny Park, SE5 8AF London, United KingdomDepartment of Experimental Psychopathology, University of Groningen, Grote Kruisstraat 2/1, 9712 TS Groningen, The Netherlands

r t i c l e i n f o

rticle history:eceived 26 January 2013eceived in revised form 8 July 2013ccepted 12 July 2013

eywords:lexithymiamygdalarainingulate

a b s t r a c t

Alexithymia is a personality trait characterized by difficulties in the experience and cognitive processingof emotions. It is considered a risk factor for a range of psychiatric and neurological disorders. Func-tional neuroimaging studies investigating the neural correlates of alexithymia have reported inconsistentresults. To integrate previous findings, we conducted a parametric coordinate-based meta-analysisincluding fifteen neuroimaging studies on emotion processing in alexithymia. During the processing ofnegative emotional stimuli, alexithymia was associated with a diminished response of the amygdala, sug-gesting decreased attention to such stimuli. Negative stimuli additionally elicited decreased activationin supplementary motor and premotor brain areas and in the dorsomedial prefrontal cortex, possiblyunderlying poor empathic abilities and difficulties in emotion regulation associated with alexithymia.

motionusiform gyrusnsula

eta-analysisirror neuron systemeuroimaging

Positive stimuli elicited decreased activation in the right insula and precuneus, suggesting reduced emo-tional awareness in alexithymia regarding positive affect. Independent of valence, higher (presumablycompensatory) activation was found in the dorsal anterior cingulate possibly indicating increased cogni-tive demand. These results suggest valence-specific as well as valence-independent effects of alexithymiaon the neural processing of emotions.

recuneus© 2013 Elsevier Ltd. All rights reserved.

ontents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17752. Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1776

2.1. Systematic literature search . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17762.2. Data extraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17762.3. Data-analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1776

3. Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17773.1. Negative stimuli . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17773.2. Positive stimuli . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1779

4. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1779

4.1. Valence-independent emotional processing. . . . . . . . . . . . . . . . . . . . . . .

4.2. Valence-specific emotional processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4.2.1. Negative emotional stimuli . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4.2.2. Positive emotional stimuli . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

∗ Corresponding author at: Department of Neuroscience, Neuroimaging Center, UMCG-ax: +31 50 363 88 75.

E-mail addresses: [email protected] (J. van der Velde), [email protected]. Bruggeman), horton [email protected] (P. Horton), [email protected] (S.G.

149-7634/$ – see front matter © 2013 Elsevier Ltd. All rights reserved.ttp://dx.doi.org/10.1016/j.neubiorev.2013.07.008

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1779 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1780

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1780 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1782

O&O, P.O. Box 196, 9700 AD Groningen, The Netherlands. Tel.: +31 50 363 87 92;

nl (M.N. Servaas), [email protected] (K.S. Goerlich), [email protected]), [email protected] (A. Aleman).

Page 2

J. van der Velde et al. / Neuroscience and Biobehavioral Reviews 37 (2013) 1774–1785 1775

4.3. Methodological issues and limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17835. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1783

Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1783Appendix A. Supplementary data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1783

. . . . . .

1

afrr1iappiasFf(tt

Brp2(oceaerr

chifttb(dde(2arbsafa2ip2

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. Introduction

Alexithymia (“no words for feelings”) is a personality trait char-cterized by difficulties in identifying, analyzing and verbalizingeelings, restricted imaginal capacities and limited emotional expe-ience (Sifneos, 1973; Vorst and Bermond, 2001). Its prevalenceate lies around 10% in the general population (Salminen et al.,999). In the past, there has been some debate on whether alex-

thymia should be conceptualized as a distinct clinical type ors a dimensional personality construct. However, recent studiesrovided strong support in favor of alexithymia as a dimensionalersonality construct (Mattila et al., 2010; Parker et al., 2008). Alex-

thymia is considered to be a risk factor for various psychiatricnd psychosomatic disorders, including substance abuse, depres-ion and schizophrenia (Taylor et al., 1997; van’t Wout et al., 2007).urthermore, individuals with alexithymia report lower life satis-action (Mattila et al., 2007) and are more likely to commit suicideHintikka et al., 2004). Therefore, it is of great clinical importanceo gain more insight in the neural basis underlying this personalityrait.

Difficulties in emotion processing are at the core of alexithymia.oth the ability to experience and cognitively process emotions iseduced. For example, individuals with alexithymia show impairederformance in remembering emotional words (Luminet et al.,006), problems during the identification of facial expressionsGrynberg et al., 2012; Parker et al., 1993) and impaired higherrder mentalizing (Swart et al., 2009). Hence, theories on neuralorrelates of alexithymia mainly focus on brain areas involved inmotion processing. One theory states that alexithymia might bessociated with a right hemisphere deficit or a left hemisphere pref-rence (Bermond et al., 2005; Buchanan et al., 1980) because theight hemisphere plays an important role in the perception andegulation of emotional behavior (Adolphs et al., 2000).

Lane et al. (1997) hypothesized a central role for the anterioringulate cortex (ACC) in alexithymia. According to their ‘blindfeel’ypothesis, the conscious experience of emotion is compromised

n individuals with alexithymia, assumed to result from a dys-unction in the ACC (Lane et al., 1997). In addition to the ACC,he insula is another relevant brain region in generating emo-ional experience. This structure receives information from internalodily states and integrates these into a subjective feeling stateCraig, 2009). Furthermore, subcortical areas, such as the amyg-ala and striatum, are proposed to underlie emotion processingifficulties in alexithymia because of their role in the detection ofmotional significance and the generation of emotional feelingsBermond et al., 2006; Goerlich et al., 2013; Kano and Fukudo,013; Larsen et al., 2003; Moriguchi and Komaki, 2013; Taylornd Bagby, 2004; Wingbermühle et al., 2012). Thus, alexithymia-elated difficulties in perceiving and experiencing emotions maye associated with dysfunction of the ACC, insula, amygdala andtriatum. According to two recent reviews, this decrease in limbicnd paralimbic activation is associated with a decrease in pre-rontal activation when individuals with high alexithymia scoresre presented with external emotional stimuli (Kano and Fukudo,

013; Moriguchi and Komaki, 2013). Especially, altered activation

n the orbitofrontal cortex and medial prefrontal cortex is pro-osed to underlie alexithymia (Bermond et al., 2006; Larsen et al.,003; Wingbermühle et al., 2012) because of their involvement

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1783

in the cognitive control of emotions, including emotion regula-tion (Ochsner and Gross, 2005) and emotional decision making(Rogers et al., 2004). In fact, lesions in these regions have beenshown to result in restricted capacities for the cognitive processingof emotions (Glascher et al., 2012). Difficulties in the restrictedimaginal capacities in alexithymia, on the other hand, are thoughtto be related to reduced activation in the posterior cingulate cor-tex (Aleman, 2005; Bermond et al., 2006; Kano and Fukudo, 2013;Larsen et al., 2003; Moriguchi and Komaki, 2013; Wingbermühleet al., 2012), because of its role in emotional memory (Maddock,1999) and the imagination of future events (D’Argembeau et al.,2008).

To date, neuroimaging studies have tried to identify these pro-posed brain regions as neural correlates of alexithymia. In 2010,Pouga and colleagues compared alexithymia-related brain activa-tion in the medial frontal gyrus, cingulate gyrus and amygdalaacross studies. They identified lower activation in the medial frontalgyrus and the amygdala in alexithymia, while results on the cin-gulate gyrus were mixed (i.e. both higher and lower activationwas associated with alexithymia). Furthermore, results regardingother brain areas in alexithymia are also inconsistent across stud-ies. Most existing reviews on the neural correlates of alexithymiahighlight the variability in findings across studies but do not inte-grate these findings (Aleman, 2005; Bermond et al., 2006; Larsenet al., 2003; Taylor and Bagby, 2004; Wingbermühle et al., 2012).Furthermore, only a selected group of brain regions are mentionedin these reviews, leaving other possible neural correlates unre-marked.

The aim of the present study was to integrate findings from theliterature and identify brain regions underlying emotion processingdifficulties in alexithymia across studies. Therefore, studies exam-ining the neural correlates of alexithymia during the processingof either positive or negative emotional stimuli were includedin this meta-analysis. Positive and negative emotional processingpresumably differ in their neural correlates (Wager et al., 2003).Furthermore, previous studies have indicated that there appearsto be a valence-specific effect on the neural correlates of emo-tion processing in alexithymia (Berthoz et al., 2002; Kano et al.,2003; Pollatos and Gramann, 2011; Reker et al., 2010). For exam-ple, studies investigating alexithymia reported decreased activityin the amygdala for negative, but not for positive stimuli (Kugelet al., 2008; Reker et al., 2010). Moreover, Berthoz et al. (2002)reported decreased activation in the ACC for negative stimuliwhile activation in this area was increased for positive stimuliin alexithymia. Besides these indications of possible differencesbetween the neural correlates of positive and negative emotionprocessing in alexithymia, a recent meta-analyses indicated thatemotional valence modulates neural abnormalities in depression(Groenewold et al., 2012). Given the fact that alexithymia is signif-icantly related to depression (Honkalampi et al., 2000), one couldhypothesize that valence might also modulate alexithymia-relatedbrain activation. Therefore, brain activation associated with alex-ithymia was examined for the processing of negative and positivestimuli separately. A novel parametric coordinate-based meta-

analysis (PCM) approach (Costafreda, 2012) was employed as thismethod allowed for the inclusion of both whole brain and regionof interest (ROI) studies as well as for the inclusion of differentthresholds across studies.

