This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Formative Research, is properly cited. The complete bibliographic information, a link to the original publication on https://formative.jmir.org, as well as this copyright and license information must be included.
Knowledge of mental distress and resilience factors over the time span from before to after a stressor is important to be able to leverage the most promising resilience factors and promote mental health at the right time. To shed light on this topic, we designed the RESIST (Resilience Study) study, in which we assessed medical students before, during, and after their yearly exam period. Exam time is generally a period of notable stress among medical students, and it has been suggested that exam time triggers mental distress.
In this paper, we aim to describe the study protocol and to examine whether the exam period indeed induces higher perceived stress and mental distress. We also aim to explore whether perceived stress and mental distress coevolve in response to exams.
RESIST is a cohort study in which exam stress functions as a within-subject natural stress manipulation. In this paper, we outline the sample (N=451), procedure, assessed measures (including demographics, perceived stress, mental distress, 13 resilience factors, and adversity), and ethical considerations. Moreover, we conducted a series of latent growth models and bivariate latent change score models to analyze perceived stress and mental distress changes over the 3 time points.
We found that perceived stress and mental distress increased from the time before the exams to the exam period and decreased after the exams to a lower level than before the exams. Our findings further suggest that higher mental distress before exams increased the risk of developing more perceived stress during exams. Higher perceived stress during exams, in turn, increased the risk of experiencing a less successful (or quick) recovery of mental distress after exams.
As expected, the exam period caused a temporary increase in perceived stress and mental distress. Therefore, the RESIST study lends itself well to exploring resilience factors in response to naturally occurring exam stress. Such knowledge will eventually help researchers to find out which resilience factors lend themselves best as prevention targets and which lend themselves best as treatment targets for the mitigation of mental health problems that are triggered or accelerated by natural exam stress. The findings from the RESIST study may therefore inform student support services, mental health services, and resilience theory.
Approximately 1 in 5 young people experience mental distress in the form of anxiety and depression [
A recent meta-analysis based on 122,356 medical students from 43 countries showed that the prevalence rate for depressive symptoms was 27.2% (range of individual studies: 1.4%-73.5%) [
The RESIST study is designed to capture (1) a period of moderate stress during the university term several months before exams, (2) a period of high stress during exam time, and (3) a period of what we expected to be low or moderate stress after exams (ie, during the summer vacation for many students). In addition to perceived stress and mental distress, we assessed 8 putative
With the RESIST study, we intend to shed light on which RFs lend themselves best as prevention targets (before the stressor) and which lend themselves best as treatment targets (at times of stress) for the mitigation of mental health problems that are triggered or accelerated by a natural stressor. Therefore, the RESIST study may lay the foundations necessary to inform student support services, mental health services, as well as resilience and transdiagnostic mental health theory. Given that our design relies strongly on the assumption that stress and mental distress levels increase during the exam period, we here conduct proof-of-principle analyses to investigate whether this is indeed the case.
RESIST is a cohort study with 3 occasions and a within-subject (natural) stress manipulation (ie, the exam period). Occasion 1 took place in a nonexam period during the university term (February and March 2018). Occasion 2 took place during the end-of-year exam period (approximately April to June 2018, depending on the timing of the exam period). Occasion 3 took place after the exam period, at the end of the term for year 6 students (for whom exams are earlier; approximately end of May to mid-July 2018), and in the summer vacation or autumn for year 1-5 students (approximately mid-August to mid-October 2018;
Study design. The Figure depicts the measures that have been assessed on the 3 occasions.
We recruited first- to sixth-year Cambridge medical students (from a cohort of approximately 1464 students). The inclusion criterion was that students had to be aged at least 18 years. Participants received monetary reimbursement for partaking (web-based vouchers: £5 [US $6.75] for occasion 1, £7 [US $9.50] for occasion 2, and £5 [US $6.75] for occasion 3). Participants who completed all 3 occasions were additionally enrolled in a prize draw (prize: five £50 [US $67.50] web-based vouchers). The maximum possible sample size we could have included was 800 participants, as we had a limited amount of money that we could spend on participant reimbursement. As a minimum sample size, we aimed for 225 participants. This is because we calculated that for a Gaussian regression-based model (minimum sample size = [((p × (p − 1))/2) × 5] [
Sample size overview (N=451).
