Neural processing of sensory and emotional-communicative information associated
with the perception of vicarious pain☆
E. Vachon-Presseaua,b,⁎, M. Royd, M.O. Martelc, G. Albouyb, J. Chenb,g, L. Budellb,g, M.J. Sullivanc,
P.L. Jacksonf, P. Rainvilleb,e,g
aDepartment of psychology, Université de Montreal, Quebec, Canada
bCentre de recherche de l'Institut universitaire de gériatrie de Montréal (CRIUGM), Montreal, Quebec, Canada and Centre de recherche en neuropsychologie et cognition (CERNEC),
Université de Montréal, Montreal, Quebec, Canada
cDepartment of Psychology, McGill University, Montreal, Quebec, Canada
dDepartment of Psychology, Columbia University, USA
eDepartment of Stomatology, Faculty of Dentistry, Université de Montréal, Montreal, Quebec, Canada
fÉcole de psychologie and CIRRIS and CRULRG, Université Laval, Quebec, Canada
gGroupe de recherche sur le système nerveux central (GRSNC), Université de Montréal, Montreal, Quebec, Canada
a b s t r a c t a r t i c l e i n f o
Accepted 17 June 2012
Available online 23 June 2012
Functional magnetic resonance imaging
Inferior parietal lobule (IPL)
Inferior frontal gyrus (IFG)
aging study, 20 participants viewed pain-evoking or neutral images displaying either sensory or emotional-
communicative information. The pain images displayed nociceptive agents applied to the hand or the foot (sen-
sory information)or facialexpressions ofpain(emotional-communicative information)and werematched with
their neutral counterparts. Combining pain-evoking and neutral images showed that body limbs elicited greater
activity in sensory motor regions, whereas midline frontal and parietal cortices and the amygdala responded
more strongly to faces. The pain-evoking images elicited greater activity than their neutral counterparts in the
bilateral inferior frontal gyrus (IFG), the left inferior parietal lobule (IPL) and the bilateral extrastriate body
area. However, greater pain-related activity was observed in the rostral IPL when images depicted a hand or
foot compared to a facial expression of pain, suggesting a more specific involvement in the coding of somato-
motor information. Posterior probability maps enabling Bayesian inferences further showed that the anterior
IFG (BA 45and 47) wastheonlyregionshowing nointrinsic probability ofactivationbytheneutralimages, con-
sistent with a role in the extraction of the meaning of pain-related visual cues. Finally, inter-individual empathy
traits correlated with responses in the supracallosal mid/anterior cingulate cortex and the anterior insula when
pain-evoking imagesofbodylimbs or facialexpressionswerepresented,suggestingthat these regions regulated
the observer's affective-motivational response independent from the channels from which vicarious pain is
© 2012 Elsevier Inc. All rights reserved.
An evolutionary perspective predicts that individuals who are capa-
ble of decodingpain inothersbenefitfrominformation that serves both
self-oriented (e.g., identifying environmental threats) and social pur-
poses (e.g., social binding and altruism), giving them an advantage
over competitors (Williams, 2002). Pain in others can be perceived
and estimated through several channels that extend beyond verbal re-
ports: it can be explicitly witnessed through viewing a noxious event,
impending injury or tissue damage, or it can be signaled through non-
verbal emotional/communicative pain behaviors, such as facial expres-
sion. While the first channel provides objective cues about the sensory
terpret the meaning of communicative cues to infer the pain experi-
enced by the expresser (for a review see Hadjistavropoulos et al.
(2011)). In this study, we distinguished the difference between the in-
fluences of sensory and emotional-communicative information on the
neural correlates of vicarious pain perception.
response to pain involves the supracallosal mid/anterior cingulate cor-
tex (mACC) and the anterior insula (AI), which are regions involved in
self-pain, negative affect and/or motor preparation (Jackson et al.,
2005; Jackson et al., 2006; Morrison et al., 2004; Singer et al., 2004). It
is believed that such shared neural representations between an
NeuroImage 63 (2012) 54–62
☆ Etienne Vachon-Presseau, Mathieu Roy, Marc-Olivier Martel, Geneviève Albouy,
Jen-I Chen, Leslie Budell, Micheal Sullivan, Philip Jackson and Pierre Rainville have no
financial or other relationships that might lead to a conflict of interest.
⁎ Corresponding author at: Centre de recherche de l'Institut universitaire de
gériatrie de Montréal, 4545 Chemin Queen-Mary, Mtl, (Qc), Canada, H3W 1W4.
E-mail address: email@example.com (E. Vachon-Presseau).
1053-8119/$ – see front matter © 2012 Elsevier Inc. All rights reserved.
Contents lists available at SciVerse ScienceDirect
journal homepage: www.elsevier.com/locate/ynimg
emotional state perceived in others and the corresponding state
mapped in the observer is the core of the theoretical framework of em-
pathy (Decety and Jackson, 2004). As such, self-related factors such as
dispositional empathy (Jackson et al., 2005; Singer et al., 2004), exper-
tise (Cheng et al., 2007; Decety et al., 2010), the affective link between
individuals (Singer et al., 2006), perceived group membership (Hein
et al., 2010) and racial bias (Avenanti et al., 2010; Xu et al., 2009)
have been shown to modulate the vicarious pain response in the ACC
and/or the AI. Besides the motivational/affective responses, the em-
pathic neural response also involves a cognitive component that per-
mits top-down influence (Lamm et al., 2007; Singer and Lamm, 2009)
and perspective taking that are necessary for altruistic behaviors (de
Vignemont and Singer, 2006; Decety and Jackson, 2004).
Animal studies have revealed that a population of neurons in the F5
to the inferior parietal lobule(IPL)) respond whena monkey generates a
goal-oriented action or creates an internal representation of a motor act
made by another agent (Cattaneo and Rizzolatti, 2009). It has been hy-
pothesized that such internal representations permit the simulation of
served in others. Accordingly, in humans, a fronto-parietal analog to this
mirror neuron system is believed to permit an understanding of the ac-
tion and the intention of others (Rizzolatti et al., 2001), learning by imi-
tation and communication with others (Rizzolatti and Craighero, 2004),
and participating in the decoding of another person's mind (Zaki et al.,
2009). This system could represent one of the gateways that enable the
resonance with other people's states and promote the shared represen-
tations (Hétu and Jackson, 2012). The presence of a vicarious pain
response in the posterior IFG (BA 44) and IPL (BA 40) has been consis-
tently reported when participants viewed a somatic representation of a
painful agent as opposed to abstract cues signaling pain in others (for a
review see Lamm et al. (2010)). Vicarious responses are also found in
response to the observation of facial pain expressions in both the IFG
and IPL, with stronger activation of the IFG when participants attend to
the meaning of the expression to evaluate the other's pain, and stronger
activation in the IPL when participants are asked to discriminate the fa-
cial movements composing the same expressions (Budell et al., 2010).
