Introduction

On March 11, 2020, the World Health Organization declared the coronavirus disease 2019 (COVID-19) global pandemic [1]. As of March 17, 2024, there have been more than 774,954,393 COVID-19 cases globally and 7,040,264 deaths [2]. To slow the rising numbers of COVID-19 infections and deaths, extraordinary scientific efforts have been made to develop vaccines against COVID-19 infection and distribute them in many countries. However, by the end of 2022, the pandemic has yet to be resolved [3]. Multiple epidemic waves of COVID-19 have been reported in many countries [4, 5], and countries worldwide must be prepared for the possibility of future waves [6].

Medical students, pre-clinical medical students, and students in clinical rotations represent a population that is vulnerable to infectious disease exposure, especially during the COVID-19 pandemic. In many countries, the medical education of students in their pre-clinical years has been changed to online activities to maintain social distancing and avoid new outbreaks of COVID-19 infection, and medical students have been advised to stay at home [7, 8]. However, in many countries, medical students in the clinical years are required to be involved in caring for patients owing to the shortage of professional health workers in hospitals because of increasing numbers of patients with COVID-19 infection [9], even though students may not have appropriate and complete training or sufficient clinical experience to be able to protect themselves and handle complex clinical situations. Therefore, both the Medical Schools Council (MSC) in the United Kingdom and the American Association of Medical Colleges (AAMC) in the United States have published a guideline for the participation of medical students as health care providers during the COVID-19 pandemic in which this group is recommended to work as volunteers [10, 11].

The effects of the COVID-19 pandemic on mental health outcomes most likely differ among different populations [3]. The emotional response to the pandemic might be stronger for groups who are vulnerable to infection, such as health care providers [12]. The ongoing pandemic has been a tremendously challenging situation for health care providers who are exposed to patients with COVID-19. Their work not only places them at high risk of becoming infected but also can lead to a fear of contagion and spread of the virus to loved ones [13]. This difficult situation has led to the development of mental health problems such as distress, anxiety, depression, insomnia, post-traumatic stress disorder (PTSD), denial, and fear among health workers [14, 15]. Furthermore, exposure to COVID-19 infection among health care providers is associated with a high risk of experiencing depression and anxiety [12]. Health care providers who are in charge of caring for patients with COVID-19 have demonstrated increased levels of distress, anxiety, and depression [16]. As future physicians, medical students may have also experienced tremendous pressure during the ongoing COVID-19 pandemic, which is associated with a high risk of depression and anxiety.

Mental health issues surrounding mental illness are prevalent among medical students [17, 18], and these might be exacerbated to become a serious issue during the global COVID-19 pandemic. Mental health problems jeopardize the well-being of medical students, leading to learning problems, which can have an impact on delivering quality patient care in the future. This is a serious problems to which medical schools and institutions should devote greater attention by developing strategies to help medical students [17]. In a cross-sectional study conducted in Bangladesh during the COVID-19 pandemic among 425 medical students, 55% experienced anxiety and 44% experienced depression [19]. Another cross-sectional study conducted among medical students in Libya revealed that 65% developed anxiety and 78% developed depression [20]. A study in the United States found that 66% of medical students had developed anxiety during the COVID-19 pandemic [16]. Two meta-analysis studies conducted during the COVID-19 pandemic revealed that 28% of medical students had developed anxiety [21] and 31% had depression [22] respectively; additionally, further meta-analysis conducted in 2023 found that 38% of medical students had developed anxiety and 41% had depression during the COVID-19 pandemic [23]. Given the high prevalence of anxiety and depression among medical students in developing and developed countries, continued efforts must be made to collect and analyze data regarding the effects of the COVID-19 pandemic on mental health outcomes among medical students to obtain a complete picture of this phenomenon globally and address this knowledge gap.

An evaluation of the psychological status will help clarify approaches for targeted psychological intervention during the ongoing COVID-19 pandemic, prompting the further development of medical and health public services. Therefore, updated evidence is crucial regarding the global mental health situation in the population of medical students during the COVID-19 pandemic. In this study, we aimed to investigate updated estimates of the prevalence of anxiety and depression among medical students during the COVID-19 pandemic, and discuss associated stressors.

Materials and methods

This systematic review (study protocol registered on PROSPERO-CRD42021252968) was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) [24].

