Int J Public Health (2016) 61:797–807

DOI 10.1007/s00038-016-0846-4

REVIEW

Setting-based interventions to promote mental health

at the university: a systematic review
A. Fernandez . E. Howse . M. Rubio-Valera . K. Thorncraft .
J. Noone . X. Luu . B. Veness . M. Leech . G. Llewellyn .
L. Salvador-Carulla

Received: 24 September 2015 / Revised: 12 October 2015 / Accepted: 9 June 2016 / Published online: 30 June 2016
 Swiss School of Public Health (SSPH+) 2016

Abstract evaluated structural and organizations strategies to promote

Objectives Universities are dynamic environments. But mental health at the university was selected.
university life presents challenges that may affect the Results 19 papers were included. The majority of the
mental health of its community. Higher education institu- studies were targeting the students, with only four aiming
tions provide opportunities to promote public health. Our to promote employees’ mental health. The most promising
objective is to summarise the current evidence on strategies strategies to promote mental wellbeing included changes in
to promote mental health at the university, following a the way students are taught and assessed. On the other
setting-based model. hand, social marketing strategies had not impact on mental
Methods We conducted a systematic literature review health. There is inconclusive evidence related to the
following standard methods. Published literature that effectiveness of policies to promote mental health.
Conclusions Universities should invest in creating sup-
portive physical, social and academic environments that
Electronic supplementary material The online version of this
promote student and staff mental wellbeing. However, the
article (doi:10.1007/s00038-016-0846-4) contains supplementary
material, which is available to authorized users.

A. Fernandez (&)
L. Salvador-Carulla M. Rubio-Valera

Mental Health Policy Unit-Brain and Mind Centre University of Spanish Network for Preventive Activities and Health Promotion
Sydney, Faculty of Health Sciences, University of Sydney, in Primary Care (redIAPP), Barcelona, Spain
Sydney, Australia
e-mail: [email protected] M. Rubio-Valera
School of Pharmacy, Universitat de Barcelona, Barcelona, Spain
L. Salvador-Carulla
e-mail: [email protected]
K. Thorncraft
University Library, University of Sydney, Sydney, Australia
E. Howse
M. Leech
G. Llewellyn
e-mail: [email protected]
Healthy Sydney University, The University of Sydney, Sydney,
Australia
J. Noone
E. Howse Ageing, Work and Health Research Unit, Faculty of Health
e-mail: [email protected] Sciences, University of Sydney, Sydney, Australia
e-mail: [email protected]
M. Leech
e-mail: [email protected]
X. Luu
G. Llewellyn Student Support Services, The University of Sydney, Sydney,
e-mail: [email protected] Australia
e-mail: [email protected]
M. Rubio-Valera
Department of Research and Teaching, Sant Joan de Deu X. Luu
Foundation, Barcelona, Spain Discipline of Behavioural and Social Sciences in Health, Faculty
e-mail: [email protected] of Health Sciences, The University of Sydney, Sydney, Australia

123
798 A. Fernandez et al.