The objective of this study was to clarify the relative implications of
this fronto-parietal network in the processing of somatosensory cues
compared with facial pain expressions.
In the present study, we sought to identify the differences in brain
activity between decoding facial expressions communicating pain
and explicitly witnessing pain through a noxious agent. First, we
aimed to replicate the robust activation found in the putative
human mirror neuron system when body limbs are observed
(Grosbras and Paus, 2006; Kret et al., 2011; Peelen and Downing,
2007; van de Riet et al., 2009) and to contrast this effect to the activa-
tion expected from facial expressions in the amygdala and the medial
prefrontal and posterior cingulate cortex which are thought to be in-
volved in the attribution of intentions and feelings in others (Adolphs
et al., 1994; Grosbras and Paus, 2006; Henson et al., 2003; Kret et al.,
2011; van de Riet et al., 2009; Zaki et al., 2009). Second, consistent
with the notion of sensorimotor resonance, we hypothesized that
the IPL would be solicited more when others' pain states are inferred
from sensory rather than emotive-communicative cues. This would
be crucial to clarify how the brain creates the internal mapping of
pain states observed in others. Third, based on our prior work show-
ing that ventral-lateral frontal areas are associated with the explicit
evaluation of pain in others (Budell et al., 2010), we used posterior
probability maps to test the prediction that the anterior IFG would
be solicited in sensory and facial pain-evoking conditions, but not
by their neutral counterparts. Finally, we hypothesized that between-
individual empathic traits of the participants would be positively corre-
lated to activityinthemACCandtheAIelicited by thepain-evokingim-
ages depicting either somatosensory or emotional-communicative
The participants were 20 healthy (10 female) volunteers between
21 and 53 years old (mean: 36; SD: 10) with no history of neurological
or psychiatric disorders. None reported having chronic pain, and their
mal ranges. All subjects gave written consent prior to the experiment,
and the study was approved by the research ethics committee of the
Institut Universitaire de Gériatrie de Montréal. Each participant filled a
standard MRI safety form prior to the study, screening for previous sur-
geries, metal implants or pregnancy, to ensure his or her safety.
Assessment of empathy
The Empathy Quotient (EQ; (Baron-Cohen and Wheelwright, 2004))
and the Interpersonal Reactivity Index (IRI; (Davis, 1980)) were used to
assess trait empathy. The EQ is a self-report questionnaire of 60 items
that are summed to an overall score while the IRI is a 28 item question-
empathic concern (EC), the perspective taking (PT) and the personal
distress (PD) factors from the IRI were included in our analyses (as in
our previous study (Vachon-Presseau et al., 2011)). Because the four
scores of interest (EQ, EC, PT and PD) showed a high correlation, a prin-
cipal component analysis was used for data reduction (Avenanti et al.,
2009a; Vachon-Presseau et al., 2011). The 4 scores were loaded onto
one factor, referred to as the global empathy score, which accounted
for 47% of the variance and was positively correlated with the EQ
(r=.83), the EC (r=.65), the PT (r=.74) and negatively correlated
with the PD (r=−.49).
a nociceptive agent applied to the right hand (n=8) or the right foot
(n=8) that were presented from an allocentric perspective or a facial
expression of pain (n=8). No stimulus involved blood or actual tissue
damage. Matching neutral images presenting the right hand (n=8)
or right foot (n=8) in similar but innocuous situations, and neutral fa-
cial expressions by the same actor (n=8), were included as a control
condition for non-specific visual and attention processes. Fig. 1A illus-
tratesexamples of pain-evokingand neutralimages fromeachcategory
(similartoJacksonet al. (2005) for bodylimbsand identical toSimonet
al. (2006) for facial expressions). We deliberately chose to use two cat-
egories of images representing the sensorimotor component of pain in
ordertotest ifbrain activitygenerated by sensoryinformation wasspe-
cific to images of the hand or foot or generalized to body limbs. The im-
portance of generalization is based on one of our previous studies,
which showed that witnessing pain-evoking images of a hand or foot
(generalized to body limbs), but not faces, facilitated the nociceptive
withdrawal reflex of the lower limb in an observer (Vachon-Presseau
et al., 2011).
In the present study, the experimental protocol included two func-
tional scans of thermal pain administered to the leg of the participant
and two functional scans where the participants observed images
depicting pain-evoking experiences. Here, only the data regarding the
impact of the sensory or emotional-communicative information on
brain processing are presented. Other data regarding self-pain will be
presented elsewhere. In the two functional scans, the participants
were shown images and asked to determine if they depicted pain or
neutral experiences. At the beginning of each trial, a fixation cross
appeared for 3, 4 or 5 s on the screen, followed by a two-second
E. Vachon-Presseau et al. / NeuroImage 63 (2012) 54–62
image. At the image offset, participants were asked to indicate if the
keys of the response box on a computerized left/right decision task dis-
played for 3 s using E-Prime (Psychology Software Tools Inc.; http://
ly paid attention to the stimuli. Each trial ended with the fixation cross
reappearing on the screen indicating that the next trial was beginning.
The participants were instructed on how to perform the experimental
task and were successful in a brief training session prior to entering
the scanner. In the scanner, each of the 48 stimuli was shown six
times for a total of 288 trials. The images were arranged in two series
of 144 trials presented in two successive functional runs of 18 min.
The order of stimuli was pseudo-randomized according to the category
sex of the model.
