key: cord-0998465-g1f3wie3
authors: Broks, V. M. A.; Stegers-Jager, K. M.; van der Waal, J.; van den Broek, W. W.; Woltman, A. M.
title: Medical students' crisis-induced stress and the association with social support
date: 2022-03-21
journal: nan
DOI: 10.1101/2022.03.18.22272605
sha: 7f769757e9f64850242118685d10f43a9ae0d076
doc_id: 998465
cord_uid: g1f3wie3

Background: Medical schools are challenged to guard student wellbeing due to the potential negative impact of the COVID-19 outbreak on top of the already high prevalence of mental distress. Whereas social support is generally associated with less crisis-induced stress, it is unknown whether this applies to medical students during the COVID-19 outbreak. Objectives: The impact of the COVID-19 outbreak on perceived stress of medical students was assessed by comparing their perceived stress levels during the outbreak to both their own baseline and the previous cohort's pre-COVID-19 stress levels. Then, the association between social support and COVID-19 induced stress was assessed. Methods: Dutch Year-1 medical students of cohort 2019 (n=99) completed the 14-item Perceived Stress Scale (PSS-14) at two time points: baseline (pre-COVID-19) and final measurement (COVID-19). Social support - emotional-informational support and club membership - was assessed during the final measurement. PSS and social support scores were compared to similar measurements of cohort 2018 (n=196). Students' baseline stress levels, gender and study performance were controlled for when comparing two cohorts. Results: Stress levels did not differ statistically significant between both pre-COVID-19 measurements of cohort 2018 and baseline cohort 2019. During the COVID-19 outbreak, cohort 2019 showed significantly higher stress levels compared to baseline (paired t-test: t=6.07, p<.001) and compared to cohort 2018 (linear regression: B=4.186, p<.001). Only during the COVID-19 outbreak, higher levels of social support - i.e. emotional-informational support (B=-0.75, p<.001) and club membership (B=-3.68, p<.01) - were associated with lower levels of stress. Conclusions: During the COVID-19 outbreak, the perceived stress of medical students was higher - especially for students with lower levels of social support. Our results suggest that medical schools should optimize social support to minimize crisis-induced stress.

The prevalence of mental distress, i.e. anxiety-, depression-, or burnout symptoms, in medical students is 48 high compared to their age-matched peers [1] [2] [3] . Approximately a quarter to one-third of medical 49 students shows symptoms of depression [4, 5] , and roughly 40% shows burnout symptoms [6] . These 50 mental problems can be caused by stress [7] . A recent stressor in the shape of a crisis, is the COVID-19 51 outbreak. The outbreak's potential negative impact on mental wellbeing combined with the already high 52 prevalence of mental problems in medical students, exacerbates the challenge medical schools face to 53 guard their students' wellbeing [8] . Research regarding factors related to higher stress levels during a crisis 54 -or in other situations in which stressors increase -will enable medical schools to limit the negative impact 55 of such crises on student wellbeing. Social support is possibly one of the factors that is associated with 56 crisis induced stress [9] [10] [11] . Therefore, the present study has two objectives. The first objective is to 57 investigate whether the COVID-19 outbreak impacted perceived stress of medical students. The second 58 objective is to investigate the association between social support and COVID-19 induced stress for medical 59 students. 60 61

The COVID-19 outbreak has disrupted everyday life, which negatively impacted the mental wellbeing of 62 the general population [12] [13] [14] . Compared to the general population, especially students reported mental 63 health problems during the outbreak [15] . A possible explanation is that student life and its social aspects 64 were affected by measurements regarding social distancing, lockdown and if necessary quarantine [16] , 65 including the transition to online education [17] . However, for medical students, studies show mixed 66 results regarding the impact of the COVID-19 outbreak on wellbeing. A systematic review reports that 67 anxiety levels in medical students did not increase during the outbreak [18] . Whereas other studies report 68 higher levels of burnout symptoms and stress for medical students during the outbreak [19] , especially 69 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted March 21, 2022. ; https://doi.org/10.1101 https://doi.org/10. /2022 controlling for baseline measurement, gender and study performance (Table 2 -Model 1). Significant 137 differences between measurements are shown: *** p<.001. 138 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) measures general distress and someone's ability to cope with this stress. The questionnaire focuses on 153 feelings during the last month, therefore, every item starts with "In the last month…", for example "In the 154 last month, how often have you been upset because of something that happened unexpectedly?". The 155

items are scored on a 5-point Likert scale ranging from 0 (never) to 4 (very often). The maximum score on 156 the PSS-14 is 56. 157 158 Social support 159 Social support was assessed with questions regarding emotional-informational support and club 160 membership to cover different types of social support that are relevant to students in the Netherlands 161

