Physiology & Behavior 140 (2015) 54–60

Contents lists available at ScienceDirect

Physiology & Behavior journal homepage: www.elsevier.com/locate/phb

Perceiving one's body shapes empathy Delphine Grynberg a,⁎, Olga Pollatos b a b

Research Institute for Psychological Sciences, Universite Catholique de Louvain, Louvain-la-Neuve, Belgium Health Psychology, Institute of Psychology and Education, University of Ulm, Germany

H I G H L I G H T S • • • •

Interoception and empathy share neural circuits such as the anterior insula. No prior studies explore whether interoception modulates affective and cognitive empathy. We showed that interoception predicts higher cognitive and affective empathy. Interoception is relevant to understanding and sharing of other people's emotions.

a r t i c l e

i n f o

Article history: Received 17 September 2014 Received in revised form 9 December 2014 Accepted 10 December 2014 Available online 11 December 2014 Keywords: Interoception Interoceptive sensitivity Empathy Pain Compassion

a b s t r a c t Background: Empathy is a basic human ability with affective and cognitive facets and high interindividual variability. Accurately detecting one's internal body signals (interoceptive sensitivity) strongly contributes to the awareness of oneself and is known to interact with emotional and cognitive processes. This study investigated whether interoceptive sensitivity (i.e., heartbeat perception task) shapes affective and cognitive empathy. Methods: Ninety-three participants were asked to report the valence of their feelings, as well as the degree of compassion, arousal, and distress they felt in response to pictures depicting other people in pain or in nonpain situations. Participants also had to estimate how painful the situation was. Results: Main results showed that greater interoceptive sensitivity enhanced the estimated degree of pain (cognitive empathy), as well as arousal and feelings of compassion (affective empathy), in response to painful pictures. Conclusions: The accurate perception of bodily states and their representation shape both affective and cognitive empathy. This perception enables us to feel more compassion for another person and to evaluate the pain that they experience as being more intense. © 2014 Elsevier Inc. All rights reserved.

1. Introduction 1.1. Empathy: definition In the eighteenth century, Hume and Smith used the word “sympathy” to refer to the ability to feel as the other person feels. Later, Lipps [1] used the term “Einfühlung” to describe a form of projection that artists could use to imagine what it would be like to be someone else or to be an inanimate object. The term Einfühlung (which literally means “feeling into”) was later translated into “empathy” by Titchener [2]. For Lipps, sympathy refers to the ability to project oneself onto another (“inner imitation”) in order to feel as the other person does. The projection is based on the perception of an emotional gesture of another person, which directly activates the same emotion in the perceiver (i.e., emotional ⁎ Corresponding author at: 10, place du Cardinal Mercier, 1348 Louvain-la-Neuve, Belgium. E-mail address: [email protected] (D. Grynberg).

http://dx.doi.org/10.1016/j.physbeh.2014.12.026 0031-9384/© 2014 Elsevier Inc. All rights reserved.

sharing). Nowadays, most definitions of empathy include an affective component [3], but there is still no agreement about the exact nature of the feeling. de Vignemont and Singer [4] argue that empathy occurs when the observation of the other's emotion induces a similar affective state in the observer. For Decety and Lamm [5], the affective response has to result from an “emotional sharing”, while Baron-Cohen and Wheelwright [6] suggest that the affective response does not have to be similar, but has to be adjusted and appropriate (e.g., feel sad for a friend who is angry). Finally, Batson [7] states that affective empathy encompasses any emotion that focuses on the well-being of the other person and thus refers to empathic concern, which is similar to compassion. It has been suggested that, in addition to the affective dimension, empathy also refers to the ability to take the perspective of others [8] and to understand others' feelings or, more generally, others' points of view [9]. There is now general agreement that the cognitive component of empathy relates to taking the perspective of others in order to understand and predict others' various mental states (e.g., thoughts, emotions).

D. Grynberg, O. Pollatos / Physiology & Behavior 140 (2015) 54–60

Notably, most research now suggests that empathy is a multidimensional concept that involves an affective and a cognitive subcomponent [3,10]. Evidence has shown that these subcomponents are associated and that they influence each other [3,11,12]. On the basis of these previous studies, we therefore define empathy as a multidimensional construct that involves both affective and cognitive components that refer, respectively, to the ability to share another person's emotional states and to infer that person's experiential states [4,13]. 1.2. Empathy: simulation and perception–action models According to the most influential model, i.e. the simulation model, empathy results from the simulation of another person's state [14] such that there is a match between the other's emotional state and the neural/body representations of this state in the empathizer. This model has been supported by several neuroimaging studies. They revealed the presence of a “shared neural circuit” involving the anterior insula (AI) and the anterior cingulate cortex (ACC), among other brain regions, which activate when one experiences pain and when one observes someone in pain [15,16]. The shared neural circuit is also proposed by the perception–action model, which is based on the existence of mirror neurons that are activated during the execution and observation of an action and on the existence of motor representations that allow us to understand, imitate, or prepare actions [17]. From these findings, some theorists [18] have postulated that a similar mechanism might underlie the empathic responses [18,19]. The perception–action model suggests that there might be a match between the representations of the empathizer's emotion and the target such that there is an unconscious and automatic activation of neural representations of emotional states that are similar to the emotional states of those observed. In other words, the attention allocated to the other's emotional state activates the representation of the emotional concept associated with this emotion [20]. Supporting this hypothesis, several studies have shown a neural overlap during the observation/decoding of either the experience or the expression (by imitation or spontaneously) of an emotion [15,21–25]. For instance, this shared neural circuit has been demonstrated during empathy for disgust, such that the AI is activated during the experience of disgust and when observing someone expressing disgust [25]. This overlap has been suggested to underlie the ability of the empathizer to experience the same feeling as the target's emotion. The empathizers might then introspect their feeling in order to understand it, and then attribute this feeling to the other. 1.3. Empathy and interoceptive sensitivity The perception–action model and the role of the shared neural circuit in empathy are supported by previous findings showing that at an intrapersonal level, the AI is a key structure underlying the ability to represent the state of the body and to perceive changes arising from the body as feelings and sensations, referred to as interoceptive sensitivity (IS [26]). Furthermore, at an interpersonal level, many studies revealed that empathy for pain tasks leads to greater activation in the AI/ACC [15,22,27] (see [28] for a meta-analysis), and some found that greater activation in the AI/ACC is associated with reports of stronger compassion responses (i.e., affective empathy) and higher ratings of pain intensity (i.e., cognitive empathy) in response to people experiencing pain [15,16]. Taken together, these findings thus suggest that there may be an interdependence between IS and empathy, specifically empathy for pain, such that this shared circuit enables individuals to activate their own body representations of pain when observing someone in pain, leading to stronger empathic responses. This hypothesis is in line with Singer et al.'s [13,15] assumption that impaired access to one's own emotional state may be associated with impaired simulation of the other's emotional state within the AI, leading to lower empathy. Specifically, they argue that because the AI/ACC allows mapping of the representation