Search strategy

Five electronic databases were used to retrieve relevant studies. The reviewing author searched the CINAHL, EMBASE, MEDLINE, PubMed, and Web of Science databases from December 1, 2019 to March 15, 2023 to identify studies published between 2020 and 2023. To avoid missing pertinent research, grey literature search using Google Scholar and manually searching by examining included studies from prior systematic reviews or meta-analysis studies were conducted. The search language was limited to English. The MeSH terms used to develop the search included: medical students; OR medical undergraduate; OR medical postgraduate; OR education, medical, undergraduate; AND mental disorder; OR mental health; OR affective disorder; OR mood disorder; OR depressive disorder; OR depression; OR anxiety; OR stress, psychological; OR depress*; OR anxiety*; OR mental wellbeing; AND COVID-19; OR coronavirus disease 2019; OR pandemic; OR 2019-nCoV; OR SARS-CoV-2; OR COV-19. Details of the search strategy are presented in Additional File 1: Appendix 1. Additionally, the references of the identified articles were searched manually and appropriate articles were reviewed.

Eligibility criteria

The following inclusion criteria were applied: a) studies conducted among students of medicine; b) cross-sectional study, cohort study, or case–control study; c) provided outcomes of anxiety and depression; d) studies conducted related to COVID-19 pandemic; and e) published in English language. The following exclusion criteria were applied: a) review studies; and b) studies that did not provide the full text. Medical students included students who were enrolled in the M.D. program (Doctor of Medicine) and M.B.B.S. programs (Bachelor of Medicine and Bachelor of Surgery).

Study selection and data extraction

Two authors (LYK and SID) independently screened all titles and abstracts according to the defined eligibility criteria, after removing duplicates using EndNote software. The authors used the PRISMA flow diagram to report study eligibility, and they independently followed the selection process for all studies and retrieved the full texts of those studies that passed the first-level screening. A full-text review was performed for each study, and data extraction was conducted in duplicate by two reviewers. The fields extracted were author, publication year, country where the study was conducted, continent of the study country, sample size, sex of participants, age, number and percentage of anxiety and depression among medical students, instruments used to evaluate anxiety and depression, and study period. All discrepancies were resolved in consultation with a third reviewer (LCJ).

Study risk of bias assessment

Accurately judging and choosing the appropriate tool for each included study is an important step in analyzing the methodological quality (risk of bias) of each identified study and exploring whether the study is of low quality or has a high risk of bias [25]. For each reviewed source, two authors (LYK and SID) assessed the risk of bias using the Joanna Briggs Institute (JBI) critical appraisal checklist for studies reporting prevalence data to assess the level of the evidence; the overall quality of each study [26]. Nine dimensions include sample frame, participant recruitment, sample size, subjects and setting description, representativeness, valid methods for the identification, a standard and reliable measurement way, valid assessment of mental problems, appropriate statistical analysis, and response rate. All discrepancies were resolved in consultation with a third reviewer (LCJ).

Statistical analysis

The pooled prevalence of anxiety and depression among medical students during the COVID-19 pandemic was calculated using a random-effects model with REML mothod. Anxiety and depression in this study referred to “anxiety symptoms” and “depressive symptoms” because most included studies used self-rated questionnaires or instruments; an actual illness could not be diagnosed using the screening tools alone. The selected studies reported the dichotomous variable of anxiety and depression as being present or absent in medical students, according to the study authors’ defined cutoff score for the selected screening instruments. Anxiety and depression were further divided into categories according to the scale of the instruments: anxiety, moderate and severe (anxiety MS) and depression, moderate and severe (depression MS). The heterogeneity of each variable in the pooled estimate was indicated with I2 using a random-effects model. For I2, 25%–49% indicated low heterogeneity, 50%–74% moderate heterogeneity, and > 75% indicated high heterogeneity [27, 28]. Funnel plots and forest plots were generated for all analyses. Sensitivity analysis was performed to confirm stability and reliability. We used leave-one-out meta-analysis to identify the influence of each study on the effect-size estimates. Both Begg’s test and Egger’s tests were used to identify the potential risk of publication bias. If various levels of heterogeneity were identified among studies and in subgroup analysis, meta-regression was used to explore the heterogeneity. Moderator variables for subgroup analyses and meta-regression were chosen post hoc. A value of p < 0.05 was considered statistically significant. All statistical analyses were conducted using Stata version 17.0 (StataCorp LLC, College Station, TX, USA).