[38]. To measure emotional-informational support, four items of the MOS social support scale were 162 selected from the subscale emotional-informational support [39] (S1 Table) . For the selection of the items, 163 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted March 21, 2022. ; https://doi.org/10.1101 https://doi.org/10. /2022 the following was taken into account: an equal distribution of both emotional and informational items, 164 item-scale correlations and applicability to our sample of medical students. The selected items focus on 165 the availability of different forms of social support: for example, having someone to listen to you, to offer 166 you advice or to share your thoughts with. The items are scored on a 5-point Likert scale ranging from 0 167 (never) to 4 (always). The maximum score on emotional-informational support is 16. Alpha reliability 168 measures were computed in order to assess internal consistency for this adjusted scale (S1 Table) . The performed for both cohorts to study whether emotional-informational support and club membership 187 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted March 21, 2022. ; https://doi.org/10.1101 https://doi.org/10. /2022 were complementary to each other in relation to stress levels. To this end, groups based on social support 188 were formed by crossing the presence of club membership (yes/no) with emotional-informational support 189 (high/low). Emotional-informational support was considered low when students' scores were in the 25 th 190 percentile of the complete sample. 191 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted March 21, 2022. ; https://doi.org/10.1101/2022.03.18.22272605 doi: medRxiv preprint

Cohort characteristics 194 The response rate was lower in cohort 2019 compared to cohort 2018 (24% vs. 48%, Table 1 ). Proportions 195 of female students and of students who acquired all obtainable credits until the final measurement were 196 comparable between cohorts (see Table 1 ). Though emotional-informational support scores remained the 197 same (Table 1) , a chi-square test showed that the percentage of students that was member of a club was 198 significantly lower in cohort 2019: 56% in cohort 2019 compared to 68% in cohort 2018 (χ2=4.150, df=1, 199 p<0.05). The mean baseline stress level was comparable between cohorts, but the mean stress level 200 during the final measurement was significantly higher for cohort 2019 than for cohort 2018 (t=4.134, 201 df=158.69, p<.001). 202 CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted March 21, 2022. ; https://doi.org/10.1101/2022.03.18.22272605 doi: medRxiv preprint

A paired t-test showed that the perceived stress levels of cohort 2019 significantly increased from 26.05 208 to 30.65 (t=6.07, df=98, p<.001), whereas the stress levels of cohort 2018 did not significantly differ 209 between the baseline and final measurement ( Fig 1B) . This indicates that the perceived stress levels of 210 students significantly increased during the COVID-19 outbreak. 211

To assess the effect of the COVID-19 outbreak, we not only compared stress levels within cohorts, we also 212 compared the stress levels during the final measurement between two cohorts of medical students. With 213 respect to the control variables, a higher baseline stress level and not having obtained all obtainable 214 credits were related to higher stress levels during the final measurement ( Fig 1B) . Compared to cohort 2018, the perceived stress 218 levels for students of cohort 2019 -during the COVID-19 outbreak -were on average approximately 4 219 units higher on the Perceived Stress Scale (ranging from 0 to 56). 220

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) social support, perceived stress levels were still significantly higher during the COVID-19 outbreak. In 230 addition, when a student experienced more emotional-informational support, perceived stress levels 231

were lower (B E-I support =-0.417, 95%-CI: -0.659 --0.177, p<.001, Table 2 -Model 2). Also, students who were 232 member of a club had significant lower stress levels compared to their fellow students who were not 233 member of a club (B club member =-1.636, 95%-CI: -3.197 --0.077, p<.05, Table 2 informational support increased (Fig 2A) . For cohort 2018 a slight decrease in stress level was visible too, 241 however, this decrease was not statistically significant. Regarding club membership, only cohort 2019 242 demonstrated a significant difference in stress level between students who were a club member or not 243 ( Fig 2B) . The results indicate that for cohort 2019, emotional-informational support and club membership 244 were complementary to each other in relation to stress levels since the effects exist beside each other. In 245 line, post-hoc analysis showed that for cohort 2019, emotional-informational support and club 246 membership were complementary to each other in relation to stress levels. Students with only one of the 247 two types of social support -i.e. club membership or high emotional-informational support -showed 248 significantly higher stress levels compared to students with both types of social support (Fig 3) . For cohort 249 2018, differences in stress levels based on social support were not present. 250 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted March 21, 2022. ; https://doi.org/10.1101 https://doi.org/10. /2022 Discussion 278 279 The present study demonstrates that the perceived stress levels of medical students were significantly 280 higher during the COVID-19 outbreak -compared to their own pre-COVID-19 baseline stress levels as well 281 as to stress levels of the previous cohort. During the outbreak, students experiencing less emotional-282 informational support or not being member of a club, showed higher perceived stress levels compared to 283 their fellow students with more emotional-informational support or who were member of a club. The 284 results indicate that during times of crisis, social support is associated with less perceived stress of medical 285 students. 286

