Introduction The concept of empathy has changed during the last two decades, as researchers tried
to refine its definition and understand the mechanisms underlying that ability. Likewise, the
means to assess empathy in individuals have also changed, making it difficult to interpret and
compare results of studies (Cuff, Brown, Taylor & Howat, 2014).
Regarding its definition, the main debate revolved around determining whether it
consisted of “feeling together with” others, sharing their emotions vicariously; understand the
emotional state of others, through emotion recognition and perspective-taking; or both
(Reniers, Corcoran, Drake, Shryane & Völlm, 2011).
The current perspective in literature states that, in fact, empathy should be considered
as a multidimensional construct, comprising one cognitive component, related to perceiving
and comprehending what others feel, and one affective component, consisting of an emotional
response to other people`s emotions (Jolliffe & Farrington, 2006; Dziobek et al., 2008;
Reniers et al., 2011; Cuff et al., 2014; Vachon & Lynam, 2015).
Neuroimaging studies support that perspective, indicating that there are different
regions of the brain associated to each component. Activation of insula, amygdala and inferior
frontal gyrus seems to be related to affective empathy, whereas activation of ventromedial and
dorsomedial prefrontal cortex, and cingulate gyrus, happens when a cognitive empathy
response is involved (Shamay-Tsoory, Aharon-Peretz & Perry, 2009; Dvash & Shamay-
Tsoory, 2014). There`s even evidence that greater grey matter density in those regions is
correlated with higher scores in measures of affective and cognitive empathy, respectively
(Eres, Decety, Louis & Molenberghs, 2015).
Understanding empathy in human beings is essential, since that ability is very
important to an adequate social functioning, having been associated with social competence
and prosocial behavior (Eisenberg & Miller, 1987; Cecconello & Koller, 2000; Rameson,
Morelli & Lieberman, 2012) and inhibition of antisocial behavior (Jolliffe & Farrington,
2004). Lack of empathy have been associated to bullying and aggressive behaviors, indicating
that interventions with individuals who present that kind of behavior should focus on trying to
improve their level of empathy (Miller & Eisenberg, 1998; van Noorden, Haselager, Cillessen
& Bukowski, 2015; Vachon, Lynam & Johnson, 2014).
Nonetheless, a systematic review points to the fact that several studies have failed to
observe correlation between lack of empathy and aggressive behavior, which could indicate a
problem with chosen measures of empathy (Vachon et al., 2014), since many instruments do
not contemplate the concept in its multidimensionality, or even assess other variables that are
not directly related to empathy, like non-emotional perspective-taking, social confidence, and
tendency to fantasize (Baron-Cohen & Wheelwright, 2004; Vachon et al., 2014).
Another problem comes out when studying Social Cognition, because there`s an
overlapping between the concept of Theory of Mind (ToM) and the cognitive dimension of
empathy. ToM was once defined as the ability to attribute mental states to other people
(Baron-Cohen et al., 1985) but, likewise empathy, ToM is also currently comprehended as a
multidimensional construct, consisting of a cognitive dimension, referring to the ability to
attribute thoughts, intentions and beliefs to other people, amongst other mental states; and an
affective dimension, related to inferring and understanding emotions of others (Kalbe et al,
2010; Shamay-Tsory & Aharon-Peretz, 2007), with neuroimaging studies also supporting that
division, by identifying shared and non-shared neural correlates (Schlaffke et al., 2015).
As may be noticed, the definition of affective ToM and cognitive empathy is
practically the same. Recent models based upon neuroimaging studies point to the possibility
that cognitive empathy may comprise both ToM dimensions (Dvash & Shamay-Tsoory,
2014), but since cognitive ToM theoretically involves other processes that are not directly
related to empathy, more research is necessary in order to clarify if those constructs are the
same or just work together. Some authors even debate about reconstructing the whole ToM
concept, since literature about it diverge when it comes to stablishing which exactly are its
subcomponents, which leads to the adoption of different methods of measurement and
inconsistent results (Schaafsma, Pfaff, Spunt & Adolphs, 2015).
Those considerations are essential when studying empathy, because, due to that
overlapping, there are many researches about ToM that should be included in reviews about
empathy and usually aren’t. For example, there`s evidence that children who perform worse
in ToM tasks are usually more rejected by peers (Villanueva, Clemente & Garcia, 2000) and
suffer more frequently with bullying and general violence inflicted by other children (Gini,
2006; Renouf et al., 2010), whereas good performance in ToM is associated with popularity
amongst colleagues (Slaughter, Imuta, Peterson & Henry, 2015) and social acceptance
(Slaughter, Dennis & Pritchard, 2002).
Sex differences have been pointed out regarding empathy, with females usually
scoring more than males in self-report scales (Wakabayachi et al., 2006; Rueckert & Naybar,
2008; Sampaio, Guimarães, Camino, Formiga & Menezes, 2011). However, it has been
argued that females might not be actually more empathic than males, but rather, tend to
perceive or report themselves as being more empathic in scales, since no significant
differences were found when neuroimaging and anatomical measurements, or behavioral
assessment, were included in studies (Derntl et al., 2010; Michalska, Kinzler & Decety,
Age effect on empathy has also been studied, with results indicating that empathy
tends to increase throughout an individual`s lifespan, until around age of 50 – 60, and then,
decrease again in old age, assuming an inverted U-shape (O’Brien, Konrath, Grühn & Hagen,
2012). Other studies, though, indicate that those differences might be limited to empathizing
with negative emotions (Blanke, Rauers & Riediger, 2015), or even that they are context-
dependent, because older people tend to empathize more with individuals or topics that are
relevant at their age, and the same would apply to younger people (Richter & Kunzmann,
2011; Wieck & Kunzmann, 2015). A longitudinal study revealed no significant differences in
empathy levels of participants throughout 12 years, although differences were found when
comparing groups of older and younger participants (Grühn, Rebucal, Diehl, Lumley &
Labouvie-Vief, 2008), which, according to the authors, might indicate that there isn`t an
actual effect of age on empathy, but instead, sampling biases. Other researchers sustain that
differences in methods and instruments used to assess empathy might be accounted for that
divergence in literature (Sun, Luo, Zhang, Li & Li, 2017). In any ways, further studies are
necessary in order to better investigate that aspect.
In this scenario, a possible solution for better understanding empathy in human beings
is to refine instruments that might properly assess that construct in its multidimensionality. Of
course, behavioral observation and anatomical measurements are important in studies,
especially to assess the affective component of empathy, but improving self-report scales is
also important, since researchers and clinical psychologists will not always have access to
such elaborate methods.
Aware of this necessity, Vachon and Lynam (2015) developed a new instrument, the