current body of evidence is scarce and more research is environmental, economic, social, organizational and cul-
needed to recommend what are the best strategies. tural circumstances, and, consequently, the main aim is to
improve ‘‘the place’’ where the person lives, studies and/or
Keywords Mental health promotion
works. Setting-based approaches move from individual-
Setting-based approach
University based to population-based interventions and are more
concerned with changing the structural and organizational
factors that impact on health rather than the individual risk
Introduction factors (Dooris 2009). In this scenario, ‘health promoting
universities’ have become a major principle, which
Mental health is a resource for daily living, a state of encourages the development of strategies with a ‘‘focus on
wellbeing that allows individuals to cope with stress, think, population, policy and environment’’ (Tsouros et al. 1998).
feel, interact with others, work productively and enjoy life Consequently, setting-based approaches aim to integrate
(WHO 2004). Consequently, the promotion of mental health-related aspects into the routine life of the university
health is vital to enable individuals to flourish and con- and to create an environment that promotes health and
tribute to society. Good mental health results from productivity, as well as connectedness and global
individual skills as well as community assets (Kobau et al. wellbeing.
2011; WHO 2004). This paper presents the result of a systematic review of
Distinct yet inter-related populations of students and the published literature evaluating the impact of setting-
staff make universities a unique setting for the promotion based interventions that explicitly aimed to promote and
of mental health. First, university students are considered a improve the mental wellbeing of university students and
high-risk population due to their age, as most mental health employees.
problems have their onset before 24 years of age (Reavley
and Jorm 2010). Second, students are exposed to personal
and academic stressors that can negatively impact on their Methods
mental health (Cvetkovski et al. 2012; Hamaideh 2011;
O’Reilly et al. 2014; Tupler et al. 2015). Lastly, employees We conducted a systematic review following the Preferred
at higher education institutions are at greater risk for Reporting Items for Systematic Reviews and Meta-Anal-
mental health problems compared to workers in other yses (PRISMA) guidelines (Moher et al. 2009). The
industries (Kinman and Wray 2014). protocol of the systematic review was published in
Although several reviews have focused on the effec- PROSPERO, a database of prospectively registered sys-
tiveness of individual-based interventions to promote the tematic reviews in health and social care (CRD
mental wellbeing of students (Conley et al. 2013, 2015; 42014011839).
Davies et al. 2014; Reavley and Jorm 2010) and the
effectiveness of workplace interventions to promote mental Search strategy
wellbeing (Czabala et al. 2011), much less is known about
strategies conducted in the university, following a setting- PsycINFO, Medline (including Premedline), Cinahl, Web
based model. The setting-based model can be framed as a of Science, Scopus, ERIC, Embase, A? Education,
socio-ecological and salutogenic approach. It recognises APAFT via Informit and Business Source Premier datasets
that health is largely determined by people’s were used. To focus on the most up-to-date information,
the search was confined to the literature published during
the period of January 2004 to June 2015. The search
X. Luu strategies used are available as an online resource. In
St Andrew’s College, The University of Sydney, Sydney,
addition, the reference list of the selected studies and
Australia
systematic reviews related to the topic of interest was
B. Veness reviewed to retrieve studies not identified in the electronic
Royal Prince Alfred Hospital, Camperdown, New South Wales, search.
Australia
We structured the search strategy around these concepts:
e-mail: [email protected]
• Population/setting: university students and staff.
M. Leech
Sancta Sophia College, University of Sydney, Sydney, Australia • Intervention: any structural and/or organizational strat-
egy to promote mental health.
G. Llewellyn
L. Salvador-Carulla • Outcome: any improvement in mental health or/and
Centre for Disability Research and Policy, Faculty of Health mental wellbeing.
Sciences, University of Sydney, Sydney, Australia

