Determinants of empathy, depression and loneliness in Serbian medical students

Abstract

Introduction

Medical students appear to be at risk for psychological problems. Therefore, we examined factors influencing depression and loneliness in this specific group, focusing on socio-demographic, psychological, somatic variables, and empathy as a protective factor.

Methods

We included 507 medical students at the Faculty of Medicine in Belgrade, Serbia. Each participant completed a written questionnaire comprising a socio-demographic questionnaire, the Toronto Empathy Questionnaire (TEQ), the UCLA Loneliness Scale, and the Patient Health Questionnaire (PHQ-9).

Results

The sample was predominantly female (75.1%) with a mean age of 22.11 (SD 1.6). Higher empathy scores were observed among female students, younger participants, those with lower academic performance, and students who attended psychotherapy. Higher loneliness and depression were reported by students with psychiatric problems, those using psychiatric medication, and those with unstable socio-economic status. Loneliness was positively associated with depression and negatively associated with empathy in our regression model.

Conclusions

Significant differences in psychological variables exist among medical students in relation to the analyzed socio-demographic, health, and behavioral characteristics, highlighting the need for tailored interventions aimed at fostering social connection and empathy.

Introduction

Medicine is one of the most demanding study programs, and mental disorders are more common among medical students than is generally accepted [1]. While higher rates of depression and suicide among doctors compared to the general population have been consistently documented [2], the determinants of psychological well-being in medical students have been less explored. Several factors may contribute to psychological distress among medical students, including academic burden, exposure to patients’ suffering, limited social life, poor family relationships, sleep deprivation, and long-distance romantic relationships [3]. Since all factors occur during a vulnerable developmental period in late adolescence, it might be worth considering medical students as an at-risk population. Indeed, medical students report higher levels of depression and anxiety compared to their peers [4]. Furthermore, depression in medical students seems to be a global phenomenon, with a prevalence of 27.2% in a study across 47 countries [1]. In the U.S. higher suicidality risk among future doctors is associated with loneliness, which is more prevalent in medical students compared to senior faculty doctors [5]. Loneliness among students has also been associated with increased anxiety, stress, maladaptive coping strategies, sleep problems, lower life satisfaction, and suicidal ideation [6]. Importantly, loneliness does not necessarily imply social isolation, as individuals may feel lonely despite having social contacts [7]. Empathy is a crucial aspect of the medical profession, associated with better patient outcomes, increased job satisfaction, and lower burnout [8]. Therefore, empathy may be a protective factor in the medical profession.

This study aims to determine the levels of empathy, loneliness, and depression among medical students in Serbia and to explore differences in these characteristics in relation to various socio-demographic, psychological, and somatic factors.

Methods

Study design and participants

This cross-sectional observational study was conducted from December 2022 to August 2023 at the largest medical faculty in the country—the Faculty of Medicine, University of Belgrade, Serbia. Participants were third- and fifth-year medical students. At this faculty, the first three years cover preclinical subjects, while the remaining three years focus on clinical subjects. The sample represented a convenience sample consisting of students who voluntarily agreed to participate during the study period. The required sample size was calculated using an expected prevalence of depression among medical students of 27.2%, based on previous literature [1], with a 95% confidence level and a 5% margin of error. The minimum required sample size was estimated to be 305 participants.

Instruments

Participants completed a written questionnaire comprising a socio-demographic questionnaire, the Toronto Empathy Questionnaire (TEQ), the UCLA Loneliness Scale, and the Patient Health Questionnaire (PHQ-9). The questionnaires were administered in Serbian. The TEQ, UCLA Loneliness Scale and PHQ-9 were all previously translated to Serbian and used in research [9–11].

The sociodemographic questionnaire included questions regarding sex, age, year of study, grade point average, socioeconomic status, desired specialization, somatic problems, psychological disorders, use of medication, psychological problems in family members, and psychotherapy attendance. Desired specializations were divided into two categories in accordance with previous literature [12]: people-oriented specializations, which primarily focus on interpersonal interaction and communication with patients (family practice, internal medicine, obstetrics/gynecology, pediatrics, physical medicine and rehabilitation, and psychiatry), and technique-oriented specializations, which predominantly focus on medical procedures, techniques, and instruments (anesthesiology, dermatology, emergency medicine, otolaryngology, pathology, radiology, and surgery).

Empathy was measured using the Toronto Empathy Questionnaire (TEQ) [13], a single-factor scale consisting of 16 items scored on a Likert-type scale ranging from 1 (never) to 5 (always). The Cronbach’s α coefficient for the TEQ in our sample was 0.77, indicating satisfactory reliability.