55

of physiological activity predicted to be associated with an emotional response to a specific event, AI/ACC activation simulates how another person will emotionally respond to the same event. Although the hypothesis about the role of IS in empathy is theoretically and empirically driven, the empirical evidence remains indirect, and so far scant direct evidence is available to suggest that empathizing for someone depends on the level of IS in the empathizer. In relation to the affective dimension of empathy, previous research showed that higher IS is associated with a better understanding of one's own emotional experience and with a more intense emotional experience. Indeed, Herbert and colleagues [29] showed that greater IS is associated with lower levels of alexithymia. Furthermore, IS contributes to emotional feelings in terms of arousal: greater IS is associated with higher ratings of arousal [30,31] and with greater heart rate deceleration in response to emotional stimuli. Taken together, these studies indicate that IS and emotional experience are associated with each other. In fact, IS, the subjective experience of emotions, and cardiovascular arousal are underlain by similar brain regions (e.g., AI, ACC [30,32]). Thus, increased activation in these regions could explain the associations between high IS, low levels of alexithymia, and more intense emotional experience. To our knowledge, only four studies have investigated the links between IS and either affective or cognitive empathy. At a behavioral level, Terasawa et al. [33] found an association between greater IS and a lower intensity threshold of emotional facial expression, in response to which participants reported feeling an emotion. In contrast, Handford et al. [34] failed to find an association between IS and performance in decoding complex mental states expressed by eye gazes. At a neural level, Ernst et al. [35] showed that the activity of the AI during an affective empathy task was enhanced when participants were required to attend to their heartbeats for a short period. Finally, Fukushima et al. [36] used heartbeat-evoked potential (HEP) in order to investigate the association between empathy and the brain activity associated with the processing of the cardiovascular system (HEP) [37]. They showed that the amplitude of HEP was higher when participants evaluated the valence of emotional gazes and was positively correlated to higher empathic concern trait scores. Therefore, although constituting valuable first explorations of possible functional interdependence between IS and empathy, these studies present some limitations: (1) they focused on only one of the two dimensions of empathy state (either affective or cognitive); or (2) they did not directly investigate the level of IS (i.e., performance) of the participants; or (3) they did not investigate whether the effects are specific to empathy or are associated with all forms of affective experiences (i.e., arousal and distress feelings). In view of these limitations, it thus appears crucial to further explore whether IS modulates empathy responses. At a theoretical level, this study supports the perception–action model such that the activation of an extensive embodied representation of pain in individuals with high levels of IS would allow them to better understand the emotional state of the target and to share it. 1.4. Hypotheses In this study, we aimed to investigate the influence of IS (i.e., heartbeat perception task) on affective and cognitive empathy for someone in pain and on other forms of affective responses. Regarding affective empathy, we focused on compassion, which refers to other-oriented and regulated feelings in response to someone in a negative situation [38]. In terms of affective but not empathic responses, we focused on feelings of distress, which are self-oriented and non-regulated feelings [38]. Greater abilities to regulate emotional responses (measured by an index encompassing emotional control, emotional and behavioral inhibition, and attentional focus) are associated with higher ratings of empathic concern, whereas lower scores on this index are associated with higher reports of distress [38,39]. Therefore, because Fustos et al. [40] showed that higher IS is associated with better regulation strategies, we hypothesized that higher IS is associated with reports of higher compassion and lower feelings of distress

