https://orcid.org/0000-0002-6999-079X
Figures
Abstract
Background
Medical students have high levels of stress, which is associated with higher incidents of burnout, depression, and suicide compared to age-matched peers. Mindfulness practices have been shown to reduce stress among medical students.
Purpose
The purpose of this systematic review and meta-analysis was to examine if mindfulness interventions have an overall effect on stress outcomes in the high-stress population of medical students globally, particularly given the wide variety of interventions. Any intervention designed to promote mindfulness was included.
Methods
A comprehensive literature search was completed to include multiple databases, ancestry, and hand-searching and 35 studies were included. Standardized mean difference effect sizes (ES) were synthesized across studies using a random-effects model for changes in stress levels in medical students ≥ 18. Moderator analyses were performed to explore variations in effects by participant and intervention characteristics.
Results
Mindfulness interventions significantly improved stress among medical students in both the two-arm studies (d = 0.370, k = 19, n = 2,199, 95% CI 0.239–0.501, p < .001) and one-arm pre-post studies (d = 0.291, k = 30, n = 18 (two cohorts from Dyrbye et al), 95% CI 0.127–0.455, p = 0.001). Moderator analyses found trends in less hours and less required practice resulted in better improvement in stress.
Conclusions
This study further confirms that despite a wide variety of mindfulness interventions for medical students around the world, they produce an overall small-to-moderate effect on stress reduction. Future research looking at the most effective protocols for high-stress medical students would be beneficial.
Citation: Sperling EL, Hulett JM, Sherwin LB, Thompson S, Bettencourt BA (2023) The effect of mindfulness interventions on stress in medical students: A systematic review and meta-analysis. PLoS ONE 18(10): e0286387. https://doi.org/10.1371/journal.pone.0286387
Editor: Jakob Pietschnig, Universitat Wien, AUSTRIA
Received: September 13, 2022; Accepted: May 16, 2023; Published: October 5, 2023
Copyright: © 2023 Sperling et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All data is available in both the articles included in the meta-analyses, and at the University of Missouri public repository MOspace at https://hdl.handle.net/10355/94381.
Funding: The authors are grateful for a grant from the University of Missouri's Graduate Professional Council which supported this research.
Competing interests: The authors have declared that no competing interests exist.
Introduction
The onset of stress begins very early in medical training and continues to be high throughout medical school and residency [1, 2]. In a recent study, 27.2% of medical students had depression and 11.1% had suicidal ideation [3]. Despite lower baseline levels of suicidality, suicide rates are higher among medical students only months after beginning medical school compared to age-matched peers [3, 4]. Further research shows that 34% of medical students experience emotional exhaustion, 34% feel high levels of depersonalization towards others, and 46.6% feel low levels of a sense of personal accomplishment [5]. More than two-thirds of medical students report sleep deprivation, lack of control over their schedule, and lack of confidence in their knowledge and skills [6]. This alarming trend continues throughout the lengthy process of becoming a physician, suggesting medical training contributes to the emergence of mental and physical health conditions [4].
The negative effects of chronic stress are well known to cause physiologic and/or psychological illness and disease [7–16]. Mindfulness interventions have been consistently shown to reduce stress in the general population are relatively simple to implement, low-cost, have negligible risks, and require low time commitments [17]. A recent systematic review evaluating the cost-effectiveness of mindfulness found that mindfulness interventions were more effective and less expensive than standard care for individuals with depression, among other conditions [18]. Mindfulness has not been consistently shown to assist with stress reduction in medical students [19–22], although a recent systematic review of meta-analyses demonstrates mindfulness was more effective than passive controls in healthcare professionals/trainees and allied health students, but not better than active controls in allied health students [23]. Medical students were not examined separately [23]. Our meta-analysis aims to add to the evidence on mindfulness-based interventions among medical students internationally, to assess the range of interventions being utilized, to determine if certain intervention or participant characteristics changed the degree of effectiveness, and to compile best practices from international sources.
Mindfulness is generally intended to help practitioners focus on the present moment, noticing their thoughts and emotions without judgement [17]. There are many types of mindfulness practices, or ways to achieve mindfulness. Common types of mindfulness practices include seated or walking meditation, yoga, breathwork, and mantras, which can reduce anxiety and depression, improve emotional coping, and increase self-awareness [19–22]. Objective physiological measures have also provided support for mindfulness, including decreased circulating levels of C-reactive protein and proinflammatory cytokines [24–27]; increased telomere length and telomerase activity [24, 25]; as well as reduced proinflammatory cytokines in those with depression and anxiety [28]. In medical students specifically, stress-related epigenetic expression of SLC6A4, a serotonin transporter gene, has been found to increase after mindfulness interventions [29], while serum cortisol decreased [30].
Past reviews and meta-analyses on mindfulness interventions have conflicting findings which this meta-analysis aims to correct. Some systematic reviews and meta-analyses have found mixed or no impact of mindfulness interventions on stress-related outcomes among medical students [19–22, 31]; while others found a reduction in stress outcomes [20]. Past reviews did not search gray literature [19] which this one did in order to reduce publication bias. One review did not assess stress specifically among medical students, limiting generalizability [21]. This meta-analysis corrects for these issues and adds 17 new studies for analysis. Table 1 lists the previous and current reviews and which articles are included in each per the PRISMA guidelines (see Table 1). We also explore topics previous reviews have not considered: do intervention effects vary based on country of study, year of students in school, duration of mindfulness practice, or with study design?
Methods
This systematic review and meta-analysis followed the Preferred Reporting Items for Systematic Reviews and MetaAnalyses (PRISMA) guidelines [32].