Loneliness was assessed using the UCLA Loneliness Scale [14]. The scale consists of 20 items, scored on a 4-point Likert-type scale ranging from 1 (never) to 4 (always), with higher scores indicating greater loneliness. The Cronbach’s α coefficient for the UCLA Loneliness Scale was 0.89, indicating high internal consistency.

The Patient Health Questionnaire (PHQ-9) is a self-administered questionnaire comprising 9 items, each scored from 0 to 3 based on how the respondent felt in the previous 2 weeks [15]. The Cronbach’s α coefficient for the PHQ-9 was 0.80, indicating high internal consistency.

Statistical analysis

Data were analyzed using descriptive and inferential statistics. Means and standard deviations were used for continuous variables, while frequencies and percentages were used for categorical variables. Normality was tested using the Kolmogorov–Smirnov test. Student’s t-test was used for normally distributed variables, while Mann–Whitney tests were used for non-normally distributed variables. A one-way analysis of variance (ANOVA) was used for comparisons involving more than two groups after verification of normality and homogeneity of variances. Pearson’s correlation was used for parametric variables, and Spearman’s correlation for nonparametric variables. Multiple regression analysis was used to explore predictors of loneliness. A p-value of less than 0.05 was considered statistically significant.

Ethical considerations

This study was approved by the Ethics Committee of the Faculty of Medicine, University of Belgrade (approval number 17/VI-13). All procedures involving human participants were performed in accordance with the ethical standards of the institutional research committee and with the Declaration of Helsinki. The study was anonymous, and informed consent was obtained from each participant prior to participating in the study.

Results

The sample consisted of 507 medical students (Mean age = 22.11 ± 1.6), with 126 (24.9%) males and 381 (75.1%) females. Of the total sample, 303 (59.8%) were in the preclinical years and 204 (40.2%) were in the clinical years. Regarding desired specialization, 281 (55.4%) preferred people-oriented specializations (e.g., psychiatry), 173 (34.1%) preferred technique-oriented specializations (e.g., anesthesiology), and 53 (10.5%) were unsure. A one-way ANOVA with post-hoc LSD tests revealed no significant differences in TEQ, UCLA, and PHQ-9 scores based on desired specializations (F (2,504) = 1.32, p = 0.27).

Somatic health problems were reported by 37 (7.3%) participants, with the most common being Hashimoto’s thyroiditis (1.4%, n = 7), asthma (1%, n = 5), insulin resistance (0.8%, n = 4), and polycystic ovary syndrome (0.6%, n = 3). Psychological disorders were reported by 48 (9.5%) participants, with anxiety being the most common (6.7%, n = 34), followed by depression (3.4%, n = 16). The most used medications were metformin (1.9%, n = 10), propranolol (1.1%, n = 6), and antidepressants (2.4%, n = 12). Among psychiatric disorders in immediate family members, depression was the most common (6.5%, n = 33).

We analyzed the effects of socio-demographic and health parameters on measures of empathy, depression, and loneliness (presented in Table 1). Female students demonstrated higher empathy and depression scores compared to male students. Higher empathy scores were observed among preclinical students, students with lower academic performance, and those attending psychotherapy. Higher loneliness and depression scores were observed among students with psychiatric disorders, psychiatric medication use, and unstable socioeconomic status.

Table 1.