56

D. Grynberg, O. Pollatos / Physiology & Behavior 140 (2015) 54–60

for people expressing pain. Furthermore, we hypothesized that higher IS is associated with reports of higher arousal, but not valence, in response to pain-related stimuli, as shown in previous studies (e.g., [41]). Finally, because some studies found that greater activation in interoceptive circuits (AI/ACC) is associated with higher ratings of pain intensity (i.e., cognitive empathy) in response to people experiencing pain [15, 16], we hypothesized that higher IS modulates the cognitive dimension of empathy and is associated with reports of higher pain intensity ratings. 2. Methods 2.1. Participants Participants were recruited on campus and were screened for health status using a questionnaire about the participants' medical history. They were included only if they did not have a history of axis 1 disorders, in particular anxiety disorders or depression according to the Diagnostic and Statistical Manual of Mental Disorders (4th Ed.). Drug use (except contraceptives) was not allowed. On the basis of these criteria, three students were excluded from the study, and 93 (17 male; MAge = 23.00, SDAge = 4.10) were included. All participants gave their written informed consent and received 10€ for their participation. The experiment took place at the University of Potsdam, Department for Emotion and Motivation Psychology, between April and September 2011. It was conducted in accordance with the Declaration of Helsinki and with the approval of the local ethics committee. 2.2. Material 2.2.1. Stimuli One hundred pictures (50 pain situations, 50 non-pain situations) from the Potsdam Pain Battery served as stimuli. The Potsdam Pain Battery consists of pictures of pain situations and non-pain situations collected at the University of Potsdam (unpublished data); the stimuli have been evaluated using a student sample of 140 participants (MAge = 23.23, SDAge = 3.66; 30 males), who evaluated the pictures on a scale ranging from 0 (no-pain) to 9 (pain). Pain pictures depict painful situations, such as accidents or injuries, and human faces displaying painful expressions, whereas non-pain pictures consist of comparable situations in which no person is harmed (Fig. 1). We tried to match all pain pictures with corresponding non-pain pictures (e.g., an accident in the kitchen matched with a scene also located in

the kitchen) whenever possible. The pictures were reliably categorized as either pain or non-pain, as they were evaluated as expressing pain (M = 6.49; SD = 0.31; min = 6.16; max = 6.83) or no pain (M = 1.29; SD = 0.17; min = 1.09; max = 1.54), respectively. We used painful pictures for two reasons. First, pain is partly emotional. According to the International Association for the Study of Pain [42], pain is “an unpleasant sensory and emotional experience associated with actual or potential tissue damage, or described in terms of such damage”. Second, observation of pain in another person permits the investigation of affective (e.g., compassion) and cognitive dimensions (i.e., evaluation of the intensity of the pain) of empathy. We measured the intraclass correlation coefficient among the 140 participants who took part in the development of the battery in order to test interrater reliability. The two-way random “absolute agreement” model revealed good interrater reliability for pain (.87) and for non-pain (.84) pictures. All participants (i.e., those who validated the battery and participants in the present study) were confronted with the pictures for the first time. 2.2.2. Subjective ratings Valence and arousal: Participants had to report the valence of their feelings and the degree of arousal they felt in response to the person in pain or non-pain situations. The ratings were provided by using a nonverbal pictorial self-report scale, the Self-Assessment Manikin [43]. The ratings given by participants were converted to a numerical scale, such that the scores ranged from 1 (very unpleasant or low arousing) to 9 (very pleasant or high arousing). Estimation of the degree of pain: Participants had to estimate how painful the situation depicted was from 1 (not painful at all) to 9 (extremely painful). Compassion and distress responses: Because pain pictures may induce compassion as well as distress responses [27], we decided to measure both. The ratings were provided by using a German translation of two adjectives derived from a study by Batson and colleagues [44]. Participants were asked how compassionate (“anteilnehmend”; compassion) and how distressed (“bekümmert”; distress) they felt in response to the person in pain or non-pain depicted in the picture on a nine-point scale ranging from 1 (not at all) to 9 (extremely). 2.3. Heartbeat perception task The participants were seated in a sound-attenuated chamber and were fitted with physiological recording equipment for heart rate

Fig. 1. Pictures used for pain and non-pain stimuli.

D. Grynberg, O. Pollatos / Physiology & Behavior 140 (2015) 54–60

57

(electrocardiography). Signals were sampled at 500 Hz and analyzed by a computer-based data acquisition system (AcqKnowledge; BIOPAC Systems Inc., Santa Barbara, CA). We measured the IS of the participants by using a heartbeat detection task [45]. In this task, participants are asked to count their own heartbeats silently and to verbally report the number of counted heartbeats at the end of the counting phase. In accordance with the Mental Tracking Method [45], there were four heartbeat counting phases (varying in length: 15, 25, 35, and 45 s). The beginning and the end of the counting intervals were acoustically signaled by a beep. IS was estimated as the mean heartbeat perception score according to the following transformation [45]:

Table 1 Descriptive data of valence, arousal, estimated intensity of pain, and feelings of compassion and distress in response to pain and non-pain pictures.

Σð1–ðjrecorded heartbeats–counted heartbeatsjÞ=recorded heartbeatsÞ:

predict valence (FModel(3, 87) = 0.40; p N .05) and distress responses (FModel(3, 87) = 0.72; p N .05). Across all models, sex and age made no significant unique contribution (all ps N .05). Higher IS is thus likely to predict reports of higher arousal and compassion, as well as evaluation of the pain as being more intense.

1

/

4

The mean IS score across all participants was .75 (SD = .11; min = .48; max = .98).

Pain pictures

Valence Arousal Estimated intensity of pain Compassion Distress

Non-pain pictures

F(1, 92)

Mean

SD

Mean

SD

3.11 5.75 6.94 6.26 5.62

0.94 1.31 1.01 1.43 1.20

6.27 3.56 1.52 4.12 1.45

1.32 1.59 0.72 1.90 0.44

358.64⁎ 113.68⁎ 1641.50⁎ 70.72⁎ 1221.20⁎

⁎ p b .001.