Inclusion & exclusion criteria
Inclusion criteria for the systematic review were (1) mindfulness component inclusive of sitting meditation, walking meditation, breathing exercises, body scanning, yoga, etc.; (2) medical students, including programs training medical doctors (MD), doctors of osteopathy (DO), or international equivalent (Bachelor of Medicine, Bachelor of surgery (MBBS), etc); (3) primary study of either single-arm pre/post design, randomized controlled trial (RCT), or two-arm non-randomized design; (4) written in or translated into English; (5) 18 years of age or older; and (6) reporting at least one stress outcome measure. Studies must have reported sufficient data to calculate effect sizes. Corresponding authors from three studies were emailed to request data appropriate for the meta-analysis, and one response was received, which was then able to be included in the meta-analysis [33]. Eligible stress outcomes were determined using the definition of stress from the National Cancer Institute [34]. Examples of stress outcomes include The Perceived Stress Scale (PSS) and the stress subscale from the Depression, Anxiety, Stress Scale (DASS; see Table 2 for all outcome measures). Burnout, anxiety, depression, compassion, compassion fatigue, perfectionism, and coping are separately defined; therefore, these outcomes were not included. No restrictions were placed on study location or date. Data of physiological measures of stress were collected, coded, and included in the appropriate meta-analysis alongside the self-reported measures if available (Fig 1). Studies were excluded if they had combined populations for which separate data were not available (e.g. medical and dental students), or were not primary studies (e.g. systematic reviews).
Search strategies
A research librarian was consulted to maximize the quality of the search. English language filters were used. The search was carried out between February 3 and May 4, 2022. The following 12 databases were searched: EBSCOhost (inclusive of Academic Search Elite; Alt Healthwatch; Business Source Elite; CINAHL; eBook Collection; ERIC; Funk & Wagnalls New World Encyclopedia; Health Source–Consumer Edition; Health Source–Nursing/Academic Edition; Library, Information Science & Technology Abstracts; MAS Ultra–School Edition; Military & Government Collection; Newspaper Source; Primary Search; OpenDissertations; APA PsycInfo; MAS Reference eBook Collection; Primary Search Reference eBook Collection; and MEDLINE); EBSCO Discovery, in the categories Complementary & Alternative Medicine, Health & Medicine, Nursing & Allied Health, Psychology, and Public Health; PubMed; Trip; Web of Science; Sociological Abstracts; SveMed+; Education Full Text; and Education Index Retrospective (available titles between 1923–1983). EBSCOhost and PubMed also search unpublished dissertations and conference abstracts. Publication Finder was searched for the following journals: Journal of Mind and Medical Sciences, International Journal of Medical Students, and Mindfulness, which were chosen due to the high probability that they might contain research on mindfulness in medical students. Research registries were searched including ClinicalTrials.gov and ISRCTN (condition categories Mental & Behavioral Disorders).
A structured search strategy was used, beginning with Boolean terms. If needed, these were then refined to MeSH terms. A Title/Abstract search was then done. Search terms included mindfulness, meditation, yoga, and medical students (see S1 Table.). The number of duplicates, irrelevant articles, non-English language articles, and books were recorded for reproducibility. A bibliographic search was done on six relevant published reviews [19–22, 35, 36]. Finally, hand searching strategies were used. The articles collected from the database searches were compared to see if there were common, frequently occurring journal names, which were then searched; these included the BMC Medical Education, Education in Medicine Journal, Medical Education Online, and Wisconsin Medical Journal. The search range for the hand search was set for 2014–2022. This date range was chosen due to when the topic of mindfulness in medical students was observed to become common: EBSCOhost and PubMed databases did not return more than two articles per year between 1977–2012, and only four articles in 2013. Article retrieval numbers increased to ten or more from 2014 onward.
Data management
Zotero and Excel were used for data management. An Excel spreadsheet was used for tracking the search source, date, and terms.
Data extraction and coding
A comprehensive codebook was developed per guidelines in Cooper [37]. The codebook draft was blind-reviewed and refined by two researchers. A coder trained in meta-analytic methods extracted data from eligible studies. Categories of variables were study identification, basic information (research design, funding, source of article), population characteristics and demographics (program type, location, mean age, gender, race/ethnicity), reliability and validity (blinding, attrition reporting, randomization), intervention details (name, length of intervention, duration of sessions, total sessions, location of sessions, delivery mode), outcomes (name, self-reported or physiologic, time point, direction of effect for outcome measure), and results (sample size, raw means, standards deviations, standard error, difference in means, and direction of effect for the intervention). The timepoints chosen for coding were the pre-intervention measures and the post-measure closest to the final day of the study, to provide consistency; most studies did not report other follow-up time points.
Data analysis
Data analysis was conducted using Comprehensive Meta-Analysis Software (CMA). A random effects model was chosen a priori to account for expected heterogeneity among studies. Eligible studies included participants and interventions with a wide variety of characteristics, necessitating the use of the random effects model, which accounts for between-study and within-study variance [38].
Standardized mean difference effect sizes (ES, Cohen’s d) were calculated in CMA. Most were done by subtracting the mean of the control group from the mean of the treatment group and dividing by the pooled standard deviation (SD); two studies required use of difference in means and standard deviation; and two studies required use of difference in means and confidence intervals. Cohen’s d was chosen because of recent research showing Hedges’ g can result in more biased meta-estimation than Cohen’s d [39]. A positive ES is indicative of improved stress outcome measures and decreased stress, while a negative ES is indicative of worsened stress outcome measures and increased stress among medical students. In the one case where more than one self-reported stress outcome measure was used, they were combined in CMA by taking the average of the means [38]. Effect sizes were classified as small (≤ 0.20), medium (= 0.50), or large (≥ 0.80) [32]. Heterogeneity of the effects were assessed using Cochrane’s Q, I2, and T. Subgroup analysis and meta-regression were used to assess possible moderating effects of dichotomous or continuous variables, respectively. Cochrane’s Risk of Bias tool version 2 (RoB2) was used to evaluate the risk of bias across studies (Fig 2) [40].
