The preliminary validation of the Greek version of the Toronto Composite Empathy Scale (TCES) demonstrated good validity and reliability (Cronbach's α = 0.895, Sig. from Hotelling’s T-Squared Test < 0.000) among medical students and could be further used in case of samples of medical students. The students that participated in our study had a moderately high empathy. As presented below, this is not surprising when measuring empathy among medical students, especially in a Western-oriented country, and given that altruism may be an important motivation for medical students to study medicine.
Furthermore, we examined differences in empathy scores by gender, class year, and carrier intention, love for animals, interest in medical ethics, belief in God, having an ill person in the family. For the most part, the finding of this study support the findings of previous literature. We found clear gender differences in empathy among the participants in the study. Moreover, the students that participated in our study demonstrated empathy decline over the years as medical studies progress. Note, however, that we found a positive correlation between empathy and factors that has not been widely explored or discussed in previous literature relative to the topic of interest, such as love for animals, interest in medical ethics, belief in God, having an ill person in the family. To help readers grain some deeper insight into subjects related to our findings, we attempted to present below a comprehensive summary of previous research on these topics.
Decline of students’ empathy during medical studies
The decline of empathy during medical studies
Research has shown a striking decline in empathy over the years as medical studies progress [9, 49, 56, 75–84]. More specifically, several studies have shown a steep decrease in empathy among students during their third (clinical) year of medical school, namely, as they begin their clinical training and hence empathic communication is critical [85, 29, 77, 86]. Mirani, Shaikh, Tahir found that empathy levels were “highest in first-year students and lowest in final-year students” [87]. Importantly, it is suggested that general empathy declines less than clinical empathy [88]. In this study, in consistency with previous literature, we found a statistically significant important empathy decline among medical students over the course of medical school.
The pattern of decline in empathy
The pattern of decline in empathy as medical studies progress may be different for males and females or for the various components (dimensions) of empathy as well. To cite just one example, Quince et al. demonstrated no change in the cognitive aspect of empathy during the studies (neither for men nor for women). Furthermore, women’s affective empathy demonstrated no change, while men's affective empathy declined slightly during the studies [89]. The decline in empathy during medical studies may be viewed as emotional neutralisation, namely, cynicism [51]. Hojat et al. stated that during the medical studies there may be “escalation of cynicism and atrophy of idealism” [77]. Moreover, note that there may be conflicting patterns in empathy change demonstrated by studies conducted within the same region or between wider geo-sociocultural regions (i.e., between the West and the Far East in the U.S.). A probable explanation for this phenomenon may be an interplay between the general and specific aspects of empathy [36]. At any rate it should be stated that while the decreasing trend of student’s empathy over the course of medical school is a multi-factorial and complex process [32], it “is not as indiscriminate (patternless) as once thought” [36].
Challenging the empathy decline
Studies failed to support the hypothesis of a strong and generalisable trend of decline in empathy over the course of medical school [23, 36, 53, 81, 90–94] or found only declines in some aspects of empathy [95]. Surprisingly, authors reported an increase in (clinical) empathy during medical studies [96]. Handford et al. found an increase in the behavioural component of empathy [95]. Interestingly, Chatterjee et al. found that empathy “fell from the first to the third semester, then more or less plateaued, and then rose again in the seventh semester” [43]. Interestingly, it has been argued that this decline may be due to differences in methodology used in research. Note that relative studies are based largely on studies using self-administered tools [23]. The decline in empathy may reflect self-representation changes [84]. Importantly, Ferreira-Valente et al. who contested the reliability of the decline of empathy as medical students begin their clinical training, conducted a review and concluded that the decline in empathy may be due to differences across the conducted studies in design, tools used and sample sizes [97]. The authors identified differences between cross-sectional and longitudinal studies [97]. The fact that physicians often are fully aware of the behavioural changes in themselves and their colleagues [49], may affect study results.
Factors affecting empathy decline
Culture and the educational background of students at admission may affect the trajectory of empathy levels across the years of medical education. While the decline of empathy levels as medical students progress has been noticed in many studies conducted in the United States, Europe or China, this was not the case with studies conducted in other contexts (i.e. Japan, Iran, Ethiopia, Portugal). This may be due to different cultural contexts or students’ educational backgrounds at admission [17, 36]. Note, however, that in a study, Pakistani medical students showed lower levels of empathy as compared to medical students in Western countries [98]. Empathy is a locally construed global construct [36].