Results

Study selection

Using the search strategy, we identified 3,144 articles in five electronic databases. After the removal of duplicates, the titles and abstracts of 1,728 articles were screened; 225 articles were determined to satisfy the eligibility criteria. After a full-text review, ninety-five articles were deemed ineligible. The remaining 130 articles were included in the final analysis. Details of the study selection process are presented in a PRISMA flow diagram (Fig. 1). The PRISMA checklist is presented in Additional File 2: Appendix 2.

Fig. 1
figure 1

Flow chart for selection of reviewed articles. *Consider, if feasible to do so, reporting the number of records identified from each database or register searched (rather than the total number across all databases/registers). **If automation tools were used, indicate how many records were excluded by a human and how many were excluded by automation tools. From: Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ 2021;372:n71. doi: 10.1136/bmj.n71

Full size image

Study characteristics

All included articles were cross-sectional studies. Three of the 130 studies included multiple sets of data; therefore, a total of 135 sets of data were included in the final analysis. A total of 132,068 medical students were included. Twenty-eight studies were conducted in China [29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56]. Sixteen studies were conducted in India [57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72]. Eleven studies were conducted in Pakistan [73,74,75,76,77,78,79,80,81,82,83]. Ten studies were conducted in Saudi Arabia [84,85,86,87,88,89,90,91,92,93]. Nine studies were conducted in United States [16, 94,95,96,97,98,99,100,101]. Five studies were conducted in France [102,103,104,105,106] and Turkey [107,108,109,110,111]. Four studies each were conducted in Brazil [112,113,114,115], Malaysia [116,117,118,119], and Peru [120,121,122,123]. Three studies were conducted in Indonesia [124,125,126] and Thailand [127,128,129]. Two studies each were conducted in Bangladesh [19, 130], Egypt [131, 132], Korea [133, 134], Morocco [135, 136], Nepal [137, 138], Spain [139, 140], and multi-countries [141, 142]. One study each was conducted in Germany [143], Greece [144], Iran [145], Iraq [146], Italy [147], Japan [148], Jordan [149], Kazakhstan [150], Mexico [151], Slovak [152], Sudan [153], Tunisia [154], United Arab Emirates [155], and Vietnam [156]. Furthermore, according to continent or region, seventy-six studies were conducted in Asia, nineteen studies in the Middle East, eleven studies were conducted in Europe, ten studies in North America, eight studies in South America, and Five study was conducted in Africa. The average age of included participants ranged from 17 to 26.1 years. The prevalence of anxiety in 100 studies ranged from 3.8% to 96.1%, and the prevalence of depression in 97 studies ranged from 7.5% to 99.6%. The prevalence of anxiety MS in 82 studies ranged from 2.3% to 73.7%, and the prevalence of depression MS in 78 studies ranged from 1.2% to 90.2%. The study periods of the included publications were between 2019 and May 2022. A summary of study characteristics is presented in Table 1.

Table 1 Summary of selected studies on mental health outcomes among medical students during COVID-19 outbreak
Full size table

Risk of bias in studies

The JBI scale was used to appraise the quality of the 130 cross-sectional studies. The results of quality assessment are presented in Additional File 3: Appendix 3.

Prevalence of anxiety and depression among medical students

Anxiety

One hundred studies that included 41,620 participants reported anxiety using the Generalized Anxiety Disorder-7 (GAD-7), the Depression, Anxiety and Stress Scale (DASS), Beck Anxiety Inventory (BAI), Zung Self-rating Anxiety Scale (SAS), and Hospital Anxiety and Depression Scale (HADS), State-Trait Anxiety Inventory (STAI), and Patient Health Questionnaire (PHQ). The pooled prevalence of anxiety among medical students during the COVID-19 pandemic was 45% (95% confidence interval [CI], 40%–49%, I2 = 99.65%, p < 0.001). The forest plot is displayed in Fig. 2a.

Fig. 2
figure 2

Pooled prevalence of anxiety and depression among medical students during COVID-19 pandemic. a Pooled prevalence of anxiety among medical students during COVID-19 pandemic. b Pooled prevalence of anxiety (moderate and severe) among medical students during COVID-19 pandemic

Full size image

Anxiety MS (moderate and severe)

Eighty-two studies that included 17,495 participants reported anxiety MS using the GAD-7, DASS-21, BAI, STAI, Zung SAS, Visual Analogue Scale (VAS), and HADS. The pooled prevalence of anxiety MS among medical students during the COVID-19 pandemic was 28% (95% CI, 24%–32%, I2 = 99.61%, p < 0.001). The forest plot is displayed in Fig. 2b.