Our finding that stress levels of medical students during the COVID-19 outbreak were higher compared to 287 pre-COVID-19 stress levels, corresponds with previous studies reporting higher stress levels during the 288 outbreak [19] [20] [21] . We found that students' stress levels increased compared to both their own baseline 289 stress levels as well as compared to the stress levels of a previous cohort, including controls for gender 290 and study performance. Also, we could take into account potential fluctuations during the academic year 291 regardless of the pandemic, as reported in a previous study [40] . The mean stress level reported in the 292 present study during the outbreak was higher than the mean stress level of Dutch medical students after 293 increased performance standards [36] . Yet, the currently reported stress levels did not exceed stress levels 294 of medical students in the US after the implementation of a new curriculum [41] . The reason why the 295 COVID-19 outbreak elevated stress levels goes beyond the scope of the present study, but one might 296 speculate about a mix of academic uncertainty [8, 17] , online education [20], blurred study-home 297 boundaries and social isolation [42] . Students' stress levels are subjected to a variety of factors and 298 thereby context-specific. This context-specificity of students' stress levels illustrates the importance of 299 baseline measurements to make valid comparisons. Such baseline measurements could also serve future 300 research on the long-term effects of this COVID-19 induced stress. 301 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted March 21, 2022. ; https://doi.org/10.1101 https://doi.org/10. /2022 In line with the stress-buffering model [10, 11] , we found that only during the COVID-19 outbreak, higher 302 levels of social support were associated with lower stress levels for medical students. This shows that also 303 during a crisis in which social contacts are strained by social distancing measures, social support is 304 associated with lower stress levels. This mechanism is also corroborated by a recent study in which 305 medical students report that their own most effective strategies to lower the negative impact of stress 306 include de-stress through friends and family [43] . Studies report that in times of the COVID-19 outbreak, 307 social support is positively associated to wellbeing in adolescents [26], college students [27] , and the 308 general population [24] . These present study adds to these findings in two ways. First, by providing 309 evidence that this also applies to medical students; a group that already experiences higher levels of 310 mental distress compared to their peers [1] [2] [3] . Second, by showing that the association between social 311 support and stress level also applies in times of crisis where social support is strained due to social 312 distancing measurements [12, 13] . Even though the present study did not reveal lower levels of 313 emotional-information support during the COVID-19 outbreak, the number of students who reported 314 being member of a club did decrease from 68% to 56%. This decrease is possibly the result of measures 315 taken to limit the spread of the coronavirus, thereby making it impossible for clubs to gather with their 316 members. A potential explanation why emotional-informational support did not decrease is that students 317 moved back home to their parents' house where emotional-informational support was still available to 318 them since close family is a source of social support for Dutch students [38] . 319

The present study illustrates how two forms of social support -emotional-informational support and club 320 membership -are complementary to each other in relation to reported stress levels during the COVID-19 321 outbreak. A possible explanation for this finding can be found in a theoretical model for mechanisms 322 linking social support to health, where primary and secondary social resources are distinguished, 323 described as intimates and knowledgeable others respectively [44] . According to this model, the two 324 sources of social support each have their own attributions to buffering the impact of stressors. Intimates 325 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted March 21, 2022. ; https://doi.org/10.1101 https://doi.org/10. /2022 buffer the stress by companionate presence, offering care and instrumental assistance whereas 326 knowledgeable others buffer stress by enabling ventilation and by role modelling. Emotional-327 informational support can be considered a primary source -i.e., intimates -whereas club membership 328 can be considered a secondary source of social support -i.e., knowledgeable others. This possibly explains 329 the added value of both emotional-informational and club membership towards each other. Yet, it should 330 be noted that club membership reflects more than the level of social support, as previous research also 331 showed its link to socioeconomic status. For example, it was shown that adolescent girls with lower 332 socioeconomic status were less likely to participate in a sports club [45] . Therefore, it is possible that a 333 lack of club membership also reflects a lower socioeconomic status, which in turn negatively affects self-334 reported health [46] . In line, during the COVID-19 outbreak it was found that university workers living in 335 smaller homes reported higher levels of anxiety-depression-and stress symptoms [47] . Also, for college 336 students during the COVID-19 outbreak, economic disadvantage was associated with higher stress levels 337 [48] . Perhaps, the higher levels of stress found in the present study for students without club membership 338 are also partially explained by their socioeconomic status, but more research is needed. 339