Results
The comprehensive database search identified 2,203 potential studies. A bibliography search and hand searching of journals identified 7 more potential studies. After removal of duplicates, 1,518 citations were screened for inclusion. The PRISMA flow diagram for study selection is shown in Fig 1 [32]. The 35 studies selected for inclusion in this systematic review are described in Table 2. The total sample size across the data sets included in the meta-analysis is 2,199 (intervention groups n = 954; control groups n = 1,245).
Study and participant characteristics
The date of study publications ranged from 2003–2022. Nine of the studies were RCTs, and nine were non-randomized studies with separate intervention and control groups, with the remaining 17 single-arm, pre-post design. The studies included 14 (k = 15) conducted in the United States and 21 (k = 33) conducted in international medical programs. A total of 12 countries were represented outside of the U.S.: Australia (n = 4), Brazil (n = 1), Canada (n = 4), Chile (n = 1), Germany (n = 2), Grenada (n = 1), India (n = 1), Jamaica (n = 1), Malaysia (n = 2), the Netherlands (n = 1), Spain (n = 1), and Thailand (n = 2). The mean age of participants ranged from 18.87 to 25.2. Female participants made up between 47% to 77.3% of the samples in the studies. Nine of the 35 studies reported data on race or ethnicity. Study characteristics can be found in Table 2.
Risk of bias
There were some concerns of risk of bias in the RCTs of this meta-analysis as evaluated with the RoB2 (Fig 2) [40]. One important risk of bias was due to non-adherence or possible non-adherence (i.e. not reported) of participants to the study protocols, which could have affected study outcomes [41–45]. Only four studies reported protocol adherence; two reported an average attendance of less than 50% [42, 46]. One study had a high drop-out rate (29%) and thus there was bias in the direction of the experimental hypothesis; as such it was coded as ‘some concern of bias.’ [47] Secondly, all studies had self-reported outcome measures, which increases risk of bias, and were subsequently coded as having ‘some concerns.’
Intervention characteristics
To understand the diversity of interventions, articles were combed for intervention characteristics. Of the 35 eligible studies for systematic review, there were 18 different mindfulness intervention protocols used, demonstrating the expected heterogeneity as programs are modified for time, convenience, and resource availability. Many universities developed their own programs, some which are listed as mandatory curriculum, some elective. There was wide variety in how much mindfulness training students received; the duration of the intervention varied from 1 week to 16 weeks, and the frequency of sessions from every other week to three times per day. The durations of the individual sessions varied from approximately 2 minutes to 4 hours, and the total number of sessions from 1 to 56.
Meditation was included in all interventions. Many interventions also had didactic instruction on the theories behind mindfulness and mind-body medicine, or principles of neuroscience such as neuroplasticity [33, 41–44, 48–59]. Breathing exercises and/or breath awareness were also commonly taught [33, 41, 43, 49, 56, 60–63], as well as body scanning and visualization [33, 41, 44, 49, 56, 61, 64]. Yoga was part of many of the interventions [33, 47–49, 56, 60, 65–68]. Three reported instructing students in walking meditation [49, 56, 63], and two described instruction in self-acceptance or gratitude [42, 44]. Several described an educational intervention on emotional reactivity and non-judgement [41, 44, 53, 59]. Three studies reported a journaling component [54, 65, 69]. Two described instruction on the Buddhist principle of loving-kindness [42, 63]. Two reported instructions on biofeedback [64, 69], and one had communication exercises [61].
Several interventions were delivered by distance learning, including through DVD [44, 70], CD [45], mobile app [46] and online content [71]. Three studies had a hybrid format with both asynchronous and in-person components [55, 59, 72]. Many programs required after-hours mindfulness or meditation practice [33, 49, 51, 54–56, 62, 63, 71].
Outcome measures
Nine different stress outcome measures were represented in all coded studies. The majority of studies (n = 23) used the Perceived Stress Scale (PSS), but the meta-analysis included a total of five different self-reported outcome measures (see Table 2). Warnecke et al. qualified for the meta-analysis and had two self-reported stress outcome measures, the PSS and the stress subscale of the DASS-21, which were combined in CMA by averaging the effect sizes per standard procedure. Dyrbye et al. had two cohorts, from 2014 and 2015; these were treated as separate samples [50]. Waechter et al. had three intervention groups, of which both the mindfulness and yoga groups was used in the meta-analysis [47] with a composite variable calculated by CMA.
Three studies investigated biomarkers instead of or in addition to self-reported outcome measures. Parshad et al. measured eleven cardiovascular biomarkers in healthy medical students before and after six weekly 1-hour yoga sessions and found statistically significant improvement in six of the eleven markers [66]. Stoffel et al. measured stress-related epigenetic expression of SLC6A4, a serotonin transporter gene, and found it increased after a 3-month mindfulness intervention (p < .0001) [29]; self-reported measures lacked enough detail to be included in the meta-analysis. Turakitwanakan et al. measured serum cortisol as a biomarker of stress, and found lower levels of serum cortisol after a four-day meditation retreat (p < .05); this study also had a self-reported stress outcome measure which also lacked sufficient information to be included in the meta-analysis [30].
Summary effect sizes
Two separate meta-analyses were done, one for two-arm studies including RCTs, and one for one-arm pre-post studies, so that all effect sizes could be observed (see Table 2). The mindfulness interventions from the two-arm studies (k = 19, n = 18) had a statistically significant, small to moderate positive effect on reducing stress outcomes in medical students from pretest to posttest, compared to the controls (d = 0.370; 95% CI 0.239–0.501, p < .001) (Fig 3). Effect sizes ranged from -0.011 to 0.863 with a positive effect size indicating decreased stress and a negative effect size indicating increased stress following mindfulness interventions (Table 2 and Fig 3). As expected, there was statistically significant heterogeneity of effects noted across studies (Q = 31.947, I2 = 46.786%, T = 0.181, p < .001).
