Edited by: Gennady Knyazev, State Scientific-Research Institute of Physiology and Basic Medicine, Russia
Reviewed by: Christopher Chad Woodruff, Northern Arizona University, United States; Sien Hu, State University of New York at Oswego, United States; Kai Yuan, Xidian University, China
†These authors have contributed equally to this work
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
Neural substrates of empathy are mainly investigated through task-related functional MRI. However, the functional neural mechanisms at rest underlying the empathic response have been poorly studied. We aimed to investigate neuroanatomical and functional substrates of cognitive and affective empathy. The self-reported empathy questionnaire Cognitive and Affective Empathy Test (TECA), T1 and T2∗-weighted 3-Tesla MRI were obtained from 22 healthy young females (mean age: 19.6 ± 2.4) and 20 males (mean age: 22.5 ± 4.4). Groups of low and high empathy were established for each scale. FreeSurfer v6.0 was used to estimate cortical thickness and to automatically segment the subcortical structures. FSL v5.0.10 was used to compare resting-state connectivity differences between empathy groups in six defined regions: the orbitofrontal, cingulate, and insular cortices, and the amygdala, hippocampus, and thalamus using a non-parametric permutation approach. The high empathy group in the Perspective Taking subscale (cognitive empathy) had greater thickness in the left orbitofrontal and ventrolateral frontal cortices, bilateral anterior cingulate, superior frontal, and occipital regions. Within the affective empathy scales, subjects with high Empathic Distress had higher thalamic volumes than the low-empathy group. Regarding resting-state connectivity analyses, low-empathy individuals in the Empathic Happiness scale had increased connectivity between the orbitofrontal cortex and the anterior cingulate when compared with the high-empathy group. In conclusion, from a structural point of view, there is a clear dissociation between the brain correlates of affective and cognitive factors of empathy. Neocortical correlates were found for the cognitive empathy dimension, whereas affective empathy is related to lower volumes in subcortical structures. Functionally, affective empathy is linked to connectivity between the orbital and cingulate cortices.
Empathy is the ability to understand the thoughts and feelings of others, and to respond to these feelings in an appropriate way. The construct of empathy can be divided into cognitive and affective dimensions (
In the past decades, there has been increasing interest in studying the neural basis of empathy with the emergence of magnetic resonance imaging (MRI) techniques in social neuroscience. Recent functional MRI (fMRI) studies have reported limbic structures such as the amygdala, the anterior insula, and the anterior cingulate cortex to be part of the neural bases of affective empathy (
Similarly, there are few published structural MRI studies addressing the neuroanatomical substrate of empathic ability measured by questionnaires. The majority of these studies used regions selected
To the best of our knowledge, no previous studies have investigated structural and functional dissociations of affective and cognitive empathy in the same sample of subjects. The aim of the present study was to investigate the structural and functional substrates of empathy in the same sample of healthy young persons. We were interested specially in differences within regions of the limbic system. We hypothesized that cognitive and affective empathy would present distinct regional cortical thickness patterns in neocortical regions and subcortical volumetric differences. Similarly, individuals would differentiate in their functional brain connectivity depending on empathy levels (low empathy and high empathy groups).
Fifty-six volunteers were recruited from advertising the study between students of the first course of the Nursing Bachelor of the University of Barcelona, Campus Clinic. These students were also invited to recruit friends or relatives of similar age and education. The inclusion criterion was that individuals would be between 18 and 35-years old. The exclusion criteria were: (1) presence of neurological or psychiatric disorders, (2) MRI incompatibilities such as metal implants that could not be extracted, (3) claustrophobia, (4) meeting DSM-IV criteria for substance abuse or dependence within the past year, and (5) current use of psychoactive medication.
Fourteen subjects were excluded due to the following reasons: 1 male met criteria of substance dependence, 1 male and 2 females were on psychoactive medication, 2 males and 3 females did not respond/did not show up on the day of the scan, 1 male and 3 females had a history of neurological or psychiatric disorders, and 1 female had MRI incompatibilities. Finally, 42 participants (22 females and 20 males) were included in the study. Additionally, for resting-state analysis 1 male was excluded due to excessive head motion.