Page 3

1 Biobe

2

2

btswmgOi2Twt

maanottmroas(taidLaaweiEcpcdrtr

2

siabMowystom(

776 J. van der Velde et al. / Neuroscience and

. Methods

.1. Systematic literature search

Studies investigating the neural basis of alexithymia, publishedefore April 2013, were identified through a systematic litera-ure search in the PubMed and Web of Knowledge databases. Theearch keywords were “alexithymia” OR “alexithymic” combinedith MRI OR magnetic resonance imaging OR fMRI OR functionalagnetic resonance imaging OR PET OR positron emission tomo-

raphy OR neuroimaging OR brain imaging OR imaging OR BOLDR blood oxygen level dependent. In addition; the references of the

ncluded manuscripts and previously published reviews (Aleman,005; Bermond et al., 2006; Lane et al., 1997; Larsen et al., 2003;aylor, 2000; Taylor and Bagby, 2004; Wingbermühle et al., 2012)ere manually scanned to obtain eligible studies not identified

hrough the initial search.Inclusion criteria were (a) the use of fMRI or PET, (b) the assess-

ent of brain activation during emotion processing related tolexithymia, (c) the use of image subtraction to identify brain areasssociated with either negative or positive emotional stimuli versuseutral stimuli or versus a neutral baseline and, (d) the assessmentf alexithymia using the Toronto Alexithymia Scale (TAS-20) orhe cognitive dimension of the Bermond-Vorst Alexithymia Ques-ionnaire (BVAQ). Both the TAS-20 and the BVAQ are self-report

easures frequently applied to assess alexithymia in current neu-oimaging studies. The TAS-20 assesses the cognitive dimensionf alexithymia while the BVAQ assesses both the cognitive andffective alexithymia dimensions. Because the cognitive dimen-ion of the BVAQ and the TAS-20 total score are highly correlatedVorst and Bermond, 2001), studies applying either the TAS-20 orhe cognitive dimension of the BVAQ were included in this meta-nalysis. The affective dimension of the BVAQ could not be includedn the current meta-analysis, since only one study examined thisimension (Pouga et al., 2010). Neuroimaging studies applying theevel of Emotional Awareness Scale (LEAS) were excluded from thenalyses as well because this scale was not developed to examinelexithymia per se (Lane et al., 1998). Furthermore, the LEAS is onlyeakly or non-significantly correlated with the TAS-20 (Lumley

t al., 2005) and the BVAQ (Fantini-Hauwel et al., 2012). Stud-es were further excluded if (a) the articles were not written innglish, (b) no original data was presented (e.g. review, letter oromment), (c) only an abstract was published or (d) the study sam-le consisted of patients only or patients and healthy controls wereombined in one sample. Two researchers (JV and MS) indepen-ently performed the data search and selection. Discrepancies wereesolved by consensus. When necessary, corresponding authors ofhe selected papers were contacted to clarify the methods and/oresults.

.2. Data extraction

The following data were extracted from each study: (1) sampleize, (2) contrasts examined (negative versus neutral/baseline, pos-tive versus neutral/baseline, positive or negative correlation withlexithymia), (3) field of view (Region of Interest (ROI) or wholerain), (4) the Automated Anatomical Labeling (AAL) (Tzourioazoyer et al., 2002). In case a ROI analysis was performed based

n a mask created according to the AAL system and no informationas given on the coordinates, the AAL region was fed into the anal-

ses. When the ROI region was not defined according to the AALystem, the region was relabeled according to the AAL system by

he authors in consultation with a trained anatomist. Informationn the included ROIs and the AAL labels can be found in supple-entary Table 1, (5) normalization template (MNI or Talairach),

6) location information (the sets of x, y, z brain coordinates for

havioral Reviews 37 (2013) 1774–1785

significant findings or an empty x, y, z set for non-significant find-ings), (7) study-specific statistical threshold, (8) information onstatistical significance (p, Z, T, F) and the corresponding value and,(9) the applied smoothing kernel. The data were extracted by thefirst author and checked independently by another researcher (MS).

2.3. Data-analysis

To summarize the results reported across studies, software wasused that implements a new parametric coordinate-based meta-analysis (PCM) technique allowing the pooling of both ROI-basedand coordinate-based findings (described in detail in Costafreda(2012)). In neuroimaging meta-analyses, differences in statisticalthresholds across studies that are being summarized are frequentlyignored. However, statistical thresholds can vary substantially,from strict family-wise correction methods to uncorrected p-values. The key advantage of the PCM approach to neuroimagingmeta-analyses is that it produces valid effect-size summaries ofstudies with different statistical thresholds. Briefly, an unbiasedvoxel-wise statistical summary map is produced based on the spa-tial location and effect size of the maxima of significant as well asnon-significant findings.

First, when necessary, coordinates were transformed fromTalairach on the MNI coordinate system by using a non-linear trans-formation (Brett et al., 2001). Second, all of the effect sizes andstatistical threshold values (i.e. p, T, r or F) were converted into Z-values as a common measure of effect size. Next, Z-value summarymaps were created for each study by convolving the significantfindings of a study with a uniform kernel size of 15 mm. In otherwords, the Z-value associated with a significant finding at a spe-cific coordinate (x, y, z) was distributed across all voxels withinthe 15 mm radius of the coordinate. This kernel width was chosenbased on empirical evaluations which found that uniform kernels of15–20 mm offer optimal sensitivity in neuroimaging meta-analyses(Radua et al., 2012; Wager et al., 2004). Non-significant find-ings were represented by intervals containing the values for theunknown measurements that were below the reported statisticalthresholds. Such an interval estimate is analog to how a confi-dence interval in standard statistical estimation contains a rangeof probable values for an unknown population parameter. That is,the measurements in voxels located outside the radial distance inthe whole-brain or ROI contrast were interval estimates with inter-val boundaries, determined by the statistical threshold reported bythe study (e.g. a non-significant finding with an uncorrected thresh-old p < .001 is approximately equivalent to a Z interval of [−Inf,3.09]).

Summary maps were created for the contrast ‘positive > neutral’and ‘negative > neutral’ separately. Positive Z-values reflected apositive correlation with alexithymia or greater brain activationin high alexithymic individuals compared to low alexithymic indi-viduals, while negative Z-values reflected negative correlationswith alexithymia or greater activation in low alexithymic individ-uals compared to high alexithymic individuals. Fourth, an overallZ-value summary map was created by pooling the study levelsummary maps. The software created a pooled summary map byobtaining maximum likelihood estimates of the population meanand standard deviation of the Z-values across studies for each voxelthrough the optimization of the likelihood function of the nor-mal distribution. The contribution of each study was weighted tothe summary map by its sample size. Fifth, voxels in the sum-mary map that had a Z-mean value significantly different fromzero (i.e. the voxels showing evidence of differential brain acti-

vation) were determined by performing a two tailed t-test onthe estimated Z-mean value for each voxel. The null hypothesiswas: |�| = 0 and the alternative hypothesis (H1) was |�| /= 0. Thus,voxels with a p-value below a chosen statistical threshold were

Page 4

J. van der Velde et al. / Neuroscience and Biobehavioral Reviews 37 (2013) 1774–1785 1777

m of s

dedlmtfiwftwa

3

eerittsa

Fig. 1. PRISMA flow diagra

eemed to show significant evidence to the alternative hypoth-sis. To correct for multiple comparisons, we used a p < .05 falseiscovery rate (FDR). Finally, the clusters of voxels which were

arger than the chosen extent threshold of 100 mm3 were deter-ined. The thresholded (i.e. the Z-values (voxels) that survived

he extent and statistical thresholds) T and r (correlation coef-cient) effect size summary maps calculated from the Z-valuesere given as output. Two separate summary maps were given

or the contrasts ‘positive > neutral’ and ‘negative > neutral’. Clus-ers of voxels with a value >0 correlated positively with alexithymiahile clusters of voxels with a value <0 correlated negatively with

lexithymia.

. Results

The initial search returned 219 original titles. Of these, 165 werexcluded after reviewing the abstracts and titles. Reasons for thexclusion of studies are presented in Fig. 1. The full texts of theemaining 54 titles were examined in more detail, which resulted

n the inclusion of 15 studies in this meta-analysis. For details onhe inclusion procedure, see the flow chart in Fig. 1. Details onhe included studies are presented in Tables 1 and 2. Specific rea-ons for exclusion of studies examining alexithymia-related brainctivation can be found in supplementary Table 2.

tudy selection procedure.

3.1. Negative stimuli

Fifteen studies were identified which examined alexithymiaduring the processing of negative stimuli. Of these studies, sevenused a region of interest (ROI) approach, two a whole brainapproach and six used both a whole-brain and ROI approach. Detailson the included regions of interest per study can be found in supple-mentary Table 1. The total sample comprised 415 individuals (239male and 176 female). Demographic details of the participants areshown in Table 1.

For negative stimuli (contrast negative > neutral/baseline)greater activation in the dorsal anterior cingulate cortex (ACC) andthe right middle temporal gyrus was found in people with higherlevels of alexithymia (Fig. 2A and Table 3A). Furthermore, higherlevels of alexithymia were associated with less activation of theleft premotor cortex, bilateral fusiform gyrus, bilateral amygdala,bilateral supplementary motor area, the left dorsomedial prefrontalcortex (dMPFC), the left middle occipital gyrus, the right puta-men and the left superior parietal gyrus (Fig. 2C and Table 3B).Four significant clusters did not survive the extent thresholdof 100 mm3. These findings are presented in the supplementary

Table S3.