Sample | Occasion | ||
|
Occasion 1, n (%) | Occasion 2a, n (%) | Occasion 3a, n (%) |
Taken part on, at least, 1 occasion | 451 (100) | 275 (61.0) | 283 (62.8) |
Taken part on, at least, 2 occasions | 324 (71.8) | 275 (61.0) | 283 (62.8) |
Taken part on occasion 2 but not 3 | 41 (9.1) | 41 (9.1) | N/Ab |
Taken part on occasion 3 but not 2 | 49 (10.9) | N/A | 49 (10.9) |
Taken part on all occasions | 234 (51.9) | 234 (51.9) | 234 (51.9) |
aOn occasions 2 and 3, only participants who had already taken part on occasion 1 were invited.
bN/A: not applicable.
The students received a web-based link to the questionnaire (survey software: REDCap [Research Electronic Data Capture]) via email. To prevent double partaking, we sent personalized emails with unique links to the students. We also advertised the study during lectures. Students who had not already participated received reminder emails until the end of the study occasion. For the first occasion, the link expired after 8 weeks. We sent the link for the second occasion approximately 3 weeks before the students’ first final exam and asked the students to confirm that they will submit the survey before their last final exam of the academic year. For year 6 students, the study link for the third occasion was sent out approximately a month before the end of the summer term. For year 1-5 students, the study link for the third occasion was sent 6 weeks before the start of the new academic year (the link for the third occasion expired after 8 weeks for all students).
Before we finalized the web-based survey, we performed a user review with volunteering medical students. On the basis of this pilot study, we evaluated whether the survey was easily understandable and acceptable. We adapted small features, mainly regarding the survey layout (importantly, these data were not part of the study). On occasion 1, a total of 139 items were assessed.
We assessed 8 demographic and clinical variables: academic course, year of academic education, gender, age, ethnicity, parental educational level, psychotherapeutic treatment, and psychopharmacology intake (ie, prescribed drugs).
We assessed the stress level during the last month using a 4-item short form of the validated Perceived Stress Scale (PSS) [
We assessed general mental health using the 12-item version of the General Health Questionnaire (GHQ-12) [
We assessed 7
Details of the resilience factor measures.
RFsa | Content and psychometric information | ||
|
|
||
|
High distress tolerance |
6-item subscale of the DTSb (15 items in total) [ Self-report items assessing distress tolerance levels such acceptability of being upset Previous research found a good reliability (DTS Cronbach α=.82 to .85; 6-item tolerance subscale Cronbach α=.82 to .84) [ In RESISTc, the distress tolerance subscale had a good reliability (Cronbach α: o1d=.82, o2e=.84, o3f=.83; coefficient Ω: o1=0.82, o2=0.85, o3=0.83) |
|
|
Low ruminative reflection |
5-item reflective rumination subscale of the RRSg (22 items in total) [ Self-report items assessing ruminative reflection levels such as trying to understand why you have a negative mood or why you feel in a given way Previous research found an acceptable reliability (RRS Cronbach α=.90; 5-item reflective rumination subscale Cronbach α=.72) [ In RESIST, the reflective rumination subscale had an acceptable reliability (Cronbach α: o1=.75, o2=.76, o3=.79; coefficient Ω: o1=0.76, o2=0.76, o3=0.80) |
|
|
Low ruminative brooding |
5 item brooding subscale of the RRS (22 items in total) [ Self-report items assessing brooding levels such as why things do not work out better or why other people do not have comparable problems Previous research found an acceptable reliability (RRS Cronbach α=.90; 5-item brooding subscale Cronbach α=.77) [ In RESIST, the brooding subscale had an acceptable reliability (Cronbach α: o1=.75, o2=.79, o3=.77; coefficient Ω: o1=0.76, o2=0.80, o3=0.78) |
|
|
High self-esteem |
10 items of the RSESh [ Self-report items assessing positive self-esteem levels such as being capable of doing things well and negative self-esteem levels such as feeling useless Previous research found a good reliability (RSES Cronbach α=.88) [ In RESIST, the RSES had an excellent reliability (Cronbach α: o1=.93, o2=.94, o3=.92; coefficient Ω: o1=0.93, o2=0.94, o3=0.92) |
|
|
High cognitive reappraisal |