2.4. Procedure 3.3. Non-pain pictures After the heartbeat perception task, the empathy paradigm consisting of 100 pictures presented in randomized order was started. Order presentation was controlled by the stimulus presentation software, with no explicit rules defined for the order of display of the 100 images. A single trial always began with a fixation cross for 500 ms followed by a variable interstimulus interval of 250–500 ms before a picture was visible for 5 s (7.5 × 15 cm or 7.5 × 15 cm). Participants were then asked to rate valence, arousal, and the level of compassion and distress they felt while watching the pictures. After a 20-second rating time, the next trial started. The whole experiment lasted about 1 h. 2.5. Statistical analyses In order to examine whether ratings of valence, arousal, estimated degree of pain, and feelings of compassion and distress differed between pain and non-pain pictures, we used repeated measures analysis of variance. We conducted hierarchical regression analyses by using SPSS 17.0, with IS score as the predictor of valence, arousal, estimated degree of pain, and feelings of compassion and distress in response to pain and non-pain pictures, after controlling for age and sex (0 = male; 1 = female). Finally, we used a test for skewness and for kurtosis (cut-offs: − 1, + 1) to examine the normality of the distributions. All factors were normally distributed, as all values were close to zero for skewness (−0.71 to 0.64) and for kurtosis (−0.74 to 0.68), except for compassion in response to pain pictures (skewness = −.98; kurtosis = 1.96) and for the evaluation of pain intensity for non-pain pictures (skewness = 1.39; kurtosis = 1.18). Therefore, these two variables were transformed in order to normalize their distribution (compassion in response to pain pictures was squared because the residuals presented a left skew [skewness = 0.10; kurtosis = 0.14]; pain intensity of non-pain stimuli was log-transformed because the residuals presented a right skew [skewness = 1.03; kurtosis = −0.09]). 3. Results 3.1. Descriptive data Table 1 demonstrates that, relative to non-pain pictures, pain pictures were evaluated as more negative, more arousing, and as expressing more pain; they also led to reports of higher compassion and distress. 3.2. Painful pictures Table 2 demonstrates that a high IS score predicted reports of higher arousal, compassion, and estimation of the intensity of pain, but did not

Regression analyses revealed a significant model, with being a female predicting greater estimated intensity of pain (β = − .32; t = − 3.04; p b .005; FModel(3, 90) = 3.37; p = .02; R2 = .10). No other models were significant (ps N .43). The statistical analysis revealed that relative to male participants, female participants evaluated the pain expressed in non-pain pictures as being more intense. 4. Discussion The aim of the study was to investigate whether IS modulates empathic responses to observing someone's pain. We also aimed to investigate the effect of IS on affective responses not specific to empathy. In this study, we examined whether IS, i.e., performance in the heartbeat detection task, modulates subjective responses to pain and non-pain pictures. More specifically, participants were asked to report the valence of their feelings, as well as the degree of compassion, arousal, and distress they felt in response to pictures depicting other people in pain or in non-pain situations. Participants also had to estimate how painful the situation was. Regarding empathy, we hypothesized that relative to those with a low IS score, those with a high IS score would report higher compassion and greater evaluation of the intensity of another person's pain. Regarding affective but not empathic responses, we hypothesized higher IS to be associated with reports of higher arousal and reports of lower distress. The results of this study demonstrated that higher IS is associated with reports of higher estimation of the intensity of pain (cognitive empathy) and with higher reports of compassion (affective empathy) and arousal, but not distress, in response to someone in pain. In terms of empathy, the present findings support our hypotheses and provide new evidence for the relevance of the perception of bodily signals, i.e., IS [35,36]. Moreover, this study is in line with previous findings showing reports of higher arousal in response to emotional stimuli in participants with higher IS (e.g., [31]). At an affective level, our findings thus suggest that the influence of IS on emotional responses goes beyond arousal to also concern feelings of compassion. At a cognitive level, we confirmed that interoception may modulate pain intensity ratings (i.e., cognitive empathy) in response to people experiencing pain [15,16]. The idea that bodily states and their representations shape emotion and cognition is a core characteristic of embodied cognition theories [46,47]. According to Barsalou [46], embodied representations of motor, somatosensory, affective, and interoceptive information have an important role in emotion and cognition. Internal states and knowledge acquired from introspection (i.e., the ability to be aware of having a personal experience) and interoception are as important as external experience for affective and cognitive processes. Several studies demonstrate

58

D. Grynberg, O. Pollatos / Physiology & Behavior 140 (2015) 54–60

Table 2 Regression coefficients for relationships among sex, age, interoceptive sensitivity, and responses to pain pictures.

Arousal

Estimated intensity of pain

Compassion

Sex Age IS Sex Age IS Sex Age IS

Beta

t

R2

FModel

R2Change

.10 −.10 .25 .03 −.15 .43 −.01 .18 .44

0.99 −0.95 2.33⁎ 0.35 −1.53 4.37⁎⁎ −0.09 −1.93 4.62⁎⁎

.03

F(2, 88) = 0.28

FChange

.09 .02

F(3, 87) = 2.71⁎ F(2, 88) = 1.12

.06

F(1, 87) = 5.43⁎

.20 .06

F(3, 87) = 7.26⁎⁎ F(2, 88) = 2.70

.18

F(1, 87) = 19.08⁎⁎

.25

F(3, 87) = 9.33⁎⁎

.19

F(1, 78) = 21.36⁎⁎

Note. IS = interoceptive sensitivity. ⁎ p b .05. ⁎⁎ p b .001.