Depression

Ninety-seven studies that included 35,828 medical students reported depression using the Patient Health Questionnaire-9 (PHQ-9), DASS-21, Zung Self-rating Depression Scale (SDS, Beck Depression Inventory-II (BDI-II), and HADS. The pooled prevalence of depression among medical students during the COVID-19 pandemic was 48% (95% CI, 43%–52%, I2 = 99.66%, p < 0.001). The forest plot is displayed in Fig. 2c.

Depression MS (moderate and severe)

Seventy-eight studies that included 15,719 medical students reported depression MS using the PHQ-9, DASS-21, Zung SDS, BDI-II, HADS, and the Self-Rated Depression Scale. The pooled prevalence of depression MS among medical students during the COVID-19 pandemic was 30% (95% CI, 26%–35%, I2 = 99.58%, p < 0.001). The forest plot is displayed in Fig. 2d.

There was high heterogeneity between studies, with I2 ranging from 99.58%–99.66%.

Sensitivity analysis

We performed sensitivity analysis and confirmed the stability and reliability of the results. We used leave-one-out meta-analysis to identify influential studies. Figure 3a shows that the corresponding pooled prevalence of anxiety ranged from 44% to 45%. Figure 3b shows that the corresponding pooled prevalence of anxiety MS varied from 27% to 28%. Figure 3c shows that the corresponding pooled prevalence of depression ranged from 47% to 48%. Figure 3d shows that the corresponding pooled prevalence of depression MS varied from 30% to 31%. This was not substantially altered. The statistically similar results revealed that no single study influenced the stability of the overall prevalence estimates in the meta-analysis.

Fig. 3
figure 3

Sensitivity analysis of anxiety and depression for individual studies on the summary effect. a Sensitivity analysis of anxiety for individual studies on the summary effect. b Sensitivity analysis of anxiety (moderate and severe) for individual studies on the summary effect. c Sensitivity analysis of depression for individual studies on the summary effect. d Sensitivity analysis of depression (moderate and severe) for individual studies on the summary effect

Full size image

Publication bias

Visual inspection of the funnel plot revealed relative symmetry for anxiety and depression (Fig. 4a and c) and asymmetry for anxiety MS and depression MS (Fig4b and d). Begg’s (z = 1.25, p = 0.210) tests showed no potential risk of publication bias for the overall prevalence of anxiety. However, Egger’s (z = 3.63, p = 0.0003) tests showed a potential risk of publication bias for the overall prevalence of anxiety. Begg’s (z = 1.54, p = 0.124) tests showed no potential risk of publication bias for the overall prevalence of anxiety MS. However, Egger’s (z = 3.82, p = 0.0001) tests showed a potential risk of publication bias for the overall prevalence of anxiety MS. Both Begg’s (z = 0.60, p = 0.551) and Egger’s (z = 0.72, p = 0.469) tests showed no potential risk of publication bias for the overall prevalence of depression. Finally, both Begg’s (z = 3.56, p = 0.0004) and Egger’s (z = 2.68, p = 0.0073) tests showed a potential risk of publication bias for the overall prevalence of depression MS.

Fig. 4
figure 4

Funnel plot of anxiety and depression among medical students during COVID-19 pandemic. a Funnel plot of anxiety among medical students during COVID-19 pandemic. b Funnel plot of anxiety (moderate and severe) among medical students during COVID-19 pandemic. c Funnel plot of depression among medical students during COVID-19 pandemic. d Funnel plot of depression (moderate and severe) among medical students during COVID-19 pandemic

Full size image

Subgroup analysis

The results of subgroup analysis are presented in Table 2. In the evaluation according to continent or region, the pooled prevalence of anxiety was high in North and South America (60.7%, 95% CI, 56.0%–65.4%), intermediate in the Middle East and Africa (53.4%, 95% CI, 45.8%–61.0%) and in Europe (45.0%, 95% CI, 32.2–57.9), and low in Asia (37.3%, 95% CI, 31.1%–43.5%). The pooled prevalence of anxiety MS showed no statistical significance. The pooled prevalence of depression was high in the Middle East (61.1%, 95% CI, 50.3%–71.9%), intermediate in North and South America ( 56.5%, 95% CI, 47.7%–65.2%), and in Asia (41.7%, 95% CI, 35.4%–47.9%), and low in Europe (39.9%, 95% CI, 27.1%–52.7%). The pooled prevalence of depression MS was high in the Middle East (43.6%, 95% CI, 31.8%–55.4%), intermediate in North and South America (30.2%, 95% CI, 21.7%–38.7%), and in Asia (26.6%, 95% CI, 20.5%–32.7%), and low in Europe (23.9%, 95% CI, 18.1%–29.8%).