6-item cognitive reappraisal subscale of the ERQi (10 items in total) [ Self-report items assessing cognitive reappraisal levels such as changing the content of thoughts to achieve a less negative or more positive mood Previous research found an acceptable reliability (6-item cognitive reappraisal subscale Cronbach α=.79) [ In RESIST, the cognitive reappraisal subscale had a good reliability (Cronbach α: o1=.83, o2=.87, o3=.88; coefficient Ω: o1=0.83, o2=0.87, o3=0.88) |
|
|
Low expressive suppression |
4-item expressive suppression subscale of the ERQ (10 items in total) [ Self-report items assessing expressive suppression levels, that is, the extent to which individuals suppress positive and negative emotions Previous research found an acceptable reliability (4-item expressive suppression subscale Cronbach α=.73) [ In RESIST, the expressive suppression subscale had an acceptable reliability (Cronbach α: o1=.75, o2=.73, o3=.76; coefficient Ω: o1=0.78, o2=0.76, o3=0.78) |
|
|
Low aggression potential |
12-item BAQj [ Self-report about aggression levels including physical aggression, verbal aggression, anger, and hostility Previous research found an acceptable-to-good reliability (BAQ Cronbach α=.76-.83) [ In RESIST, the BAQ had an acceptable reliability (Cronbach α: o1=.78, o2=.80, o3=.79; coefficient Ω: o1=0.79, o2=0.81, o3=0.78) |
|
|
|
||
|
High immediate family support |
6-item abbreviated version of the PSS-Fak (20 items in total) [ Self-report about family support, such as getting emotional support and having someone who can help out solving problems Previous research found a low reliability (PSS-Fa Cronbach α=.90; 6-item abbreviated PSS-Fa Cronbach α=.69) [ In RESIST, the abbreviated PSS-Fa had a good reliability (Cronbach α: o1=.88, o2=.83, o3=.85; coefficient Ω: o1=0.88, o2=0.83, o3=0.85) |
|
|
High extended family support |
13-item KSSl [ Self-report about extended family and kinship support, such as asking relatives for advice when making decisions or confiding in relatives when having a problem Previous research found an acceptable-to-good reliability (KSS Cronbach α=.72-.89) [ In RESIST, the KSS had an excellent reliability (Cronbach α: o1=.92, o2=.91, o3=.93; coefficient Ω: o1=0.92, o2=0.91, o3=0.93) |
|
|
High family cohesion |
5-item family cohesion subscale of the SFI-IIm (36 items in total) [ Self-report about family cohesion, such as preferably spending time with the family rather than with others Previous research found a low reliability (SFI-II Cronbach α=.91; 5-item family cohesion subscale Cronbach α=.60) [ In RESIST, the family cohesion subscale had a good reliability (Cronbach α: o1=.86, o2=.84, o3=.87; coefficient Ω: o1=0.87, o2=0.85, o3=0.88) |
|
|
High positive parenting |
6-item positive parenting subscale of the APQn (42 items in total) [ Child (ie, in our study, young adult) report about positive parenting, such as positive encouragement, compliments, and praise from parents for doing a good job (ie, for the time when the participants lived with their parents) [ Previous research found an acceptable reliability (6-item positive parenting subscale Cronbach α=.72-.75) [ In RESIST, the positive parenting subscale had a good reliability (Cronbach α: o1=.87, o2=.88, o3=.88; coefficient Ω: o1=0.87, o2=0.88, o3=0.88) |
|
|
High parental involvement |
10-item parental involvement subscale of the APQ (42 items in total) [ Child (ie, in our study, young adult) report about parental involvement levels, such as doing activities together and asking about the child’s friends and school performances (ie, for the time when the participants lived with their parents) [ We collapsed separate statements for “moms” and “dads” into a single “parent” statement (eg, original: “Your mom talks to you about your friends. How about your dad?,” adaptation: “Your parents talk to you about your friends.” as done in previous studies, such as in van Harmelen et al [ Previous research found an acceptable-to-good reliability (10-item parental involvement subscale Cronbach α=.71-.83) [ In RESIST, the parental involvement subscale had a good reliability (Cronbach α: o1=.87, o2=.89, o3=.87; coefficient Ω: o1=0.87, o2=0.90, o3=0.87) |
|
|
|
||
|
High friendship support |
6-item abbreviated version of the PSS-Fro (20 items in total) [ Self-report about friendship support, such as getting moral support and having companionship Previous research found an acceptable reliability (PSS-Fr Cronbach α=.88; 6-item abbreviated PSS-Fr Cronbach α=.75) [ In RESIST, the abbreviated PSS-Fr had a good reliability (Cronbach α: o1=.80, o2=.81, o3=.79; coefficient Ω: o1=0.80, o2=0.81, o3=0.79) |
aRF: resilience factor.