that IS is a relevant variable for explaining interindividual differences in emotional and cognitive situations. Indeed, higher IS was associated with more intense feelings and higher activation of underlying brain structures during emotional stimulation [26,31,48]. Moreover, IS was also associated with cognitive functions such as selective attention or self-regulation during physical exercise [49,50]. IS may thus modulate the relationship between bodily responses and affective and cognitive variables. Taking these lines of evidence into account, we suggest that the affective and the cognitive dimensions of empathy might be modulated by differences in IS. Although the supposed association between interoception and empathy is appealing, empirical evidence has been lacking. This study is the first to demonstrate that IS modulates the affective and cognitive dimensions of empathy for pain. In relation to the embodied representations of pain, it has been shown that witnessing someone's pain leads to an enhancement of physiological responses [51] and that empathy in pain tasks leads to greater activation in the AI/ACC [15,22,27,28]. Moreover, some studies found that greater activation in the AI/ACC is associated with reports of higher compassion and of higher ratings of pain intensity in response to people experiencing pain [15,16]. This suggests that observing someone in pain might activate the visceral representations of pain in the observer. Furthermore, an association has been shown between high IS and enhanced sensitivity and decreased tolerance to pain induced by a pressure algometer [52] (for nonsignificant association between IS and thermal and distension pain sensitivity see [53,54]). This was interpreted by the authors as an activation of interoceptive representations of pain. Therefore, the present findings might be accounted for by more extensive embodied (visceral) representations of pain in high IS (versus low IS), which is reactivated when observing someone's pain, leading to greater emotional and cognitive empathy. At a theoretical level, although our findings are restricted to empathy for pain, they support the perception–action model, which suggests a match (imitation) between the state of the other person and the representations of this state in the empathizer [18]. According to this theory, there is an unconscious and automatic activation of neural representations of the state in the empathizer that are similar to those of the state that is observed. In other words, the empathizer resonates with the target's emotional state by synchronizing (imitating) his/her own representations with those of the target. The activation of these representations primes, or generates, the somatic responses associated with the state. In the present study, this would have allowed high IS scorers to better understand the emotional state of the target and to share their pain. The findings of this study also support the importance of the social context of the pain experience [55]. According to the communication model of pain [56], the way that people perceive and respond to others' pain (i.e., avoid or support) can be affected by the characteristics of the sender (e.g., catastrophizing), by the observer (e.g., accuracy of the recognition of pain), and by contextual factors (e.g., relationship between the sender and the observer) [56]. Our findings thus confirm the

importance of the observer characteristics, i.e., IS, in the pain experience [57] and these findings more broadly support its social dimension [55]. Regarding the underlying factors, our results—that people with high IS reported higher compassion and did not report more personal distress than did those with low IS—suggest that participants with higher IS may present a more effective emotion regulation capacity. It has been shown that emotion regulation is associated with greater reports of compassion and with lower reports of distress [38,39] and that reappraisal of negative pictures leads to greater modulation of neural activity in those with high IS than it does in those with low IS [40]. Although emotion regulation abilities were not measured in our study, we might hypothesize that it is because those with high IS report higher arousal and present higher emotion regulation abilities that they report more compassion than do those with low IS. Okun et al. [38] showed that compassion feelings are associated with greater reports of dispositional negative emotional intensity and emotion regulation. Alternatively, reports of higher compassion and no difference in terms of reports of personal distress in those with high IS may simply represent better distinction between oneself and others. Self/other distinction refers to the ability of two persons interacting together to preserve their individuality and to disentangle their own feelings from the feelings shared with the other [5]. Therefore, higher levels of self/other distinction allow us to distinguish between our emotions and those of others and may thus directly impact empathic responses. This allows us to know that the other person is the source of our affective state and to recognize that our perspective is different from the perspective of others. The ability to know who is actually feeling an emotion might thus impact affective responses. More precisely, higher self/other differentiation is linked with a higher level of empathic concern, and lower differentiation with a higher level of distress [5]. Regarding the association between IS and self/other differentiation, the modulation of body ownership illusions by IS [58] supports the idea that good heartbeat perceivers can more easily distinguish themselves from others (and therefore, presumably, another person's pain from their own). Therefore, it would be reasonable to expect that good self/other distinction is associated with more empathy but less personal distress, accounting for the present positive association between IS and compassion. Some methodological and statistical constraints have to be considered. First, the findings do not provide evidence about causal direction of interoception on empathy. Ernst et al. [35] investigated the causal relationship between attending to one's own heartbeats for a short period and the activity of the AI during a subsequent affective empathy task. However, they did not directly investigate the level of IS and this manipulation had no influence at a behavioral level. This suggests, therefore, that future studies are needed to better understand the causal role of IS on empathy. Second, as highlighted in the introduction, several definitions of empathy coexist, especially regarding its affective dimension. Although our hypotheses were based on the perception–action model, according to which simulation of the state of others leads to emotional sharing, we have focused on feelings of compassion and not on