Table 2 Subgroup analyses of prevalence of anxiety and depression in medical students
Full size table

With categories divided into Asia and the rest of the world, the pooled prevalence of anxiety was lower in Asia than the rest of the world (37.3%, 95% CI, 31.1%–43.5% versus 54.8%, 95% CI, 50.2%–59.4%). The pooled prevalence of depression was lower in Asia than in the rest of the world (41.7%, 95% CI, 35.4%–47.9% versus 54.9%, 95% CI, 48.5%–61.3%). The pooled prevalence of anxiety MS and depression MS was lower in Asia than in the rest of the world, but without statistical significance.

When evaluating the effect of sex, the pooled prevalence of anxiety, anxiety MS, depression, and depression MS revealed higher trends in female than male. When evaluating medical students’ year of study, the pooled prevalence of anxiety, anxiety MS, depression, and depression MS revealed higher trends in the pre-clinical years than the clinical years. However, there was no significance regarding the effect of sex and medical students’ year of study.

Meta-regression

We identified various levels of heterogeneity among studies and in subgroup analysis, and therefore meta-regression was used to explore potential sources of heterogeneity. As a potential source of heterogeneity, continent or region and survey date (the month during 2020 in which the last survey was conducted) were included in the meta-regression analysis. The results of meta-regression analysis are shown in Table 3. Continent was significantly associated with anxiety (coefficient 0.176; 95% CI, 0.094–0.257), and depression (coefficient 0.132; 95% CI, 0.040–0.224). Medical students in Asia showed a lower prevalence of anxiety and depression than those on other continents. The survey date (February to June, 2020) was significantly and positively correlated with anxiety (coefficient 0.109; 95% CI, 0.061–0.157), anxiety MS (coefficient 0.078; 95% CI, 0.038–0.117), depression (coefficient 0.079; 95% CI, 0.028–0.129), and depression MS (coefficient 0.083; 95% CI, 0.030–0.136). For every one-month increase, 10.9%, 7.8%, 7.9%, and 8.3% increase in the prevalence of anxiety, anxiety MS, depression, and depression MS, respectively. The survey date (May to December, 2020) was negatively correlated with anxiety, anxiety MS, depression, and depression MS. However, there were no statistical significance.

Table 3 Meta-regression for the prevalence of anxiety and depression in medical students
Full size table

Discussion

This systematic review and meta-analysis of 130 articles that included 132,068 participants revealed that the pooled estimates of prevalence showed that 45% of medical students have experienced anxiety and 48% have experienced depression. Furthermore, 28% of medical students have experienced anxiety MS and 30% of medical students have experienced depression MS. Moderate and severe anxiety and depression are of clinical importance as conditions for which further assistance and intervention might be needed. Our findings demonstrated that anxiety and depression are important issues for medical students during the COVID-19 pandemic, with wide variation in the prevalence among different studies.

To our knowledge, this is the most updated meta-analysis to simultaneously report anxiety and depression, as well as moderate and severe anxiety and depression, among medical students during the COVID-19 pandemic. We further investigated the characteristics of studies that that reported the influence on the prevalence of these conditions. The pooled prevalence indicated that continent was significantly associated with the prevalence of anxiety and depression in this population. Medical students’ sex and year of study (pre-clinical and clinical) were not significantly associated with the prevalence of anxiety and depression. Meta-regression analysis revealed that the date of survey was associated with mental health problems among medical students. From February to June in 2020, the prevalence of anxiety and depression among medical students increased over time. The increasing trend echoed the results of a previous study conducted among students of higher education, in which the authors believed that end-of-year examinations might account for this trend [157]. Moreover, we believe that the severity of the pandemic might have an influence on trends in anxiety and depression in this population. Further study is needed to confirm these results. These factors may have an impact on the learning of medical students and medical education.