bDTS: Distress Tolerance Scale.
cRESIST: Resilience Study.
do1: occasion 1.
eo2: occasion 2.
fo3: occasion 3.
gRRS: Ruminative Response Scale.
hRSES: Rosenberg Self-Esteem Scale.
iERQ: Emotion Regulation Questionnaire.
jBAQ: Brief Aggression Questionnaire.
kPSS-Fa: Perceived Social Support from Family Scale.
lKSS: Kinship Social Support Measure.
mSFI-II: Self-Report Family Inventory Version II.
nAPQ: Alabama Parenting Questionnaire.
oPSS-Fr: Perceived Social Support from Friends Scale.
We assessed 5
We assessed 1
Environmental childhood and youth adversity was assessed using an updated version of the 12-item Youth Trauma Scale (YTS) [
Furthermore, we assessed the psychological maltreatment and neglect subscales of the Comprehensive Child Maltreatment Scale (CCMS) [
We assessed psychotherapeutic treatment and the use of psychopharmacological drugs for the period between occasions 1 and 2.
We assessed perceived stress (Cronbach α=.74; coefficient Ω=0.75) and global stress severity in the same way as described for occasion 1, while this time specifically focusing on the exam period. Moreover, we quantified the number of exams (completed and not yet completed). The mental distress (Cronbach α=.89; coefficient Ω=0.89) and RF levels (for reliability coefficients, see
Adversity was again assessed with the updated version of the YTS [
A total of 127 items were assessed.
We assessed psychotherapeutic treatment and the use of psychopharmacological drugs for the period between occasions 2 and 3.
We assessed perceived stress (Cronbach α=.74; coefficient Ω=0.75) and global stress severity in the same way as described for occasion 1. Moreover, we asked the students whether they had stressful or significant work during the last 4 weeks. Mental distress (Cronbach α=.90; coefficient Ω=0.91) and RF levels (for reliability coefficients, see
Adversity was again assessed with the updated version of the YTS [
As occasion 2 took place during an exam period, we assumed that it could potentially be the case that missingness may not be completely random but dependent on the students’ stress level during occasion 2. Therefore, we asked all participants how stressful the exam period had been.
A total of 129 items were assessed.
Before starting the content part of the web-based survey, participants were asked to read the information sheet, which contained the major study aims, and to complete a consent form. Before completion of the survey, participants were enabled to download their consent form and the information sheet. Moreover, we provided details on how to get help and support, in case the study would bring up difficult feelings or in case a participant would want to report childhood maltreatment or a crime, in a mental health services information sheet, which could be downloaded from the web-based survey. Further details regarding participant safety considerations are provided in Supplement III in
RESIST was approved by the Cambridge Psychology Research Ethics Committee (PRE.2017.096). RESIST was funded by JF’s Medical Research Council Doctoral Training Grant and by POW’s personal research account.