D. Grynberg, O. Pollatos / Physiology & Behavior 140 (2015) 54–60

emotional sharing (i.e., the negative affect associated with painful experience). This is an important limitation, as the differences between these two affective responses have recently been highlighted at subjective [59] and neural levels [60]. Furthermore, although interoception modulates subjective reports of empathy, physiological evidence of intensity of emotional response to pain stimuli has not been recorded. Future studies should thus go beyond subjective reports and enlarge the present findings to emotional sharing and its physiological correlates. Third, studies should also investigate whether the association between IS and empathy will generalize to empathy for emotions other than pain (e.g., sadness). Furthermore, using low (arousal) pain stimuli might have been a better control condition than the pictures we used. Our results suggest that IS may impact only high negative arousal and not low arousal pictures. This thus supports the necessity for future studies to vary only the arousal of stimuli (i.e., high pain versus low pain). Finally, we did not measure the pain that was experienced by the participants, which may have primed the body representation of pain, especially among those with high IS. This could then have influenced the effect of IS on subjective responses. Valeriani et al. [61] showed, in support of the importance of one's own current body perception, that when observing someone in pain, neural activation depends on the intensity of the pain experience that is triggered in the observer. Future studies should thus investigate the effect of this possible confounding factor. 5. Conclusion This study showed that IS modulates affective and cognitive empathy and thus substantially extends, at an interpersonal level, former research underscoring the relevance of interoception for the intensity of emotions, as well as emotion and cognitive processing [31,48,62]. From the definition that empathy is characterized by affective (i.e., compassion feelings) and cognitive (i.e., identification of other's state) dimensions, we have demonstrated for the first time that both affective and cognitive empathy are shaped by interindividual differences in the ability to perceive internal signals. Disclosures No funding sources were provided. The authors report no conflict of interest. Acknowledgments We want to thank Jennifer Meyer for her help in programming the experimental design, Jürgen Füstös for his help in stimuli compilation, and Kevin Görsch, Julia Schneider, and Alexander Dreyer for their help in data acquisition. References [1] T. Lipps, Einfühlung, innere Nachahmung und Organempfindung (Empathy, inner imitation and sense-feelings), Arch. Gesamte Psychol. 1 (1903) 465–519. [2] E.B. Titchener, Lectures on the Experimental Psychology of Thought Processes, Macmillan, New York, 1909. [3] M.H. Davis, Measuring individual-differences in empathy — evidence for a multidimensional approach, J. Pers. Soc. Psychol. 44 (1983) 113–126, http://dx.doi.org/ 10.1037/0022-3514.44.1.113. [4] F. de Vignemont, T. Singer, The empathic brain: how, when and why? Trends Cogn. Sci. 10 (2006) 435–441, http://dx.doi.org/10.1016/j.tics.2006.08.008. [5] J. Decety, C. Lamm, Human empathy through the lens of social neuroscience, Sci. World J. 6 (2006) 1146–1163, http://dx.doi.org/10.1100/tsw.2006.221. [6] S. Baron-Cohen, S. Wheelwright, The empathy quotient: an investigation of adults with Asperger syndrome or high functioning autism, and normal sex differences, J. Autism Dev. Disord. 34 (2004) 163–175, http://dx.doi.org/10.1098/rstb.2002.1206. [7] C.D. Batson, Altruism in Humans, Oxford University Press, Oxford, NY, 2011. [8] G.H. Mead, Mind, Self & Society From the Standpoint of a Social Behaviorist, University of Chicago Press, Chicago, 1934. [9] R.F. Dymond, A scale for the measurement of empathic ability, J. Consult. Psychol. 13 (1949) 127–133, http://dx.doi.org/10.1037/h0061728.