Impact of COVID-19 pandemic on medical education

Consequent to the unprecedented COVID-19 pandemic, pedagogic changes have resulted in a paradigm shift in teaching and learning processes. Medical education has been disrupted owing to the closure of medical schools or barring of patient contact during the pandemic. The traditional face-to-face teaching mode had largely been replaced by online learning [158,159,160]. Alternative modes of teaching and learning, including web-based learning, role play, video vignettes, and use of both live and mannequin-based simulated patients, have been used to minimize disruption to medical education [161]. Many challenges have arisen in this shift from traditional teaching methods to online learning. Study load and workload, enhanced engagement, and technical issues might affect student and faculty satisfaction [162]. One qualitative study provided recommendations emphasizing three important axes of institutional capacity, effective learning and assessment, and human resources, which would lead to planning and implementation of successful online learning activities [163].

Associated stressors: personal, academic, environmental and cultural, and pandemic factors

The ongoing COVID-19 pandemic has had a great impact on medical students. The pooled prevalence of anxiety among medical students in our study was higher than the prevalence in a previous report [21]. One study found a 28% pooled prevalence of anxiety among medical students globally in September 2020 [21], which is much lower than our finding. A possible reason might be that half of the studies included in that report were conducted in China, and the sample sizes in those studies resulted in 89% of students being Chinese, which could limit generalization of the results. The pooled prevalence of depression among medical students in our study was higher than the prevalence in another study [22] reporting a 31% pooled prevalence of depression globally, but similar to the recent study [23] reporting a 38% pooled prevalence of anxiety and 41% pooled prevalence of depression among medical students. Notably, 28% of medical students reported anxiety MS and 30% reported depression MS. Medical institutes should make greater effort to identify those students in need of clinical intervention and provide timely assistance. Many reported stressors can induce symptoms of anxiety and depression in medical students, particularly during the COVID-19 pandemic. These can be categorized into several factors, including personal, academic, environmental and cultural, and pandemic factors.

Personal factors

In terms of personal factors, sex has been significantly associated with anxiety and depression among medical students in many studies, with most reporting that women have a higher prevalence of anxiety [16, 19, 43, 48, 65, 87, 98, 111, 114, 115, 120, 132, 135, 137, 145, 155] and depression [16, 19, 43, 48, 65, 69, 114, 115, 120, 131, 135]. Only one study reported that male students had a high risk of anxiety [32], and one study reported that male students had a high risk of depressive symptoms [164]. No difference in sex was reported for anxiety [20, 29, 35, 69, 107, 138] or for depression [20, 35, 107, 138] in numerous studies. Although our results revealed that sex was not significantly associated with anxiety or depression, women tended to have a higher prevalence of anxiety and depression. Having a history of psychiatric consultation or psychologic or mental problems was associated with anxiety and depression [132, 135, 138, 165]. Having negative thoughts or engaging in negative actions and feeling depressed was associated with a greater likelihood of anxiety [43]. Individuals who engage in negative thinking or actions, and those with anxiety levels have greater odds of experiencing some level of depression [43]. Maintaining a healthy lifestyle has been associated with less depression symptoms [43, 52].

Academic factors

Regarding academic factors, the influence of medical students’ year of study on anxiety and depression remains controversial. Although our results revealed that the pre-clinical and clinical years of study were not significantly associated with anxiety or depression, students in the pre-clinical years tended to have a higher prevalence of anxiety and depression than those in clinical years. Some studies have reported that students in the pre-clinical years of study have a higher prevalence of anxiety [16, 48, 92, 114, 120, 137] and depression [16, 48, 114, 120, 135, 137]. One study reported no significant differences in anxiety and depression between pre-clinical and clinical groups [138]. However, a past study found that students in their clinical phase of study had greater anxiety levels [155]. Medical students in clinical training, especially those participating in higher-risk unit rotations, have a greater risk of exposure to infectious diseases, which has a considerable impact on their mental health [155]. Further research is needed to confirm these results.