TEQ, UCLA and PHQ-9 scores by sociodemographic and health parameters

Variable		N(%)	Empathy (TEQ)		Loneliness (UCLA)		Depression (PHQ)
			M ± SD	p	M ± SD	p	M ± SD	p
Sex	M	126 (24.8%)	44.97 ± 7.26	0.01	42.07 ± 8.76	0.82	6.77 ± 4.52	0.01
	F	381 (74.9%)	49.35 ± 6.79		41.95 ± 9.00		8.14 ± 4.57
Year of study	Preclinical	303 (59.8%)	49.41 ± 6.49	0.01	42.22 ± 9.05	0.55	8.16 ± 4.64	0.01
	Clinical	204 (40.1%)	46.52 ± 7.74		41.63 ± 8.77		7.23 ± 4.45
Grade point average	High	271 (53.5%)	47.68 ± 7.19	0.04	42.47 ± 9.03	0.26	7.82 ± 4.68	0.25
	Low	236 (46.5%)	48.92 ± 7.09		41.42 ± 8.81		7.85 ± 4.53
Emotional status	Relationship	323 (63.5%)	47.62 ± 6.71	0.25	41.38 ± 9.35	0.25	8.31 ± 4.90	0.49
	Single	184 (36.1%)	47.89 ± 7.32		42.32 ± 8.69		7.82 ± 4.53
Somatic disorders	Yes	37 (7.3%)	47.17 ± 6.73	0.20	42.65 ± 10.14	0.70	8.00 ± 4.96	0.97
	No	468 (91.9%)	48.35 ± 7.18		41.88 ± 8.82		7.76 ± 4.54
Psychiatric disorders	Yes	49 (9.6%)	50.12 ± 8.29	0.06	47.72 ± 9.26	0.01	11.31 ± 5.34	0.01
	No	458 (90.0%)	48.06 ± 7.01		41.37 ± 8.69		7.42 ± 4.34
Psychiatric medication	Yes	15 (2.9%)	47.60 ± 6.40	0.49	49.61 ± 11.60	0.01	13.60 ± 6.87	0.01
	No	491 (96.5%)	48.27 ± 7.19		41.72 ± 8.73		7.62 ± 4.40
Psychotherapy	Yes	39 (7.7%)	50.92 ± 6.02	0.01	43.17 ± 8.67	0.68	9.85 ± 6.18	0.05
	No	468 (91.9%)	48.04 ± 7.21		41.88 ± 8.96		7.62 ± 4.40
Psychiatric disorders in the family	No	458 (90.0%)	47.98 ± 7.98	0.86	44.08 ± 8.34	0.08	10.45 ± 5.65	0.01
	Yes	49 (9.6%)	48.29 ± 7.07		41.75 ± 8.98		7.51 ± 4.37
Socioeconomic status	Stable	486 (95.5%)	48.38 ± 7.06	0.09	41.74 ± 8.84	0.01	7.62 ± 4.44	0.01
	Unstable	21 (4.1%)	45.57 ± 8.89		47.49 ± 9.49		11.76 ± 6.13

p values in bold: p < 0.05

Pearson’s correlation test showed a medium positive correlation between loneliness and depression, while other correlations exhibited weak associations.

Multiple regression was conducted with sex, age, years of study, grade point average, empathy, and depression as the predictors, with loneliness as the dependent variable (Table 2). The model was significant and explained 22% variance (F (6,490) = 23.47, R² = 0.22, p < 0.05). The results showed that depression is a positive predictor and empathy is a negative predictor of loneliness.

Table 2.

Regression model predicting loneliness scores

	B	SE of B	β	Significance
Constant	39.900	8.193		< 0.05
Age	0.091	0.287	0.016	0.752
Year of study	− 0.354	0.480	− 0.039	0.462
Grade point average	0.537	0.515	0.044	0.298
Empathy score	− 0.185	0.053	− 0.148	< 0.05
Sex	− 0.567	0.866	− 0.027	0.513
Depression score	0.861	0.080	0.442	< 0.05

Significance values in bold: p < 0.05

Discussion

We found significant differences in empathy, depression and loneliness in relation to different sociodemographic and psychological factors in our sample of medical students. Specifically, higher empathy scores were observed among female students, younger participants, those with lower academic scores, and students who attended psychotherapy. Higher loneliness and depression were reported by students with psychiatric problems, those using psychiatric medication, and those with unstable socio-economic status. In our regression model, loneliness was positively associated with depression and negatively associated with empathy.

In our study, higher empathy scores were observed among female students, which is consistent with findings reported by other authors [9]. The available literature suggests that from early development, females demonstrate greater sensitivity to emotional and social cues from their mothers. During further development, these differences appear to become more pronounced through emotional maturation, socialization, language development, and the ability to recognize others’ emotions [16]. Elkin et al. (2021) found significantly higher empathy scores among women than among men, with no statistically significant differences in cognitive empathy. The authors suggested that this finding may be partially explained by social expectations and gender stereotypes, according to which women may feel more comfortable expressing empathy, but also by possible biological differences [17]. Quince et al. also reported significantly higher affective empathy scores among female medical students across all years of medical education. However, gender differences in cognitive empathy were present only during certain years of training [18]. Nevertheless, findings on gender differences in empathy remain inconsistent across studies and should therefore be interpreted with caution. Additionally, women constituted the majority of our sample, and this should be taken into consideration when interpreting the results. The finding that females score higher on empathy may be a part of the explanation why patients treated by female doctors have better outcomes and a lower risk of complications compared to those treated by male doctors [19]. However, this association is speculative, and further research is needed to confirm whether empathy indeed contributes to these differences in patient outcomes.

We found that third-year students had higher empathy scores compared to fifth-year students, a result that has also been robustly replicated [20]. Chen et al. also reported a decline in empathy during the transition from preclinical to clinical years of medical education. As possible explanations, the authors proposed frequent rotations through different clinical departments, which may lead to the “traumatic deidealization of medicine”, reduced patient contact due to the need for rapid decision-making and the development of modern technologies, as well as empathic distancing as a potential defense mechanism [21]. Socioeconomic status has also been described as a factor for reduced empathy, albeit not consistently [22]. In our study, most participants had a stable socioeconomic status, and we did not find significant differences in empathy scores related to socioeconomic status. However, those with unstable socioeconomic status had higher loneliness and depression scores, which is a finding robustly replicated in research [23]. In anyway, the erosion of empathy during medical training represents a concern and active efforts should be made to prevent it.