The mindfulness interventions in the one-arm studies (k = 30, n = 18) also had a statistically significant, small to moderate positive effect on reducing stress outcomes in medical students from pretest to posttest (d = 0.291, 95% CI 0.127–0.455, p = 0.001); Table 2 and Fig 4. (Note the n = 18 includes the two cohorts from Dyrbye et al., 2017, which had separate information from two cohorts and thus were produced separate effect sizes). Effect sizes ranged from -0.759 to 1.134 (Table 2, Fig 4). There was statistically significant heterogeneity of effects in this meta-analysis, also as expected (Q = 112.978, I2 = 84.953%, T = 0.310, p = 0.001).
Publication bias was assessed using three methods: the ‘p-uniform’ test [73], funnel plots and [74], and Egger’s regression intercept [75]. Multiple methods were used in order to provide more confidence in our assessment of publication bias. The ‘p-uniform’ test was completed in R statistical software with the assistance of an expert in R; the funnel plots and Egger’s regression intercept were completed in CMA. The p-uniform test provides a meta-analytic point estimate corrected for publication bias [73]. It indicates that if there is no publication bias between affirmative (i.e., significant and positive) and non-affirmative (i.e., nonsignificant or negative) studies, the corrected meta-analytic point estimate for the two-arm studies would be 0.37 (95% CI [0.23–0.51]) and 0.28 (95% CI [0.07–0.49) for the one-arm studies [73]. Reassessment of both data sets was done with removal of potentially biased studies; for the one-arm meta-analysis this did not change results [64], and for the two-arm meta-analysis, there was less publication bias with the removal of one study with a large negative effect [50]. The p-uniform test also produces funnel plots, which indicated the meta-analyses for both groups of studies were robust to publication bias [73]; all of the data output from R, including the p-uniform funnel plots, are freely accessible in the data repository.
Issues in interpretating traditional funnel plots can occur with a small number of studies, as plots can look asymmetric by chance without publication bias being present [73]; however, for completeness, they are included here (Figs 5 and 6). The funnel plots showed general symmetry, and effect sizes are seen in the lower left corner of the plot representing studies with lower effect sizes, demonstrating this meta-analysis includes studies which possibly faced publication bias due to lack of significant results (Figs 5 and 6).
The Egger’s regression test in the two-arm meta-analysis (t = 0.04, p = 0.97, intercept 0.04) and one-arm meta-analysis (t = 0.47, p = 0.64, intercept 0.67) were not significant, suggesting that the data sets are unlikely to be influenced by small sample bias [75]. A one-study-removed sensitivity analysis was run in CMA for both meta-analyses and no studies were identified as having undue influence on the magnitude of the summary effect sizes. These analyses together suggest little if any publication bias in the meta-analytic data sets.
Moderator analyses
Moderator analyses were undertaken to investigate heterogeneity, and there were several significant findings (see Table 3). The average effect size associated with the absence of required home practice was larger in the two-arm studies (d = 0.586, p < .001) than that associated with studies which required home practice (d = 0.318, p < 0.001; Q-between = 1.370, p of Q = 0.242); this was also true for the one-arm studies (absence of required home practice: d = 0.426, p < 0.001; required home practice: d = 0.186, p = 0.102; Q-between = 3.820, p of Q = 0.148). In the one-arm meta-analysis the average effect size was greater in studies with mindfulness intervention delivered to clinical students as compared to preclinical students, but preclinical vs. clinical participants in two-arm studies were very similar (see Table 3). In the one-arm studies, elective curriculum (d = 0.482, p = <0.001) was more effective than mandatory curriculum (d = 0.075, p = 531, Q = 6.862, p of Q = 0.032); the opposite was found in the two-arm studies, but with a low number of studies using mandatory curriculum which reduces trust in the results and should be viewed with caution. Interventions with yoga produced better stress outcomes in the one-arm studies, but those without yoga produced better stress outcomes in the two-group studies.
Meta-regression analyses did not produce statistically significant results for total hours, total number of sessions, average individual session duration, frequency of sessions, or duration of the entire intervention in either meta-analysis (Table 3).
Discussion
This work presents a systematic review and meta-analysis on the effects of mindfulness interventions among medical students internationally. Significant findings include a small-to-moderate summary effect size of d = 0.385 compared to control groups. The absence of required home practice and less total hours in the mindfulness intervention resulted in significantly better stress outcomes. As predicted, there was wide variety in the type of mindfulness interventions, duration in weeks, duration per session, combinations of interventions, delivery methods, and effectiveness. Unexpectedly, the results of the meta-regression suggested that dosage of features of the intervention were either uncorrelated or negatively associated with the magnitudes of the effect sizes. This may partly be due to the fact that even though some interventions were very short (a day) and some long (greater than 8 weeks), the great majority of the interventions were 6–8 weeks, and as such, lack in variation in length of the interventions may have contributed to the lack of association. The data suggesting that eliminating at-home practice and reducing the hours of intervention results in improved stress outcomes may reflect the fact that medical students are exceedingly busy, and less demanding interventions may be more accessible for them. Research suggests a minimum of eight weeks is needed before changes in the brain can be visualized on MRI [76]. Alternately, it may reflect selection bias or indicate that there is an effect from simply selecting to do any mindfulness at all.
Although most of the subgroup analyses were not significant, our findings suggest that mindfulness interventions may be equally effective at reducing stress among medical students in both clinical and preclinical years, students receiving their education in the United States compared to internationally, or RCTs compared to non-randomized two-arm studies.
The meta-analytic findings suggest mindfulness training is associated with a significant, small to moderate improvement in stress among medical students compared to controls. We found a slightly higher effect size than have previous meta-analyses on mindfulness interventions, incorporating a relatively larger number of effect sizes and focusing exclusively on medical students, which more accurately reflects the influence of mindfulness on reductions in stress among this population.