Several and complex factors may drive or tackle the decrease of empathy trajectory during medical studies. Understanding the drivers of empathy decline is an important but difficult task. Not surprisingly, “determining with certainty whether it is more likely to change or to remain stable throughout medical studies has proven to be inconclusive” [99]. The lower the students’ empathic ability at the beginning of their studies the stronger the decline of their empathy levels during their studies [56, 99]. Students with high empathy levels may show not decreasing but stable trajectory during their studies [99]. Moreover, the students’ personality in the entry year may play a crucial role in forming the empathy trajectory during their studies [99].
Studies have shown relationship between medical students' personality and empathy [100–103]. Furthermore, it is suggested that medical students’ attachment styles predict empathy dimensions” [104].
Medical students’ altruistic motives for studying medicine when entering medical school are associated with greater empathy [105, 106]. The promotion of altruistic and interpersonally oriented motives may tackle the decrease of empathy trajectory during medical studies [99]. Empathy is considered a “motivated phenomenon” “in which people may choose to experience or avoid the process of understanding other people’s emotions” [107]. Findyartini et al. explored the relationship between empathy and students’ motivation types [25]. Kusurkar et al. provided four motivational profiles. They state that “there are different theories of motivation; some focus on quantity of motivation and others on quality” [108]. The authors provide four types of extrinsic motivation: integrated regulation, identified regulation, introjected regulation, external regulation [108]. Duan argues that motivation affects the aspects of empathy in different ways [109]. The author suggests that sadness of the target person causes increased cognitive empathy whereas the happiness of the target person causes increased affective empathy [109]. Further research is required to further explore the cited in the literature factors and identify factors that contribute to changes in empathy [110].
Workplace stress and fatigue secondary to workload and increasing levels of clinical responsibility are cited in literature as factors promoting empathy decline among physicians [49]. Furthermore, “sleep loss has been shown to affect some aspects of empathy” [111]. It has been argued that the decline of medical students’ empathy during their studies may be a self-protective or coping mechanism at times of transition [23, 112–114]. However, Triffaux, Tisseron, Nasello ask whether the decline of empathy among medical students is useful coping process or dehumanization [84]. Furthermore, the fact that medical curriculum focuses more mechanistic view on illness (that may reduce the patients to a disease or an object) than humanistic values may tackle empathy decline during medical studies [5]. It is arguably said that “the more doctors depend solely on technology, the more they lose their humanism” [110]. This assumption has been supported by a systematic review and meta-synthesis of qualitative studies [115].
Empathy and carrier interest / specialty interest
It is widely argued that empathy drives students to select specific specialties. Students preferring technology-oriented specialties (i.e., radiology, pathology, surgery) have lower empathy levels while students preferring people-oriented specialties (i.e., family medicine, pediatrics, internal medicine, psychiatry) have higher empathy levels [5, 9, 49, 56, 80, 84, 85, 91, 93, 101, 116]. Note, however, that some studies did not support the hypothesis of relationship between empathy specialty preference [117, 118]. As presented in Results section above, we found no statistically significant correlation between empathy and specialty interest among participants in this study. Moreover, we explored the students’ carrier intention, namely, whether they were interested in practicing medicine in the public or in the private sector, or even following academic carrier. However, we found no statistically significant correlation between empathy and carrier interest among participants in this study.
Empathy gender differences
The empathy gender bias is a strong effect observed in several empathy-related phenomena [84]. Longitudinal and cross-sectional research has consistently associated the empathy of medical students with gender [9, 43, 75, 80, 97, 119, 120]. Many studies have reported gender differences with females scoring higher or significantly higher than men [2, 9, 56, 75, 77, 79, 85, 87, 89, 93, 96, 101, 103, 121–123]. In many studies male students showed a more significant empathy decline than females during the studies. For instance, Raof and Yassin found that female students had significantly higher empathy levels than male students who showed a significant empathy decline during the studies [116]. Shashikumar et al. did not found significant difference among female students in different semesters. However, they found significant decline in empathy among male students [82]. Ye et al. found no significant gender differences in the empathy scores among medical students before early clinical contact. Note, however, that authors reported no empathy gender differences [59]. Di Lillo et al. found that although females had slightly higher empathy levels the difference was not statistically significant [124]. In the same vein were Iranian studies [92, 125]. The study of Tariq, Rasheed, Tavakol with Pakistani students did not support the hypothesis that female medical students have higher empathy levels than males [98]. Cultural context may be a factor affecting the gender differences in the empathy scores among medical students. As presented in Results section above, we found statistically significant empathy gender differences among participants in our study.