123
Setting-based interventions to promote mental health at the university… 799

Selection of studies selected was assessed using a checklist taking into account
the experimental (Jadad et al. 1996) or observational
We included any quantitative study, regardless of their approach used (Mallen et al. 2006). Methodological quality
design, so long as they contained information on the impact was not used to exclude studies.
of the intervention. This included those using experimental External validity was also assessed using the criteria
(i.e. randomized controlled trial, both explanatory and developed by Green and Glasgow (2006), including reach
pragmatic), or observational (e.g. controlled trial without and representativeness, program or policy implementation
randomization; or pre–post and time series) approaches. and adaptation, outcomes for decision making, and main-
Duplicated papers were excluded. Studies were screened tenance and institutionalization.
for inclusion in two phases and the process was carried out
in duplicate by reviewers (AF read all the studies and BV, Data extraction and analysis
EH, JN, MR and XL made the duplicate). In Phase 1,
records were selected by reviewing the title and/or the Key features of the papers were extracted using a prede-
published abstract. In Phase 2, the full-text article was fined template. Key features included aim, intervention
reviewed. In case of any disagreement, researchers dis- components, study design, participants (i.e. students and/or
cussed the problems, trying to reach consensus. When staff) and sample size, the main outcome measures reported
consensus was not reached, a third co-author read the by the authors, and main results. Point estimates and con-
paper. fident intervals were extracted if available.
Articles were included if: Results are presented in a narrative way following the
different strategies that can be used from a setting-based
(a) The study setting was the university; studies under-
approach, if available (i.e. policies; built environment;
taken in colleges were included if the term ‘college’
social marketing; academic- and curriculum-based strate-
referred to a tertiary education institution or a
gies; and/or; and community coalitions).
residential college;
(b) The intervention followed a setting-based approach.
Setting-based interventions used the following meth-
Results
ods (Dooris 2009):
The electronic search strategy identified 21,427 records.
Policies; After removing records that were duplicated, 12,914 papers
Strategies to improve the built environment; were evaluated by two reviewers. A total of 12,490 papers
Social marketing; were excluded by reviewing title and abstract. Of the 424
Academic- and curriculum-based strategies; and/or full-text articles assessed for eligibility, 19 were finally
Community coalitions. included (Fig. 1).
Consequently, we excluded all interventions whose Four studies evaluated the impact of policies on mental
main aim was to directly modify individual health. Two assessed social marketing strategies and thir-
behaviours. teen were focused on academia-based strategies. The
majority of the 19 studies were targeting the students’
(c) It included any global measure of mental wellbeing,
wellbeing, with only four targeting the promotion of
mental health, wellness or mental health related
employees’ mental health.
quality of life. Studies using condition-specific out-
Online Supplementary Table 1 summarises the papers
come measures (such as depression or anxiety) were
included in this review.
also included as a proxy for mental health and
wellbeing;
Quality assessment
(d) It was written in English;
(e) It was published after the year 2004.
Internal validity
All the articles that met the criteria were included for
quality assessment and data extraction. Table 1 summarises the main aspects related to the internal
validity of the studies.
Quality assessment In terms of internal validity, the quality of the evidence
is very low. There is only one cluster randomised trial
The quality of each study was addressed taking into (published in two different papers), but its risk of bias is
account internal validity (risk of bias), as well as the high due to a high attrition rate of participants. Only two
external and ecological validity of the study. Each study studies incorporated a clear control group, while six had a

123
800 A. Fernandez et al.

Fig. 1 Flow diagram

Records identified through Additional records identified
database searching through other sources
(n = 21,477) (n =1)

Records after duplicates removed

(n = 12,914)

Records screened Records excluded

(n = 12,914) (n =12,490)

Full-text articles assessed for Full-text articles excluded, with

eligibility reasons (n=405)
(n =424)
145 - Type of paper (no outcome
evaluation)
124 – No population-based
intervention
97- No Mental Health Outcome
28- No primary prevention
4- Not conducted within a university
setting
Papers included as a primary studies in the qualitative 5- Not found
synthesis (n =19) 2- Language not English

historical cohort that was used as a control group. The risk of their disciplinary and professional expertise and interest.
of bias is very high; as a result, it cannot be inferred that the Finally, some of the papers missed relevant outcomes that
differences between the groups were due to the may be related to the process of decision making, such as
interventions. information about the costs of implementing the program, or
outcomes such as performance (when assessing academic-
External validity based strategies) or levels of absenteeism due to illness
(when evaluating occupational health and safety programs
The external validity of the studies can be considered low to for staff). Nonetheless, the reviewers do acknowledge that
moderate. The main strength of the body of evidence these data could have been presented in other papers that may
reviewed is the fact that some of them were real programs have been missed and not included in this review (Table 2).
that were conducted in real life settings, which indicate that
such strategies were institutionalised and maintained after Policies
the study concluded. However, the majority of the papers
presented the experience of one college or just one school, This section includes those setting-based approaches that
which limits the generalization of the intervention to other have used the implementation of a policy as a strategy to
groups or populations. Indeed, half of the papers were con- promote mental health. By policy, we mean an institutional
ducted in health-related disciplines (i.e. medicine, plan that defines procedures and guides action.
psychology or nursing), which may have affected the find- In 1984, the University of Illinois implemented a policy
ings, as these students tend to be more engaged in the that mandated any students who made a suicide threat, or
promotion of mental health and wellbeing due to the nature attempt, to receive four sessions of professional assessment

123
 Table 1 Assessment of the risk of internal bias
First author Type of study Random. Allocation Complete data Appropriate sampling Justification Control Control of confounding
(year) of group
measurement