After the scanning session, the participants were asked to rate the
pain perceived in the images presented in the experiment using a
computerized visual-analog scale (VAS). The images were presented
for 2 s while the scale was displayed for 12 s and labeled with the ver-
bal anchor “neutral” at 0 (left extremity) and “extreme pain” at 100
(right extremity). This allowed us to verify that the estimated pain
perceived in the images were comparable between the categories.
fMRI acquisition and analyses
Imaging was performed on a 3.0 T whole-body scanner (Siemens
TRIO), using a 12-channel head coil, at the Centre de Recherche de
l'Institut Universitaire de Gériatrie de Montréal (CRIUGM) in Montréal,
QC, Canada. Blood oxygenation level-dependent (BOLD) signal was ac-
quired using a standard T2*-weighted gradient-echo EPI sequence
(TR=3000 ms, TE=30 ms; flip angle=90°; FOV=220×220 mm2;
matrix=40 interleaved, axial slices per whole-brain volume at
3.4 mm thickness; in-plane resolution of 3.4×3.4 mm for isotropic
voxels; 360 volumes). Structural images were acquired using a high-
2.99 ms; flip angle=9°; FOV=256 mm; matrix=256×256; 1×1×
1.2 mm voxels; 160 slices per whole-brain volume). All data
preprocessing and analysis were performed using SPM 8 (Statistical
Parametric Mapping, Version 8; Wellcome Department of Imaging
Neuroscience, London, UK, http://www.fil.ion.ucl.ac.uk/spm/software/
spm8/) executed inMatlab 7.8.(Mathworks, Sherborn,Massachusetts).
Offline preprocessing of functional images included realignment of
functional time series, co-registration of each subject's functional and
anatomical data, spatial normalization to an EPI template conforming
to the Montreal Neurological Institute space (2×2×2 mm), and spatial
smoothing (8 mm FWHM Gaussian kernel).
The analysis of the fMRI data, based on a mixed effects model, was
conducted in 2 serial steps, accounting for fixed and random effects.
For each subject, changes in brain regional responses were estimated
sequence (TR=2.3 ms;TE=
by a general linear model. Six trial types were modeled: 3 categories
(hand, foot, or face) and 2 pain levels (pain-evoking image or neu-
tral). Each type of trial was modeled as a delta function representing
stimulus onset. The ensuing vector was convolved with the canonical
hemodynamic response function and used as a regressor in the
individual design matrix. Movement parameters estimated during re-
alignment (translations in x, y, and z directions and rotations around
x-, y-, and z-axes) and a constant vector were also included in the ma-
trix as variables of no interest. A high-pass filter was implemented
using a cut-off period of 128 s to remove the low-frequency drifts
from the time series. Serial correlations in the fMRI signal were esti-
mated using an autoregressive (order 1) plus white noise model
and a restricted maximum likelihood (ReML) algorithm.
and foot) and faces to detect differences between body and facial ex-
pressions irrespective of the pain content in the images. Second, a con-
trast between pain-evoking and neutral images was performed across
all categories. Third, an interaction, [pain-evoking vs. neutral images]
by [images of body limbs vs. facial expressions], was performed to de-
termine if pain-evoked activity differed when perceived by sensory or
emotional-communicative cues. To test if this effect was specific to
each limb, the interactions, [pain-evoking vs. neutral images] by [im-
ages of hand vs. facial expressions] and [pain-evoking vs. neutral im-
ages] by [images of foot vs. facial expressions], were performed
independently. These linear contrasts generated statistical parametric
maps [SPM(T)]. These summary statistical images were then further
spatially smoothed (Gaussian kernel 6 mm FWHM) and entered in a
second-level analysis, corresponding to a random effects model and ac-
countingfor inter-subject variance. One-sample t-tests were run on the
data of all the subjects. Finally, to assess the relationship between brain
activity during pain processing and individual empathy, we regressed
each individual's within-subject contrast images for pain-evoking im-
ages (body limbs and facial expressions) against their individual empa-
The resulting set of voxel values for each contrast constituted a map
of the t statistic [SPM(T)], with the threshold at pb0.001 (uncorrected
for multiple comparisons). Cluster-based statistics were used to define
significant activations based on intensity and spatial extent. Clusters
considered significant at the whole brain level based on the conserva-
tive family-wise error correction(FWE pb.05) wereused for inferential
statistics. Significant clusters were further explored to localize peaks of
activation (Z≥3.81), as reported in the Tables. Additionally, regions of
interest (ROI) were defined anatomically using the WFU PickAtlas soft-
ware toolbox (Maldjian et al., 2003) and significant activation was
assessed in these structures after family-wise error correction (FWE;
pb.05) accountingfor thevolumeoftheROIs.Based onourhypotheses,
theIFG (BA44, 45 and 47) and theIPL (BA 40)were targeted asROIsfor
the interaction term [pain-evoking vs. neutral images] by [images of
body limbs vs. facial expressions]. The ACC (BA 24) and the insula
Fig. 1. A. Examples of pain-evoking and neutral images from each category. B. The estimated perceived pain scores (± standard error of the mean) are equivalent between
E. Vachon-Presseau et al. / NeuroImage 63 (2012) 54–62
were targets for the regression of the pain-evoking images against the
individual empathy score. In this analysis, the insula did not survive
FWE correction over the anatomically defined ROI. Because the ante-
rior region of this structure (AI) is well documented for its associa-
tion with empathy for pain, we used a less stringent threshold
(pb.005 uncorrected) to detect activity in the AI over a small volume
of interest (10 mm) centered on a peak previously shown to be asso-
ciated with vicarious painandempathy [x, y, z: 32,18, 6 from Jackson
et al. (2005)].
Finally, in the random-effects analyses, posterior probability maps
(PPMs) enabling Bayesian inferences were generated (Friston and
Penny, 2003). This type of analysis permitted the determination of the
intrinsic probability of activation by the neutral images in the clusters
of activity from the contrast [pain-evoking vs. neutral]. To test this,
the PPM and effect size were computed for the pain-evoking and neu-
tral images to verify which areas have a low probability of generating
activations in the neutral condition. The target areas were determined
usinga 10-mmvolumeofinterest aroundtheactivationpeaksobtained
with the contrast [pain-evoking vs. neutral]. This permitted us to statis-
tically test if the activity in these regions was specific to the pain-
evoking condition by showing a very low probability of activation
when neutral images were presented.