Importantly, it is arguable that further research is necessary to explore the relationship between gender and clinical empathy levels [43]. Christov-Moore et al. state that further investigation of sex/gender differences in empathetic skill may be informative for understanding the nature of empathy [1]. Moreover, it is argued that “gender differences in response to medical humanities programs require further study” [6].
Difference by gender may be more evident for some components of empathy. Quince et al. found statistically significant gender differences which however were not same in affective and cognitive empathy. As regards affective empathy, gender differences were found in all six years while cognitive empathy gender differences were found for four of the six years [89]. Many studies argued that the difference by gender is “more evident for the caring component than the sharing component” [44]. To cite just one example, Ishikawa et al. found that “female students manage to maintain their patient-centered attitude while presumably going through comparable experiences in their medical training” [44].
Among the reasons cited in the literature for the empathy gender differences are included: traditional gender roles [43], intra- and inter- cultural gender-related personality characteristics [53], the fact that self-reporting tools used in research methodology may induce biases in favor of gender-based stereotypes [126], or even “phylogenetic and ontogenetic roots” [1]. Shashikumar et al. stated that female students are by nature more caring and loving [82]. Christov-Moore et al. state that “there is also a small but growing literature on gender differences in the ability to recognize emotional body language” and that “females are faster and more accurate than males in recognizing facial expressions” [1]. Note that the understanding of the mechanisms underlying different subtypes of empathy is increasing [127]. Some studies highlighted the gender-related differences between the neurobiological underpinnings of empathy [1]. Schulte-Rüther et al. suggested that “females recruit areas containing mirror neurons to a higher degree than males during both SELF- and OTHER-related processing in empathic face-to-face interactions. This may underlie facilitated emotional "contagion" in females” [128]. Note, however, that in the neuroscience of empathy there may be methodological and conceptual pitfalls [129]. It is not clear whether or not there is significant difference in the activation of the various brain structures between men and women.
Ultimately, empathy gender differences may be due to the method used in the research. Some authors suggest that, when a self-report method is used to assess empathy, women’s gender-role model is activated and hence they respond more empathically [130]. It is suggested that differences in general emotional responsiveness may be the reason for gender differences in self-reported empathy [131].
Religiosity / Spirituality
Empathy and religiosity seem to be linked as both underline values such as altruism, sympathy, helping and caring for others. Málaga, Gayoso and Vásquez found higher levels of empathy according to religious beliefs among medical students (“practicing a religious denomination is related to a higher level of empathy”) [123].
However, there is not a clear positive correlation between religiosity and empathy in the literature. Damiano, DiLalla et al. found that “spirituality openness was related to empathy only in nondepressed students. Empathy of students with higher levels of depression was generally lower and not affected by spirituality openness” [132]. Empathy may have a mediating role between emotional intelligence and religiosity [133]. Damiano et al. found “that meaning of life and previous mental health treatment but not Religiosity were positively related to empathy” [134]. At any rate, the relationship between religiosity and empathy needs further research. Note that in this study we found a (statistically significant) positive correlation between trust in God and empathy among participants in our study.
The factor age
In this study, we found no statistically significant correlation between participants’ age and empathy among participants in our study. Studies have shown a negative association between age and (cognitive) empathy. For instance, it is argued that a large subset of elderly people experience a decline in their social understanding abilities with advancing years [135] [“Happé, Winner, and Brownell reported that theory of mind performance was superior in the elderly [136]. Yet, in direct contrast to these findings, Maylor, Moulson, Muncer, and Taylor report a decline in theory of mind abilities with advancing years” [137]]. Studies have found reduced neural responses in brain activity associated to empathy in older adults [138, 139]. Chen et al. found that “the neural response associated with emotional empathy lessened with age, whereas the response to perceived agency is preserved” [138]. Many studies have suggested that while cognitive empathy declines with advancing age, there is no age-related difference in affective empathy [140].