Some studies have reported that worrying about academic delays was a stressor leading to higher levels of anxiety [29, 120, 166] and depression [120]. One study reported that students with a higher grade point average (GPA) experienced less anxiety and depression [145]. A study in Japan found that the sudden shift to online education was associated with greater odds of having generalized anxiety and being depressed [148]. However, one study reported the mental health of medical students improved after the transition from traditional to online learning during the quarantine period [150]. In that study, 65.2% of those concerned about the shift to online education felt that this learning mode was less effective than in-person education [148]. Previous studies have reported the utility and equivalent effectiveness of online learning in comparison with offline or in-person learning [167]. Medical students who said that they would request food assistance and mental health care resources from the university in the case of future COVID-19 outbreaks also had greater odds of having generalized anxiety and being depressed [148]. Being unable to fully concentrate on their studies during the COVID-19 pandemic was found to be associated with a significantly higher risk of anxiety among medical students [19]. One study compared undergraduate and graduate students and found that graduate students had a higher risk of anxiety [43]. The authors explained that this may be owing to increased pressure regarding job-seeking or completion of a thesis required for graduation.

Environmental and cultural factors

Worrying about the economic effects of the pandemic and influence of the pandemic on daily life was found to be related to anxiety levels [29]. Living in an urban area, living with parents, and family financial stability have been identified as protective factors against anxiety [29, 120] and depressive symptoms [120]. One study reported that financial hardship during periods of lockdown and social distancing were associated with anxiety and depression [115]. However, another study reported that family financial stability was not associated with anxiety and depressive symptoms [20]. Students living alone had a higher prevalence of anxiety [20, 89]. A study showed that quarantine at home with family was not associated with anxiety and depression [138]. Another study showed that living away from the family during the pandemic had no association with depression or anxiety [69]. Having accurate knowledge was found to be a protective factor against anxiety [32] and depression [32, 81]. One study reported that a lower anxiety level might be associated with greater exposure to mass media and social media and stressed the importance of information during a pandemic [168]. Fear of being assaulted or insulted on the way to the hospital or at home was related to a significantly greater risk of anxiety among medical students [19]. Studies reported the strongly significant association of perceived social support with reduced levels of anxiety and depression [48, 165, 169], as well as improved quality of life [169]. Thus, provision of effective social support is paramount to lowering psychological stress during a global crisis [19].

Pandemic factors

Finally, pandemic factors such as social distancing and isolation from family members might worsen anxiety disorders [20, 166]. Having relatives or acquaintances with COVID-19 infection is a risk factor for increased anxiety [16, 29]. However, one study reported that having family members or friends who were infected with COVID-19 was not statistically associated with anxiety and depressive symptoms [20]. Possible COVID-19 exposure or having contact with patients who have COVID-19 infection is associated with anxiety [138, 155] and depression [138], as is experiencing COVID-19 symptoms [145]. Living in a location with a high prevalence of COVID-19 infections was found to be associated with a higher prevalence of severe anxiety and depression [48, 135]. One study compared the prevalence of anxiety disorders between students at two universities in China and found that the prevalence was significantly higher in Wuhan, which was far more severely affected by COVID-19 than Beijing [43]. More than 25 days of confinement was associated with severe anxiety and depressive symptoms [135]. One study found that most students reported being worried about transmitting COVID-19 to a family member or friends, and 65% of them worried about catching the virus themselves [155]. Another study reported that transmitting the virus to family members was not associated with anxiety or depression [19]. Students who were highly or moderately concerned about becoming infected were at higher risk (3.5-fold, 1.5-fold, respectively) of anxiety, as compared with students with no concerns regarding contracting COVID-19 infection [19]. Students who were highly or moderately concerned about contracting COVID-19 infection were also at higher risk (2.75-fold, 1.96-fold, respectively) of having depressive symptoms compared with their counterparts who were unconcerned [19]. One study reported that students with a low degree of concern about COVID-19 had a high risk of depressive symptoms [164].

Other factors impact on anxiety and depression of medical students

When organized geographically, continent was also significantly associated with mental health problems among medical students in our study. Students in North and South America and Africa had the highest prevalence of anxiety whereas those in Asia had a lower prevalence of depression and depression MS. These results might be explained by the severity of the pandemic in different countries and regions. The COVID-19 pandemic has been more severe on the North and South American continents than on other continents, with the highest number of confirmed cases and deaths in the United States [2]. Another reason might be cultural influences. Mental disorders are considered a social stigma in many countries, especially Asian countries. Self-report measures of mental disorders might not reflect the real situation among medical students in Asian countries, which might explain the difference in the prevalence of anxiety and depression among different continents.