59

[10] S.G. Shamay-Tsoory, The neural bases for empathy, Neuroscientist 17 (2011) 18–24, http://dx.doi.org/10.1177/1073858410379268. [11] M. Stel, A. van Knippenberg, The role of facial mimicry in the recognition of affect, Psychol. Sci. 19 (2008) 984–985, http://dx.doi.org/10.1111/j.1467-9280.2008.02188.x. [12] R.W. Levenson, A.M. Ruef, Empathy: a physiological substrate, J. Pers. Soc. Psychol. 63 (1992) 234–246, http://dx.doi.org/10.1037/0022-3514.63.2.234. [13] T. Singer, H.D. Critchley, K. Preuschoff, A common role of insula in feelings, empathy and uncertainty, Trends Cogn. Sci. 13 (2009) 334–340, http://dx.doi.org/10.1016/ j.tics.2009.05.001. [14] A. Goldman, In defense of the simulation theory, Mind Lang. 7 (1992) 104–119, http://dx.doi.org/10.1111/j.1468-0017.1992.tb00200.x. [15] T. Singer, B. Seymour, J. O'Doherty, H. Kaube, R.J. Dolan, et al., Empathy for pain involves the affective but not sensory components of pain, Science 303 (2004) 1157–1162, http://dx.doi.org/10.1126/science.1093535. [16] M.V. Saarela, Y. Hlushchuk, A.C. Williams, M. Schurmann, E. Kalso, et al., The compassionate brain: humans detect intensity of pain from another's face, Cereb. Cortex 17 (2007) 230–237, http://dx.doi.org/10.1093/cercor/bhj141. [17] G. Rizzolatti, G. Luppino, The cortical motor system, Neuron 31 (2001) 889–901, http://dx.doi.org/10.1016/S0896-6273(01)00423-8. [18] S.D. Preston, F.B. de Waal, Empathy: its ultimate and proximate bases, Behav. Brain Sci. 25 (2002) 1–20, http://dx.doi.org/10.1017/S0140525X02000018. [19] V. Gallese, The ‘shared manifold’ hypothesis — from mirror neurons to empathy, J. Conscious. Stud. 8 (2001) 33–50, http://dx.doi.org/10.1080/10481880903231910. [20] A.J. Hofelich, S.D. Preston, The meaning in empathy: distinguishing conceptual encoding from facial mimicry, trait empathy, and attention to emotion, Cogn. Emot. 1–10 (2011), http://dx.doi.org/10.1080/02699931.2011.559192. [21] L. Carr, M. Iacoboni, M.C. Dubeau, J.C. Mazziotta, G.L. Lenzi, Neural mechanisms of empathy in humans: a relay from neural systems for imitation to limbic areas, Proc. Natl. Acad. Sci. U. S. A. 100 (2003) 5497–5502, http://dx.doi.org/10.1073/ pnas.0935845100. [22] P.L. Jackson, A.N. Meltzoff, J. Decety, How do we perceive the pain of others: a window into the neural processes involved in empathy, NeuroImage 24 (2005) 771–779, http://dx.doi.org/10.1016/j. neuroimage .2004.09.006772. [23] C. Keysers, D.I. Perrett, Demystifying social cognition: a Hebbian perspective, Trends Cogn. Sci. 8 (2004) 501–507, http://dx.doi.org/10.1016/j.tics.2004.09.005. [24] C. van der Gaag, R.B. Minderaa, C. Keysers, Facial expressions: what the mirror neuron system can and cannot tell us, Soc. Neurosci. 2 (2007) 179–222, http:// dx.doi.org/10.1080/17470910701376878. [25] B. Wicker, C. Keysers, J. Plailly, J.P. Royet, V. Gallese, et al., Both of us disgusted in my insula: the common neural basis of seeing and feeling disgust, Neuron 40 (2003) 655–664, http://dx.doi.org/10.1016/S0896-6273(03)00679-2. [26] S. Wiens, Interoception in emotional experience, Curr. Opin. Neurol. 18 (2005) 442–447, http://dx.doi.org/10.1097/00019052-200508000-00015. [27] C. Lamm, C.D. Batson, J. Decety, The neural substrate of human empathy: effects of perspective-taking and cognitive appraisal, J. Cogn. Neurosci. 19 (2007) 42–58, http://dx.doi.org/10.1162/jocn.2007.19.1.42. [28] C. Lamm, J. Decety, T. Singer, Meta-analytic evidence for common and distinct neural networks associated with directly experienced pain and empathy for pain, Neuroimage 54 (2011) 2492–2502, http://dx.doi.org/10.1016/j.neuroimage.2010.10.014. [29] B.M. Herbert, C. Herbert, O. Pollatos, On the relationship between interoceptive awareness and alexithymia: is interoceptive awareness related to emotional awareness? J. Pers. 79 (2011) 1149–1175, http://dx.doi.org/10.1111/j.14676494.2011.00717.x. [30] H.D. Critchley, S. Wiens, P. Rotshtein, A. Ohman, R.J. Dolan, Neural systems supporting interoceptive awareness, Nat. Neurosci. 7 (2004) 189–195, http://dx.doi.org/10.1038/ nn1176. [31] O. Pollatos, B.M. Herbert, E. Matthias, R. Schandry, Heart rate response after emotional picture presentation is modulated by interoceptive awareness, Int. J. Psychophysiol. 63 (2007) 117–124, http://dx.doi.org/10.1016/j.ijpsycho.2006.09.003. [32] A.D. Craig, How do you feel? Interoception: the sense of the physiological condition of the body, Nat. Rev. Neurosci. 3 (2002) 655–666, http://dx.doi.org/10.1016/S09594388(03)00090-4. [33] Y. Terasawa, Y. Moriguchi, S. Tochizawa, S. Umeda, Interoceptive sensitivity predicts sensitivity to the emotions of others, Cogn. Emot. 28 (2014) 1435–1448, http:// dx.doi.org/10.1080/02699931.2014.888988. [34] C. Handford, J. Lemon, M.C. Grimm, U. Vollmer-Conna, Empathy as a function of clinical exposure—reading emotion in the eyes, PLoS One 8 (2013) e65159. [35] J. Ernst, G. Northoff, H. Boker, E. Seifritz, S. Grimm, Interoceptive awareness enhances neural activity during empathy, Hum. Brain Mapp. 34 (2013) 1615–1624, http://dx.doi.org/10.1002/hbm.22014. [36] H. Fukushima, Y. Terasawa, S. Umeda, Association between interoception and empathy: evidence from heartbeat-evoked brain potential, Int. J. Psychophysiol. 79 (2011) 259–265, http://dx.doi.org/10.1016/j.ijpsycho.2010.10.015. [37] O. Pollatos, R. Schandry, Accuracy of heartbeat perception is reflected in the amplitude of the heartbeat-evoked brain potential, Psychophysiology 41 (2004) 476–482, http://dx.doi.org/10.1111/1469-8986.2004.00170.x. [38] M.A. Okun, S.A. Shepard, N. Eisenberg, The relations of emotionality and regulation to dispositional empathy-related responding among volunteers-in-training, Personal. Individ. Differ. 28 (2000) 367–382, http://dx.doi.org/10.1016/S0191-8869(99) 00107-5. [39] D. Derryberry, M.K. Rothbart, Arousal, affect, and attention as components of temperament, J. Pers. Soc. Psychol. 55 (1988) 958–966, http://dx.doi.org/10.1037// 0022-3514.55.6.958. [40] J. Fustos, K. Gramann, B.M. Herbert, O. Pollatos, On the embodiment of emotion regulation: interoceptive awareness facilitates reappraisal, Soc. Cogn. Affect. Neurosci. 8 (2013) 911–917, http://dx.doi.org/10.1093/scan/nss089.