To examine the relative contribution of the included studies tothe summary findings, the distance � was calculated voxel-wisebetween the coordinates of the significant findings in the summary

Page 5

1778

J. van

der V

elde et

al. /

Neuroscience

and Biobehavioral

Review

s 37

(2013) 1774–1785

Table 1Details of studies included in the meta-analysis.

Study N N male/Nfemale

Mean age (SD) Mean alexithymia scores (SD) Questionnaire Covariates Scanner method Smoothingkernel (mm)

Reker et al. (2010) 33 0/33 24.8 (3.4) 37.9 (7.7) TAS-20 BDI, STAI fMRI 3 T 6 × 6 × 6Lee et al. (2011) 38 0/38 23.2 (3.47) 43.93 (9.98) TAS-20 None fMRI 1.5 T 7 × 7 × 7Eichmann et al. (2008) 22 14/8 27.9 (7.9) 43.9 (10.9) TAS-20 None fMRI 3 T 6 × 6 × 6Kugel et al. (2008) 21 13/8 26.5 (3.9) 43.2 (13.7) TAS-20 BDI, STAI fMRI 3 T 6 × 6 × 6Kano et al. (2003) 24 (12 HA; 12 LA) 24/0 HA: 23.2 (2.4) LA: 22.8 (1.7) HA: 64.2 (3.6) LA: 40.5 (5.7) TAS-20 None PET 12 × 12 × 12Mériau et al. (2006) 23 0/23 27.1 (4.7) 40.3 (6.7) TAS STAI, PANAS PA fMRI 1.5 T 8 × 8 × 8Heinzel et al. (2010) 60 (30 HA; 30 LA) 60/0 HA: 26.6 (4.2) LA: 27.1 (4.8) HA: 59.06 (5.43) LA: 33.32 (5.62) TAS-20 BDI fMRI 1.5 T 8 × 8 × 8Berthoz et al. (2002) 16 (8 HA; 8 LA) 16/0 21.5 (NS) HA: 57.37 (8.9) LA: 33.62 (7.3) TAS-20 None fMRI 3 T 7 × 7 × 7Bird et al. (2010) 18 18/0 35.0 (12.8) 50.3 (14.5) TAS-20 None fMRI 1.5 T 10 × 10 × 10Moriguchi et al. (2007) 30 (16 HA; 14 LA) 5/25 HA: 20.2 (1.0) LA: 20.8 (0.89) HA: 66.1 (4.5) LA: 34.1 (3.7) TAS-20 None fMRI 1.5 T 6 × 6 × 6Noll-Hussong 19 7/12 48.79 (12.25) 44.37 (8.56) TAS-20 None fMRI 3 T 8 × 8 × 8Kano et al. (2007) 45 34/11 22 (2) 47.8 (11) TAS-20 None PET 12 × 12 × 12Strigo et al. (2013) 12 0/12 24.8 (6.1) 37.4 (7.2) TAS-20 None fMRI 3 T 4 × 4 × 4Pouga et al. (2010) 34 (13 HA; 12 LA; 9 intermediate) 34/0 21.26 (2.39) 48.06 (10.85) TAS-20: 47.23 (12.53) BVAQ-B BDI, STAI fMRI 3 T 6 × 6 × 6Mantani et al. (2005) 20 (10 HA; 10 LA) 14/6 HA: 25.9 (3.3) LA: 23.7 (3.0) HA: 61.9 (4.0) LA: 37.9 (3.9) TAS-20 None fMRI 1.5 T 10 × 10 × 10

Abbreviations: BDI: Beck Depression Inventory; BVAQ: Bermond-Vorst Alexithymia Questionnaire; fMRI: functional magnetic resonance imaging; HA: high alexithymia group; LA: low alexithymia group; N: number of participants;NS: not specified; PANAS PA: Positive And Negative Affect Schedule, positive affect; PET: positron emission tomography; SD: standard deviation; STAI: State Trait Anxiety Inventory; TAS: Toronto Alexithymia Scale.

Table 2Overview of task paradigms applied in the included studies.

Study Faces IAPS Other Valence Task Whole brain/ROI

Sad Angry Fear Happy Positive Negative

Reker et al. (2010) X X X X Rating valence of neutral mask CombinedLee et al. (2011) X X X Rating arousal and valence

after each sessionROI

Eichmann et al. (2008) X X X X Rating valence of neutral mask ROIKugel et al. (2008) X X X X Rating valence of neutral mask ROIKano et al. (2003) X X X X X Gender decision Whole brainMériau et al. (2006) X X X Gender or emotional valence

decisionCombined

Heinzel et al. (2010) X X X X X Passive viewing CombinedBerthoz et al. (2002) X X X Passive viewing (rating valence

after scan)ROI

Bird et al. (2010) Empathy (painful stimulation other) X Rating level of unpleasantness ROIMoriguchi et al. (2007) Empathy (pictures of painful situations) X Rating pain intensity of others CombinedNoll-Hussong et al. (2013) Pictures of painful situations X Rating subjective intensity of

painROI

Kano et al. (2007) Pain X Rating visceral sensation Whole brainStrigo et al. (2013) Anticipation of pain and pain X Rating various variables after

scanROI

Pouga et al. (2010) Watching fearful actions X Oddball-task CombinedMantani et al. (2005) Imagining past and future events X X Rating vividness and emotional

intensityCombined

Abbreviations: IAPS: International Affective Picture System; ROI: region of interest.

Page 6

Biobe

mreS

3

lttrsD

ashphrgfin(

4

cfilopncbrwrs

4

mnetiavte(stieoos(iw

J. van der Velde et al. / Neuroscience and

ap and the local maxima reported by the included studies. Theesults revealed that the significant findings were not solely influ-nced by studies implementing the same task paradigms (see Table4).

.2. Positive stimuli

Seven studies were included that examined the neural corre-ates of alexithymia during the processing of positive stimuli. Ofhese studies, three used a ROI-based approach, one investigatedhe whole brain and three applied both methods (for details on usedegions of interest per study, see supplementary Table 1). The totalample consisted of 196 participants (141 male and 55 female).emographic details can be found in Table 1.

For positive stimuli (contrast positive > neutral/baseline) higherlexithymia was associated with increased activation of the dor-al ACC and middle cingulate cortex, in part overlapping theigher activity in the dorsal ACC found during negative emotionalrocessing (Fig. 2A and Table 3C). Lower activation relative toigher levels of alexithymia was found in the precuneus, cuneus,ight posterior and anterior insula and in the left superior temporalyrus (Fig. 2B and Table 3D). Several studies contributed to thesendings which can be found in supplementary Table S4. Seven sig-ificant clusters did not survive the extent threshold of 100 mm3

cf. supplementary Table S3).

. Discussion

The aim of the present meta-analysis was to identify the neuralorrelates of alexithymia during emotion processing. Independentrom valence, alexithymia was associated with higher activationn the dorsal anterior cingulate cortex (ACC) and middle cingu-ate cortex. In addition, valence-specific effects of alexithymia werebserved, with lower activation in the amygdala, fusiform gyrus,remotor areas and the dorsomedial prefrontal cortex (dMPFC) foregative stimuli, and lower activation of the right insula and pre-uneus for positive stimuli. Most of the results were influencedy studies applying various task paradigms suggesting that theseegions show atypical activation patterns in alexithymia across aide range of emotion processing tasks. Visual inspection of the

esults did not indicate a right hemisphere deficit or a left hemi-phere preference in alexithymia.

.1. Valence-independent emotional processing

The finding of increased activation in the dorsal ACC andiddle cingulate cortex in alexithymia was consistent for both

egative and positive stimuli. In their ‘blindfeel’ hypothesis, Lanet al. (1997) proposed that a dysfunction of the ACC underlieshe problems in the conscious experience of emotions observedn alexithymia. The dorsal ACC is involved in various cognitivend emotional processes (Etkin et al., 2011) and increased acti-ation in this area reflects higher cognitive demand during bothhese processes (Koch et al., 2012; Mulert et al., 2005; Taylort al., 2003; Urry et al., 2009). In their reviews, Kano and Fukudo2013) and Moriguchi and Komaki (2013) proposed that whentimuli have a physical context (e.g. pain stimulation), activa-ion in several somatosensory areas (including the dorsal ACC)ncreases in individuals with alexithymia. They suggested that thisnhanced activation might be associated with the amplificationf physical sensations in individuals with alexithymia. However,nly two studies contributing to the finding of increased dor-

al ACC activation applied paradigms including physical stimulie.g. painful visceral stimulation or pictures of hands and feetn painful situations) (Kano et al., 2007; Moriguchi et al., 2007),

hile the other contributing studies (Berthoz et al., 2002; Heinzel

havioral Reviews 37 (2013) 1774–1785 1779

et al., 2010; Mériau et al., 2006; Pouga et al., 2010) did notinclude stimuli with a physical context. These latter studies appliedemotional picture viewing tasks which did not require explicitjudgment of emotional valence (e.g. gender decision, oddball-task, passive viewing) but all required some form of cognitiveprocessing and attention, including passive viewing (Pessoa et al.,2002). Given the involvement of the dorsal ACC in both cogni-tive and emotional processing (Etkin et al., 2011), the increasein dorsal ACC activation may be associated with increased cog-nitive demand in individuals with alexithymia. This hypothesisis supported by an EEG-study of Pollatos and Gramann (2011),in which alexithymia was associated with higher N2 amplitudesduring emotion processing, reflecting higher levels of cognitivedemand (Pollatos and Gramann, 2011). This indicates that indi-viduals with high scores on alexithymia may call upon neuralresources to a stronger extent to attend and understand the emo-tional stimulus.