The primary articles provide a rich mine of ideas for implementing mindfulness into medical schools. Mindfulness has been implemented by integrating it into the preclinical curriculum in the form of elective or mandatory courses, such as the 11-week “Embodied Health” elective at Boston University School of Medicine [48], or the “Stress Management and Resilience Training (SMART)” course at the Mayo Clinic School of Medicine, which was required of all first and second-year medical students [50]. The University of Ottawa incorporated peer-led sessions, which may reduce faculty workload [61]. If space and instructors are a limitation, distance-mediated courses were also successful, such as the “Mindful-Gym” DVD [44, 70], or the “Mindfulness Training Program” which emailed 5-minute guided meditations daily [71]. One program simply had the students come early each morning to practice 20 minutes of breath-based meditation before classes began; this could likely be offered to students in most settings [62].
There was a wide variety of activities and combinations of activities that participants engaged in during the interventions which are primed for future investigation. Moderator analyses also suggest future areas of research, such as direct comparisons of at-home vs. no home practice, session durations, program lengths, and other program and participant characteristics; clear conceptualization of moderator variables is necessary to facilitate future examination of characteristics. In addition, further studies looking at the effect of social support, [77], coping strategies [78], exercise habits [79, 80], sleep habits [81], pre-existing mental health conditions [82], and personality traits [83] would be appropriate, as all of these have been shown to individually influence medical students’ wellbeing.
There are several important limitations of this study, beginning with a relatively small number of studies included in the meta-analysis. The wide variety in the type and duration of interventions, while not unexpected, affects our ability to draw conclusions. We were unable to convert the effect size to an original metric due to the heterogeneity of outcome measures used to capture stress. Gray literature was included to the extent possible with the available databases, but some may have been missed. This also prevents a comparison of published and unpublished literature to fully address publication bias. There was limited data on the number of sessions students attended out of the total possible sessions, with only four studies in the systematic review reporting this number. Many of the studies were English-language translations, but some studies in other languages may have been missed, contributing to language and country bias. Finally, a number of potentially confounding factors were not address in the primary studies, such as if students were restricted from doing exercise besides that prescribed by the intervention, or if students had access to quiet practice space.
Most of the studies did not use random assignment, and there is need for more RCTs on medical students’ stress. Many of the studies were conducted by the researchers who created the intervention, delivered the intervention, and/or were at the same institution that created the intervention, and therefore results may exhibit unconscious bias. A majority of the studies had elective courses as the intervention, and all of the randomized studies were initiated with students volunteering for inclusion; therefore, sampling bias is likely present, as students more likely to engage in mindfulness self-selected into the studies.
The studies conducted in the United States included primarily white individuals. Further research focusing on and incorporating Blacks, Hispanics, Asians, and other non-white and underrepresented groups of medical students into research on stress and stress prevention is greatly needed. There was also particularly limited information on diverse socioeconomic groups, religious preferences, LGBTQIA2S+ orientation, or gender identity, though membership in a minority group has been shown to lead to significantly higher stress levels than in peers [84–86].
Only stress outcome measures were included. Other studies have found that mindfulness also has effects on reducing anxiety and depression [21, 23] and enhancing such variables as empathy [21, 48] and wellbeing [60, 63], but that was not assessed here.
Conclusions
With more than half of medical students experiencing unmanaged stress, at least one in ten experiencing suicidality, and the COVID-19 pandemic exacerbating these issues, many of our physician trainees are clearly undergoing a mental health crisis. The effect of chronic stress on medical students is serious and life-threatening. This meta-analysis confirms the benefit of mindfulness in medical students and describes the wide variety of interventions that medical schools and preceptors have undertaken to address stress. Students who are educated on and given opportunities to engage with these techniques may experience less psychological and physiological symptoms of chronic stress. Further effort to incorporate mindfulness and other beneficial techniques into medical training is warranted.
Acknowledgments
The authors would like to sincerely thank Jo-Ana Chase, PhD, expert consultant on meta-analytic techniques; Rebecca Graves, research librarian, Sinclair School of Nursing; and Johannie Spaan, PhD, expert consultant for R statistical programming
References
- 1. Brooks SK, Gerada C, Chalder T. The specific needs of doctors with mental health problems: Qualitative analysis of doctor-patients’ experiences with the Practitioner Health Programme. J Ment Health. 2017; 26(2):161–166. pmid:27841030
- 2. Hill MR, Goicochea S, Merlo LJ. In their own words: Stressors facing medical students in the millennial generation. Med Educ Online. 2018; 23(1). pmid:30286698
- 3. Rotenstein LS, Ramos MA, Torre M, et al. Prevalence of depression, depressive symptoms, and suicidal ideation among medical students: A systematic review and meta-analysis. JAMA. 2016;316(21):2214–2236. https://www.doi.org.10.1001/jama.2016.17324 pmid:27923088
- 4. Kalmoe MC, Chapman MB, Gold JA, Giedinghagen AM. Physician suicide: A call to action. Mo Med. 2019;116(3): 211–216. pmid:31527944
- 5. Cecil J, McHale C, Hart J, Laidlaw A. Behaviour and burnout in medical students. Med Edu Online. 2014;19(1):25209. pmid:25160716
- 6. Mazurkiewicz R, Korenstein D, Fallar R, Ripp J. The prevalence and correlations of medical student burnout in the pre-clinical years: A cross-sectional study. Psychol Health Med. 2012; 17(2):188–195. pmid:21781020
- 7. Cole KR, Shields RK (2019). Age and cognitive stress influences motor skill acquisition, consolidation, and dual-task effect in humans. J Mot Behav 51(6): 622–639. pmid:30600778
- 8. Cooper ML, Pasini S, Lambert WS et al. Redistribution of metabolic resources through astrocyte networks mitigates neurodegenerative stress. Proc Natl Acad Sci. 2020;117(31): 18810–18821. pmid:32690710
- 9. Eriksson PS, Perfilieva E, Björk-Eriksson T et al. Neurogenesis in the adult human hippocampus. Nat Med. 1998; 4(11), 1313–1317. pmid:9809557
- 10. Fernandes HJR, Patikas N, Foskolou S, et al. Single-cell transcriptomics of Parkinson’s disease human in vitro models reveals dopamine neuronspecific stress responses. Cell Rep. 2020; 33(2): 108263. pmid:33053338
- 11. Ghadri JR, Sarcon A, Diekmann J, et al. Happy heart syndrome: role of positive emotional stress in takotsubo syndrome. Eur Heart J 2016;37(37): 2823–2829. pmid:26935270
- 12. Jabre P, Belpomme V, Azoulay E, et al. Family presence during cardiopulmonary resuscitation. N Engl J Med. 2013;368(11): 1008–1018. pmid:23484827
- 13. Lowry CA, Jin AY. Improving the social relevance of experimental stroke models: Social isolation, social defeat stress and stroke outcome in animals and humans. Front Neurol. 2020;11: 427. pmid:32477259
- 14. Mender I, Zhang A, Ren Z, et al. Telomere stress potentiates sting-dependent anti-tumor Immunity. Cancer Cell. 2020; 38(3): 400–411.e406. pmid:32619407
- 15. Orosz E, van de Wiele N, Emri T, et al. Fungal Stress Database (FSD)—a repository of fungal stress physiological data. Database. 2018: bay009. pmid:29688353
- 16. Stengel ST, Fazio A, Lipinski S, et al. Activating transcription factor 6 mediates inflammatory signals in intestinal epithelial cells upon endoplasmic reticulum stress. Gastroenterology. 2020;159(4): 13571374.e1310. pmid:32673694
- 17. Kabat-Zinn J. Mindfulness based stress reduction (MBSR). Constructivism in the Human Sciences. 2003;8(2): 73–107.