Studies comparing medical and non-medical students found that non-medical students had higher anxiety levels [45, 155] and depression [45, 55] than medical students. Possible reasons for these findings might be that medical students have better knowledge of the virus that causes COVID-19, the disease prognosis, and transmission and control measures. Our finding is in line with those of other studies stressing the importance of information during a pandemic [168]. In a comparison of medical and dental students, dental students reported higher levels of anxiety [155] as well as higher levels of anxiety and depression [170, 171]. This might be because dental students are in very close proximity to patients when providing dental care, which may increase the likelihood of exposure to highly transmissible respiratory viruses, thereby increasing the potential risk of COVID-19 infection owing to the nature of the dental field itself [172].

Implications and recommendations

Medical students have been identified as having a high risk of developing mental health problems [17]. Amid the COVID-19 crisis, some medical students might be in a challenging position. To limit the use of personal protective equipment and to ensure the safety of medical students, clerkships and clinical activities have been suspended by some medical schools [173]. In contrast, medical students in some regions have been urged to participate in patient care, sometimes even earning credit toward their degree under certain circumstances [174]. According to one study, higher rates of depression, suicidal ideation, and stigmatization around depression have been noted in these students [175]. Depressed medical students might feel less respected because their coping skills may be viewed as inadequate or they may be considered less able to handle their responsibilities. Therefore, these individuals might feel that seeking mental health counseling services for depression is risky, making them less likely to seek support or treatment for depression [175]. It is important to safeguard the mental health of medical students with an effective plan to support their wellness as well as their education. Interventions targeting the many predisposing psychological factors in COVID-19-related anxiety and depression symptoms among medical students should be adopted by medical college administrators and policy makers to ameliorate psychological distress, which may negatively impact students’ academic performance [101, 176]. Several strategies had been investigated [177, 178, 179, 180]. Implications and recommendations for different levels, including medical students, medical schools and institutions, and policymakers, are presented in Table 4.

Table 4 Implications and recommendations
Full size table

Study strengths and limitations

Our study has several strengths. The search strategy was comprehensive. As far as we know, our study was the most updated analysis and comprehensive review on this topic. Moreover, our implications and recommendations would be helpful for medical students, schools and institutions, and policymakers developing strategies to promote mental health for medical students when encountering the pandemic. Our review also had several limitations. Firstly, although the identified articles involved a large number of participants, meta-analysis and quantitative analysis revealed a variety of heterogeneity in the data. Some studies have also reported high heterogeneity in anxiety and depression among medical students [21, 22]. Publication bias is possible, especially because more than half of studies were conducted in Asia.

Secondly, the search strategies influenced the articles searched as well as the number of articles retrieved. For instance, the use of wildcards, adjacencies, and truncations in a keyword search might identify more relevant studies. Moreover, including the outcome measures (sometimes depression or anxiety was a secondary outcome and might not be discussed in the abstract or title) or using more or different key terms (such as “mental health” or “mental wellness”) in the search strategy might have an influence on the search results. Therefore, the possibility remains that some unidentified or unpublished articles and gray literature were not included in our study. Moreover, because the search was limited to English language, studies in other languages were not included. Furthermore, medical students were limited to M.D. and M.B.B.S. programs, so some different programs for medical students might not be included in the study.

Thirdly, many other factors that might have an influence on heterogeneity of the data, such as family history, family income, emotional trauma, residential area, smoking, and substance use, were not included in this study. Although our results revealed the prevalence of anxiety and depression among medical students in some countries and regions, data were lacking for other countries and regions.

Fourthly, the identification of anxiety symptoms or depressive symptoms using a self-report questionnaire is likely to be subjective so response bias is possible, although the questionnaires used had good validity and reliability. Further objective measures might be needed to confirm the diagnoses. Lastly, longitudinal follow-up studies are required to track the evolution of symptoms and measure the long-term mental health impacts of the COVID-19 pandemic among medical students worldwide.

Conclusions

This systematic review and meta-analysis highlighted that medical students are at high risk of developing anxiety and depression during the COVID-19 pandemic. A substantial proportion of medical students have experienced adverse pandemic-related psychological impacts. Appropriate strategies are needed to meet the psychological needs of this population and protect their mental health status. Future studies are necessary to assess the appropriateness of management strategies to treat and prevent mental health disorders among medical students during the COVID-19 pandemic.