To include both participants with incomplete and complete data, we used a full information maximum likelihood (FIML) estimator. The use of this estimator has been shown to function well in longitudinal structural equation models. For example, Kievit et al [
We conducted a series of latent growth models (LGMs) to explore the mean change trajectory of perceived stress and mental distress over the 3 occasions. We fixed the slope loading of occasion 2 to 1, expecting this occasion to have the highest level of perceived stress (and mental distress), and the slope loading of occasion 3 to 0, expecting this occasion to have a lower level of perceived stress (and mental distress) than occasion 2. Hence, the slope loading of occasion 1 was freely estimated and provides an indication of where the (scaled) mean level lies in comparison to occasion 2 (fixed to 1) and occasion 3 (fixed to 0). We conducted the LGMs with invariant residual variances for the 3 occasions (M1). To test whether our latent growth model is significantly different from a
We conducted a series of bivariate latent change score models (BLCSMs; as described by Kievit et al [
All analyses were conducted in R version 3.5.1 (The R Foundation) [
Students were approximately uniformly distributed over all 6 academic years, with percentages ranging from 12.4% (56/451) to 20.6% (93/451) per year. A total of 57.4% (259/451) of the students were female (1.3% [6/451] preferred not to answer) and 58.3% (263/451) were White. Most students were between 18 and 23 years of age and had parents with higher education after secondary school. About 13.5% (61/451) of the students received psychotherapeutic treatment and 10.9% (49/451) received psychopharmaceutic treatment in the 6 months before occasion 1.
Demographic and clinical characteristics for the overall sample (N=451).
Characteristics | Sample size per answer category, n (%) | |||
|
||||
|
First year | 93 (20.6) | ||
|
Second year | 83 (18.4) | ||
|
Third year | 66 (14.6) | ||
|
Fourth year | 88 (19.5) | ||
|
Fifth year | 56 (12.4) | ||
|
Sixth year | 65 (14.4) | ||
|
||||
|
Female | 259 (57.4) | ||
|
Male | 185 (41.0) | ||
|
Prefer not to say | 6 (1.3) | ||
|
||||
|
18-20 | 170 (37.7) | ||
|
21-23 | 196 (43.5) | ||
|
24-26 | 65 (14.4) | ||
|
≥27 | 17 (3.8) | ||
|
||||
|
White | 263 (58.3) | ||
|
Non-White | 184 (40.8) | ||
|
||||
|
No | 390 (86.5) | ||
|
Yes | 61 (13.5) | ||
|
||||
|
No | 402 (89.1) | ||
|
Yes | 49 (10.9) | ||
|
||||
|
|
|||
|
|
Yes | 359 (79.6) | |
|
|
No | 88 (19.5) | |
|
|
Unknown | 4 (0.9) | |
|
|
|||
|
|
Yes | 369 (81.8) | |
|
|
No | 77 (17.1) | |
|
|
Unknown | 4 (0.9) |
aOne student did not answer this question. Due to missingness, some percentages may not add up.
bThree students did not answer this question. Due to missingness, some percentages may not add up.
cFour students did not answer this question. Due to missingness, some percentages may not add up.
The mean levels suggest that perceived stress and mental distress increased from the time before the exams to the exam period and decreased after the exams to a lower level than before the exams.
Perceived stress and mental distress levels for the 3 occasions.
Measure | Occasion 1 | Occasion 2 | Occasion 3 | ||||
|
Score, mean (SD) | Participants, n | Score, mean (SD) | Participants, n | Score, mean (SD) | Participants, n | |
PSSa | 10.42 (2.77) | 451 | 11.61 (2.77) | 274 | 9.89 (2.67) | 282 | |
GHQ-12b | 25.40 (5.82) | 445 | 27.39 (6.09) | 273 | 23.31 (5.93) | 282 |
aPSS: Perceived Stress Scale.
bGHQ-12: General Health Questionnaire, 12-item version.
Gender and ethnicity as well as psychopharmacological medication and global stress severity (ie, stress slider) on occasion 1 were identified as predictors for missing data patterns (see test results in Supplement VI in
The LGM showed that, on average, students experience most perceived stress during exams (occasion 2: slope loading fixed to 1) and least perceived stress after the exams (occasion 3: slope loading fixed to 0); before the exams, they experienced more perceived stress than after the exams, but less than during the exams (occasion 1: estimated slope loading=0.29;
The left panel depicts the perceived stress (sum score mean level) trajectory and the right panel depicts the mental distress (sum score mean level) trajectory. The faded gray lines indicate person-level trajectories. The red line indicates the group-level sum score trajectory, which was averaged across the students. The dotted black line represents the group-level sum score for occasion 1. This was done solely to enhance the comparison with the other occasions. o1: occasion 1; o2: occasion 2; o3: occasion 3.