Joffe (2008) Time series na na na Yes Yes Partially No

Grawitch et al. Cross-sectional na na na Not explained Yes No Partially
(2007)
Pignata and Cross-sectional na na na Survey sent to all the staff; only 14 % Yes No No
Winefield responded
(2013)
Pignata et al. Pre–post na na na Only 13 % of the original sample answered Yes No Partially. They adjust for
(2014) time 2 survey baseline psychological strain
Reavley et al. Cluster Yes Not 54.1 % were lost at na na na na
(2014a) randomised possible follow-up 9 (wave 3)
trial
Reavley et al. Cluster Yes Not 27.9 %were lost at na na na na
(2014b) randomised possible follow-up (wave 3)
trial
Becker et al. Academia-based na na na Not explained No No No
(2008) interventions
Shek (2012) Pre–Post na na na Not explained Yes No No
Setting-based interventions to promote mental health at the university…

Shek (2013) Pre–Post na na na Not explained No Yes Not known

Hassed et al. Non-randomized na na na 148 of an eligible cohort of 270 (61 %) Yes No No
(2009) controlled trial
Bergen-Cico Pre–post na na na Yes Yes Yes Partially
et al. (2013)
Foster et al. Non-randomized na na na Less than 10 % of the potential sample No Yes No
(2014) controlled trial completed both pre and post tests.
Bughi et al. Pre–post na na na 104 students, they select the last 32 for the Yes No No
(2006) pre–post analysis
Bloodgood et al. Non-randomized na na na All the students in the 2 cohorts. All accepted Yes Partially No
(2009) controlled trial
Rohe et al. Time series na na na All the students were invited to participate Yes Partially No
(2006) (98 % consented).
Reed et al. Time series na na na All students were invited, 58 % responded Yes Partially Partially
(2011)
Slavin et al. Cross-sectional na na na All students invited; 84 % response rate Yes Partially No
(2014)
Tucker et al. Time series na na na Only 34.5 % of total in class 2013 and 32.7 % Yes Partially No
(2015) of total in class 2014 participated
Jones and Time series na na na All students invited; response rate oscillated Yes Partially No
Johnston between 79 % and 83 %
(2006)
801

123
802 A. Fernandez et al.

Table 2 Assessment of the risk of external bias

First Author (year) Reach and representativeness Implementation Outcomes for decision Maintenance and
making institutionalisation

Joffe (2008) Only 1 university Yes Partially Yes

Grawitch et al. (2007) Only 1 university No No No
Pignata and Winefield (2013) Only 1 university No No No
Pignata et al. (2014) 13 universities No No No
Reavley et al. (2014a) 9 campuses within 1 university No Partially No
Reavley et al. (2014b) 9 campuses within 1 university No Partially No
Becker et al. (2008) Only 1 university Yes Partially Yes
Shek (2012) Only 1 university Partially No Partially
Shek (2013) Only 1 university Partially No Partially
Hassed et al. (2009) Only 1 school within 1 university (medical school) Yes Partially Yes
Bergen-Cico et al. (2013) 1 college within 1 university (psychology school) Yes Partially No
Foster et al. (2014) Only 1 university Partially No No
Bughi et al. (2006) 2 universities (medical schools) No Partially No
Rohe et al. (2006) 1 medical school Partially Partially Yes
Bloodgood et al. (2009) 1 medical school Yes Yes Yes
Reed et al. (2011) 12 US Medical School Campuses No Yes No
Slavin et al. (2014) 1 medical school in 1 university Yes Yes Yes
Tucker et al. (2015) 1 medical school in 1 university Yes Yes Yes
Jones and Johnston (2006) 1 nurse school in 1 university Yes Yes Yes