Although the results of the current meta-analysis showedincreased activation in the dorsal ACC, various studies onalexithymia reported decreased ACC activation during emotionprocessing. Pouga et al. (2010) and Kano and Fukudo (2013) sug-gested that these discrepancies regarding ACC activation may bedue to differences in task paradigms and stimuli between stud-ies. Notably, most studies reporting lower activation in the dorsalACC during emotion processing in association with alexithymia(Kano et al., 2003; Karlsson et al., 2008; Moriguchi et al., 2007)applied tasks which require more cognitive processing of the emo-tional stimuli, such as rating the valence of emotional video clips(Karlsson et al., 2008) or empathizing (Moriguchi et al., 2007). Ifpassive emotional processing already requires additional neuralresources in individuals with alexithymia, tasks which require evenmore emotional awareness might become too strenuous. There-fore, we propose an inverted U-shape of dorsal ACC activationin alexithymia to explain the differences between ACC activa-tion findings in alexithymia literature. Such an inverted U-shapepattern in dorsal ACC activation has previously been shown inpatients with obsessive compulsive disorder (OCD) during the N-back task (Koch et al., 2012). When the task load increased from1-back to 2-back, dorsal ACC activation increased in this group ofpatients compared to healthy controls (i.e. compensation). How-ever, when the task load increased even further (3-back), dorsalACC activation dropped and became lower in comparison to healthycontrols. This inverted U-shape hypothesis might also explain thediscrepancy between the current finding of increased dorsal ACCactivation and the study of McRae et al. (2008) who reportedthat individuals with high trait emotional awareness (which isnegatively associated with alexithymia) showed increased dor-sal ACC activity when viewing high arousing emotional stimuli.Processing of high arousing stimuli requires more allocation ofattention and more complex emotional processing which mightbe too demanding for individuals with alexithymia. However,another possible explanation for this discrepancy between ourfindings and the findings of McRae et al. (2008) could be thatindividuals with high trait emotional awareness are more focusedon high arousing stimuli, which is reflected in increased brainactivation, while the attention of individuals with alexithymiais less directed toward these stimuli for which they have tocompensate.

The current meta-analysis showed that alexithymia is asso-ciated with increased dorsal ACC activation, possibly reflectingeither the amplification of physical sensations or the need tocall upon neural resources to a greater extent, depending on

the type of emotion processing task. Further studies are neededto assess the association between cognitive load and emotionalprocessing in alexithymia in order to verify the inverted U-shapetheory.

Page 7

1780 J. van der Velde et al. / Neuroscience and Biobehavioral Reviews 37 (2013) 1774–1785

Fig. 2. Brain regions with altered activation during emotion processing in alexithymia. (A) Increased activation in alexithymia during the processing of positive (blue) andnegative (red) stimuli. (B) Decreased activation in alexithymia during the processing of positive stimuli. (C) Decreased activation in alexithymia during the processing ofn te of

N

4

4

iipitsepatisfo2u(dttstibiaTg

egative stimuli. Numbers represent the sagittal (x), coronal (y) or axial (z) coordinaeurological Institute template brain.

.2. Valence-specific emotional processing

.2.1. Negative emotional stimuliThe results of this meta-analysis revealed that alexithymia

s associated with decreased activation in several brain regionsnvolved in the processing of negative emotional stimuli. For exam-le, activation in a system important for emotional attention

ncluding the amygdala, fusiform gyrus and middle occipital cor-ex, was lower in individuals with alexithymia. The amygdala is atructure widely known for its role in emotion processing (Phillipst al., 2003) and becomes activated when emotional stimuli areresented. Via a feedback loop through the fusiform gyrus, themygdala influences the occipital cortex directing visual attentionoward emotional stimuli (Vuilleumier, 2005). As a consequence,ndividuals pay more attention to emotional stimuli than to neutraltimuli (Hodsoll et al., 2011). However, when the feedback systemails (e.g. because of amygdala lesions), attention toward emotionalr novel stimuli declines (Anderson and Phelps, 2001; Jacobs et al.,012). Such a decline in emotional attention is also seen in individ-als with alexithymia as they are less distracted by negative wordsMueller et al., 2006) and are impaired in the recall of emotionalistractors (Suslow et al., 2003). Suslow et al. (2003) showed thathis impairment in recall was particularly related to difficulties inhe identification of feelings. An fMRI-study examining the uncon-cious processing of surprised faces reported lower activation inhe fusiform gyrus in alexithymia, also related to the difficulties indentifying feelings (Duan et al., 2010). Moreover, an associationetween decreased activation in the amygdala and difficulties in

dentifying feelings has been suggested by several other reviewss well (Moriguchi and Komaki, 2013; Wingbermühle et al., 2012).herefore, the lower activation found in the amygdala, fusiformyrus and occipital cortex may be related to reduced emotional

each slice. Results are displayed at p < .05 FDR corrected and overlaid on a Montreal

attention in individuals with alexithymia, which might underlietheir difficulties in identifying emotions. This finding might seemcounterintuitive with the finding of increased dorsal ACC activa-tion as discussed above. However, if attention is less automaticallydirected toward emotional stimuli by a decrease in activation of theamygdala, fusiform gyrus and occipital cortex, tasks in which indi-viduals are explicitly instructed to keep their attention toward theemotional stimuli might require activation from a region involvedin top-down attention, such as the dorsal ACC, to a strongerextent.

Besides the above mentioned regions, lower activation inthe dorsal premotor cortex, the superior/inferior parietal cor-tex and supplementary motor area was found in alexithymiaduring the processing of negative emotional stimuli. Althoughnot part of the classic mirror neuron system (MNS) (Rizzolattiand Craighero, 2004), these areas appear to have mirror neu-ron properties (Gazzola and Keysers, 2009; Molenberghs et al.,2012; Mukamel et al., 2010). This means that these areas areactivated during the observation, imitation and execution ofmotor actions (Molenberghs et al., 2009), but also of emotionalfacial expressions (Carr et al., 2003). Therefore, these regions areimportant during emotion processing and empathy (Lamm et al.,2011), the ability to understand other people’s feelings, which isimpaired in individuals with alexithymia (Grynberg et al., 2010;Guttman and Laporte, 2002). Poor empathic skills in individualswith alexithymia might be related to the decreased activation inbrain areas with mirror properties found in this meta-analysis. Ina study directly investigating the MNS in alexithymia, individuals

with and without alexithymia were presented with a classic MNStask (observation and execution of motor actions). Individualsin the high alexithymia group showed increased activation inthe premotor cortex and the superior/inferior parietal cortex

Page 8

J. van der Velde et al. / Neuroscience and Biobehavioral Reviews 37 (2013) 1774–1785 1781

Table 3Summary of significant findings of neural correlates associated with alexithymia during negative or positive emotional processing.

Cluster region Volume(mm3)

Anatomical label (based onthe AAL atlas)

T maxscore

r maxscore

MNI coordinates df

x y z

(A) Positive correlations with alexithymia for the contrast negative > neutral/baselineDorsal anterior cingulate cortex 15,144 R Anterior cingulate gyrus

L Anterior cingulate gyrus21.6416.93

.590

.5672−2

4034

1012

99

No Aal label 16.70 .567 0 30 8 7L Superior frontal gyrus 10.13 .533 0 46 20 8R Middle cingulate gyrus 8.77 .523 14 28 32 8L Middle cingulate gyrus 4.20 .469 0 24 34 8

R Middle temporal gyrus 128 R Middle temporal gyrus 4.58 .498 60 −42 0 7

(B) Negative correlations with alexithymia for the contrast negative > neutral/baselineL Premotor cortex 5320 No Aal label −17.65 −.515 −26 −8 46 7

L Precentral gyrus −16.60 −.515 −26 −8 48 8L Superior frontal gyrus −16.44 −.511 −24 −10 52 8L Middle frontal gyrus −16.44 −.511 −26 −8 50 8

L Fusiform gyrus 18,024 L Fusiform gyrus −14.21 −.471 −42 −58 −22 8L Fusiform gyrus −9.36 −.461 −36 −30 −12 7L Inferior temporal gyrus −9.36 −.461 −44 −36 −18 7L Cerebellum −8.42 −.517 −30 −40 −26 7L Cerebellum −8.42 −.517 −30 −40 −28 7L Hippocampus −5.61 −.463 −36 −16 −16 7L Middle temporal gyrus −5.61 −.463 −46 −10 −20 7L Lingual gyrus −4.71 −.473 −22 −54 −12 7L Cerebellum −4.43 −.464 −36 −42 −38 7L Cerebellum −4.43 −.464 −38 −38 −40 7L Cerebellum −4.43 −.464 −34 −44 −42 7L Cerebellum −4.43 −.464 −26 −40 −40 7

R Amygdala 1944 R Amygdala −12.48 −.410 24 0 −20 9

L Suppl motor area 176 L Supplementary motorarea

−8.23 −.473 −2 0 50 8

L Middle cingulate gyrus −8.22 −.475 0 −2 46 8

L Dorsomedial prefrontal cortex 6488 L Medial superior frontalgyrus

−7.43 −.416 −8 44 30 8

L Superior frontal gyrus −6.54 −.417 −10 56 42 8L Middle frontal gyrus −5.54 −.418 −18 50 28 8

R Suppl motor area 2048 R Superior frontal gyrus −6.77 −.458 20 −4 54 7R Superior frontal gyrus −6.05 −.461 22 −4 56 8R Suppl motor area −6.05 −.461 12 −4 54 8

L Amygdala 1720 L Amygdala −6.68 −.367 −28 −2 −26 9

L Superior parietal gyrus 4752 L Superior parietal gyrusL Superior parietal gyrus