Written informed consent was obtained from all participants after full explanation of procedures. The study was approved by the ethics committee of the Hospital Clinic of Barcelona. Subjects of this study were participants of an ongoing study funded by the Spanish Ministry of Science and Innovation (PSI2014-58004-P).
To exclude the presence of psychiatric disorders, the Mini International Neuropsychiatric Interview (
Briefly, the Perspective Taking scale assesses the intellectual ability of putting oneself in someone else’s place. The Emotional Understanding scale measures the ability of acknowledging and understanding the emotional states, intentions, and impressions of others. Within the affective scales, Emotional Distress is the ability of sharing others’ negative emotions, such as pain (
Scores were transformed into T scores as recommended in the TECA manual (
Magnetic resonance images were acquired with a 3T scanner (MAGNETOM Trio, Siemens, Germany), using an 8-channel head coil. The scanning protocol included high-resolution three-dimensional T1-weighted images acquired in the sagittal plane (TR = 2,300 ms, TE = 2.98 ms, TI = 900 ms, 240 slices, FOV = 256 mm; matrix size = 256 × 256; 1 mm isotropic voxel) and a resting-state 10-min-long functional gradient-echo echo-planar imaging sequence (240 T2∗ weighted images, TR = 2.5 s, TE = 28 ms, flip angle = 80°, slice thickness = 3 mm, FOV = 240 mm). Subjects were instructed to keep their eyes closed, not to fall asleep, and not to think anything in particular.
Cortical thickness was estimated using the automated FreeSurfer stream (version 6.0
Six subcortical volumes (amygdala, hippocampus, nucleus accumbens, thalamus, caudate, and putamen) and estimated total intracranial volume (eTIV) were obtained via whole-brain segmentation (
Basic functional image preprocessing, using AFNI
For connectivity analysis, based on previous literature we defined 6 regions of interest: the bilateral orbitofrontal, cingulate, and insular cortices, amygdala, hippocampus, and thalamus. The corresponding masks were extracted from the Brainnetome Atlas, which is built on functional and anatomical images
Regarding head motion parameters, an exclusion cut-off was established for mean interframe head motion at ≥0.3 mm translation or 0.3° rotation; and for maximum interframe head motion at ≥1 mm translation or 1° rotation. As described in the participants section, we excluded 1 male participant due to excessive head movement (maximum rotation: 3.06°).
In order to remove the effects of head motion and other non-neural sources of signal variation from the functional data, we used an Independent Component Analysis (ICA)-based strategy for Automatic Removal of Motion Artifacts (ICA-AROMA,
As quality control measure to assess the efficacy of ICA-AROMA in reducing relationship between signal variation and motion, we performed correlations between framewise head displacement (
Demographic and volumetric statistical analyses were conducted using IBM SPSS Statistics 25.0 (2011; IBM Corp, Armonk, NY, United States). We tested for group differences in demographics between females and males and between groups of high and low empathy for each test scale using the Mann-Whitney
Intergroup cortical thickness comparisons were performed using a vertex-by-vertex general linear model with FreeSurfer. The model included cortical thickness as a dependent factor and the low/high groups of empathy from each subscale as independent factors. Scores in the Vocabulary subtest from the Wechsler Adults Intelligence Scale-IV (
Group differences between groups of high and low empathy in subcortical volumes were tested with the Hotelling’s T-squared distribution test for multivariate ANOVA and
The first eigenvariate of the time series of all voxels included in each of the six masks described above (see Resting-state images section) was extracted with the fslmeants tool
There were no significant differences in demographical variables between gender groups. Females scored significantly higher than males in all the empathic scales except for the Perspective Taking scale. However, no gender differences were found between groups of low and high empathy (
Demographical and empathy variables.