(Joffe 2008). Using a time series study, the rates of suicide employees with the opportunity to increase their knowledge
before and after the policy was implemented were com- and skills through opportunities for advanced training and
pared. These rates were also compared with the rates of continuing education); and (c) occupational health and safety
other universities. The rate of suicides before the policy programs (i.e. that focus on prevention, wellness programs,
(1976–1983) was 6.91 per 100,000 enrolled students. After health screenings and stress management). In comparison,
the policy was implemented (1984–2005) it was reduced to some policies and practices were associated with decreased
3.78 per 100,000 enrolled students. This represented a mental wellbeing; these included policies and practices that
reduction of 45.3 %. However, while the program of focused on recognition (i.e. practices and policies that provide
mandated assessment was associated with a 72.2 % monetary and nonmonetary rewards for employees) and
reduction in suicide amongst undergraduate students, there work-life balance for employees (e.g. flexible work arrange-
was an increase of 94.6 % among graduate and profes- ments). A case study conducted in an Australian university
sional students, casting doubt on the efficacy of the (Pignata and Winefield 2013) found no differences in psy-
intervention for different groups within the population. The chological health outcomes between employees who reported
overall decrease in the rates of suicide at this university that stress-reduction interventions were implemented in their
contrasted with an increasing rate of suicide at similar workplace and those who were not aware of these interven-
universities that did not have this policy. Although the tions. One year later, the same authors published a similar
absence of an appropriate control group and lack of study conducted in 13 universities (Pignata et al. 2014). In this
adjustment for confounds could bias the study findings, its second study, employees who reported an awareness of stress-
external validity is moderate. reduction interventions implemented at their university scored
Three papers were included that evaluated the impact of lower on psychological strain. However, these studies only
different workplaces’ policies and practices on employees’ examined employees’ perceptions of the employer’s healthy
mental health. A descriptive cross-sectional study conducted workplace practices, and did not evaluate the real existence or
in the USA (Grawitch et al. 2007) examined the relationship quality of specific programs.
between satisfaction with different workplace practices and
employee outcomes. The workplace practices that were cor- Social marketing
related with an improvement in mental health included:
(a) Employee involvement policies; (b) growth and develop- Social marketing is defined as ‘the application of com-
ment practices (i.e. programs and policies that provide mercial marketing technologies to the analysis, planning,

123
Setting-based interventions to promote mental health at the university… 803

execution, and evaluation of programs designed to influ- self-understanding, resilience, interpersonal relationships
ence the voluntary behaviour of target audiences to and self-leadership. Pilot evaluations of the program (a one
improve their personal welfare and that of society’(Dono- group pre–post study and a non-randomized controlled
van 2011). This review only included one study, published trial) suggested that the course was effective in increasing
across two separate papers, which met the inclusion connectedness, hope and general positive youth develop-
criteria. ment qualities. However, detailed information is not
The intervention (MindWise) was designed to be whole- provided in the report. Consequently, these results should
of-campus and to be run over two academic years. The be taken with caution.
researchers randomized the campuses to a control or The ‘Health Enhancement Program (HEP)’is a manda-
intervention condition. The intervention included key tory course delivered to all first-year medical students at
messages about mental health that were delivered through Monash University, Australia (Hassed et al. 2009). It
social media (such as Facebook and Twitter), e-mails, consisted of eight lectures providing information on: the
posters, campus events, factsheets/booklets and mental link between physical and mental health, mind–body
health first aid training courses. The student results medicine, behaviour change strategies, mindfulness-based
(Reavley et al. 2014b) indicated that there was no impact therapies, and the importance of healthy behaviours. The
on mental wellbeing outcomes. Similarly, the results of the lectures were supported by six 2 h tutorials with 15–16
staff groups (Reavley et al. 2014a) showed no benefit in participants. Compared with pre-test measures, post-test
terms of improved mental health outcomes, but did show results were significantly better for outcomes related to
an increase in knowledge and recognition of depression psychological quality of life, depression symptomatology,
and risky alcohol consumption. In spite of using an hostility and global mental health. Although, the absence of
experimental design, the risk of bias of this study is high a control group increases the risk of bias, the external
due to the contamination between the intervention and the validity of the study is moderate to high.
control group (i.e. students can attend different campuses Using a non-randomized controlled trial, Bergen-Cico
during the same year) and by the high level of attrition. Its et al. (2013) assessed the effectiveness of a brief mind-
external validity is low to moderate as the strategy was not fulness-based stress reduction elective course for
maintained long term. psychology students in a US university. They found that
after 6 weeks, students in the intervention group increased
Academic-based strategies their scores for mindfulness and self-compassion, but there
was no effect on anxiety levels. The risk of bias of this
This section focuses on the interventions that aimed to study is moderate because of the potential confounders.
improve the mental wellbeing of university students The external validity was determined to be low to
through academic-based strategies, including mandatory moderate.
courses on topics related to mental wellbeing, different
assessment strategies, and curriculum design. Curriculum infusion