- 18. Zhang L, Lopes S, Lavelle T, Jones KO, Chen L, Jindal M, et al. Economic evaluations of mindfulness-based interventions: A systematic review. Mindfulness. 2022 Oct;13(10):2359–78. pmid:36061089
- 19. Daya Z, Hearn JH. Mindfulness interventions in medical education: A systematic review of their impact on medical student stress, depression, fatigue and burnout. Med Teach. 2018;40(2):146–153. pmid:29113526
- 20. Hathaisaard C, Wannarit K, Pattanaseri K. Mindfulness-based interventions reducing and preventing stress and burnout in medical students: A systematic review and meta-analysis. Asian J Psychiatr. 2022;69:102997. pmid:34995839
- 21. McConville J, McAleer R, Hahne A. Mindfulness training for health profession students—The effect of mindfulness training on psychological well-being, learning and clinical performance of health professional students: A systematic review of randomized and non-randomized controlled trials. Explore. 2017;13(1):26–45. pmid:27889445
- 22. Witt K, Boland A, Lamblin M, et al. Effectiveness of universal programmes for the prevention of suicidal ideation, behaviour and mental ill health in medical students: a systematic review and meta-analysis. Evid Based Ment Health. 2019;22(2). pmid:30918000
- 23. Goldberg SB, Riordan KM, Sun S, Davidson RJ. The empirical status of mindfulness-based interventions: A systematic review of 44 meta-analyses of Randomized Controlled Trials. Perspect Psychol Sci. 2022 Jan;17(1):108–30. pmid:33593124
- 24. Black DS, Slavich GM. Mindfulness meditation and the immune system: a systematic review of randomized controlled trials: Mindfulness meditation and the immune system. Ann NY Acad Sci. 2016 Jun;1373(1):13–24.
- 25. Dunn TJ, Dimolareva M. The effect of mindfulness-based interventions on immunity-related biomarkers: a comprehensive meta-analysis of randomised controlled trials. Clinical Psychology Review. 2022 Mar;92:102124. pmid:35078038
- 26. Oken BS, Fonareva I, Haas M, Wahbeh H, Lane JB, Zajdel D, et al. Pilot Controlled Trial of Mindfulness Meditation and Education for Dementia Caregivers. The Journal of Alternative and Complementary Medicine. 2010 Oct;16(10):1031–8. pmid:20929380
- 27. O’Toole MS, Bovbjerg DH, Renna ME, Lekander M, Mennin DS, Zachariae R. Effects of psychological interventions on systemic levels of inflammatory biomarkers in humans: A systematic review and meta-analysis. Brain, Behavior, and Immunity. 2018 Nov;74:68–78. pmid:29630988
- 28. Sanada K, Montero-Marin J, Barceló-Soler A, Ikuse D, Ota M, Hirata A, et al. Effects of Mindfulness-Based Interventions on Biomarkers and Low-Grade Inflammation in Patients with Psychiatric Disorders: A Meta-Analytic Review. IJMS. 2020 Apr 3;21(7):2484. pmid:32260096
- 29. Stoffel M, Aguilar-Raab C, Rahn S, et al. Effects of mindfulness-based stress prevention on serotonin transporter gene methylation. Psychother Psychosom. 2019;88(5):317–319. pmid:31461722
- 30. Turakitwanakan W, Mekseepralard C, Busarakumtragul P. Effects of mindfulness meditation on serum cortisol of medical students. J Med Assoc Thai. 2013;96 (Suppl. 1):S90–S95. pmid:23724462
- 31. Yogeswaran V, El Morr C. Effectiveness of online mindfulness interventions on medical students’ mental health: a systematic review. BMC Public Health. 2021;21(1):2293. pmid:34920715
- 32. Page MJ, McKenzie JE, Bossuyt PM, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. Br Med J. Published online March 29, 2021:71. pmid:33782057
- 33. Lampe LC, Müller-Hilke B. Mindfulness-based intervention helps preclinical medical students to contain stress, maintain mindfulness and improve academic success. BMC Med Educ. 2021;21(1):145. pmid:33663478
- 34.
National Cancer Institute at the National Institutes of Health. Stress. Available at https://www.cancer.gov/publications/dictionaries/cancer-terms/def/stress. Accessibility verified June 7, 2022.