Males ( |
Females ( |
Test stat | Total sample ( |
||
---|---|---|---|---|---|
Age, median (IQR) | 22.5 (8.0) | 19.0 (2.0) | 150.5001 | 0.071 | 19.0 (5.0) |
Education, years, median (IQR) | 13.5 (6.0) | 13.00 (2.0) | 184.5001 | 0.353 | 13.0 (4.0) |
Vocabulary test∗, median (IQR) | 38.0 (8.0) | 38.0 (6.0) | 224.0001 | 0.714 | 38.0 (6.0) |
TECA total score, mean (SD) | 114.2 (12.0) | 132.4 (13.4) | 4.6412 | <0.001 | 123.7 (15.6) |
TECA total, low/high (%) | 10 (50.0) / 10 (50.0) | 6 (27.3) / 16 (72.7) | 0.204 | 0.116 | 16 (38.1) / 26 (61.9) |
Perspective taking, mean (SD) | 31.4 (4.5) | 33.4 (4.3) | 1.4922 | 0.144 | 32.4 (4.4) |
Perspective taking, low/high (%) | 8 (40.0) / 12 (60.0) | 6 (27.3) / 16 (72.7) | 0.515 | 0.293 | 14 (33.3) / 28 (66.7) |
Emotional understanding, mean (SD) | 31.9 (5.7) | 35.6 (4.2) | 2.4402 | 0.019 | 33.9 (5.2) |
Emotional understanding, low/high (%) | 9 (45.0) / 11 (55.0) | 7 (31.8) / 15 (68.2) | 0.527 | 0.288 | 16 (38.1) / 26 (61.9) |
Empathic distress, mean (SD) | 19.5 (6.5) | 28.0 (5.7) | 4.4942 | <0.001 | 24.0 (7.4) |
Empathic distress, low/high (%) | 16 (80.0) / 4 (20.0) | 13 (59.1) / 9 (40.9) | 0.190 | 0.129 | 29 (69.0) / 13 (31.0) |
Empathic happiness, median (IQR) | 32.0 (7.0) | 36.5 (7.0) | 333.0001 | 0.004 | 34.0 (7.0) |
Empathic happiness, low/high (%) | 11 (55.0) / 9 (45.0) | 9 (40.9) / 13 (59.1) | 0.537 | 0.273 | 20 (47.6) / 22 (52.4) |
Whole-brain cortical thickness comparisons showed that subjects grouped in the high Perspective Taking scores (cognitive empathy) showed thicker cortex in left lateral and medial orbitofrontal gyrus, lateral pars opercularis extending to pars triangularis, and inferior frontal gyrus, as well as in bilateral medial superior frontal, anterior and middle cingulate gyrus, and lateral and medial occipital regions (
Cortical thickness differences between groups of low and high empathy within the Perspective Taking subscale.
Subcortical volumetric analyses showed that participants grouped in the high Empathic Distress scale had significantly higher bilateral thalamus volumes than the low-empathy group, although the multivariate test was not statistically significant (
For the resting-state permutation testing analyses, no head motion parameters (e.g., framewise displacement, rotation, and translation) were considered as covariates since there were no significant differences between high and low empathy groups.
Significant differences were found in one of the affective empathy scales, the Empathic Happiness. Subjects classified as having low empathy had increased connectivity between the bilateral orbitofrontal and the anterior cingulate regions (x,y,z MNI coordinates: 12,45,3; 67 voxels in the cluster; max
Resting-state connectivity group differences in the Empathic Happiness subscale. ACC, anterior cingulate cortex (MNI coordinates). Represented in yellow, the orbital mask from the Brainnetome atlas and in green the cluster that reached statistical significance at
The novelty of this study is the characterization of distinct neuroanatomical and functional correlates of cognitive and affective empathy in the same sample of healthy young adults. Overall, our findings showed that the orbitofrontal and cingulate cortices were related to both empathic dimensions. Higher cognitive empathy was associated with orbitofrontal thickening extending to the ventrolateral prefrontal cortex and bilateral superior frontal, cingulate, and occipital cortices. On the other hand, high negative affective empathy was linked to higher bilateral thalamus volumetry. The low positive affective empathy group had higher connectivity at rest between the bilateral orbitofrontal and cingulate cortices.