Latent growth model summary.
Model | Slope |
Slope |
Slope |
Intercept mean | Slope mean | Residual variance, occasion 1 | Residual variance, occasion 2 | Residual variance, occasion 3 | Intercept slope covariance | ||||||||||
|
|||||||||||||||||||
|
M1a | 0.29 | 1.00 | 0 | 9.92 | 1.79 | 3.78 | 3.78 | 3.78 | 0.45 | |||||||||
|
M2b | 0.52 | 1.00 | 0 | 10.54 | 0 | 3.68 | 3.68 | 3.68 | −1.35 | |||||||||
|
M1c | 0.29 | 1.00 | 0 | 9.91 | 1.79 | 3.75 | 3.75 | 3.75 | 0.41 | |||||||||
|
|||||||||||||||||||
|
M1 | 0.60 | 1.00 | 0 | 23.21 | 4.08 | 16.85 | 16.85 | 16.85 | −8.71 | |||||||||
|
M2 | −4.41 | 1.00 | 0 | 25.44 | 0 | 37.50 | 37.50 | 37.50 | −1.65 | |||||||||
|
M2c | 0.67 | 1.00 | 0 | 25.26 | 0 | 16.74 | 16.74 | 16.74 | −18.07 |
aM1: freely estimated trajectory model.
bM2: no-change trajectory model.
cVariance for the latent slope constrained to >0 to render it nonnegative.
Latent growth model fit.
Model | AICa | BICb | CFIc | TLId | RMSEAe | SRMRf | Chi-square ( |
BICwg (%) | AICwh (%) | |
|
||||||||||
|
M1i | 4718.05 | 4746.83 | 0.98 | 0.96 | 0.07 | 0.04 | 6.5 (2) | 100 | 100 |
|
M2j | 4811.92 | 4836.59 | 0.47 | 0.47 | 0.27 | 0.18 | 102.3 (3) | 0 | 0 |
|
M1k | 4718.06 | 4746.84 | 0.98 | 0.96 | 0.07 | 0.04 | 6.5 (2) | 100 | 100 |
|
||||||||||
|
M1 | 6295.92 | 6324.66 | 0.94 | 0.91 | 0.09 | 0.05 | 9.2 (2) | 100 | 100 |
|
M2 | 6366.44 | 6391.08 | 0.32 | 0.32 | 0.24 | 0.19 | 81.7 (3) | 0 | 0 |
|
M2k | 6364.52 | 6389.17 | 0.34 | 0.34 | 0.24 | 0.18 | 79.8 (3) | 0 | 0 |
aAIC: Akaike information criterion.
bBIC: Bayesian information criterion.
cCFI: confirmatory fit index.
dTLI: Tucker-Lewis fit index.
eRMSEA: root mean square error of approximation.
fSRMR: standardized root mean square residual.
gBICw%: weight percentage for the Bayesian information criterion (compared to the respective other model); the higher the weight, the more in favor is the model.
hAICw%: weight percentage for the Akaike information criterion (compared to the respective other model); the higher the weight, the more in favor is the model.
iM1: freely estimated trajectory model.
jM2: no-change trajectory model.
kVariance for the latent slope constrained to >0 to render it nonnegative.