The mean scores (M±SD) of pain perceived in the images were
65.4±17.8 for the hand, 67.8±11.2 for the foot, and 70.9±13.6 for
the face category in the pain-evoking images (Fig. 1B) while the neu-
tral images yielded mean scores of b5.0 in each category. An ANOVA
on the scores of the pain perceived revealed no significant differences
between pain-evoking images of the three different categories
fMRI data: main effect of image category
The first analysis was performed to identify brain regions differen-
tially activated by the images depicting body limbs and those
depicting facial expressions irrespective of the pain content of the im-
ages (i.e., combining pain-evoking and neutral images). As illustrated
in Fig. 2A and presented in Table 1A, the results revealed that the ob-
servation of body limbs, compared to facial expressions, elicited
stronger activity in the fusiform gyrus (FG, BA 20/37), the para-
hippocampal gyrus (BA 36), the extrastriate body area (EBA, BA 37),
the superior parietal lobule (BA 7), the rostral part of the IPL (BA
40), the pre‐central gyrus (BA 4), the post central gyrus (BA 1/2/3),
the middle occipital gyrus (BA 19) and the cerebellum. In contrast,
as shown in Fig. 2B and Table 1B, more activity was evoked by images
of facial expressions in the ACC (BA 32), the medial prefrontal cortex
(BA 10), the posterior cingulate cortex (BA 31), the primary visual
cortex (BA 17), the superior temporal sulcus (STS, BA 22) and the
amygdala. These findings robustly replicated a whole body of litera-
ture showing that images depicting body limbs elicited stronger ac-
tivity in sensorimotor regions, while facial expressions most
strongly recruited the midline cortical structures and subcortical lim-
fMRI data: The IFG was the only structure that exclusively responded to
A contrast between pain-evoking and neutral images was first
performed to identify regions more responsive to witnessing others'
pain, regardless of the material used to induce this effect (hand, foot
or facial expression). As illustrated in Fig. 2C and presented in
Table 2, pain-evoking vs. neutral images yielded peaks of activation
in the bilateral IFG (BA 45 and BA 47), the left rostral IPL (BA 40)
and the left EBA (BA 37). Fig. 3 illustrates the mean activity estimates
associated with each category of images in a sphere of interest
(10 mm radius) around these peaks. From these, bilateral IFGs were
Fig. 2. A. Images depicting body limbs that produced stronger activity in sensory-motor regions compared with images depicting facial expressions. B. The contrast between regions ac-
tivated by facial expressions compared to body limbs, such as the medial prefrontal cortex, the precuneus and the superior temporal sulcus involved in the theory of mind. C. Cerebral
activity from pain-evoking vs. neutral images was observed in the inferior frontal gyrus (IFG), the left inferior parietal lobule (IPL), and the extrastriate body area (EBA). D. The rostral
IPL responded more strongly to images of body limbs in pain than to images of facial pain expressions. Functional data are shown over the mean structural image of all participants on
a 3D rendering of the brain to illustrate the location of peak activity (pb.001 uncorrected). The error bars represent the standard error of the mean. **pb.01; ***pb.001.
E. Vachon-Presseau et al. / NeuroImage 63 (2012) 54–62
the only brain regions responding to the pain-evoking images but not
to the neutral images. The lack of activity during the presentation of
neutral images is supported by the calculation of the posterior prob-
ability map (Friston and Penny, 2003), as inferred by Bayesian statis-
tics, showing that the probability of activation of the IFG is very low
when the image was neutral (left IFG: 2%; right IFG: 0%). This con-
trasts with the other pain-activated structures showing weaker but
clearly significant responses to neutral images (left IPL: 100%; left
EBA: 100%; right EBA: 100%). This result suggests that the IFG is spe-
cifically sensitive to the pain depicted in the images without being
sensitive to other features present in the neutral image.
fMRI data: The IPL coded the sensory cues depicted in the pain-evoking
TheIPL coded thesensorycuesdepicted inthepain-evokingimages.
The present study aimed to identify specific regions differently pro-
cessing observed pain based on sensory or emotive-communicative
cues. Fig. 2D illustrates that, as predicted, when testing the interaction
between [pain-evoking images vs. neutral images] and [hand and foot
vs. faces], the left rostral IPL (BA 40) was most strongly recruited
when pain-evoking images depicted body limbs [x, y, z: −60, −26,
34; Z=3.59; FWE in ROI (BA 40) p=.05]. This effect could have been
specific to images of the hand or foot. The analysis was therefore also
performed independently with the contrasts [[pain-evoking hand vs.
neutralhand] vs. [pain-evokingfacevs.neutralface]] and[[pain-evoking
the glass brain (Fig. 2D; uncorrected pb.001), both analyses resulted
in increased brain activity limited to the rostral IPL (hand vs. face: x, y,
z: −60, −26, 36; Z=3.40; and foot vs. face: x, y, z: −64, −26, 28;
Z=3.54). This indicates that the IPL codes somatic information about
pain in others. The same interaction term examining the effect of pain
(pain-evoking vs. neutral) across hand vs. foot, foot vs. hand, face vs.
hand, or face vs. foot, yielded no significant effect when corrected for
in response to pain-evoking images.
fMRI data: Trait empathy correlated with the activity in the ACC and the
AI independently from the channels in which vicarious pain is perceived.
Trait empathy correlated with the activity in the ACC and the AI
independent from the channels in which vicarious pain is perceived.
Clusters considered significant in the whole brain (FWE pb.05) when contrasting A. [body limbs vs. a facial expression] and B. [a facial expressions vs. body limbs].
k Brain area BA MNI coordinates Local peak z-value
A. Cluster analysis corrected for the whole brain: hand and foot vs. face
417 Pre‐central gyrus
947 Post central gyrus
Inferior parietal lobule
20,886 Inferior parietal lobule
Post central gyrus
Inferior temporal gyrus
Superior parietal lobule
Middle occipital gyrus19
B. Cluster analysis corrected for the whole brain: face vs. hand and foot
1425 Medial prefrontal cortex
Anterior cingulate cortex
Superior temporal sulcus
Posterior CC and precuneus
Primary visual cortex
Montreal Neurological Institute (MNI)
aSignificant at a peak level (FWEb.05).
Clustersconsideredsignificantwhencontrasting [pain-evoking vs. neutral pictures]andcorrecting for A.thewhole brain (FWEpb.05), orB.ROIsdelineating a bilateral IFG (FWE pb.05).
k BRAIN AREA BA MNI COORDINATESLOCAL PEAK z-value
A. Cluster analysis corrected for the whole brain: pain‐evoking pictures vs. neutral pictures
727Inferior parietal lobule
B. ROI correction: pain-evoking pictures vs. neutral pictures
Inferior frontal gyrus pars orbitalis
Inferior frontal gyrus pars triangularis
aSignificant at a peak level.
bMarginally significant at a cluster level.