An interesting finding of our study was that students with a lower grade point average had higher empathy scores. Previous studies found that empathy was not associated with academic performance among medical students but was associated with higher clinical competence [24]. A possible explanation for our finding is that students with a higher grade point average might be more competitive and perfectionistic, and these traits have been associated with reduced empathy and altruism [25]. However, further research is needed to explore these associations, as not all aspects of competitiveness are found to reduce empathy (i.e., competitiveness to develop oneself is associated with increased altruism, whereas competitiveness to win and surpass is not) [25].

Higher levels of empathy were found in students who attended psychotherapy. Indeed, there is evidence that psychotherapy can increase empathy [26]. Furthermore, various psychological interventions, such as mindfulness and meditation techniques, appear to increase compassion [27] and may be useful in programs aimed at promoting empathy in medical settings.

Loneliness was associated with the presence of psychiatric disorders, the use of psychotropic medication, and unstable socio-economic status, all of which have been robustly demonstrated in previous research [28]. However, unlike previous studies in US and Chinese medical students [5, 29], we did not find gender differences in loneliness. This is surprising considering that we replicated gender differences in both empathy and depression. One possible explanation is that, since female students strongly outnumber male students at Serbian medical faculties, the female experience is common, and female students are not isolated. It is also possible that certain cultural factors, yet to be explored, underlie the lack of gender difference in loneliness. Therefore, the absence of gender differences in loneliness in our sample should be interpreted cautiously.

Higher depression scores were detected in female students, younger (preclinical) students, those with psychiatric disorders, those who use psychiatric medication, those who go to psychotherapy, students with a family history of psychiatric disorders, as well as those with unstable socio-economic status. These results have all been noted in previous studies [30].

Loneliness was positively predicted by depression in our regression model. Accordingly, previous studies have shown that in the student population, social and emotional loneliness positively predicted depressive symptoms [31] and that loneliness predicted the development of major depressive disorder, even after accounting for the subclinical level of depressive symptoms at baseline [32]. In fact, loneliness is emerging as an important risk factor in various populations for not just depression, but a plethora of negative physical and psychological outcomes [33, 34].

In our regression model, loneliness was negatively predicted by empathy. Research has shown that people who report feeling lonely often avoid prosocial behavior to prevent negative emotional outcomes, but when prosocial behavior is accompanied by a reward, they become more motivated to engage in it [35]. This may have implications for building effective measures to prevent and reduce loneliness. Indeed, providing emotional support to others predicts lower levels of loneliness and perceived stress on both the same day and the following day. Also, providing emotional support may have benefits not only for the recipient, but also for the provider, such that empathy and gratitude may contribute to well-being over time. Taken together, empathy and emotionally engaged interpersonal relationships may play a protective role against loneliness [36]. It is important to note although our regression model was statistically significant, the percentage of explained variance was relatively modest, suggesting that additional factors contributing to loneliness were not assessed in our study.

Conclusions

In conclusion, specific sociodemographic, psychological, and somatic factors are associated with empathy, loneliness, and depression symptoms in medical students. Importantly, empathy and depression significantly predicted loneliness in our sample. Also, students with psychiatric disorders and unstable socio-economic status expressed higher levels of loneliness. Therefore, tailored interventions aimed at fostering social connection and empathy might improve both mental health and the practice of medicine among medical professionals.

Author contributions

NB contributed to the study design, data collection, data analysis, and writing of the manuscript. VP contributed to data analysis and manuscript writing. TS contributed to data analysis, interpretation of results, writing and revising of the manuscript, MPS contributed to data analysis, interpretation of results, writing and revising of the manuscript, MI contributed to data analysis, interpretation of results, writing and revising of the manuscript, ML contributed to data analysis, interpretation of results, writing and revising of the manuscript, MN contributed to study design, data analysis, writing of the manuscript.

Funding

No funding was received for this study.

Data availability

The datasets generated during the current study are not publicly available but are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This study was approved by the Ethics Committee of the Faculty of Medicine, University of Belgrade (approval number 17/VI-13). All procedures involving human participants were performed in accordance with the ethical standards of the institutional research committee and with the Declaration of Helsinki. The study was anonymous, and informed consent was obtained from each participant prior to participating in the study.

Consent to publish

Not applicable.

Competing interests

The authors declare that they have no competing interests.