The LGM showed that, on average, students experience most mental distress during exams (occasion 2, slope loading fixed to 1) and least mental distress after exams (occasion 3, slope loading fixed to 0); before the exams, they experienced more mental distress than after the exams, but less than during the exams (occasion 1, estimated slope loading=0.60;
The BLCSM showed that perceived stress and mental distress on occasion 1 are significantly positively associated, which indicates that students with higher perceived stress report on average higher mental distress (
Bivariate Latent Change Score Models. The upper-left panel depicts occasions 1-2, the upper-right panel depicts occasions 1-3, and the lower-left panel depicts occasions 2-3. Green arrows represent positive associations, red arrows represent negative associations, black arrows represent fixed parameters, and blue arrows represent estimated intercepts and variances. Double-headed arrows represent covariances and variances, and single-headed arrows represent intercepts, regressions, and autoregressions. Solid lines indicate a significant association (
The BLCSM showed that changes in perceived stress and mental distress from occasion 1 to 3 are significantly positively associated, which indicates that individuals with a greater decrease in perceived stress report, on average, also a greater decrease in mental distress (
The BLCSM showed that perceived stress and mental distress on occasion 2 are significantly positively associated, which indicates that individuals with higher perceived stress report on average higher mental distress (
Both perceived stress and mental distress were lower before the exams (ie, during the regular university term) than during the exam period, but higher before the exams than after the exams. Higher mental distress during term time was, on average, associated with a greater increase in perceived stress from the term time to the exam period, when controlling for perceived stress levels during the term time. Hence, students who already had mental health problems before the exam period were most prone to develop increased levels of stress during the exam period. Higher perceived stress during the exam period was, on average, associated with less recovery of mental distress after the exam time, when controlling for mental distress levels during exams. Thus, students who reported high stress during the exam period were less successful (or quick) in recovering from mental distress. Overall, we found that higher mental health problems before the exams increase the risk of developing more perceived stress during the exams, and higher perceived stress during the exams in turn increases the risk of a less successful (or quick) recovery of mental distress after exams.
Future analyses on the RESIST data are primarily set out to shed light on which RFs lend themselves best as prevention targets (before the stressor) and which as treatment targets (at times of stress) for the mitigation of mental health problems that are triggered or accelerated by natural exam stress. Therefore, the RESIST study may lay the foundations necessary to inform student support services, as well as mental health services, as well as resilience and transdiagnostic mental health theory.
We aim to publish all articles that are based on RESIST data in peer-reviewed journals, ideally under an open access agreement. Alongside the manuscripts, we aim to release the related and anonymized data on the Cambridge Data Repository.
Contains 8 supplements (ie, Supplement I: Additional information for family-related questionnaires; Supplement II: Items of the adapted version of the Youth Trauma Scale; Supplement III: Participant safety considerations; Supplement IV: Latent Growth Models with varying residual variances; Supplement V: Demographic and clinical characteristics for the sample with data for at least 2 occasions; Supplement VI: Missingness predictors and analyses results when including the auxiliary variables; Supplement VII: Analyses results when excluding one potentially influential case; Supplement VIII: Model summaries and exact coefficients of the Bivariate Latent Change Score Models).
bivariate latent change score model
Comprehensive Child Maltreatment Scale
full information maximum likelihood
General Health Questionnaire
latent growth model
Perceived Stress Scale
Research Electronic Data Capture
Resilience Study
resilience factor
Youth Trauma Scale
This manuscript is based on the related ethics proposal coauthored by AVH and POW. The authors want to thank Pascal Schlechter (former MPhil student) for his help with the anonymization of the data and with the participant payments. They also want to thank all participants as well as the students who participated in the user review (ie, the pilot run). Moreover, the authors thank Dr Diana Wood and Dr Robbie Duschinsky, who additionally approved the ethics application as external scientists, and Dr Thelma Quince, who commented on an early sketch of the ethics application. The application was approved on December 20, 2017, by the Cambridge Psychology Research Ethics Committee (approval number: PRE.2017.096).
Data were collected and managed using REDCap [
JS received support from the National Institute for Health Research, Collaboration for Leadership in Applied Health Research and Care – East of England, at the Cambridgeshire and Peterborough NHS Foundation Trust. RAK was funded by Medical Research Council Programme Grant (grant SUAG/047 G101400) and a Radboudumc Hypatia Fellowship. AVH was funded by the Royal Society Dorothy Hodgkin Fellowship and the Leiden Social Resilience and Securtity Programme. JF was funded by the Medical Research Council Doctoral Training/Sackler Fund and the Pinsent Darwin Fund.
JS and RAK served as statistical advisers. AVH served as a topic-related adviser. POW obtained approval for the study from the Cambridge Medical School, helped recruit the participants, and served as principal investigator. AVH and POW approved the ethics protocol before submission. JF designed the study, wrote the ethics protocol, created the study materials (eg, the web-based survey), recruited the participants, conducted the analyses, and wrote the first draft of this manuscript. All authors commented on several previous versions and approved the final version of this manuscript. AVH and POW are the joint last authors of this manuscript.
JS disclosed consultation for Ieso Digital Health. None of the authors declared competing interests that affect the manuscript.