According to our structural and functional results, the orbitofrontal and anterior cingulate cortices seem to be key structures involved in empathy. However, cortical thickness was better able to discriminate between individuals with high and low empathy than resting state functional connectivity. The fact that the structural neuroanatomical information was more informative than functional connectivity is compatible with the notion that empathy was measured as a personality trait (e.g., “to understand how another person feels is something really easy to me”) rather than a state (
Group differences in cortical thickness between groups of high and low empathy within the Perspective Taking subtest were observed in both medial and lateral orbital cortices. Previous structural MRI studies also pointed to the dorsomedial prefrontal cortex as an anatomical substrate of empathy in healthy subjects using cortical thickness measures both studying cortical parcellations (
In pathological conditions, cortical thickness correlates in lateral and medial prefrontal cortices have been described. For example, in pathological narcissism, which is characterized by arrogant behavior and lack of empathy, volumetric reductions, and cortical thinning in the right dorsolateral prefrontal cortex have been found (
Structurally, we also found thicker bilateral cingulate cortex in the group with high perspective-taking empathy than in the low-empathy group. Positive correlations between cognitive empathy and the anterior cingulate (
In the present study, group connectivity differences were found between the bilateral orbitofrontal with the anterior cingulate in the scale evaluating positive affective empathy. Within the limbic system, the cingulate cortex has been reported as a hub region, defined as a region that integrates different brain processes (
In our results, we also found regional thickness differences in bilateral occipital cortex involving the pericalcarine and lingual gyri using whole-brain cortical thickness analysis. Since most of the structural studies performed with healthy subjects have focused on regions selected
In the current study, analyses of volumetric subcortical structures showed that participants with high Empathic Distress scale scores had significantly higher bilateral thalamus volumes compared with those in the low-empathy group. Previous MRI structural studies have also found a link between the thalamus and both affective and cognitive empathy. One study investigated the neural correlates of both empathy and alexithymia in a sample of healthy participants and reported the thalamus, together with other structures including left amygdala, hippocampus and parahippocampal gyrus as significant correlates of both constructs (
Global TECA scores did not reveal any significant differences between groups either in cortical or subcortical structures or in functional connectivity, thus emphasizing the importance of differentiating between the cognitive and the affective empathy, with different underlying neural bases. Sex differences were found in all TECA subscales and global scores as previously reported in the literature (
Previous studies mainly focused on
The main limitation is the small sample size. Unlike in the whole-brain approach used in the structural analysis, we decided to select
In conclusion, we found that structural differences between individuals with high and low empathy are more marked than functional ones. Cognitive empathy had clear correlates with cortical structures, namely medial and lateral prefrontal cortices and associative occipital ones. For affective empathy, only a link with the thalamus was observed. However, in the absence of neuroanatomical differences in positive affective empathy, individuals with low empathy showed increased orbital functional connectivity with the anterior cingulate.
The study was approved by the ethics committee of the Hospital Clinic of Barcelona. Subjects of this study were participants of an ongoing study funded by the Spanish Ministry of Science and Innovation (PSI2014-58004-P). Written informed consent was obtained from all participants after full explanation of procedures.
CJ contributed in the design of the study. CU, AP-D, and AA contributed to the analysis of the data. CU, AP-D, AA, HB, BS, and CJ contributed to the interpretation of the data, revised the manuscript critically for important intellectual content, and approved the final version of the manuscript. AP-D wrote a first draft. CU and CJ modified the first draft of the article.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
We thank Dr. Antonio Guillamon for facilitating the MRI data for the study. We appreciate the cooperation of the participants. We are also indebted to the Magnetic Resonance Imaging core facility of the IDIBAPS for the technical support, especially to C. Garrido and G. Lasso and we acknowledge the CERCA Programme/Generalitat de Catalunya.
The Supplementary Material for this article can be found online at: