Perspective Chapter: Having Heart – The Different Facets of Empathy

Abstract

This chapter discusses the history of the various definitions of empathy and states two widely accepted current definitions for affective and cognitive empathy. The neural basis of different or overlapping cortical regions used by affective or cognitive empathy as well as sympathy/compassion are summarized, with the ventromedial prefrontal cortex as a probable common denominator for these emotions. Longitudinal studies of allopathic and osteopathic medical students confirm that women have higher affective and cognitive empathy scores than men, via the use of the Balanced Emotional Empathy Scale (BEES) and the Jefferson Scale of Empathy (JSE), respectively. During undergraduate medical education, BEES and JSE scores drop after the completion of the first basic science year and after the first year of clinical rotations. Students with higher empathy scores tend to enter primary care residencies, whereas students with lower scores are more likely to enter technical or procedure-oriented specialties. The ability to partially blunt an affective empathic response to an emotionally charged patient situation helps to ensure the health care provider can devote all their attention to the patient vs. the provider being caught up in their own emotions. Affective blunting may also be helpful in preventing burnout, especially among women health care workers.

1. Introduction

The study of “empathy” is complex, since this emotion is multifaceted and conflated with the terms sympathy and compassion. Accordingly, this chapter reveals how the word empathy has been defined over the past 100 years, and how researchers have explained empathy from a social point of view and, more recently, how it applies to the health professions. Accordingly, recently developed survey instruments that measure affective and cognitive empathy have been used to study empathy in individuals in the medical field. These scales also confirm the sociological evidence showing women have increased empathic behavior as compared to men. To determine the different types of empathy and associated emotions, numerous neuroimaging or lesions studies have parsed out the various central nervous system (CNS) regions that are used by cognitive or affective empathy, and sympathy/compassion. Other portions of the chapter show how affective and cognitive empathy scores change as allopathic or osteopathic medical students go through their 4 years of medical training. These empathy scores also help to reveal what residency specialty those students will most likely enter. Finally, information is provided how empathy is related to burn out by health care professionals.

2. Definitions of empathy

In a review on empathy, Engelen and Röttger-Rössler [1] stated, “Almost anybody writing in the field would declare that there is no accepted standard definition of empathy—either among the sciences and humanities or in the specific disciplines.” Ergo, there have been multiple definitions of empathy over the past century [2, 3]. At this point in time, it is generally accepted that empathy can be divided into two broad categories: affective or vicarious empathy and cognitive or role-taking empathy. This leads to the dilemma on whether these two aspects of empathy, i.e. understanding what another is thinking vs. understanding what the other is feeling, are separate or not, since the first is a cognitive function intimately relying on Theory of Mind (ToM) as part of the process [4, 5], whereas the latter is a more archaic emotional function that uses phylogenetically older parts of the CNS. Both of these aspects of empathy depend on the person recognizing, or identifying, the other person is equal in some respect, i.e. the person being viewed needs to be considered as yourself in whatever particular situation the other person is in.

Originally, the English word empathy was translated from the German word “Einfühlung” which means “feeling into.” This German word was used to describe the feelings one gets by observing things of beauty, e.g. artwork and nature, and appreciating their esthetics. The translation into “empathy” (from the ancient Greek empatheia (passion); composed of “en” (in) and “pathos” (feeling)) was coined by psychologist Edward Titchener in 1909 [6]. Titchener was familiar with, and influenced by, the work of David Hume who stated, “the minds of men are mirrors to one another” [7, 8]. As time advanced, various researchers, e.g. Lockwood [9], basically defined “empathy” as being able to vicariously experience and understand the affect another person is emoting. Therefore, Einfühlung is not only an emotional, cognitive state, where you understand what another feels, but it also evokes an affective, autonomic reaction within a person, a “gut feeling” if you will, by feeling what the other person is experiencing.

Engelen and Röttger-Rössler [1] define empathy as, “A social feeling that consists in feelingly grasping or retracing the present, future, or past emotional state of another.” In their definition, empathy is a vicarious emotion where the ability to separate feeling from comprehending is not clearcut. In 2012, H. Walter [10] probably proposed the most complex definition of empathy. He suggests that affective empathy is, “(a) An affective state that is (b) elicited by the perceived or imagined, or inferred state of the affective state of another; (c) is similar (isomorphic) to the other’s affective state; (d) is oriented toward the other; and (e) includes at least some cognitive appreciation of the other’s affective state comprising perspective-taking, self—other distinction, and knowledge of the causal relation between the self and the other’s affective state.” Walter also indicates that cognitive empathy “Refers to the ability to understand the feelings of others without necessarily implying the empathizer is in an affective state.” (Italics via Walter.) For further historical insights about the study of empathy, see these other excellent reviews [11, 12, 13].

3. The development of empathy survey instruments

The use of empathy throughout the development of human cultures promotes harmonious social interactions [11]. Since effective patient care is largely dependent on social interactions between the physician and the patient, its importance in the health professions became an issue of concern and considerable research over the past four decades. In this regard, social scientists took an interest in studying empathy and contributed their own definitions and measurement scales. In 1983, M.H. Davis [14] defined empathy using a multidimensional approach, i.e. both the affective and cognitive components, as the “reactions of one individual to the observed experiences of another.” In 1980, he developed the widely used Interpersonal Reactivity Index (IRI) which parses out four aspects of what he defines as the social aspects of empathy, i.e. social functioning, self-esteem, emotionality, and sensitivity to others [15]. Consequently, the IRI scale examines perspective-taking, which assesses the tendencies to spontaneously adopt the psychological viewpoint of others; Fantasy, which assesses the tendency to transpose themselves imaginatively into the feelings and actions of fictional characters in books, movies, and plays; Empathic Concern assesses other-oriented feelings of sympathy and concern for unfortunate others; and Personal Distress measures self-oriented feelings of personal anxiety and unease in tense interpersonal settings.

A large advancement in the field of cognitive empathy took place in 2002, when Dr. M. Hojat and colleagues [16] developed an excellent definition of what they term as “clinical empathy.” This definition of cognitive empathy is used in the context of health care professions education and patient care as, “A predominantly cognitive (rather than an affective or emotional) attribute that involves an understanding (rather than a feeling) of pain and suffering of the patient combined with a capacity to communicate this understanding and with an intention to help.” The four key terms in this definition are italicized by Hojat and colleagues to underscore their significance in the construct of patient care. Since that time, it has been accepted, almost worldwide, as the definitive definition of the use of cognitive empathy in health professions. This definition was the result of their development of the widely used “Jefferson Scale of Physician Empathy” [17]. (Note: The name has now been shortened to the “Jefferson Scale of Empathy” (JSE)).

Before Hojat’s seminal work on empathy, Hogan, in 1969, developed his scale that measures the cognitive aspects of empathy [18]. He defined empathy as, “The intellectual or imaginative apprehension of another’s condition or state of mind.” Even earlier, in 1949, Dymond devised an empathy scale and defined cognitive, role-taking empathy as, “An empathic person can imaginatively take the role of another and can understand and accurately predict that persons’ thoughts, feelings and actions” [19]. In contrast to scales examining cognitive empathy, or a combination of cognitive and affective empathy, in 1996, A. Mehrabian developed the Balanced Emotional Empathy Scale (BEES) that measures the affective aspect of empathy [20]. He defined emotional, i.e. affective or vicarious empathy, via work by Stotland [21], as, “a vicarious emotional response to the perceived emotional experiences of others.” In the context of personality measurement, it describes individual differences in the tendency to have emotional empathy with others [20, 22, 23]. Therefore, the BEES measures both components of emotional empathy, i.e. the vicarious experience of others’ feelings as well as the positiveness, adaptability, and affiliative aspects of a person in a balanced way [20].

4. Compassion and sympathy

“Compassion” is another word frequently used by health care professionals, and measurement scales have been developed [24]. In this regard, the definition of clinical empathy by Hojat and colleagues encompasses the meaning of compassion. Compassion, which etymologically means “to suffer with” [25], can be defined as, “A deep awareness of the suffering of another coupled with the wish to relive it.” and is incorporated within Hojat’s phrase “intention to help” [16]. Yet Post and colleagues and others [2, 26] argue that compassionate care is an intensification of the affective dimension of empathy regarding patient suffering. Therefore, it would seem difficult to have compassion for another being that is suffering without having some kind of effect upon the observer of the suffering individual.

“Sympathy” is another term frequently used by the general population as well as some health care professionals and social scientists, and scales have been devised to measure it [27]. Decety and Chaminade [28] define sympathy as, “The affinity, association, or relationship between persons wherein whatever affects one similarly affects the other.” Also, sympathy and pity are not the same. Whereas pity is feeling sorry for another, sympathy infers that you have a favorable impression of the other person. Both pity and sympathy seem to occur among individuals or groups with whom the observer is familiar, but it is not evoked from others who are not considered as a part of your group or whom you cannot identify with [1]. Studies by Post and colleagues [2] and Sinclair et al. [29] indicate that the use of sympathy by health care workers or friends of patients invokes a negative emotional reaction from the patient who discerns a sense of pity or misfortune accompanied by the feeling that they are being treated unfairly. Later in the chapter we’ll see that while some researchers equate affective empathy to sympathy [30], this is not entirely correct. The distinction between empathy and sympathy has been described by Hein and Singer as “feeling as and feeling for the other,” respectively [31]. Indeed, the development of the Adolescent Measure of Empathy and Sympathy (AMES) by Vossen and colleagues [32] shows that sympathy is more closely related to cognitive empathy vs. affective empathy, and that affective empathy and sympathy appear to be two different emotions. Therefore, sympathy, in contrast to compassion, does not necessarily evoke a need to help another individual in distress.

Recall, that the original definition of Einfühlung states that it can be induced by a feeling of awe or joy by viewing something that is esthetically beautiful to the beholder. This seems distant and somewhat unrelated to how the intertwined terms of affective and cognitive empathy, and sympathy and compassion are used today. This stresses the importance of each researcher to carefully define what they are studying so that equitable comparisons can be made among studies. For additional definitions of empathy, see the thorough review by Cuff and colleagues [3].

In summary, this chapter will use the definitions proposed by Mehrabian [20] for affective empathy, “A vicarious emotional response to the perceived emotional experiences of others.” Cognitive empathy will be used as it relates to health care professionals, so the definition by Hojat and colleagues [16] for what they term as “clinical empathy” will be used: “A predominantly cognitive (rather than an affective or emotional) attribute that involves an understanding (rather than a feeling) of pain and suffering of the patient combined with a capacity to communicate this understanding and with an intention to help.”

5. The neural basis of empathy

Studies using the JSE [17] to measure cognitive empathy or the BEES [20] to measure affective empathy show sex differences with women having higher JSE or BEES scores than men. For representative studies, see the following references [16, 20, 33, 34, 35, 36, 37]. These empathic sex differences have a neural basis. Briefly, during the emergence of humans, prosocial behavior, including empathy, gave an evolutionary advantage to those who possessed this trait. Additionally, paramount in the development of empathy was the ability for humans to distinguish “self” from “other.” For through reviews on these concepts, see [11, 38, 39]. Commensurate with self-other distinction, the ability to respond to pain or distress in others evoked the more phylogenetically archaic/emotional brain to provide a “low-road,” i.e. a bottom-up, vicarious response that evoked the desire to help the one in distress [10]. This was particularly true in females who needed to care for their helpless infants when they were perceived as being in distress. Indeed, this affective empathic trait is phylogenetically preserved across many species of animals because they all possess the archaic neural structures necessary for an affective emotional response to others in pain [38, 40]. These cortical areas, e.g. the inferior frontal gyrus, insular cortex, and parts of the cingulate gyrus, are known to be cytoarchitecturally more primitive (as defined by Broadmann areas) and are the first to mature in the human cerebral cortex [41]. It is important to note that an affective empathic response activates what is called the “pain axis” within the CNS [42, 43]. As such, the pain axis utilizes the phylogenetically older regions of the CNS, e.g. the amygdala, anterior insular cortex, and various regions of the cingulate cortex. For a more thorough review, see [44].

After the development of the low road vicarious empathic response, a “high-road,” i.e. a top-down empathic response, developed which utilized higher cognitive functions, including ToM that is closely related to cognitive empathy [4, 5, 10]. Thus, cognitive empathy uses cortical regions that developed later in primate phylogeny and includes the ventromedial prefrontal cortex, the temporoparietal junction, and the posterior aspect of the superior temporal sulcus [10, 12, 41].

5.1 Theory of mind and empathy

In a review by Walter [10], his figure 1 illustrates the CNS components that are involved with the “high road” (i.e. top-down, cognitive empathy) and the “low road” (i.e. bottom-up, affective empathy) and how these relate to cognitive ToM; with ToM being important in self-other distinction and the ability to represent and understand the mental states of others [4, 5]. During an empathic response, the affective/vicarious aspects of empathy arrive in the CNS before cognitive empathy is recruited. Intervening between affective and cognitive empathy is the use of ToM to put a self-other distinction on the experience. In brief, Walter proposes that the “lynch pin” between ToM and affective empathy is the use of the ventromedial prefrontal cortex involved with cognitive empathy. Other CNS regions are also implicated, ToM primarily uses the dorsomedial prefrontal cortex (supplemented by the superior temporal sulcus and the temporoparietal junction), while the affective aspects of empathy utilize the anterior insula and the medial cingulate cortex (supplemented by the inferior frontal gyrus, amygdala, and the secondary somatosensory cortex). For another more recent meta-analysis of ToM and empathy, see Schurz et al. [5].

5.2 CNS regions used by affective or cognitive empathy and sympathy/compassion

Table 1 reviews the regions of the CNS that play a role in affective empathy, cognitive empathy, and sympathy/compassion. The studies (see the references within Table 1) used either fMRI, transcranial magnetic stimulation, positron emission tomography (PET) scans, or various CNS lesions. Each of these methods has its own documented drawbacks, but collectively, the data show an overlap in regions that are consistently activated with each of the aforementioned emotional states. Note, that many of the studies are meta-analyses evaluating and summarizing numerous studies.

Affective/vicarious empathy
Cortical region	Studies referenced
Right & left anterior insular cortex	[10, 45, 46, 47, 48, 49] | [41, 50, 51] | [12]
Anterior and mid-cingulate gyrus	[10, 45, 46, 47, 49] | [50] | [12]
Inferior frontal gyrus	[46, 48, 49] | [41, 50, 51]
Right temporal pole	[41, 50, 51]
Right ventromedial prefrontal cortex	[51, 52]
Dorsolateral prefrontal cortex	[52]
Dorsomedial prefrontal cortex	[50]
Right dorsal anterior cingulate cortex	[48, 49]
Supplementary motor area	[48]
Right orbitofrontal cortex	[50]
Cognitive/role-taking empathy
Cortical region	Studies referenced
Ventromedial prefrontal cortex	[10, 45, 53] | [41, 51, 52]
Superior temporal gyrus/cortex	[10, 45, 53] | [41] | [12, 54]
Temporoparietal junction	[10, 45, 46, 53] | [41, 50]
Medial prefrontal cortex	[10, 46, 53] | [12, 54]
Temporal poles	[53] | [12, 54]
Orbitofrontal cortex	[48] | [54]
Dorsal anterior medial cingulate cortex	[48]
Posterior cingulate cortex	[10, 53]
Anterior insula	[48]
Supplementary motor area	[48]
Inferior frontal cortex	[54]
Sympathy/compassion
Cortical region	Studies referenced
Ventral tegmental area	[46]
Striatum	[46]
Nucleus accumbens	[46]
Amygdala	[28, 55]
Lateral or medial orbitofrontal cortex	[46]
Sympathy/compassion
Cortical region	Studies referenced
Ventromedial anterior cingulate cortex	[46] | [55]
Posterior medial frontal cortex	[56]
Inferior frontal gyrus	[28]
Right superior frontal gyrus	[28]
Ventromedial prefrontal cortex	[28]
Right temporal pole	[28]
Anterior insula	[55]

Table 1.

CNS regions implicated in affective or cognitive empathy and sympathy/compassion.

Bold font = Meta-analysis.

Italic font = Lesion study.

Regular font = fMRI, PET scan, or transcranial magnetic stimulation.

Regarding affective empathy, meta-analyses and lesion studies show that the right and left anterior insular cortex, the anterior and medial cingulate cortex, and the inferior frontal gyrus are heavily recruited. Several other lesion studies show that the ventromedial prefrontal cortex and the right temporal pole are also implicated in the affective aspect of empathy. The only region that was exclusive for affective empathy was the dorsolateral prefrontal cortex that was implicated in a lesion study [52].

The most frequently activated regions involved with cognitive empathy via meta-analyses or lesion studies include the ventromedial prefrontal cortex (which may overlap with the medial prefrontal cortex), the posterior superior temporal gyrus, and the temporoparietal junction that is involved with ToM [4]. Regions exclusive for cognitive empathy were those areas involved with ToM, i.e. temporoparietal junction and the posterior superior temporal gyrus. Another potential exclusive region was the posterior cingulate cortex which blends into the precuneus of the parietal lobe [10].

With the above being said, most of the cognitive empathy regions that overlap with affective empathy areas comes from a meta-analyses study [48] that mentions the left anterior insular cortex, anterior and middle cingulate cortex, and the supplementary motor cortex. Two lesion studies [51, 52] also show overlap via the ventromedial prefrontal cortex. A single fMRI study [54] indicates an overlap with the inferior frontal gyrus.

Two studies [28, 46] examined regions of the CNS activated by sympathy/compassion using a meta-analyses of fMRI data or PET scans. Although there is overlap with regions involved with both affective and cognitive empathy, there are also many regions involved with the CNS reward system that are exclusive to sympathy/compassion. These reward areas include the ventral tegmental area, the striatum, nucleus accumbens, and the amygdala. Overlap occurs with regions implied for affective empathy via the inferior frontal gyrus, parts of the cingulate cortex, the temporal poles, and the ventromedial prefrontal cortex. Cognitive empathy and sympathy/compassion overlap in the temporal poles, the orbitofrontal cortex, and the medial and ventromedial prefrontal cortex.

When examining affective vs. cognitive empathy vs. sympathy/compassion, there are two regions that seem to be involved in all three; these are the ventromedial prefrontal cortex and the temporal poles. However, the implication of the right temporal pole for affective empathy came from three lesion studies [41, 50, 51], whereas their involvement with cognitive empathy was via two fMRI studies [12, 54], and their involvement with sympathy/compassion was from a single study using a PET scan [28]. The fMRI study by Singer [12] examined both cognitive and affective empathy, but the superior temporal pole was only activated via cognitive empathy. The study by Schulte-Rüther and colleagues [54] only examined cognitive empathy. Therefore, the involvement of the right superior temporal pole with affective empathy must be viewed with caution, since lesion studies can, by nature, involve more areas of the cerebral cortex than the more “focused” studies using fMRI or PET scans. This leaves the prefrontal cortex, especially the right ventromedial region, as the only area being involved in affective and cognitive empathy as well as sympathy/compassion, since it is implicated in lesion studies [51, 52] (both of which used the BEES to show a decline in affective empathy scores due to the lesions), meta-analyses [10, 45], fMRI, and PET scans [28, 55].

In contrast to the above, a robust multi-level kernel density analysis (MKDA) study by Fan and colleagues [48] does not include the ventromedial prefrontal cortex as a region that is activated by both cognitive and affective empathy. Instead, Figure 3 by Fan et al. [48] shows that the left anterior insula was the only region to be activated by both types of empathy (sympathy/compassion was not examined). The right anterior insula was activated by affective empathy, while the left anterior mid-cingulate cortex was activated by cognitive empathy. Other regions that were involved with affective empathy (with an uncorrected MDKA threshold) are the right dorsal anterior cingulate cortex, the right dorsomedial thalamic nucleus, and the midbrain (most likely involving the periaqueductal gray). For cognitive empathy, the left medial orbitofrontal cortex and the left dorsomedial thalamic nucleus were involved. Their description of a core empathy network shows how the exteroceptive (affective) as well as the interoceptive (cognitve) aspects of empathy are interrelated [48]. The bottom-up, affective empathy areas use CNS networks that are involved with more primitive CNS regions involved with pain, disgust, and fear, whereas the cognitive aspects of empathy are using phylogenetically and cytoarchitecturally more recently developed CNS regions for a top-down empathic response.

In summary, mapping the CNS regions used by affective and cognitive empathy and sympathy/compassion is much like a Venn diagram with many sites potentially overlapping for all three emotions. Yet, each of these emotions have distinct regions that are reported to be activated using a variety of techniques, including lesion studies. Affective empathy activates the anterior insular cortex, the anterior and medial cingulate cortex, and the inferior frontal gyrus. Cognitive empathy, which uses areas involved with ToM, activates the ventromedial prefrontal cortex, the superior temporal cortex, the temporoparietal junction, and the medial prefrontal cortex. Sympathy/compassion is associated with the reward system and activates the ventral tegmental area, the striatum, amygdala, nucleus accumbens, and the medial orbitofrontal cortex. In many studies, it appears that the ventromedial prefrontal cortex is involved in all three emotions, and as Walter [10] suggests, it may be the common denominator among these emotions. In an imagined situation where a person is exposed to an emotionally charged setting, e.g. caring for a severely injured person in the Emergency Department, affective empathic responses enter the CNS of the health care provider first, these are modified by higher cortical regions involved with ToM, and then a cognitive empathic response is elicited.

6. Varied other factors can influence empathy

Studies have shown that many things can modify the empathy of an individual. Although an in-depth discussion of these factors is beyond the scope of this chapter, some of these can be briefly mentioned. Reviewed in a previous work [44] is how various CNS regions, especially those involved with the pain axis, are anatomically different between the sexes. In addition to anatomical sexual dimorphism, there is a sexually dimorphic recruitment of empathy-related cortical areas within the CNS, e.g. the amygdala, superior temporal sulcus, temporoparietal junction, and the inferior frontal gyrus [54, 57]. Walter [10] and others [58, 59] discuss the genetic aspects of empathy, especially in regard to the phylogenetically ancient hormones/neurotransmitters oxytocin and vasopressin, as well as reviewing how various genes and the environment may be implicated in empathic responses. Derntl et al. [60] reveal how trait empathy changes with the menstrual cycle; and another study shows how gonadal hormones influence empathy [61]. Schulte-Rüther et al. also show sex differences in how empathy is processed [62].

A study by Thirioux and colleagues [63] shows the time course of how various higher CNS regions are recruited in relation to cognitive empathy and sympathy, and how they relate to the mirror neuron system and the mentalizing/ToM network. Other studies show that as children develop, there are changes in which regions are used for cognitive and affective empathy as well as how strongly they are activated. In brief, younger-aged people use a more bottom-up response to an emotionally charged situation, and as they age, a more top-down approach is utilized as the cerebral cortex, especially the prefrontal cortex, matures after the teenage years [64, 65]. In a chapter by Newton [44], studies are reviewed that show how physicians empathically respond to noncompliant patients and to patients who are disparate from themselves. Finally, even an over-the-counter drug as ubiquitous as acetaminophen can reduce a person’s affective empathy for seeing another person in physical or social pain [66].

7. Affective and cognitive empathy among allopathic and osteopathic medical students

Newton [35] and Newton and colleagues [34, 67, 68] have performed a series of longitudinal studies examining how cognitive and/or affective empathy changes as allopathic or osteopathic medical students go through their 4 years of undergraduate medical education. These studies used the BEES and the JSE surveys to determine affective and cognitive empathy scores, respectively [17, 20]. These data were then compared to a population norm established by Mehrabian [20] and verified by Newton et al. [34], or to a recently published nationwide norm of osteopathic medical students in the US [33]. The longitudinal studies by Newton and colleagues also examined the residency specialty the students desired as they went through their 4 years of training, as well as their residency choice for their first postgraduate year after obtaining their medical degree.

The longitudinal empathy study performed at an allopathic school (University of Arkansas for Medical Science (UAMS)) showed that affective empathy, i.e. BEES scores, significantly dropped after the completion of the first year of basic science courses, as well as, unexpectedly, significantly dropping after the completion of their first year of clinical rotations [34, 69]. BEES scores did not significantly change after students completed their second basic science year. It was proposed this was most likely due to the students knowing they passed the first year, so there was a probable decrease in anxiety and/or they had refined their study techniques.

These data also confirmed the sexually dimorphic aspect of affective empathy with BEES scores of women being significantly higher than male scores [20, 34, 68]. Regarding desired specialty choice while in medical school, 23 residency specialties were segregated by Newton and colleagues [69] into “Core” and “Non-Core” groups. Core specialties are those that are generally considered as primary care specialties, i.e. those with a large amount of patient contact and continuity of care. The Core specialties are family medicine, internal medicine, pediatrics, Ob/Gyn, and psychiatry. Eighteen Non-Core specialties are more procedure or technically-oriented and have little or no patient contact or continuity of care, e.g. diagnostic radiology, emergency medicine, anesthesiology, or any of the surgical specialties. Men or women who desired to enter the Core specialties were better able to maintain their BEES scores vs. those allopathic students who desired Non-Core specialties [69]. Women who desired the Non-Core specialties had a 17.3% drop in BEES scores after completing their first year of clinical rotations, with some of them having BEES scores which approached the male norm. The greatest 4-year decline was for men in Non-Core specialties (38.7%) with women BEES scores in Non-Core specialties dropping by 29.3% over their 4 years of undergraduate medical training. Women desiring Core specialties had the smallest 4-year decline in BEES scores (13.0%), suggesting they were better able to maintain their affective empathy. (Note: only affective empathy was studied at UAMS since the JSE had not yet been developed).

Since the development of the JSE in 2001 [17], Newton and Vaskalis [67, 68] have completed data collection for another 7-year longitudinal empathy study examining both affective and cognitive empathy in osteopathic medical students at the Campbell University Jerry M. Wallace School of Osteopathic Medicine (CUSOM). These data have been presented in abstract form at the International Association of Medical Science Educators (IAMSE) annual meeting in 2015–2018 and 2022. [67, 68, 70]. (Manuscripts in preparation.) One of the main hypotheses of the CUSOM longitudinal study was to determine if osteopathic education, with its strong emphasis on the osteopathic philosophy of “mind, body and spirit” [71], combined with an emphasis on effective patient-physician communication skills, would result in a moderation of the drop in BEES scores with commensurate increases in JSE scores. Importantly, it is now possible to compare the CUSOM JSE data to students at other osteopathic schools since the publication of the ongoing nationwide Project on Osteopathic Medical Education and Empathy (POMEE) study data by Hojat and colleagues [33].

The osteopathic data from the Newton and Vaskalis studies [67, 68, 70] follow the same trends as what Newton and colleagues found with the allopathic students [34, 69]. Women have higher BEES and JSE scores than men, women entering into Core specialties better maintain their BEES and JSE scores than women who enter the Non-Core specialties, and male BEES and JSE scores tend to have larger drops in scores than those seen in women. Once again, the largest drops in BEES and JSE scores occurred after the completion of the first basic science year and the completion of the first clinical rotation year. The timing of the drops in CUSOM JSE scores closely match the reductions seen in JSE scores at other allopathic and osteopathic medical schools, e.g. see [72, 73, 74, 75].

7.1 BEES and JSE scores vs. residency specialty match

The UAMS Longitudinal Empathy Study showed the different residencies selected by allopathic graduates as determined by their BEES scores [76]. Out of the 23 possible types of residency specialties, 15 had seven or more students who entered each of those specialties. (Eight additional specialties were not selected by enough students to establish statistical power.) The BEES scores of the male or female students who entered into the Core residencies placed the five Core residencies in the top six positions. In descending order, they were Ob/Gyn, pediatrics, psychiatry, family medicine, and internal medicine. Anesthesiology was in the fifth position out of the 15 residencies selected. All of the BEES scores for the Core specialties were designated by Mehrabian [20] as “Average” (50th percentile) when compared to the population norm. (See table 1 for the Mehrabian designations in [35].) At the bottom of the 15 specialties were those male or female students who selected pathology or orthopedic surgery. Both specialties were ranked at the 16th percentile when compared to a normal population and had “Moderately Low” BEES scores. These data indicate that students with higher BEES scores preferentially entered the Core specialties that have considerable patient contact and continuity of care. Conversely, those students with lower BEES scores tended to enter the more technically oriented, Non-Core specialties with little or no patient contact or continuity of care.

BEES scores for osteopathic residency choice [35] resemble what was found in the allopathic study [76]. In the osteopathic longitudinal study, only 11 of the possible 23 residency specialties were selected by eight or more students. The top four positions, in descending order, were occupied by the Core specialties of pediatrics, family medicine, internal medicine, and Ob/Gyn. All of these were designated as “Average” according to Mehrabian [20], with the exception of men entering pediatrics where their combined BEES score was ranked at the 69th percentile and designated as “Slightly High.” In contrast to the allopathic data, psychiatry was eighth out of 11 slots and was designated as “Slightly Low” (31st percentile). The bottom two rankings were for anesthesiology and diagnostic radiology, both at the 16th percentile and had “Moderately Low” BEES scores. It is currently unknown why the BEES scores for allopathic students entering into anesthesiology placed the specialty in the fifth slot (out of 15) vs. the BEES scores for the osteopathic students that placed anesthesiology in the 10th slot out of the 11 possible selections. This was the largest discrepancy found for residency choice between the allopathic vs. the osteopathic graduates.

JSE scores were slightly different for those residency specialties entered by the osteopathic graduates [35]. The five Core specialties were in the top 6/11 positions: In descending order, they were pediatrics, family medicine, Ob/Gyn, psychiatry, and internal medicine. Orthopedics (only selected by men) and anesthesiology were the bottom two positions. What is concerning is that when compared to the POMEE norms for third- and fourth-year osteopathic students [33], only women CUSOM graduates who entered into psychiatry had a percentile score that was above the 50th percentile (59th); all other percentiles for the men and women in Core specialties were between the 29–44th percentiles. Percentile JSE scores for the bottom two selected specialties were even lower and were between 12th and 21st percentiles. So, overall, only 10 women out of a cohort of 345 CUSOM graduates had JSE scores that were above the 50th percentile when compared to the POMEE norm.

8. Is it possible to blunt affective empathy while increasing cognitive empathy?

One goal for any program that teaches health professionals should be to empower the students with the ability to have a certain amount of detached affective empathy. This is where the affective response is blunted so that the health care provider can attend to the patient without allowing the visceral reaction to an emotionally charged situation mentally distract them from providing effective care [44]. Yet, over blunting affective empathy can lead to the development of a health professional who has inadequate social skills and will appear and act in a distant, detached manner from the patient. In this regard, it has been shown that establishing an empathic bond of trust with patients leads to a lower incidence of malpractice suits, increased patient compliance, and better health outcomes [77, 78, 79, 80, 81]. Ergo, the converse would be detrimental to patient satisfaction and compliance. Therefore, being able to adequately quell the affective response allows the health care provider to respond in a calm, reassuring fashion to the patient by using ToM and cognitive empathy, thereby establishing a bond of trust with the patient.

In reference to the above, a further analysis of the data from the CUSOM longitudinal study plotted BEES scores vs. JSE scores [70]. Scatter plots were used (Figure 1) to determine which students raised or maintained their cognitive empathy, via JSE scores, while decreasing their BEES (Those points enclosed in the box). JSE scores had to be at or above the 75th percentile for the CUSOM cohort, while BEES scores had to decrease by −0.50 to −1.49 s.d. off the population norm established by Mehrabian [20]. BEES scores in this range would be considered as “Slightly Low” to “Moderately Low”. Decreasing BEES scores by ≥ −1.5 s.d. would place those students in the “Very Low” (7th percentile) to “Very Extremely Low” (0.6th percentile) portion of the population. These latter students would correspond to the scatter plot points to the far left and lower left in Figure 1. The BEES scores between −25 and − 75 suggest these individuals would have a very low affective response and may be perceived by patients as a callous, indifferent health care provider. In a similar fashion, those individuals who have BEES scores >80, i.e. ≥ +1.5 s.d., (top 7% of a normal population) may have an affective response that is too great and result in ineffective patient care. These students are represented by the upper right data points in Figure 1. In conclusion, only a small subset of students just before graduation (points enclosed in the box in Figure 1.) met the criteria of blunting affective empathy while raising or maintaining their cognitive empathy scores.

Figure 1.

Scatter plot of the CUSOM graduating classes of 2017–2019.

When examining all five data collection points, only 20/169 men and 21/176 women, i.e. < 12% of the cohort, qualified as blunting affective empathy while raising or maintaining cognitive empathy scores [70]. Furthermore, when examining the data from the first year through the end of the fourth year, 32 students qualified at only one of the five data collection time points, eight at two time points, and only one person qualified for four of the possible five time points. Out of the 41 qualifying osteopathic students, 24 entered Core residencies, while 17 selected Non-Core residencies. These data are disturbing in that the efforts to teach effective communication skills, via standardized patient encounters, osteopathic manipulation labs, mock mass-casualty incidents, and didactic sessions dealing with subjects such as end-of-life, giving bad news, spirituality, or even a session on empathy, did not have a substantial impact on the student empathy scores.

9. Empathy and burnout

It is possible that the large number of CUSOM students who have high BEES scores put themselves at risk of burnout, since it takes considerable cognitive effort to control affective empathic responses [82, 83, 84]. Many of the students/graduates with higher BEES and JSE scores prefer to enter into the Core, people-oriented, specialties [35, 76, 85]. Thus, having an increased amount of affective and cognitive empathy takes a toll on the health care provider, especially women, who have a greater rate of burnout than men [86, 87, 88]. Many women osteopathic graduates prefer to enter the Core specialties, especially pediatrics, Ob/Gyn, and family medicine at a rate of almost twice that of men [35, 85]. For the allopathic students at UAMS, only pediatrics and Ob/Gyn were predominated by women, once again at an almost 2:1 margin. Any allopathic or osteopathic student with blunted BEES scores may be better able to perform in the Non-Core, technical, or procedure-oriented specialties without an increased rate of burnout, with the caveat of not becoming too hardened. As an example, it has been shown that physicians who, by necessity, induce pain in their patients during a procedure blunt their affective response to the pain they are causing in their patients [89, 90]. This allows them to concentrate on the patient vs. thinking about the pain they are causing. For a more detailed discussion, see [44].

10. Conclusions

This chapter points out that over the years, there have been many definitions of empathy, and some of them incorporate sympathy or compassion into the definitions. Over the past few decades, there has been a consensus there are two basic types of empathy: an affective or vicarious empathy and a cognitive or role-taking empathy. Each of these two types of empathy employ some unique CNS regions which makes each type distinct. However, like a Venn diagram, there is overlap in many of the regions, and this may be due to the type of imaging being used, e.g. fMRI, PET scans, transcranial magnetic stimulation, and/or the empathy- or sympathy-inducing scenarios being used by the investigators. Other studies have used patients with various CNS lesions to determine regions used by empathy. Although some authors feel that sympathy/compassion is the same as affective empathy, various studies show that “reward” regions of the CNS were consistently, uniquely activated when the subject is in a sympathy study. Consequently, this author feels that sympathy is not the same emotion as vicarious/affective empathy. One region that was recruited by both affective and cognitive empathy, as well as sympathy/compassion, was the ventromedial prefrontal cortex. Thus, this region may be a key moderator of emotional responses.

Virtually, all studies show that women have significantly higher affective (BEES) and cognitive (JSE) empathy scores. Longitudinal studies by the author and colleagues at an allopathic and osteopathic medical school indicate that both affective and cognitive empathy drops after the completion of the first basic science year of study and after the first clinical rotation year. These studies also show that those students with higher BEES and/or JSE scores preferentially enter into the Core, people-oriented, specialties (i.e. family and internal medicine, pediatrics, Ob/Gyn, and psychiatry), while those students with lower empathy scores tend to gravitate toward the Non-Core, technical or procedure-oriented, specialties, e.g. diagnostic radiology, emergency medicine, anesthesiology, and surgical specialties.

Finally, the osteopathic longitudinal empathy study shows that educational efforts to boost effective communication skills and to partially blunt affective empathy are currently inadequate [70]. However, several other researchers are attempting to enhance cognitive empathic communication skills [91, 92, 93, 94] to increase patient satisfaction and compliance [95]; while others ask for additional neuroscience studies to help improve medical student empathy [96, 97]. Nevertheless, medical students or recent graduates who have high BEES and JSE scores may be at risk for increased rates of burnout, especially women. As stated in a previous chapter [44], “Physicians walk a fine empathic line to ensure they can relate to the patient without becoming too hardened themselves.”

Acknowledgments

Dr. Newton thanks the UAMS classes of 2001–2004 and the CUSOM classes of 2017–2019 for voluntarily participating in their respective Longitudinal Empathy Studies. He is also extremely grateful to Dr. Albert Mehrabian, emeritus professor of psychology, UCLA, for permission to use the BEES for educational purposes.