−6.21−6.21

−.484−.484

−24−20

−54−62

4240

77

L Inferior parietal gyrus −6.21 −.484 −22 −50 54 7L Precuneus −5.96 −.481 −12 −66 48 7L Middle occipital gyrus −5.53 −.452 −22 −60 40 7

L Middle occipital gyrus 4000 No Aal label −6.21 −.432 −40 −94 −2 7L Middle occipital gyrus −6.21 −.432 −36 −94 −2 7L Inferior occipital gyrus −6.21 −.432 −32 −84 0 7

R Fusiform gyrus 2216 R Fusiform gyrus −5.97 −.440 44 −44 −22 8

L Middle occipital gyrus 224 L Middle occipital gyrus −5.87 −.478 −24 −86 16 7

R Putamen 216 No Aal label −5.11 −.453 32 −12 14 7R Putamen −4.93 −.422 30 −10 14 8

(C) Positive correlations with alexithymia for the contrast positive > neutral/baselineDorsal anterior cingulate/Middlecingulate cortex

3040 L Anterior cingulate gyrusL Anterior cingulate gyrusR Anterior cingulate gyrus

14.8314.3313.15

.516

.516

.518

−6−42

121210

262426

444

L Middle cingulate gyrus 13.15 .518 0 14 34 4R Middle cingulate gyrus 13.15 .518 2 10 32 4

(D) Negative correlations with alexithymia for the contrast positive > neutral/baselinePrecuneus/Cuneus 7400 L Cuneus

L Precuneus−27.96−27.96

−.550−.550

−10−10

−68−58

2628

33

R Precuneus −21.18 −.550 4 −70 34 3L Calcarine −14.42 −.546 −4 −70 22 3R Calcarine −11.59 −.544 2 −62 20 3R Cuneus −10.94 −.543 4 −74 24 3L Posterior cingulate gyrus −9.71 −.544 −2 −52 26 4

Page 9

1782 J. van der Velde et al. / Neuroscience and Biobehavioral Reviews 37 (2013) 1774–1785

Table 3 (Continued)

Cluster region Volume(mm3)

Anatomical label (based onthe AAL atlas)

T maxscore

r maxscore

MNI coordinates df

x y z

R Precuneus −7.74 −.541 4 −56 32 4R Middle Cingulate gyrus −7.69 −.540 2 −52 34 4

L Superior temporal gyrus 256 No Aal labelL Superior temporal gyrus

−16.42−16.42

−.586−.586

−44−46

−12−14

−16−10

33

L Middle temporal gyrus −16.42 −.586 −46 −12 −16 3

R Anterior insula 4744 R Inferior frontal gyrus,opercular

−16.26 −.577 50 10 4 4

R Rolandic operculum −16.26 −.577 50 4 2 4R Insula −16.26 −.577 40 4 −6 4R Insula −15.44 −.575 38 4 −6 3R Putamen −15.44 −.575 34 4 −4 3

R Posterior insula 2016 R Insula −10.30 −.612 36 −14 10 4R Rolandic operculum −9.05 −.617 50 −2 4 4R Insula −8.55 −.614 38 −10 2 3Heschl’s gyrus −8.55 −.614 44 −14 6 3

us

A : right

ieiatgmdpr

ap(vdaGiairit(tirrsadariai(tarBad

Superior temporal gyr

bbreviations: Aal: anatomical automatic labeling; df: degrees of freedom; L: left; R

n comparison to the low alexithymia group, while performingqually well on the MNS task (Moriguchi et al., 2009). Accord-ng to the authors, the increased neural response represented

compensation mechanism. Furthermore, they suggested thathis compensation mechanism might fail during empathy andeneral emotion processing in alexithymia. The results of thiseta-analysis are in line with this interpretation and suggest that

ecreased activation in dorsal premotor cortex, superior/inferiorarietal cortex and supplementary motor areas identified may beelated to poorer empathic abilities in alexithymia.

The dorsomedial prefrontal cortex (dMPFC), found to be lessctivated in alexithymia, is involved in various cognitive emotionalrocesses, including social cognition and emotion regulationAmodio and Frith, 2006; Ochsner and Gross, 2005). Lower acti-ation in the dMPFC in alexithymia has previously been shownuring a social cognition task (Moriguchi et al., 2006). Furthermore,n EEG-study on emotion regulation in alexithymia (Pollatos andramann, 2012) reported reduced P3 and slow wave amplitudes

n low alexithymic individuals during reappraisal, but not in highlexithymic individuals, indicating that the top-down regulation inndividuals with alexithymia is compromised. Although emotionegulation was not investigated explicitly in the studies includedn this meta-analysis, it is conceivable that implicit regulationakes place when individuals are presented with negative stimuliGyurak et al., 2011). Impairments in emotion regulation arehought to be one of the core deficits in alexithymia. For example,ndividuals with alexithymia report difficulties with emotionegulation (Venta et al., 2012) and make less use of cognitiveeappraisal (Swart et al., 2009), the reinterpretation of emotionaltimuli in such a way that it reduces their emotional impact (Johnnd Gross, 2004). During reappraisal, prefrontal areas including theMPFC become activated (Diekhof et al., 2011) and down-regulatectivation in limbic regions (Ochsner and Gross, 2005). Indeed,eappraisal success has been found to correlate with activationn the dMPFC (Wager et al., 2008). Hence, reduced activity in thisrea might lead to less successful emotion regulation in alex-thymia, resulting in the experience of more negative emotionsDe Gucht et al., 2004; Yelsma, 2007). However, one might arguehat individuals with alexithymia do not need to down-regulatemygdala activation, since activation in this area is already low in

esponse to negative emotional stimuli (see Section 3.1). Vorst andermond (2001) divided the alexithymia construct into a cognitivend affective dimension (Vorst and Bermond, 2001). The cognitiveimension refers to the processing of emotions at a cognitive

−8.55 −.614 44 −8 −4 3

.

level, such as identifying, analyzing and verbalizing emotions. Incontrast, the affective dimension refers to the level of subjectiveemotional experience. It might be that individuals with highemotional experience but low cognitive processing capacities areunable to down-regulate their negative affect through reappraisal,resulting in an even higher experience of negative emotions. Incontrast, individuals with both low emotional experience and lowcognitive emotional processing capacities might not need to down-regulate their emotions, since emotional experience is already low.Unfortunately, due to the low number of neuroimaging studiesexamining the affective dimension, the variation on this dimensioncould not be taken into account in the current meta-analysis. Futureresearch is needed to investigate the neural correlates of the twoalexithymia dimensions, especially during emotion regulation, toexamine this hypothesis and the role of the dMPFC in alexithymia.

4.2.2. Positive emotional stimuliThe results of the analysis on emotion processing of positive

stimuli in alexithymia should be considered preliminary due tothe low number of included whole brain studies. Nonetheless, theresults revealed lower activation in the right anterior and pos-terior insula and the precuneus. The insula, especially the right,is considered to be a neuroanatomical substrate for emotionalawareness (Critchley et al., 2004). It has been shown that the rightanterior insula is functionally connected to the precuneus duringthe evaluation of emotional state (Terasawa et al., 2013). Accord-ing to the authors, emotional awareness arises from transforminginteroceptive information (insula) into a subjective emotional state(precuneus). Several other studies have underlined the involve-ment of the insula and precuneus in emotional awareness (forreview see Tsuchiya and Adolphs, 2007). This holds for the expe-rience of both positive and negative emotions (Craig, 2009; Habelet al., 2005). Therefore, decreased activation in these areas in alex-ithymia suggests reduced emotional awareness of positive affect.Bagby and Parker (1997) described individuals with alexithymia tohave “a limited capacity to experience positive emotions, such asjoy, happiness and love”. Since then, several studies have reportedan association between alexithymia and lower positive affect (DeGucht et al., 2004; Yelsma, 2007) or a lower tendency to experiencepositive emotions (Luminet et al., 1999), while negative affect is

increased (De Gucht et al., 2004; Yelsma, 2007). Therefore, reducedemotional awareness during positive emotion processing mightunderlie the lower positive affect that individuals with alexithymiaexperience.

Page 10

Biobe

4

oaaisrisspsb

asaa2bsaas

ptsti

rRdrerneh2dd

irht(smtbc

aioadtrio

J. van der Velde et al. / Neuroscience and

.3. Methodological issues and limitations

The studies included in this meta-analysis differed in the meth-ds that were applied. Although part of these differences, suchs sample size, smoothing kernel and statistical threshold areccounted for by the PCM method (Costafreda, 2012), differencesn task paradigms, included covariates and variation in alexithymiacores introduce unavoidable heterogeneity which might haveeduced the sensitivity. Furthermore, the modest number of stud-es included in the meta-analysis, especially regarding the positivetimuli might have limited the statistical power to detect moreubtle effects of alexithymia on brain activation during emotionrocessing. Therefore, the results regarding the positive stimulihould be considered preliminary and no definite conclusions cane drawn on brain regions which did not reach significance.

By applying the new parametric coordinate-based meta-nalysis (PCM) method (Costafreda, 2012), we were able to includetudies with a ROI approach as well as studies with a whole brainpproach. Including ROI studies can provide valuable informationnd increases sensitivity for certain brain areas (Groenewold et al.,012). However, by including ROI-studies, the power for certainrain areas increases, which might slightly increase the chance ofignificant findings in these areas in comparison to areas whichre not examined with a ROI approach. In addition, it should becknowledged that strategies of ROI selection may differ betweentudies.