Mandatory courses included in curricula Curriculum infusion can be defined as ‘a pedagogy that
involves integrating health issues into academic courses
Becker et al. (2008) conducted a one group pre–post study with the aim of changing attitudes and behaviours’ (Riley
to evaluate the impact that a mandatory course ‘Health in and McWilliams 2012). We found two papers that can be
Modern Society’ had on the mental wellbeing of enrolled classified under this category.
students. It was a two-credit general education course that Bughi et al. (2006) evaluated the effect that a psycho-
used active learning techniques; such as debates/round-ta- educational lecture on stress management had on the
bles, problem-based learning, and simulated-scenarios/case mental wellbeing of third and fourth year medical students.
studies; and covered areas related to lifestyle, mental This was infused into a 1-month rotation in Diabetes/En-
health, and sexual health. Post-test measures compared docrine service at a tertiary referral centre and a one group
with pre-test measures showed a statistically significant pre–post study approach was used to evaluate its impact.
increase in mental health knowledge and a non-statistically After 1 month, the prevalence of reported stress decreased
increase in mental health wellness behaviours. The absence by 46.7 %. However, the absence of a control group and
of a control group also affects the validity of the findings. other information related to its external validity precludes
Similarly, Shek (2012, 2013) evaluated the impact of a any strong conclusion.
mandatory leadership course on students’ wellbeing from More recently, Foster et al. (2014) tested the effective-
the University of Hong Kong. The course focused on ness of infusing one brief intervention in a general first-
positive youth development and addressed areas related to year mandatory course with a non-randomized controlled

123
804 A. Fernandez et al.

study. It consisted of a manual that was offered to the and mental health problems of their students. A variety of
students during the first tutorial of a unit of study and changes were, therefore, explored. Firstly, the grading
explained how to cope with daily stressors. Some of the system for preclinical courses was changed from a 5-in-
students, in addition, received a daily SMS that acted as a terval system to a 2-interval pass/fail system. Secondly,
prompt to help students cultivate positive emotions. There contact hours during the first 2 years of the curriculum
was no impact on the level of wellbeing for students in the were reduced by nearly 10 %. Thirdly, the School intro-
control group and in the manual-only intervention group. duced more longitudinal electives. Fourth, learning
However, the students who received the manual and the communities were established that were composed of
SMS demonstrated improved wellbeing, suggesting that students and faculty staff who shared common interests.
the daily SMS was a key factor related to the success of the Finally, a required course based on mindfulness and
intervention. The fact that less than 10 % of the partici- resilience was included in the curriculum. The researchers
pants concluded the study and the absence of found lower levels of depression symptoms, anxiety
randomizations together with its low external validity are symptoms and reported stress in those students exposed to
major limitations of this study. the changed curriculum in comparison with the students in
older cohorts who were not exposed to changes. However,
Assessment strategies another study found contrary results (Tucker et al. 2015).
The authors also looked at a traditional medical course vs
Rohe et al. (2006) compared two medical school classes: a new curriculum in terms of medical students’ wellbeing,
the class of 2005 (the last class under a 5-interval grading stress and performance. The new curriculum included a
system), and the class of 2006 (the first class under a change in the traditional structure of co-occurring disci-
2-interval grading system of pass/fail). Both classes took pline-based courses, which was substituted with ten
the same array of courses. However, students covered by successive interdisciplinary courses followed by a cap-
the 2-interval grading system reported lower levels of stone course to further integrate basic sciences and
perceived stress. Bloodgood et al. (2009) found similar clinical information with clinical correlations. Learning
results. They followed students during four semesters and activities based on problems and small-group sessions
compared those in the 5-interval grading system, with were increased. They also changed the traditional five-
others in a 2-interval grading system. Students that interval grading system, replacing it with an honour-
changed to a pass/fail grading system had higher levels of s/pass–fail system. When comparing the two classes,
wellbeing for the three first semesters evaluated, although students in the new curriculum class scored higher than
not for the fourth. Results also suggested that the change students in the traditional class in depression scores and
was not associated with lower levels of academic per- perceived stress, and lower in quality of life. There was no
formance in those subjects. The internal validity of these difference in the students’ average exam scores, sug-
studies is compromised by the attempted comparison of gesting that there was no impact on academic
two cohorts in different temporal moments. However, performance.
their external validity is moderate to high, as these are Jones and Johnston (2006) explored the impact of a
strategies currently implemented that have also taken into student-centred, problem-based curriculum vs traditional
account other possible outcomes (i.e. academic curriculum on student wellbeing and performance, using a
performance). sample of nursing education students in Scotland and a
Reed et al. (2011) conducted a survey in seven times-series design. At week 25, students in the new
medical schools in the USA. The study found that stu- course reported fewer academic concerns, clinical con-
dents in medical schools using grading scales with three cerns and personal problems. They also reported lower
or more categories had higher levels of stress, burnout, levels of distress. However, these same students had a
emotional exhaustion and depersonalization. The cross- greater level of absence due to sickness. Similar patterns
sectional design of the survey, however, limits its were found at week 50 except for academic worries: the
conclusions. difference observed previously had disappeared. In terms
of academic performance, those in the new course per-
Changes in the curriculum formed less well on a comparative essay, but better in the
exam.
Broad curriculum changes were explored by the Saint The results of these studies are threatened by the pos-
Louis University School of Medicine (USA) (Slavin et al. sibility of selection bias, as the cohorts are from different
2014). In 2009 and 2010, the School introduced a series of years and may be exposed to different historical and con-
curriculum changes that aimed to alter the structural textual events. However, these three studies show moderate
conditions of the curriculum that contribute to the stress to high external validity.