60

D. Grynberg, O. Pollatos / Physiology & Behavior 140 (2015) 54–60

[41] O. Pollatos, W. Kirsch, R. Schandry, On the relationship between interoceptive awareness, emotional experience, and brain processes, Cogn. Brain Res. 25 (2005) 948–962, http://dx.doi.org/10.1111/j.1467-6494.2011.00717.x. [42] H. Merskey, N. Bogduk, Part III: pain terms, a current list with definitions and notes on usage, Classification of Chronic Pain, Second Edition, IASP Task Force on TaxonomyIASP Press, Seattle, 2012, pp. 209–214. [43] M.M. Bradley, P.J. Lang, Measuring emotion: the Self-Assessment Manikin and the semantic differential, J. Behav. Ther. Exp. Psychiatry 25 (1994) 49–59, http:// dx.doi.org/10.1016/0005-7916(94)90063-9. [44] C.D. Batson, S. Early, G. Salvarani, Perspective taking: imagining how another feels versus imagining how you would feel, Personal. Soc. Psychol. Bull. 23 (1997) 751–758, http://dx.doi.org/10.1177/0146167297237008. [45] R. Schandry, Heart beat perception and emotional experience, Psychophysiology 18 (1981) 483–488, http://dx.doi.org/10.1111/j.1469-8986.1981.tb02486.x. [46] L.W. Barsalou, Grounded cognition, Annu. Rev. Psychol. 59 (2008) 617–645, http:// dx.doi.org/10.1146/annurev.psych.59.103006.093639. [47] P.M. Niedenthal, Embodying emotion, Science 316 (2007) 1002–1005, http:// dx.doi.org/10.1126/science.1136930. [48] B.D. Dunn, H.C. Galton, R. Morgan, D. Evans, C. Oliver, et al., Listening to your heart. How interoception shapes emotion experience and intuitive decision making, Psychol. Sci. 21 (2010) 1835–1844, http://dx.doi.org/10.1177/0956797610389191. [49] E. Matthias, R. Schandry, S. Duschek, O. Pollatos, On the relationship between interoceptive awareness and the attentional processing of visual stimuli, Int. J. Psychophysiol. 72 (2009) 154–159, http://dx.doi.org/10.1016/j.ijpsycho.2008.12.001. [50] B.M. Herbert, P. Ulbrich, R. Schandry, Interoceptive sensitivity and physical effort: implications for the self-control of physical load in everyday life, Psychophysiology 44 (2007) 194–202, http://dx.doi.org/10.1111/j.1469-8986.2007.00493.x. [51] G. Hein, C. Lamm, C. Brodbeck, T. Singer, Skin conductance response to the pain of others predicts later costly helping, PLoS One 6 (2011) e22759. [52] O. Pollatos, J. Fustos, H.D. Critchley, On the generalised embodiment of pain: how interoceptive sensitivity modulates cutaneous pain perception, Pain 153 (2012) 1680–1686, http://dx.doi.org/10.1016/j.pain.2012.04.030.

[53] B. Horing, H. Kugel, V. Brenner, S. Zipfel, P. Enck, Perception and pain thresholds for cutaneous heat and cold, and rectal distension: Associations and disassociations, Neurogastroenterol. Motil. 25 (2013) e791–802, http://dx.doi.org/10.1111/nmo.12207. [54] N. Werner, S. Duschek, M. Mattern, R. Schandry, The relationship between pain perception and interoception, J. Psychophysiol. 23 (2009) 35–42, http://dx.doi.org/ 10.1027/0269-8803.23.1.35. [55] T. Hadjistavropoulos, K.D. Craig, S. Duck, A. Cano, L. Goubert, et al., A biopsychosocial formulation of pain communication, Psychol. Bull. 137 (2011) 910–939, http:// dx.doi.org/10.1037/a0023876. [56] T. Hadjistavropoulos, K.D. Craig, A theoretical framework for understanding selfreport and observational measures of pain: a communications model, Behav. Res. Ther. 40 (2002) 551–570, http://dx.doi.org/10.1016/S0005-7967(01)00072-9. [57] L. Goubert, K.D. Craig, T. Vervoort, S. Morley, M.J. Sullivan, et al., Facing others in pain: the effects of empathy, Pain 118 (2005) 285–288, http://dx.doi.org/10.1016/ j.jpain.2005.03.004. [58] M. Tsakiris, A. Tajadura-Jimenez, M. Costantini, Just a heartbeat away from one's body: interoceptive sensitivity predicts malleability of body-representations, Proc. R. Soc. B Biol. Sci. 278 (2011) 2470–2476, http://dx.doi.org/10.1098/rspb.2010.2547. [59] T. Singer, O.M. Klimecki, Empathy and compassion, Curr. Biol. 24 (2014) R875–878, http://dx.doi.org/10.1016/j.cub.2014.06.054. [60] O.M. Klimecki, S. Leiberg, C. Lamm, T. Singer, Functional neural plasticity and associated changes in positive affect after compassion training, Cereb. Cortex 23 (2013) 1552–1561, http://dx.doi.org/10.1093/cercor/bhs. [61] M. Valeriani, V. Betti, D. Le Pera, L. De Armas, R. Miliucci, et al., Seeing the pain of others while being in pain: a laser-evoked potentials study, Neuroimage 40 (2008) 1419–1428, http://dx.doi.org/10.1016/j.neuroimage.2007.12.056. [62] N.S. Werner, K. Jung, S. Duschek, R. Schandry, Enhanced cardiac perception is associated with benefits in decision-making, Psychophysiology 46 (2009) 1123–1129, http://dx.doi.org/10.1111/j.1469-8986.2009.00855.x.