One hypothesis regarding the lateralization of emotionrocessing in alexithymia could not be statistically examined inhe current meta-analysis. Future studies should examine a hemi-phere by task interaction in alexithymia to gain more insight intohe right hemisphere deficit or left hemisphere preference in alex-thymia.

Another limitation of the current meta-analysis is that only brainegions associated with cognitive alexithymia were examined.ecently, studies have indicated that the cognitive and affectiveimension of alexithymia might have different underlying neu-al correlates (Goerlich et al., 2012; Larsen et al., 2003; Pougat al., 2010). Unfortunately, this could not be assessed in the cur-ent meta-analysis, since only one included study examined theeural correlates of the affective alexithymia dimension (Pougat al., 2010). Because the two different dimensions have beenypothesized to underlie different pathologies (Moormann et al.,008), further research on the neural correlates of these separateimensions might give valuable insights into alexithymia and theevelopment of psychopathology.

All studies included in the current meta-analysis assessed alex-thymia through self-report measures. These measurements rely oneflecting one’s own emotions which is limited in individuals withigh scores on alexithymia. Therefore, several authors proposedo examine alexithymia through observer-based questionnairesLundh et al., 2002; Suslow et al., 2001). Unfortunately, only onetudy (Moriguchi et al., 2007) included an observer rated measure-ent to confirm the presence of alexithymia. This limits our results

o self-reported alexithymia. We encourage future studies to com-ine self-reports with observer-rated measurements to get a moreomprehensive measure of alexithymia.

Furthermore, alexithymia is related to feelings of depression andnxiety (Hendryx et al., 1991). Unfortunately, only half of the stud-es included in the current meta-analysis controlled for the effectsf depression and anxiety (by including these factors as a covari-te or by excluding individuals with high scores on anxiety and/orepression). Therefore, it is not possible to completely eliminate

he effects of these mood states on the current findings. Futureesearch should further examine the relationship between alex-thymia, depression and anxiety and the effect of this relationshipn brain activation.

havioral Reviews 37 (2013) 1774–1785 1783

As pointed out in the introduction, recent studies stronglysuggest to conceptualize alexithymia as a dimensional construct.Most studies included in this meta-analysis adopted this view anddefined alexithymia as a continuous personality dimension. How-ever, cut-off scores for both the TAS-20 (≥61, Bagby et al., 1994) andthe BVAQ-B (≥56, Deborde et al., 2008) have been formed to indi-cate clinical meaningful levels of alexithymia. Only a few studiesincluded in this meta-analysis (Kano et al., 2003, 2007; Moriguchiet al., 2007) included a substantial number of individuals reachingthe clinically relevant alexithymia scores. Therefore, it remains anopen question whether the currently identified neural correlatesgeneralize to clinical alexithymia.

5. Conclusion

This meta-analysis identified several neural correlates of alex-ithymia. Specifically, in alexithymia the amygdala, a key node ofthe emotional perception/attention system, is less activated duringnegative emotional processing. While performing simple emo-tional processing tasks, such a deficit might be compensated byhigher activation in the dorsal anterior and middle cingulate cor-tex. However, this compensation mechanism might fail when taskdifficulty increases (i.e. when individuals with alexithymia showperformance reductions). When presented with physical emotionalstimuli, increased activation in the dorsal anterior cingulate cor-tex might reflect hypersensitivity toward the physical informationof these stimuli. Furthermore, lower activation was found in MNSrelated brain regions and the dorsomedial prefrontal cortex duringthe processing of negative emotional stimuli. This may underlie theimpaired empathic capabilities and the difficulties in emotion reg-ulation observed in alexithymic individuals. A decrease in insulaand precuneus activation was identified during the processing ofpositive stimuli, possibly reflecting reduced emotional awareness,which may help explain the lower positive affect experienced bythese individuals.

Acknowledgements

SGC is supported by a National Institute of Health Research(NIHR) Academic Clinical Lectureship. PH is supported by the NIHRBiomedical Research Unit in Dementia at South London and Maud-sley, NHS Foundation Trust (SLaM) and the Institute of Psychiatry,King’s College Lodon. AA is supported in part by a VICI grant fromN.W.O., grant nr. 435-11-004.

Appendix A. Supplementary data

Supplementary data associated with this article canbe found, in the online version, at http://dx.doi.org/10.1016/j.neubiorev.2013.07.008.

References

Adolphs, R., Damasio, H., Tranel, D., Cooper, G., Damasio, A.R., 2000. A role forsomatosensory cortices in the visual recognition of emotion as revealed bythree-dimensional lesion mapping. Journal of Neuroscience 20, 2683–2690.

Aleman, A., 2005. Feelings you can’t imagine: towards a cognitive neuroscience ofalexithymia. Trends in Cognitive Sciences 9, 553–555.

Amodio, D., Frith, C., 2006. Meeting of minds: the medial frontal cortex and socialcognition. Nature Reviews Neuroscience 7, 268–277.

Anderson, A.K., Phelps, E.A., 2001. Lesions of the human amygdala impair enhancedperception of emotionally salient events. Nature 411, 305–309.

Bagby, R.M., Parker, J.D., Taylor, G.J., 1994. The twenty-item Toronto Alexithymia

Scale-I. Item selection and cross-validation of the factor structure. Journal ofPsychosomatic Research 38, 23–32.

Bagby, R.M., Parker, J.D.A., 1997. Affect dysregulation and Alexithymia. In: Taylor,G.J., Bagby, R.M., Parker, J.D.A. (Eds.), Disorders of Affect Regulation. CambridgeUniversity Press, Cambridge, p. 31.

Page 11

1 Biobe

B

B

B

B

B

B

C

C

C

C

D

D

D

D

D

E

E

F

G

G

G

G

G

G

G

G

G

H

H

784 J. van der Velde et al. / Neuroscience and

ermond, B., Bleys, J.W., Stoffels, E.J., 2005. Left hemispheric preference andalexithymia: a neuropsychological investigation. Cognition and Emotion 19,151–160.

ermond, B., Vorst, H.C.M., Moormann, P., 2006. Cognitive neuropsychology of alex-ithymia: implications for personality typology. Cognitive Neuropsychiatry 11,332–360.

erthoz, S., Artiges, E., Van De Moortele, P., Poline, J., Rouquette, S., Consoli, S.,Martinot, J., 2002. Effect of impaired recognition and expression of emotionson frontocingulate cortices: an fMRI study of men with alexithymia. AmericanJournal of Psychiatry 159, 961–967.

ird, G., Silani, G., Brindley, R., White, S., Frith, U., Singer, T., 2010. Empathic brainresponses in insula are modulated by levels of alexithymia but not autism. Brain133, 1515–1525.

rett, M., Christoff, K., Cusack, R., Lancaster, J., 2001. Using the Talairach atlas withthe MNI template RID A-2231-2010. Neuroimage 13, S85.

uchanan, D.C., Waterhouse, G.J., West, S.C., 1980. A proposed neurophysiologicalbasis of alexithymia. Psychotherapy and Psychosomatics 34, 248–255.

arr, L., Iacoboni, M., Dubeau, M., Mazziotta, J., Lenzi, G., 2003. Neural mechanismsof empathy in humans: a relay from neural systems for imitation to limbic areas.Proceedings of the National Academy of Sciences of the United States of America100, 5497–5502.

ostafreda, S., 2012. Parametric coordinate-based meta-analysis: valid effect sizemeta-analysis of studies with differing statistical thresholds. Journal of Neuro-science Methods 210, 291–300.

raig, A.D., 2009. How do you feel – now? The anterior insula and human awareness.Nature Reviews Neuroscience 10, 59–70.

ritchley, H., Wiens, S., Rotshtein, P., Ohman, A., Dolan, R., 2004. Neural systemssupporting interoceptive awareness. Nature Neuroscience 7, 189–195.

’Argembeau, A., Xue, G., Lu, Z., Van der Linden, M., Bechara, A., 2008. Neural cor-relates of envisioning emotional events in the near and far future. Neuroimage40, 398–407.

e Gucht, V., Fischler, B., Heiser, W., 2004. Neuroticism, alexithymia, negative affect,and positive affect as determinants of medically unexplained symptoms. Per-sonality and Individual Differences 36, 1655–1667.

eborde, A., Berthoz, S., Wallier, J.M., Fermanian, J., Falissard, B., Jeammet, P., Corcos,M., 2008. The Bermond-Vorst Alexithymia Questionnaire cutoff scores: a studyin eating-disordered and control subjects. Psychopathology 41, 43–49.

iekhof, E., Geier, K., Falkai, P., Gruber, O., 2011. Fear is only as deep as the mindallows: a coordinate-based meta-analysis of neuroimaging studies on the regu-lation of negative affect. Neuroimage 58, 275–285.

uan, X., Dai, Q., Gong, Q., Chen, H., 2010. Neural mechanism of unconscious per-ception of surprised facial expression. Neuroimage 52, 401–407.

ichmann, M., Kugel, H., Suslow, T., 2008. Difficulty identifying feelings and auto-matic activation in the fusiform gyrus in response to facial emotion. Perceptualand Motor Skills 107, 915–922.

tkin, A., Egner, T., Kalisch, R., 2011. Emotional processing in anterior cingulate andmedial prefrontal cortex. Trends in Cognitive Sciences 15, 85–93.

antini-Hauwel, C., Boudouhka, A.H., Aciszewski, T., 2012. Adult attachment andemotional awareness impairment: a multimethod assessment. SocioaffectiveNeuroscience and Psychology 2, 10744.