123
Setting-based interventions to promote mental health at the university… 805

Discussion model, which recognises that ‘healthy structures’ (e.g. how

we design the curricula) is a pre-condition for ‘healthy
This review indicates that most setting-based mental health processes’ (e.g. how we communicate with the students)
promotion strategies at the university are based on modi- and that both structures and processes are also pre-condi-
fying the way students are taught and assessed. The risk of tions for healthy outcomes (Dooris et al. 2014).
bias of the studies reviewed is high, although some have The integration of mindfulness-based stress-reduction
acceptable external validity. There is a lack of information programs in university curricula is a major trend in health-
with regard to interventions directed at employees. The related faculties (Dobkin and Hutchinson 2013). However, as
results suggest that policies that successfully promote the identified earlier, the majority of the studies assessing cur-
mental wellbeing of university staff include the involve- riculum-related strategies to promote mental health were done
ment of employees in decision making and provide in health-related disciplines. Students from these disciplines
opportunities to increase employees’ knowledge of mental may be more willing to participate in this type of course.
health and wellbeing. Unfortunately, the number of studies Future research should focus on the impact of embedding
found was low and the quality of the evidence was poor, mindfulness or life skills courses in other disciplines.
prohibiting any strong recommendation. Another academic strategy that may be promising is
The results of this review should be considered with the changing the grading system of university courses.
following limitations in mind. Firstly, the search strategy Although evidence suggests that this strategy may improve
only included scientific literature. Consequently, we missed the mental health of students, concerns have been raised
potentially useful grey literature. This may be important for with regard to the effect that this strategy can have on
interventions which are based on policies, as these can be academic performance. A recent study in the medical
difficult to evaluate in terms of effectiveness and impact. We context suggested that the new system was associated with
recommend complementing this report with a review of the decreased performance on the preclinical examination,
grey literature in this area. Secondly, due to the large amount although no decrease in performance was apparent on other
of data retrieved and the relatively short timeframe, a general exams (McDuff et al. 2014).
assessment of the studies and literature was performed, New pedagogical methods, such as ‘Flipped’ teaching,
rather than a systematic analysis of the quality of the studies. problem-solving and student-centred teaching, may have
In spite of the above limitations, this is a comprehensive the potential to promote the mental health of students.
review of the effectiveness of setting-based interventions to However, the current body of evidence is scarce, contra-
promote mental health and wellbeing of university staff dictory, and ultimately inconclusive. More research is
and students. The process was carried out in duplicate and needed to clearly comprehend the consequences of these
followed the PRISMA guidelines. A sensitive strategy was educational methods.
adopted with no restrictive filters. This strategy led us to This review found little data on evaluating policies to
review an extensive number of papers that were manually promote mental health and wellbeing within universities.
excluded. However, we are confident that we have included This may be related to the difficulty of evaluating policies
all the information relevant for our aim. and defining mental health outcomes. In the area of mental
Mental illnesses are one of the leading causes of disease health promotion, most of the policies are related to the
burden worldwide, in low, middle and high income coun- prevention of suicide incorporating ‘restriction of means
tries (Whiteford et al. 2013). It is urgent to establish approaches’ (Silverman 2008; Stratford 2012; Washburn
strategies to prevent mental disorders and promote mental and Mandrusiak 2010). These approaches include regula-
health. Key settings such as primary care, workplace or the tions limiting access to building roofs; protocols, such as
university are ideal places to do this, as they can reach a restricting the size of window openings; and norms
large population and are well positioned to promote public restricting student access to laboratories with dangerous or
health approaches, adopting a system perspective that takes harmful chemicals. However, there are no data on the
into account that the place where people live matters. impact that these strategies have had.
The fact that the majority of the interventions utilise In recent years, there has been an increased interest in
academic-based strategies makes sense, as the review evaluating the impact that the built environment may have
focused on a specific type of setting and population—the on mental health (Evans 2003). Unfortunately, no research
university. Universities are unique places of learning where evaluating the impact of the built environment on the
the setting has a strong focus on teaching, learning and mental health of the university community was identified.
research to support students to develop, thrive and gradu- Nonetheless, a series of recommendations have emerged
ate. It is these areas in which interventions can impact the from the review of the literature that deserve further
whole student population. It also mirrors the principles of research: the creation of stress reduction spaces (Klainberg
the setting-based approach, framed into a socio-ecological et al. 2010); the design of ‘‘healing gardens’’ (Lau and