- 35. de Vibe M, Bjørndal A, Fattah S, Dyrdal GM, Halland E, Tanner-Smith EE. Mindfulness-based stress reduction (MBSR) for improving health, quality of life and social functioning in adults: a systematic review and meta-analysis. Campbell Syst Rev. 2017;13(1):1–264.
- 36. Walsh A, Lehmann S, Zabinski J, et al. Interventions to prevent and reduce burnout among undergraduate and graduate medical education trainees: A systematic review. Acad Psychiatry. 2019;43(4). pmid:30710229
- 37.
Cooper H. Research synthesis and meta-analysis. Thousand Oaks, CA: Sage Publications; 2017.
- 38.
Borenstein M, Hedges L, Higgins J, Rothstein H. Introduction to Meta-Analysis. West Sussex, UK: Wiley; 2021.
- 39. Lin L, Aloe AM. Evaluation of various estimators for standardized mean difference in meta‐analysis. Stat Med. 2021;40(2):403–426. pmid:33180373
- 40. Sterne JAC, Savović J, Page MJ, Elbers RG, Blencowe NS, Boutron I, et al. RoB 2: a revised tool for assessing risk of bias in randomized trials. British Med J. 2019;366.
- 41. Damião Neto A, Lucchetti ALG, da Silva Ezequiel O, Lucchetti G. Effects of a required large-group mindfulness meditation course on first-year medical students’ mental health and quality of life: a randomized controlled trial. J Gen Intern Med. 2020;35(3):672–678. pmid:31452038
- 42. Danilewitz M, Bradwejn J, Koszycki D. A pilot feasibility study of a peer-led mindfulness program for medical students. Can Med Educ J. 2016;7(1). http://pubmed.ncbi.nlm.nih.gov/27103950/ pmid:27103950
- 43. Erogul M, Singer G, McIntyre T, Stefanov DG. Abridged mindfulness intervention to support wellness in first-year medical students. Teach Learn Med. 2014;26(4):350–356. pmid:25318029
- 44. Phang CK, Mukhtar F, Ibrahim N, Shian-Ling K, Mohd S. Effects of a DVD-delivered mindfulness-based intervention for stress reduction in medical students: A randomized controlled study. Educ Med J. Published online 2015:14.
- 45. Warnecke E, Quinn S, Ogden K, Towle N, Nelson MR. A randomised controlled trial of the effects of mindfulness practice on medical student stress levels: Effects of mindfulness practice on student stress levels. Med Educ. 2011;45(4):381–388. pmid:21401686
- 46. Yang E, Schamber E, Meyer RML, Gold JI. Happier healers: Randomized controlled trial of mobile mindfulness for stress management. J Altern Complement Med. 2018;24(5):505–513. pmid:29420050
- 47. Waechter R, Stahl G, Rabie S, et al. Mitigating medical student stress and anxiety: Should schools mandate participation in wellness intervention programs? Med Teach. 2021;43(8). pmid:33832384
- 48. Bond A, Mason H, Lemaster C, et al. Embodied health: the effects of a mind-body course for medical students. Med Educ Online. 2013;18. pmid:23639275
- 49. Bughi S, Sumcad J, Bughi S. Effect of brief behavioral intervention program in managing stress in medical students from two southern California universities. Med Educ Online. 2006;11(1). pmid:28253798
- 50. Dyrbye L, Shanafelt T, Werner L, Sood A, Satele D, Wolanskyj A. The impact of a required longitudinal stress management and resilience training course for first-year medical students. J Gen Intern Med. 2017;32(12). pmid:28861707
- 51. Finkelstein C, Brownstein A, Scott C, Lan YL. Anxiety and stress reduction in medical education: an intervention. Med Educ. 2007;41(3):258–264. pmid:17316210
- 52. Greeson J, Toohey M, Pearce M. An adapted, four-week mind–body skills group for medical students: Reducing stress, increasing mindfulness, and enhancing self-care. Explore. 2015;11(3):186–192. pmid:25792145
- 53. Hassed C, de Lisle S, Sullivan G, Pier C. Enhancing the health of medical students: outcomes of an integrated mindfulness and lifestyle program. Adv Health Sci Educ. 2009;14(3):387–398. pmid:18516694
- 54. Kakoschke N, Hassed C, Chambers R, Lee K. The importance of formal versus informal mindfulness practice for enhancing psychological wellbeing and study engagement in a medical student cohort with a 5-week mindfulness-based lifestyle program. PLoS One. 2021;16(10):e0258999. pmid:34673830
- 55. Oró P, Esquerda M, Mas B, Viñas J, Yuguero O, Pifarré J. Effectiveness of a mindfulness-based programme on perceived stress, psychopathological symptomatology and burnout in medical students. Mindfulness. 2021;12(5):1138–1147. pmid:33437325
- 56. Rosenzweig S, Reibel DK, Greeson JM, Brainard GC, Hojat M. Mindfulness-based stress reduction lowers psychological distress in medical students. Teach Learn Med. 2003;15(2):88–92. pmid:12708065
- 57. Slavin SJ, Schindler DL, Chibnall JT. Medical Student Mental Health 3.0: Improving Student Wellness Through Curricular Changes. Acad Med. 2014;89(4):573–577. pmid:24556765
- 58. van Dijk I, Lucassen PLBJ, Akkermans RP, van Engelen BGM, van Weel C, Speckens AEM. Effects of mindfulness-based stress reduction on the mental health of clinical clerkship students: A cluster-randomized controlled trial. Acad Med. 2017;92(7):1012–1021. pmid:28121650
- 59. Zúñiga D, Torres-Sahli M, Nitsche P, et al. Reduced burnout and higher mindfulness in medical students after a self-care program during the COVID-19 pandemic. Rev méd Chile. 2021;149(6):846–855. pmid:34751343
- 60. Bansal R, Gupta M, Agarwal B, Sharma S. Impact of short term yoga intervention on mental well being of medical students posted in community medicine: A pilot study. Indian J Community Med. 2013;38(2):105. pmid:23878424
- 61. MacLean H, Braschi E, Archibald D, et al. A pilot study of a longitudinal mindfulness curriculum in undergraduate medical education. Can Med Educ J. 2020;11(4). pmid:32821298
- 62. Paholpak S, Piyavhatkul N, Rangseekajee P, et al. Breathing meditation by medical students at Khon Kaen University: Effect on psychiatric symptoms, memory, intelligence and academic achievement. J Med Assoc Thai. 2012;95(3).