azzola, V., Keysers, C., 2009. The observation and execution of actions share motorand somatosensory voxels in all tested subjects: single-subject analyses ofunsmoothed fMRI data. Cerebral Cortex 19, 1239–1255.

lascher, J., Adolphs, R., Damasio, H., Bechara, A., Rudrauf, D., Calamia, M., Paul,L., Tranel, D., 2012. Lesion mapping of cognitive control and value-based deci-sion making in the prefrontal cortex. Proceedings of the National Academy ofSciences of the United States of America 109, 14681–14686.

oerlich, K., Aleman, A., Martens, S., 2012. The sound of feelings: electrophysiologicalresponses to emotional speech in alexithymia. PLoS ONE 7, e36951.

oerlich, K.S., Witteman, J., Schiller, N.O., van Heuven, V.J., Aleman, A., Martens,S., 2013. Blunted Feelings: Alexithymia is associated with a diminished neu-ral response to speech prosody. Social Cognitive and Affective Neuroscience,epub.

roenewold, N., Opmeer, E., de Jonge, P., Aleman, A., Costafreda, S., 2012. Emotionalvalence modulates brain functional abnormalities in depression: Evidence froma meta-analysis of fMRI studies. Neuroscience and Biobehavioral Reviews 37,152–163.

rynberg, D., Luminet, O., Corneille, O., Grèzes, J., Berthoz, S., 2010. Alexithymia in theinterpersonal domain: A general deficit of empathy? Personality and IndividualDifferences 49, 845–850.

rynberg, D., Chang, B., Corneille, O., Maurage, P., Vermeulen, N., Berthoz, S., Luminet,O., 2012. Alexithymia and the processing of emotional facial expressions (EFEs):systematic review, unanswered questions and further perspectives. PLoS ONE7, e42429.

uttman, H., Laporte, L., 2002. Alexithymia, empathy, and psychological symptomsin a family context. Comprehensive Psychiatry 43, 448–455.

yurak, A., Gross, J., Etkin, A., 2011. Explicit and implicit emotion regulation: a dual-process framework. Cognition and Emotion 25, 400–412.

abel, U., Klein, M., Kellermann, T., Shah, N.J., Schneider, F., 2005. Same or different?Neural correlates of happy and sad mood in healthy males. Neuroimage 26,206–214.

einzel, A., Schfer, R., Mller, H., Schieffer, A., Ingenhag, A., Eickhoff, S., Northoff,G., Franz, M., Hautzel, H., 2010. Increased activation of the supragenual ante-rior cingulate cortex during visual emotional processing in male subjects withhigh degrees of alexithymia: an event-related fMRI study. Psychotherapy andPsychosomatics 79, 363–370.

havioral Reviews 37 (2013) 1774–1785

Hendryx, M.S., Haviland, M.G., Shaw, D.G., 1991. Dimensions of alexithymia and theirrelationships to anxiety and depression. Journal of Personality Assessment 56,227–237.

Hintikka, J., Honkalampi, K., Koivumaa Honkanen, H., Antikainen, R., Tanskanen,A., Haatainen, K., Viinamki, H., 2004. Alexithymia and suicidal ideation: a 12-month follow-up study in a general population. Comprehensive Psychiatry 45,340–345.

Hodsoll, S., Viding, E., Lavie, N., 2011. Attentional capture by irrelevant emotionaldistractor faces. Emotion 11, 346–353.

Honkalampi, K., Hintikka, J., Tanskanen, A., Lehtonen, J., Viinamäki, H., 2000. Depres-sion is strongly associated with alexithymia in the general population. Journalof Psychosomatic Research 48, 99–104.

Jacobs, R.H., Renken, R., Aleman, A., Cornelissen, F.W., 2012. The amygdala, top-downeffects, and selective attention to features. Neuroscience and BiobehavioralReviews 36, 2069–2084.

John, O.P., Gross, J.J., 2004. Healthy and unhealthy emotion regulation: personalityprocesses, individual differences, and life span development. Journal of Person-ality 72, 1301–1333.

Kano, M., Fukudo, S., 2013. The alexithymic brain: the neural pathways linkingalexithymia to physical disorders. Biopsychosocial medicine 7, 1.

Kano, M., Fukudo, S., Gyoba, J., Kamachi, M., Tagawa, M., Mochizuki, H., Itoh, M.,Hongo, M., Yanai, K., 2003. Specific brain processing of facial expressions inpeople with alexithymia: an H2 15O-PET study. Brain 126, 1474–1484.

Kano, M., Hamaguchi, T., Itoh, M., Yanai, K., Fukudo, S., 2007. Correlation betweenalexithymia and hypersensitivity to visceral stimulation in human. Pain 132,252–263.

Karlsson, H., Näätänen, P., Stenman, H., 2008. Cortical activation in alexithymia as aresponse to emotional stimuli. British Journal of Psychiatry 192, 32–38.

Koch, K., Wagner, G., Schachtzabel, C., Peikert, G., Schultz, C.C., Sauer, H., Schlsser, R.,2012. Aberrant anterior cingulate activation in obsessive-compulsive disorderis related to task complexity. Neuropsychologia 50, 958–964.

Kugel, H., Eichmann, M., Dannlowski, U., Ohrmann, P., Bauer, J., Arolt, V., Heindel,W., Suslow, T., 2008. Alexithymic features and automatic amygdala reactivity tofacial emotion. Neuroscience Letters 435, 40–44.

Lamm, C., Decety, J., Singer, T., 2011. Meta-analytic evidence for common and distinctneural networks associated with directly experienced pain and empathy forpain. Neuroimage 54, 2492–2502.

Lane, R.D., Sechrest, L., Riedel, R., 1998. Sociodemographic correlates of alexithymia.Comprehensive Psychiatry 39, 377–385.

Lane, R., Ahern, G.L., Schwartz, G.E., Kaszniak, A.W., 1997. Is alexithymia the emo-tional equivalent of blindsight? Biological Psychiatry 42, 834–844.

Larsen, J., Brand, N., Bermond, B., Hijman, R., 2003. Cognitive and emotionalcharacteristics of alexithymia: a review of neurobiological studies. Journal ofPsychosomatic Research 54, 533–541.

Lee, B.-T., Lee, H.-Y., Park, S.-A., Lim, S.-Y., Tae, W.S., Lee, M.-S., Joe, S.-H., Jung,I.-K., Ham, B.-J., 2011. Neural substrates of affective face recognition in alex-ithymia: a functional magnetic resonance imaging study. Neuropsychobiology63, 119–124.

Luminet, O., Bagby, R.M., Wagner, H., Taylor, G.J., Parker, J.D., 1999. Relation betweenalexithymia and the five-factor model of personality: a facet-level analysis. Jour-nal of Personality Assessment 73, 345–358.

Luminet, O., Vermeulen, N., Demaret, C., Taylor, G.J., Bagby, R.M., 2006. Alexithymiaand levels of processing: evidence for an overal deficit in remembering emotionwords. Journal of Research in Personality 40, 713–733.

Lumley, M., Gustavson, B., Partridge, R.T., Labouvie Vief, G., 2005. Assessing alex-ithymia and related emotional ability constructs using multiple methods:interrelationships among measures. Emotion 5, 329–342.

Lundh, L., Johnsson, A., Sundqvist, K., Olsson, H., 2002. Alexithymia, memory ofemotion, emotional awareness, and perfectionism. Emotion 2, 361–379.

Maddock, R.J., 1999. The retrosplenial cortex and emotion: new insights from func-tional neuroimaging of the human brain. Trends in Neurosciences 22, 310–316.

Mantani, T., Okamoto, Y., Shirao, N., Okada, G., Yamawaki, S., 2005. Reduced activa-tion of posterior cingulate cortex during imagery in subjects with high degreesof alexithymia: a functional magnetic resonance imaging study. Biological Psy-chiatry 57, 982–990.

Mattila, A.K., Keefer, K.V., Taylor, G.J., Joukamaa, M., Jula, A., Parker, J.D.A., Bagby,R.M., 2010. Taxometric analysis of alexithymia in a general population samplefrom Finland. Personality and Individual Differences 49, 216–221.

Mattila, A., Poutanen, O., Koivisto, A., Salokangas, R.K.R., Joukamaa, M., 2007. Alex-ithymia and life satisfaction in primary healthcare patients. Psychosomatics 48,523–529.

McRae, K., Reiman, E., Fort, C., Chen, K., Lane, R., 2008. Association between traitemotional awareness and dorsal anterior cingulate activity during emotion isarousal-dependent. Neuroimage 41, 648–655.

Mériau, K., Wartenburger, I., Kazzer, P., Prehn, K., Lammers, C.H., Van der Meer, E.,Villringer, A., Heekeren, H.R., 2006. A neural network reflecting individual dif-ferences in cognitive processing of emotions during perceptual decision making.Neuroimage 33, 1016–1027.

Molenberghs, P., Cunnington, R., Mattingley, J., 2012. Brain regions with mirrorproperties: a meta-analysis of 125 human fMRI studies. Neuroscience and Biobe-havioral Reviews 36, 341–349.

Molenberghs, P., Cunnington, R., Mattingley, J., 2009. Is the mirror neuron systeminvolved in imitation? A short review and meta-analysis. Neuroscience andBiobehavioral Reviews 33, 975–980.

Moormann, P.P., Bermond, B., Vorst, H.C.M., Bloemendaal, A.F.T., Teijn, S.M., Rood, L.,2008. New avenues in alexithymia research: the creation of alexithymia types.