A strength of the current study is that two measurements within two cohorts are included, enabling 340 controls for baseline measurements and potential fluctuations throughout the academic year described 341 in the literature [40] . Moreover, gender and study performance were controlled for, which are known to 342 be correlated with stress levels [20, 22, 36] . Even though study performance was taken into account, the 343 meaning of having obtained all possible credits up until the final measurement differed between cohorts. 344

Due to the outbreak, exams were postponed which resulted in a higher number of exams that still had to 345 be completed in the last part of the academic year. Therefore, the present study controlled for study 346 performance of students regardless of the COVID-19 outbreak, in order to rule out potential effects of 347 differences in performance level between students of different cohorts. However, this does not exclude 348 the potential stress caused by postponed exams during the outbreak as described in a previous study [49] . 349 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted March 21, 2022. ; https://doi.org/10.1101 https://doi.org/10. /2022 Further, a limitation of the present study is the lower response-rate for cohort 2019. This was possibly 350 due to the COVID-19 outbreak, which resulted in a fully online data collection instead of a combination of 351 data collection in class and online. Even though we were able to control for student characteristics in the 352 analyses, a bias may still be present due to this lower response-rate. The present study focused on Year-353 1 students, and this appears to be a relevant group since students in early stages of medical and dental 354 school seem susceptible to the negative impact of the COVID-19 outbreak [19, 50] . Whether the results 355 described in the present study will be similar for students in advanced stages of medical school needs to 356 be further investigated. This also applies to the generalizability of the results to students from other 357 schools, since the relation between social support and perceived stress may be different for medical 358 students compared to students from other schools due to the higher prevalence of mental distress in 359 medical students compared to their age-matched peers [1, 2] . 360

Even though medical schools are not able to change the current COVID-19 crisis or any future crisis, they 361 can help students get through it. Our findings suggest that in times of crisis medical students' wellbeing 362 can benefit from social support. With the COVID-19 pandemic still not being over and potential new 363 lockdowns are possible, medical schools could play a more active role themselves by further exploring the 364 (digital) options to provide different kinds of social support to its students. The first type of social support 365 entails companionate care and instrumental assistance, which can be provided through one-on-one 366 (online) mentoring with a faculty member [51], or a peer [52, 53] . The second type of social support has 367 more to do with enabling ventilation with peers and role modelling, some sort of social embeddedness. 368

Medical schools can achieve this by for example creating online communication platforms [54] , or by 369 stimulating peer relationships by promoting cooperation amongst students in the medical school program 370 [55, 56] . All in all, when implementing (online) education, medical schools should not only focus on 371 qualification but also on the social functions of education. 372 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted March 21, 2022. ; https://doi.org/10.1101 https://doi.org/10. /2022 To conclude, the present study provides solid evidence of COVID-19 induced stress in medical students, 373 especially among those with less social support. Findings of the present study are in line with an existing 374 model describing the buffering effect of social support on crisis-induced stress and therefore go beyond 375 the current COVID-19 pandemic. Medical schools can optimize social support for its students by offering 376 social support on different levels in order to minimize the negative impact of future global, national or 377 individual crises. 378 379 Acknowledgements 380 We thank Daphne Pol, who helped with data collection and data entry of cohort 2018 for our study. We 381 also wish to thank David van Klaveren for helping us with our statistical analyses. 382 383 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted March 21, 2022. ; https://doi.org/10.1101/2022.03.18.22272605 doi: medRxiv preprint S1 Someone to turn to for suggestions about how to deal with a personal problem Iemand die je suggesties kan geven over hoe je met een persoonlijk probleem om moet gaan 0 1 2 3 4

Note: Questions and scoring emotional-informational support in English, with in italic-grey the Dutch 534 version used for the data-collection. Minimum score: 0, maximum score: 16. 535

