Rapid Communication The neural correlates of understanding the other’s distress: A positron emission tomography investigation of accurate empathy S.G. Shamay-Tsoory, a,b, * H. Lester, c R. Chisin, c O. Israel, a R. Bar-Shalom, a A. Peretz, d R. Tomer, a,b Z. Tsitrinbaum, a,b and J. Aharon-Peretz a a Rambam Medical Center, P.O.B. 9602, Haifa 3109, Israel b Department of Psychology, University of Haifa, Haifa 31905, Israel c Department of Radiology, Hadassah Hebrew University Hospital, Jerusalem, Israel d GE Healthcare, POB 170, Tirat Carmel, Israel Received 2 November 2004; revised 26 April 2005; accepted 4 May 2005 Available online 28 June 2005 The purpose of the present study was to assess the relationship between brain metabolism and empathic response. Six right-handed healthy volunteers were scanned with PETand fluorodeoxyglucose twice: during an interview about neutral story themes and during an empathic response eliciting interview about a story of a character in distress. Metabolic values in the medial and superior frontal gyrus, occipitotem- poral cortices, thalamus and the cerebellum were higher during empathic response than during the neutral theme interview. Further- more, the subjects’ empathy scores were positively correlated with metabolism in the medial aspects of the superior frontal gyrus. Our results suggest that empathy consists of both affective and cognitive components and hence may involve cortices that mediate simulation of emotional processing and mental state attribution. D 2005 Elsevier Inc. All rights reserved. Keywords: Empathy; Positron emission tomography; Theory of mind; Simulation Introduction Empathy is a fundamental part of the social fabric of emotion, providing a bridge between the feelings of one person and those of another. Because of its important role in human social interaction, empathy was suggested as a biological concept with an evolutionary necessity (Brothers, 1989). Indeed, impaired empathy is a central characteristic of frontal brain lesions (Eslinger, 1998; Eslinger and Damasio, 1985; Shamay-Tsoory et al., 2003), autism (Baron-Cohen and Wheelwright, 2003; Charman et al., 1997), Asperger’s syn- drome (Shamay-Tsoory et al., 2002), antisocial personality disorders (Blair, 2001) and schizophrenia (Frith and Corcoran, 1996). Some psychodynamic theories define empathy as a process of experiencing the inner life of another while retaining objectivity (Kohut, 1982), whereas, according to the humanistic theory, empathy involves sensing another’s feelings and emotions as if one were that person (Rogers, 1957). Echoing the different emphases of these definitions of empathy, two views have been proposed explaining how we understand another’s behavior. According to the theory of mind (ToM) perspective (Premack and Woodruff, 1978), mental states attributed to other people are conceived of as unobservable, theoretical posits and something akin to a scientific theory, whereas the Fsimulation_ perspective (Gallese and Goldman, 1999) suggests that the other’s mental states are represented by tracking or matching his emotional states with resonant states of one’s own. The core difference between ToM and simulation approaches to empathy is that, while ToM views empathy as a thoroughly Fdetached_ theoretical analysis involving brain regions usually activated during mental state attribution, simulation explains empathy as an attempt to replicate the other’s affective mental state via emotion processing-related neural networks. To date, there has been only one published report directly examining the functional brain imaging of empathic judgment (Farrow et al., 2001). In agreement with research on sympathy (Decety and Chaminade, 2003) and imitation (Carr et al., 2003), these authors report mainly inferior frontal activation, as well as premotor and ventromedial prefrontal involvement. Using posi- tron emission tomography (PET), we investigated the neural correlates of processing accurate empathy in response to another’s distress in normal individuals. We adopted an experimental design (previously described by Wang et al., 1999) using [F-18] fluorodeoxyglucose (FDG) PET twice: during an interactive interview about four neutral stories and during an interactive interview about four accurate empathic response-eliciting stories. In this paradigm, similarly to mood induction paradigms (e.g. Ottowitz et al., 2004) and in accordance with the course of FDG uptake (Reivich et al., 1979), a continuous prolonged state of 1053-8119/$ - see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.neuroimage.2005.05.012 www.elsevier.com/locate/ynimg NeuroImage 27 (2005) 468 – 472 * Corresponding author. Department of Psychology, University of Haifa, Haifa 31905, Israel. E-mail address: sshamay@psy.haifa.ac.il (S.G. Shamay-Tsoory). Available online on ScienceDirect (www.sciencedirect.com).