- 63. Garneau K, Hutchinson T, Zhao Q, Dobkin P. Cultivating person-centered medicine in future physicians. EJPCH. 2013 Nov 18;1(2):468.
- 64. Kraemer K, Luberto C, O’Bryan E, Mysinger E, Cotton S. Mind-body skills training to improve distress tolerance in medical students: A pilot study. Teach Learn Med. 2016;28(2). pmid:27064724
- 65. Motz K, Graves K, Gross C, et al. Impact of a mind-body medicine skills course on medical students’ perceived stress, mindfulness and elements of emotional intelligence. BMC Complement Altern Med. 2012;12(1):1–1.
- 66. Parshad O, Richards A, Asnani M. Impact of yoga on haemodynamic function in healthy medical students. West Indian Med J. 2011;60(2): 148–152. pmid:21942118
- 67. Prasad L, Varrey A, Sisti G. Medical students’ stress levels and sense of well being after six weeks of yoga and meditation. Evid Based Complementary Altern Med. 2016;2016. pmid:28053644
- 68. Simard AA, Henry M. Impact of a short yoga intervention on medical students’ health: A pilot study. Med Teach. 2009. pmid:19877871
- 69. Williams M, Estores I, Merlo L. Promoting resilience in medicine: The effects of a mind-body medicine elective to improve medical student well-being. Glob Adv Health Med. 2020;9. pmid:32499968
- 70. Keng SL, Phang CK, Oei TP. Effects of a brief mindfulness-based intervention program on psychological symptoms and well-being among medical students in Malaysia: A controlled study. Int J Cogn Ther. 2015;8(4):335–350.
- 71. Moore S, Barbour R, Ngo H, et al. Determining the feasibility and effectiveness of brief online mindfulness training for rural medical students: a pilot study. BMC Med Educ. 2020;20(1):1–12. pmid:32252750
- 72. Kemper KJ, Yun J. Group online mindfulness training: Proof of concept. Evid Based Complementary Altern Med. Published online October 10, 2014. pmid:25305208
- 73. van Assen MALM, van Aert RCM, Wicherts JM. Meta-analysis using effect size distributions of only statistically significant studies. Psychol Methods. 2015;20(3):293–309. pmid:25401773
- 74.
Cohen J. Statistical power analysis for the behavioral sciences (2nd ed.). Mahwah, New Jersey: Lawrence Earlbaum Associates; 1988.
- 75. Egger M, Smith GD, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. Br Med J. 1997;315(7109):629–634. pmid:9310563
- 76. Gotink RA, Meijboom R, Vernooij MW, Smits M, Hunink MGM. 8-week Mindfulness Based Stress Reduction induces changes similar to traditional long-term meditation practice–A systematic review. Brain Cogn. 2016;108. https://doi.org/10.1016/j.bandc.2016.07.001
- 77. Huang Z, Zhang L, Wang J, et al. Social support and subjective well-being among postgraduate medical students: the mediating role of anxiety and the moderating role of alcohol and tobacco use. Heliyon. 2021;7(12):e08621. pmid:34988318
- 78. Abbasi S, Mubeen N, Ayub T, Khan M, Abbasi Z, Baig N. Comparison of stress levels among medical and dental students in the clinical years of training and their coping strategies. J Pak Med Assoc. 2020;70(6). pmid:32810097
- 79. Mikkelsen K, Stojanovska L, Polenakovic M, Bosevski M, Apostolopoulos V. Exercise and mental health. Maturitas. 2017;106:48–56. pmid:29150166
- 80. Tsatsoulis A, Fountoulakis S. The protective role of exercise on stress system dysregulation and comorbidities. Ann NY Acad Sci. 2006; 1083: 196–213. pmid:17148741
- 81. Rathi A, Ransing RS, Mishra KK, Narula N. Quality of sleep among medical students: Relationship with personality traits. J Clin Diagnostic Res. 2018;12(9):1–4;
- 82. Yates J, James D, Aston I. Pre-existing mental health problems in medical students: a retrospective survey. Med Teach. 2008; 30(3), 319–321, pmid:18484461
- 83. Doherty EM, Nugent E. Personality factors and medical training: a review of the literature. Medical Education. 2011;45: 132–140. pmid:21208259
- 84. Berger M, Sarnyai Z. More than skin deep: stress neurobiology and mental health consequences of racial discrimination." Stress. 2015;18.1: 1–10. pmid:25407297
- 85. Hendricks ML, Testa RJ. A conceptual framework for clinical work with transgender and gender nonconforming clients: An adaptation of the Minority Stress Model. Prof Psychol Res Pr. 2012; 43(5), 460–467.
- 86. Meyer IH. Prejudice, social stress, and mental health in lesbian, gay, and bisexual populations: Conceptual issues and research evidence. Psychol Bull. 2003;129(5):674–697. pmid:12956539
- 87. Chen A, Kumar A, Haramati A. The effect of mind body medicine course on medical student empathy: a pilot study. Medical education online [Internet]. 2016 Jun 27 [cited 2022 Feb 14];21. Available from: http://pubmed.ncbi.nlm.nih.gov/27357909/ pmid:27357909
- 88. Zheng G, Lan X, Li M, Ling K, Lin H, Chen L, et al. Effectiveness of Tai Chi on physical and psychological health of college students: Results of a Randomized Controlled Trial. Schooling CMeditor. PLoS ONE. 2015 Jul 6;10(7):e0132605. pmid:26147842