E. Vachon-Presseau et al. / NeuroImage 63 (2012) 54–62
In this study, we also sought to determine if personality traits were
differently associated with the neural correlates of empathy for vicari-
ous pain depicted by sensory or emotive-communicative cues. For this
purpose, we only used the brain signal during pain-evoking images be-
primary target areas in this analysis (mACC: x, y, z: −8, 8, 40; Z=5.64;
FWE pb.001; AI: x, y, z: −8, 8, 40; Z=5.64; FWE pb.001). Fig. 4 shows
that when corrected for an ROI targeting the supracallosal anterior and
images of the hand or foot was positively correlated with the empathy
score (peak x, y, z: 4, 2, 36; Z=3.58; FWE, p=.03). Similarly, using
the same correction also revealed a positive correlation between the
Fig. 3. The mean parameter estimates in the sphere of interest (10 mm radius) for each category of images in pain-evoking and neutral situations. The peaks were selected from the
pain-evoking vs. neutral contrast images.
Fig. 4. The enhancement of vicarious pain responses by empathic traits of the observer was independent of the channel from which the vicarious pain was witnessed. Correlation
between the subjects' empathy scores and the mean parameter estimates in anterior cingulate cortex with A. an image of a body limb in a pain-evoking situation and B. a facial pain
expression. C. Conjunction analysis of A. and B. Functional data are shown over the mean structural image of all participants or a 3D rendering of the brain to illustrate the location of
the peak (pb.001 uncorrected).
E. Vachon-Presseau et al. / NeuroImage 63 (2012) 54–62
empathy scores and brain activity in the mACC when images of facial
expressions were presented [peak x, y, z: −2, 6, 34; Z=3.77; p=.02].
Hence, as shown by a conjunction analysis illustrated in Fig. 4C, these
results suggest that trait empathy positively correlates with the
level of activity within the mACC independent of cues depicting pain
to the observer. Similarly, using a more liberal correction (pb.005
uncorrected), the conjunction analysis of brain activity elicited by
related with the anterior right insula [peak x, y, z: 4, 2, 36; Z=2.64; svc
FWE, p=.07]. Performing the same analysis on the brain activity elicit-
thy in these ROIs. No additional cluster survived the FWE correction
across the whole brain in theregression analysis of the sensory or emo-
tive cue conditions or of pain-evoking or neutral images.
the perception of pain in others. As expected, the results show that im-
cial expressions of pain rather more strongly activated the amygdala
and cortical regions associated with socio-emotional processes. The re-
sults also revealed that the IFG, the IPL, and the EBA were sensitive to
the pain evoked in the images. The novel findings of the current study
were, first, that the bilateral anterior IFG was the only structure that
uniquely responded to pain-evoking images, but not to their neutral
counterparts, and second, that the rostral IPL very clearly coded more
specifically for the pain depicted with sensory cues. These results com-
plement those of a previous study suggesting that the IFG may be in-
volved in understanding the meaning of the pain, while the IPL would
be more sensitive to the sensorimotor cues conveying information
about the pain (Budell et al., 2010). Finally, the results show that
in themACCand theAI(only atpb.005in theAI) whenobservingpain-
evoking images of body limbs or facial expressions.
Differences between body limbs and facial expressions
The results of the present study show that the ventral posterior IFG
and the rostral IPL regions were recruited more by the observation of
body limbs compared with the observation of facial expressions.
These regions are located within the core mirror neuron system
(Rizzolatti and Craighero, 2004), suggesting that attending to images
of body limbs elicited greater sensory-somatic resonance than attend-
ing to facial expressions. In addition, the EBA, a region sensitive to bio-
logical motion (Downing et al., 2001), the FG and the parahippocampal
limbs elicited stronger activity than facial expressions in the FG, but
these findings are consistent with previous reports, showing increased
activity in the FG and the EBA when watching fearful and angry body
postures compared to fearful and angry facial expressions (Kret et al.,
2011; van de Riet et al., 2009). Similarly, a study looking at neutral
hand movements compared with neutral facial expressions showed
that the hand images generated stronger activity in the EBA and the
FG, as well as the dorsal pre-motor area (Grosbras and Paus, 2006).
are processed holistically, similar to facial expressions, and share com-
municative features conveying information about social interaction
(de Gelder et al., 2010).
That facial expressions most strongly activated the amygdala con-
firms its role in the communication of a potential threat to an observer
by acting as a detector capable of enhancing the saliency of information
with emotional significance to an observer (Anderson and Phelps,
2001). Our results also show increased activity in the medial prefrontal
cortex and the precuneus when facial expressions were depicted
compared to body limbs. These midline structures are commonly rec-
ruited in the default mode network serving self-referential thoughts
(Buckner et al., 2008; Northoff et al., 2006; Raichle et al., 2001). The
same midline structures are also involved in mentalizing, a cognitive
strategy that uses a self-referential approach to understand others
(Mitchell, 2009). The observation of facial expressions generated great-
er activation in this network, possibly because the social-emotional di-
mension is emphasized in these images in contrast to images explicitly
depicting only nociceptive information. Facial expressions also yielded
stronger activity in the STS, a region believed to be involved in the imi-
tation of the action of others (Iacoboni et al., 2001; Molenberghs et al.,
2010), communication abilities between individuals (Noordzij et al.,
2009) and the attribution of emotional states to others (Frith and
Frith, 2006). Together, these main effects robustly replicate previous
findings and validate the current stimuli and experimental procedure
in eliciting vicarious pain-related brain responses in expected regions.