Page 12

Biobe

M

M

M

M

M

M

M

N

O

P

P

P

P

P

P

P

R

R

R

R

S

J. van der Velde et al. / Neuroscience and

In: Denollet, J., Vingerhoets, A.J.J.M., Nyklicek, T. (Eds.), Emotion Regulation:Conceptual and Clinical Issues. Springer, New York, pp. 27–42.

origuchi, Y., Ohnishi, T., Lane, R.D., Maeda, M., Mori, T., Nemoto, K., Matsuda, H.,Komaki, G., 2006. Impaired self-awareness and theory of mind: an fMRI studyof mentalizing in alexithymia. Neuroimage 32, 1472–1482.

origuchi, Y., Decety, J., Ohnishi, T., Maeda, M., Mori, T., Nemoto, K., Matsuda, H.,Komaki, G., 2007. Empathy and judging other’s pain: an fMRI study of alex-ithymia. Cerebral Cortex 17, 2223–2234.

origuchi, Y., Komaki, G., 2013. Neuroimaging studies of alexithymia: physical,affective, and social perspectives. Biopsychosocial Medicine 7, 8–8.

origuchi, Y., Ohnishi, T., Decety, J., Hirakata, M., Maeda, M., Matsuda, H., Komaki,G., 2009. The human mirror neuron system in a population with deficient self-awareness: an fMRI study in alexithymia. Human Brain Mapping 30, 2063–2076.

ueller, J., Alpers, G., Reim, N., 2006. Dissociation of rated emotional valence andStroop interference in observer-rated alexithymia. Journal of PsychosomaticResearch 61, 261–269.

ukamel, R., Ekstrom, A., Kaplan, J., Iacoboni, M., Fried, I., 2010. Single-neuronresponses in humans during execution and observation of actions. Current Biol-ogy 20, 750–756.

ulert, C., Menzinger, E., Leicht, G., Pogarell, O., Hegerl, U., 2005. Evidence for a closerelationship between conscious effort and anterior cingulate cortex activity.International Journal of Psychophysiology 56, 65–80.

oll-Hussong, M., Otti, A., Wohlschlaeger, A.M., Zimmer, C., Henningsen, P., Lahman,C., Ronel, J., Subic-Wrana, C., Lane, R.D., Decety, J., Guendel, H., 2013. Neuralcorrelates of deficits in pain-related affective meaning construction in patientswith chronic pain disorder. Psychosomatic Medicine 75, 124–136.

chsner, K.N., Gross, J.J., 2005. The cognitive control of emotion. Trends in CognitiveSciences 9, 242–249.

arker, J.D., Taylor, G.J., Bagby, R.M., Acklin, M.W., 1993. Alexithymia in panic dis-order and simple phobia: a comparative study. American Journal of Psychiatry150, 1105–1107.

arker, J.D.A., Keefer, K., Taylor, G., Bagby, R.M., 2008. Latent structure of the alex-ithymia construct: a taxometric investigation. Psychological Assessment 20,385–396.

essoa, L., McKenna, M., Gutierrez, E., Ungerleider, L.G., 2002. Neural processingof emotional faces requires attention. Proceedings of the National Academy ofSciences of the United States of America 99, 11458–11463.

hillips, M., Drevets, W., Rauch, S., Lane, R., 2003. Neurobiology of emotion percep-tion I: the neural basis of normal emotion perception. Biological Psychiatry 54,504–514.

ollatos, O., Gramann, K., 2012. Attenuated modulation of brain activity accompaniesemotion regulation deficits in alexithymia. Psychophysiology 49, 651–658.

ollatos, O., Gramann, K., 2011. Electrophysiological evidence of early processingdeficits in alexithymia. Biological Psychology 87, 113–121.

ouga, L., Berthoz, S., de Gelder, B., Grzes, J., 2010. Individual differences in socioaffec-tive skills influence the neural bases of fear processing: the case of alexithymia.Human Brain Mapping 31, 1469–1481.

adua, J., Mataix Cols, D., Phillips, M.L., El Hage, W., Kronhaus, D.M., Cardoner, N.,Surguladze, S., 2012. A new meta-analytic method for neuroimaging studies thatcombines reported peak coordinates and statistical parametric maps. Europeanpsychiatry 27, 605–611.

eker, M., Ohrmann, P., Rauch, A., Kugel, H., Bauer, J., Dannlowski, U., Arolt, V.,Heindel, W., Suslow, T., 2010. Individual differences in alexithymia and brainresponse to masked emotion faces. Cortex 46, 658–667.

izzolatti, G., Craighero, L., 2004. The mirror-neuron system. Annual Review of Neu-roscience 27, 169–192.

ogers, R., Ramnani, N., Mackay, C., Wilson, J., Jezzard, P., Carter, C., Smith, S., 2004.Distinct portions of anterior cingulate cortex and medial prefrontal cortex are

activated by reward processing in separable phases of decision-making cogni-tion. Biological Psychiatry 55, 594–602.

alminen, J.K., Saarijärvi, S., Toikka, T., Kauhanen, J., 1999. Prevalence of alexithymiaand its association with sociodemographic variables in the general populationof Finland. Journal of Psychosomatic Research 46, 75–82.

havioral Reviews 37 (2013) 1774–1785 1785

Sifneos, P.E., 1973. The prevalence of ‘alexithymic’ characteristics in psychosomaticpatients. Psychotherapy and Psychosomatics 22, 255–262.

Strigo, I.A., Matthews, S.C., Simmons, A.N., Oberndorfer, T., Klabunde, M.,Reinhardt, L.E., Kaye, W.H., 2013. Altered insula activation during painanticipation in individuals recovered form Anorexia Nervosa: EVIDENCE ofinteroceptive dysregulation. International Journal of Eating Disorders 46,23–33.

Suslow, T., Kersting, A., Ohrmann, P., Arolt, V., 2001. A critique of the construct“alexithymia” and its measurement – the weakness of self-report and the oppor-tunities of an objective assessment approach. Zeitschrift fur PsychosomatischeMedizin und Psychotherapie 47, 153–166.

Suslow, T., Kersting, A., Arolt, V., 2003. Alexithymia and incidental learning of emo-tional words. Psychological Reports 93, 1003–1012.

Swart, M., Kortekaas, R., Aleman, A., 2009. Dealing with feelings: characterizationof trait alexithymia on emotion regulation strategies and cognitive-emotionalprocessing. PLoS ONE 4, e5751.

Taylor, G.J., 2000. Recent developments in alexithymia theory and research. Cana-dian Journal of Psychiatry 45, 134–142.

Taylor, G.J., Bagby, R.M., Parker, J.D., 1997. Disorders of Affect Regulation:Alexithymia in Medical and Psychiatric Illness. Cambridge University Press,Cambridge.

Taylor, G., Bagby, R.M., 2004. New trends in alexithymia research. Psychotherapyand Psychosomatics 73, 68–77.

Taylor, S., Phan, K.L., Decker, L., Liberzon, I., 2003. Subjective rating of emotionallysalient stimuli modulates neural activity. Neuroimage 18, 650–659.

Terasawa, Y., Fukushima, H., Umeda, S., 2013. How does interoceptive awarenessinteract with the subjective experience of emotion? An fMRI Study. Human BrainMapping 34, 598–612.

Tsuchiya, N., Adolphs, R., 2007. Emotion and consciousness. Trends in CognitiveSciences 11, 158–167.

Tzourio Mazoyer, N., Landeau, B., Papathanassiou, D., Crivello, F., Etard, O., Delcroix,N., Mazoyer, B., Joliot, M., 2002. Automated anatomical labeling of activations inSPM using a macroscopic anatomical parcellation of the MNI MRI single-subjectbrain. Neuroimage 15, 273–289.

Urry, H.L., Van Reekum, C.M., Johnstone, T., Davidson, R.J., 2009. Individual differ-ences in some (but not all) medial prefrontal regions reflect cognitive demandwhile regulating unpleasant emotion. Neuroimage 47, 852–863.

van’t Wout, M., Aleman, A., Bermond, B., Kahn, R., 2007. No words for feelings:alexithymia in schizophrenia patients and first-degree relatives. ComprehensivePsychiatry 48, 27–33.

Venta, A., Hart, J., Sharp, C., 2012. The relation between experiential avoidance,alexithymia and emotion regulation in inpatient adolescents. Clinical ChildPsychology and Psychiatry, http://dx.doi.org/10.1177/1359104512455815 (inpress).

Vorst, H.C.M., Bermond, B., 2001. Validity and reliability of the Bermond-Vorst Alexithymia Questionnaire. Personality and Individual Differences 30,413–434.

Vuilleumier, P., 2005. How brains beware: neural mechanisms of emotional atten-tion. Trends in Cognitive Sciences 9, 585–594.

Wager, T., Davidson, M., Hughes, B., Lindquist, M., Ochsner, K., 2008.Prefrontal–subcortical pathways mediating successful emotion regulation.Neuron 59, 1037–1050.

Wager, T., Jonides, J., Reading, S., 2004. Neuroimaging studies of shifting attention:a meta-analysis. Neuroimage 22, 1679–1693.

Wager, T., Phan, K.L., Liberzon, I., Taylor, S., 2003. Valence, gender, and lateraliza-tion of functional brain anatomy in emotion: a meta-analysis of findings fromneuroimaging. Neuroimage 19, 513–531.

Wingbermühle, E., Theunissen, H., Verhoeven, W.M.A., Kessels, R.P.C., Egger,

J.I.M., 2012. The neurocognition of alexithymia: evidence from neu-ropsychological and neuroimaging studies. Acta Neuropsychiatrica 24,67–80.

Yelsma, P., 2007. Associations among alexithymia, positive and negative emotions,and self-defeating personality. Psychological Reports 100, 575–584.