Alpha reliability cohort 2018 = 0.895 536

Alpha reliability cohort 2019 = 0.907 537 538 539 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted March 21, 2022. ; https://doi.org/10.1101 https://doi.org/10. /2022 

Systematic review of depression, anxiety, and other 385 indicators of psychological distress among US and Canadian medical students

Medical student depression, anxiety and distress outside N orth A 388 merica: a systematic review

A narrative review on burnout experienced by medical students and 390 residents

Prevalence of depression, 392 depressive symptoms, and suicidal ideation among medical students: a systematic review and meta-393 analysis

Prevalence of depression amongst medical students: 395 A meta-analysis

Burnout in medical students 397 before residency: a systematic review and meta-analysis

More than a feeling: 400 A unified view of stress measurement for population science

Mental Health from Medical School to Medical Practice: Finding a Path Forward

Social support as a moderator of life stress

Social relationships and health

Stress, social support, and the buffering hypothesis

The Mental Health Consequences of COVID-19 and Physical 410 Distancing: The Need for Prevention and Early Intervention

Mental health during the COVID-19 pandemic: Effects of stay-413 at-home policies, social distancing behavior, and social resources

Impact of COVID-19 pandemic on mental 416 health in the general population: A systematic review

Stress and coping during COVID-19 pandemic: Result of an online survey

Psychosocial impact of 421 COVID-19

Priorities for addressing the impact of the COVID-423 19 pandemic on college student mental health

426 Prevalence of anxiety in medical students during the COVID-19 pandemic: a rapid systematic review 427 with meta-analysis

Medical Studies during the 429 COVID-19 Pandemic: The Impact of Digital Learning on Medical Students

Association of COVID-19 Pandemic with 432 undergraduate Medical Students' Perceived Stress and Coping

Levels of stress in medical students due to 435 COVID-19

The effects of acute stress on performance: implications for health professions 437 education

Measuring psychological resilience to disasters: are 439 evidence-based indicators an achievable goal? Environ Health

Loneliness in the UK during the COVID-19 pandemic: Cross-sectional results from the COVID-19

Stress and coping in the time of COVID-19: Pathways to resilience and 445 recovery

The effect of social support on mental 447 health in Chinese adolescents during the outbreak of COVID-19

Resilience, social support, and coping as 450 mediators between COVID-19-related stressful experiences and acute stress disorder among college 451 students in China

Factors associated with 453 resilience to and recovery from burnout: a prospective, multi-institutional study of US medical students

Meeting the challenges of medical student mental health and 456 well-being today

The associations between social support, 458 health-related behaviors, socioeconomic status and depression in medical students

Resilience among medical students: the role of 461 coping style and social support

Academic burnout and depression of Chinese medical 464 students in the pre-clinical years: the buffering hypothesis of resilience and social support. Psychol 465 Health Med

Burnout in medical students: 467 a systematic review of experiences in Chinese medical schools

Chinese medical students: the effect of resilience, personality and social support

A global measure of perceived stress

Crosswell AD, Lockwood KG. Best practices for stress measurement: How to measure 478 psychological stress in health research

Monitor Mentale gezondheid en Middelengebruik 481

English: Monitor Mental Health and Substance Use Students in Higher Education. Sub-report I. Mental 483 health of students in higher education)

The MOS social support survey

Fluctuations in Stress Over Time During the First Year of Health Science 487 Programs

Do medical student stress, health, 489 or quality of life foretell step 1 scores? A comparison of students in traditional and revised preclinical 490 curricula

Medical Student Education in the Time of COVID-19

Strategies for enhancing 494 resilience in medical students: a group concept mapping analysis

Mechanisms linking social ties and support to physical and mental health

Family support and ease of access link 499 socio-economic status and sports club membership in adolescent girls: a mediation study

Social relations or social capital? Individual and community health effects of 502 bonding social capital

Psychological Impact 504 of the Lockdown Due to the COVID-19 Pandemic in University Workers: Factors Related to Stress

Psychological Stress and Healthy Lifestyle Behaviors during COVID-19 among Students in a US Midwest 508 University

Assessing the psychological impacts of CoViD-19 in undergraduate 510 medical students

Psychological 512 impact of the coronavirus disease 2019 (COVID-19) pandemic on dental students: A nationwide study

Excuse the cat…" Reflections on online mentoring during the COVID-19 pandemic

We know what they're struggling with": student 517 peer mentors' embodied perceptions of teaching in a health professional education mentorship 518 program

Peer mentoring for medical students 520 during COVID-19 pandemic via a social media platform