Shared and unique activations to sensory and emotive-communicative
The main objective of the present study was to determine common
and distinct networks involved in pain perception by sensory or emo-
tive cues. We found that the IFG pars triangularis and orbitalis (BA 45
and BA 47), left IPL, and EBA showed increased activity when pain-
evoking images were presented compared with their neutral counter-
parts. Fig. 2C illustrates that activity in the IFG was restricted to the
pars triangularis (BA 45) and pars orbitalis (BA 47), without extending
to thepars opercularis (BA 44).Recently, Budell et al. (2010) foundthat
when participants were asked to evaluate the perceived pain or the
motor component of identical dynamic facial expression of pain, the
IFG (BA 45 and BA 47) was most strongly activated by the pain evalua-
tion task. Although the peaks of activity found in the present report are
somewhat anterior to those reported in previous studies, the combined
results suggest that the activity in the IFG, outside of BA 44 (the equiv-
alent of the monkey F5, where mirror neurons were found), specifically
tion for the neutral images was very low in the IFG but very high in the
other brain regions. This supports the idea that the IFG could be in-
volved in the extraction of meaning from the pain-evoking situation.
in detecting salient stimuli (Corbetta and Shulman, 2002) and might
therefore underlie a stimulus-driven attention process. This is particu-
larlyrelevant toourexperimentaldesignbecausewe useda paindetec-
We also found that pain perceived from sensory cues activated the
rostral IPL (BA 40) more strongly than pain perceived from an emotive
cue. It has been shown previously that the IPL (homologous to the PF in
the monkey) contains mirror neurons (Rizzolatti and Craighero, 2004),
is activated by the observation of others' actions and contains proprio-
ceptive information allowing motor planning and imitation (Iacoboni
et al., 2001). In the present study, this region responded most strongly
to pain-evoking images when body limbs were depicted, suggesting
uate the presence and/or the amount of pain. This finding is consistent
with our previous study showing that the IPL was more involved in the
evaluation of motor rather than emotional aspects of facial expressions
(Budell et al., 2010). Together, our findings point toward the IFG as the
integrator of the meaning of a pain situation depicted in an image and
the IPL as a coder of the somatic and motor-related information.
It is known that the IPL codes self-made motor acts and perceived
movements in others with a specific goal (such as grasping). In the
current study, witnessing sensory cues explicitly depicting a noxious
agent threatening a body limb might have implied movement of
withdrawal. This proposition is supported by recent studies showing
that static images that implied movement recruit the putative mirror
neuron system (Urgesi et al., 2006; Urgesi et al., 2010). It is therefore
E. Vachon-Presseau et al. / NeuroImage 63 (2012) 54–62
likely that a top-down influence permits the anticipatory representa-
tion of the withdrawal of body limbs threatened by a noxious agent.
This is particularly interesting because a previous study has shown
that mirroring activity in the IPL is organized based on the type of
movement being observed (dragging, dropping, grasping or pushing)
regardless of the effectors (i.e., movement of the hand, the foot or the
mouth) (Jastorff et al., 2010). This suggests that implied movement,
rather than the body limb, would explain the neural activations in
the IPL during pain observation.
Previous behavioral studies have shown that witnessing pain in
others could modulate the motor response in an observer. For instance,
pain observation slowed key-press responses and improved key-
task, activity in the cingulate cortex paralleled fastest reaction times by
showing increased activity during the pain-related modulation of overt
motor responses (Morrison et al., 2007a). It is possible that activity in
the putative mirror neuron system associated with pain perceived in
body limbs primed the corresponding responses in the observer. Ac-
induced by an electrical noxious shock was facilitated when images
depicting body limbs (hand or foot), but not static facial pain expres-
sions, were presented to an onlooker (Vachon-Presseau et al., 2011).
The impact of vicarious pain on motor movement seems to engage a
complex inhibitory and facilitatory network. Transcranial magnetic
stimulation of the motor cortex inhibited the motor evoked potential
of the hand while a participant observed a needle entering a congruent
hand (same hand that is stimulated) and facilitated the MEP when en-
tering the opposite hand (contralateral to the stimulated hand)
(Avenanti et al., 2005; Avenanti et al., 2009b). Together, these studies
revealed that pain observation can modulate pain-related responses
that might be driven by different mechanisms involving the spinal
cord (Vachon-Presseau et al., 2011), the cortical motor system
(Avenanti et al., 2005; Costantini et al., 2008) and the cingulate cortex
(Morrison et al., 2007a).
The affective-motivational response of the vicarious pain is independent
from the channel of perceived pain
Several studies have shown that empathy for pain involves the
mACC and the AI (e.g., (Jackson et al., 2005; Lamm et al., 2010;
Lamm and Singer, 2010; Morrison et al., 2004; Singer et al., 2004)).
It has been proposed that activity found in these regions might result
either from a shared representation mechanism (Preston and de
Waal, 2002), or from valence, arousal and withdrawal response prim-
ing induced by vicarious pain perception (Decety, 2010). The results
of our conjunction analysis demonstrated that pain-evoking images
of body limbs and facial expressions generated peaks of activity in
the mACC and the AI that correlated with empathy traits in the ob-
server. These peaks of activity are found almost at the same coordi-
nates reported by Singer et al. (2004), who showed a positive
correlation between activity in both the mACC and the AI with the
empathy traits of the participant when a signal indicated that a pain-
ful electrical stimulation was administered to a loved one. Similar cor-
relations were obtained in the cingulate cortex for the amount of pain
perceived in others (Jackson et al., 2005; Saarela et al., 2007) and so-
cial distress personality traits during a social exclusion task
(Eisenberger et al., 2003). Together these findings indicate that the
channel by which vicarious pain is induced (communicative or senso-
ry cues) have little influence on the vicarious response in brain re-
gions underlying the motivational and affective components of pain
perception. Accordingly, our results are consistent with the notion
that the activity in these structures is involved in the regulation of be-
havior as a function of individual personality factors independent of
the sub-channel from which vicarious pain is perceived. That such a
relation was only found in the pain-evoking images, but not in the
neutral ones, argues for the facilitation of brain activity by empathy
only when arousal and feelings were elicited by the images in the
Conclusion, future perspectives and limitations
This study stresses the influence of cues signaling vicarious pain on
brain activity. These findings are important because they help under-
stand how the brain extracts and processes information about the
pain in others. This study demonstrates how the activation of separate
input channels (sensory vs. emotional-communicative) converges on
a core system underlying the representation of pain in others. Impor-
tantly, one limitation of most neuroimaging studies is the small sample
size (nb30 participants) (Yarkoni et al., 2010). Here, a bigger sample
could have yielded the most robust activations in the ACC and the AI
that might have survived the FWE correction in the whole brain when
comparing the pain-evoking with the neutral images. Further studies
are needed to determinehow theneural correlatesof pain communica-
states or disorders of communication. These studies may elaborate a
model that explains the effect of chronic pain on both the expressive
derstanding the impact of pain cues on individuals exposed to high
rates of vicarious pain such as health care professionals.
This work was supported by grants from the Fonds de recherche
Québec—Santé (FRQS; to P.R. and M.J.S.) and by the National Sciences
and Engineering Research Council of Canada (NSERC; P.R.) and doc-
toral scholarships from the Canadian Institutes of Health Research
and from the FRQS (to E.V.P.).
Adolphs, R., Tranel, D., Damasio, H., Damasio, A., 1994. Impaired recognition of emotion
in facial expressions following bilateral damage to the human amygdala. Nature
Anderson, A.K., Phelps, E.A., 2001. Lesions of the human amygdala impair enhanced
perception of emotionally salient events. Nature 411, 305–309.
Avenanti, A., Bueti, D., Galati, G., Aglioti, S.M., 2005. Transcranial magnetic stimulation
highlights the sensorimotor side of empathy for pain. Nat. Neurosci. 8, 955–960.
Avenanti, A., Minio-Paluello, I., Bufalari, I., Aglioti, S.M., 2009a. The pain of a model in
the personality of an onlooker: influence of state-reactivity and personality traits
on embodied empathy for pain. NeuroImage 44, 275–283.
Avenanti, A., Minio-Paluello, I., Sforza, A., Aglioti, S.M., 2009b. Freezing or escaping?
Opposite modulations of empathic reactivity to the pain of others. Cortex 45,
Avenanti, A., Sirigu, A., Aglioti, S.M., 2010. Racial bias reduces empathic sensorimotor
resonance with other-race pain. Curr. Biol. 20, 1018–1022.
Baron-Cohen, S., Wheelwright, S., 2004. The empathy quotient: an investigation of
adults with Asperger syndrome or high functioning autism, and normal sex differ-
ences. J. Autism Dev. Disord. 34, 163–175.
Beck, A.T., Steer, R.A., Brown, G.K., 1996. Manual for the Beck Depression Inventory-II.
Psychological Corporation, San Antonio, TX.
Buckner, R.L., Andrews-Hanna, J.R., Schacter, D.L., 2008. The brain's default network:
anatomy, function, and relevance to disease. Ann. N. Y. Acad. Sci. 1124, 1–38.
Budell, L., Jackson, P., Rainville, P., 2010. Brain responses to facial expressions of pain:
emotional or motor mirroring? NeuroImage 53, 355–363.
Cattaneo, L., Rizzolatti, G., 2009. The mirror neuron system. Arch. Neurol. 66, 557–560.
Cheng, Y., Lin, C.P., Liu, H.L., Hsu, Y.Y., Lim, K.E., Hung, D., Decety, J., 2007. Expertise
modulates the perception of pain in others. Curr. Biol. 17, 1708–1713.
Corbetta, M., Shulman, G.L., 2002. Control of goal-directed and stimulus-driven atten-
tion in the brain. Nat. Rev. Neurosci. 3, 201–215.
Costantini, M., Galati, G., Romani, G.L., Aglioti, S.M., 2008. Empathic neural reactivity to
noxious stimuli delivered to body parts and non-corporeal objects. Eur. J. Neurosci.
Davis, M., 1980. A multidimensional approach to individual differences in empathy.
JSAS Catalog of Selected Documents in Psychology, 10, p. 85.
de Gelder, B., Van den Stock, J., Meeren, H.K., Sinke, C.B., Kret, M.E., Tamietto, M., 2010.
Standing up for the body. Recent progress in uncovering the networks involved in
the perception of bodies and bodily expressions. Neurosci. Biobehav. Rev. 34,
de Vignemont, F., Singer, T., 2006. The empathic brain: how, when and why? Trends
Cogn. Sci. 10, 435–441.
Decety, J., 2010. To what extent is the experience of empathy mediated by shared neu-
ral circuits? Emot. Rev. 2, 204–207.
E. Vachon-Presseau et al. / NeuroImage 63 (2012) 54–62
Decety, J., Jackson, P.L., 2004. The functional architecture of human empathy. Behav. Download full-text
Cogn. Neurosci. Rev. 3, 71–100.
Decety, J., Yang, C.Y., Cheng, Y., 2010. Physicians down-regulate their pain empathy re-
sponse: an event-related brain potential study. NeuroImage 50, 1676–1682.
Downing, P.E., Jiang, Y., Shuman, M., Kanwisher, N., 2001. A cortical area selective for
visual processing of the human body. Science 293, 2470–2473.
Eisenberger, N.I., Lieberman, M.D., Williams, K.D., 2003. Does rejection hurt? An FMRI
study of social exclusion. Science 302, 290–292.
Friston, K.J., Penny, W., 2003. Posterior probability maps and SPMs. NeuroImage 19,
Frith, C.D., Frith, U., 2006. The neural basis of mentalizing. Neuron 50, 531–534.
Grosbras, M.H., Paus, T., 2006. Brain networks involved in viewing angry hands or faces.
Cereb. Cortex 16, 1087–1096.
Hadjistavropoulos, T., Craig, K.D., Duck, S., Cano, A., Goubert, L., Jackson, P.L., Mogil, J.S.,
psychosocial formulation of pain communication. Psychol. Bull. 137 (6), 910–939.
Hein, G., Silani, G., Preuschoff, K., Batson, C.D., Singer, T., 2010. Neural responses to
ingroup and outgroup members' suffering predict individual differences in costly
helping. Neuron 68, 149–160.
Henson, R.N., Goshen-Gottstein, Y., Ganel, T., Otten, L.J., Quayle, A., Rugg, M.D., 2003.
Electrophysiological and haemodynamic correlates of face perception, recognition
and priming. Cereb. Cortex 13, 793–805.
Hétu, S., Jackson, P., 2012. The neural systems involved in motor cognition and social.
In: Schulkin, J. (Ed.), Action, Perception and the Brain: Adaptation and Cephalic Ex-
pression (New Directions in Philosophy and Cognitive Science). Palgrave
Iacoboni, M., Koski, L.M., Brass, M., Bekkering, H., Woods, R.P., Dubeau, M.C., Mazziotta,
J.C., Rizzolatti, G., 2001. Reafferent copies of imitated actions in the right superior
temporal cortex. Proc. Natl. Acad. Sci. U. S. A. 98, 13995–13999.
Jackson, P.L., Meltzoff, A.N., Decety, J., 2005. How do we perceive the pain of others? A
window into the neural processes involved in empathy. NeuroImage 24, 771–779.
Jackson, P.L., Brunet, E., Meltzoff, A.N., Decety, J., 2006. Empathy examined through the
neural mechanisms involved in imagining how I feel versus how you feel pain.
Neuropsychologia 44, 752–761.
Jastorff, J., Begliomini, C., Fabbri-Destro, M., Rizzolatti, G., Orban, G.A., 2010. Coding ob-
served motor acts: different organizational principles in the parietal and premotor
cortex of humans. J. Neurophysiol. 104, 128–140.
Kret, M.E., Pichon, S., Grezes, J., de Gelder, B., 2011. Similarities and differences in per-
ceiving threat from dynamic faces and bodies. An fMRI study. NeuroImage 54,
Lamm, C., Singer, T., 2010. The role of anterior insular cortex in social emotions. Brain
Struct. Funct. 214, 579–591.
Lamm, C., Nusbaum, H.C., Meltzoff, A.N., Decety, J., 2007. What are you feeling? Using
functional magnetic resonance imaging to assess the modulation of sensory and af-
fective responses during empathy for pain. PLoS One 2, e1292.
Lamm, C., Decety, J., Singer, T., 2010. Meta-analytic evidence for common and distinct
neural networks associated with directly experienced pain and empathy for pain.
NeuroImage 54 (3), 2492–2502.
Maldjian, J.A., Laurienti, P.J., Kraft, R.A., Burdette, J.H., 2003. An automated method for
neuroanatomic and cytoarchitectonic atlas-based interrogation of fMRI data sets.
NeuroImage 19, 1233–1239.
Mitchell, J., 2009. Inferences about mental states. Philos. Trans. R. Soc. Lond. B, Biol. Sci.
Molenberghs, P., Brander, C., Mattingley, J.B., Cunnington, R., 2010. The role of the su-
perior temporal sulcus and the mirror neuron system in imitation. Hum. Brain
Mapp. 31, 1316–1326.
Morrison, I., Lloyd, D., di Pellegrino, G., Roberts, N., 2004. Vicarious responses to pain in
anterior cingulate cortex: is empathy a multisensory issue? Cogn. Affect. Behav.
Neurosci. 4, 270–278.
Morrison, I., Peelen, M.V., Downing, P.E., 2007a. The sight of others' pain modulates
motor processing in human cingulate cortex. Cereb. Cortex 17, 2214–2222.
Morrison, I., Poliakoff, E., Gordon, L., Downing, P., 2007b. Response-specific effects of
pain observation on motor behavior. Cognition 104, 407–416.
Noordzij, M.L., Newman-Norlund, S.E., de Ruiter, J.P., Hagoort, P., Levinson, S.C., Toni, I.,
2009. Brain mechanisms underlying human communication. Front. Hum. Neurosci.
Northoff, G., Heinzel, A., de Greck, M., Bermpohl, F., Dobrowolny, H., Panksepp, J., 2006.
Self-referential processing in our brain—a meta-analysis of imaging studies on the
self. NeuroImage 31, 440–457.
Peelen, M.V., Downing, P.E., 2007. The neural basis of visual body perception. Nat. Rev.
Neurosci. 8, 636–648.
Preston, S.D., de Waal, F.B., 2002. Empathy: its ultimate and proximate bases. Behav.
Brain Sci. 25, 1–20 (discussion 20–71).
Raichle, M.E., MacLeod, A.M., Snyder, A.Z., Powers, W.J., Gusnard, D.A., Shulman, G.L.,
2001. A default mode of brain function. Proc. Natl. Acad. Sci. U. S. A. 98, 676–682.
Rizzolatti, G., Craighero, L., 2004. The mirror-neuron system. Annu. Rev. Neurosci. 27,
Rizzolatti, G., Fogassi, L., Gallese, V., 2001. Neurophysiological mechanisms underlying
the understanding and imitation of action. Nat. Rev. Neurosci. 2, 661–670.
Saarela, M.V., Hlushchuk, Y., Williams, A.C., Schurmann, M., Kalso, E., Hari, R., 2007. The
compassionate brain: humans detect intensity of pain from another's face. Cereb.
Cortex 17, 230–237.
Simon, D., Craig, K.D., Miltner, W.H., Rainville, P., 2006. Brain responses to dynamic fa-
cial expressions of pain. Pain 126, 309–318.
Singer, T., Lamm, C., 2009. The social neuroscience of empathy. Ann. N. Y. Acad. Sci.
Singer, T., Seymour, B., O'Doherty, J., Kaube, H., Dolan, R.J., Frith, C.D., 2004. Empathy for
pain involves the affective but not sensory components of pain. Science 303,
Singer, T., Seymour, B., O'Doherty, J.P., Stephan, K.E., Dolan, R.J., Frith, C.D., 2006. Em-
pathic neural responses are modulated by the perceived fairness of others. Nature
Urgesi, C., Moro, V., Candidi, M., Aglioti, S.M., 2006. Mapping implied body actions in
the human motor system. J. Neurosci. 26, 7942–7949.
Urgesi, C., Maieron, M., Avenanti, A., Tidoni, E., Fabbro, F., Aglioti, S.M., 2010. Simulating
the future of actions in the human corticospinal system. Cereb. Cortex 20,
Vachon-Presseau, E., Martel, M.O., Roy, M., Caron, E., Jackson, P.L., Rainville, P., 2011.
The Multilevel Organization of Vicarious Pain Responses: Effects of Pain Cues and
Empathy Traits on Spinal Nociception and Acute Pain. Pain.
van de Riet, W.A., Grezes, J., de Gelder, B., 2009. Specific and common brain regions in-
volved in the perception of faces and bodies and the representation of their emo-
tional expressions. Soc. Neurosci. 4, 101–120.
Williams, A.C., 2002. Facial expression of pain: an evolutionary account. Behav. Brain
Sci. 25, 439–455 (discussion 455–488).
Xu, X., Zuo, X., Wang, X., Han, S., 2009. Do you feel my pain? Racial group membership
modulates empathic neural responses. J. Neurosci. 29, 8525–8529.
Yarkoni, T., Poldrack, R.A., Van Essen, D.C., Wager, T.D., 2010. Cognitive neuroscience
2.0: building a cumulative science of human brain function. Trends Cogn. Sci. 14,
Zaki, J., Weber, J., Bolger, N., Ochsner, K., 2009. The neural bases of empathic accuracy.
Proc. Natl. Acad. Sci. U. S. A. 106, 11382–11387.
E. Vachon-Presseau et al. / NeuroImage 63 (2012) 54–62