123
806 A. Fernandez et al.

Yang 2009); and the explicit inclusion of windows in new from two Southern California universities. Med Educ Online
buildings, or, ‘‘digital windows’’ in old buildings, which 11:1–8
Conley CS, Durlak JA, Dickson DA (2013) An evaluative review of
are large plasma displays showing real-time outside scenes, outcome research on universal mental health promotion and
to increase sense of belonging, connectedness and mental prevention programs for higher education students. J Am Coll
restoration (Friedman et al. 2008). Health 61(5):286–301
Finally, the lack of studies targeting staff is surprising, Conley CS, Durlak JA, Kirsch AC (2015) A meta-analysis of
universal mental health prevention programs for higher educa-
particularly given the growing policy interest in mental tion students. Prev Sci 16(4):487–507. doi:10.1007/s11121-015-
health at the workplace (Mykletun and Harvey 2012). The 0543-1
lack of evidence related to environmental strategies is not Cvetkovski S, Reavley NJ, Jorm AF (2012) The prevalence and
specific of the university sector, but a constant in the correlates of psychological distress in Australian tertiary students
compared to their community peers. Aust N Z J Psychiatry
promotion of mental health at the workplace (Kahn-Mar- 46(5):457–467
shall and Gallant 2012). This is possibly due to the Czabala C, Charzynska K, Mroziak B (2011) Psychosocial interven-
difficulties associated with their implementation (e.g. tions in workplace mental health promotion: an overview. Health
organisational commitment to change) and evaluation. Promot Int 26:I70–I84. doi:10.1093/heapro/dar050
Davies EB, Morriss R, Glazebrook C (2014) Computer-delivered and
web-based interventions to improve depression, anxiety, and
Conclusion psychological well-being of university students: a systematic
review and meta-analysis. J Med Internet Res 16(5):e130.
To encourage the promotion of mental health at the uni- doi:10.2196/jmir.3142
Dobkin PL, Hutchinson TA (2013) Teaching mindfulness in medical
versity, classical individual strategies that build personal school: where are we now and where are we going? Med Educ
skills should be combined with a setting-based approach 47(8):768–779. doi:10.1111/medu.12200
that impacts on and improves the overall setting of the Donovan RJ (2011) The role for marketing in public health change
university, which includes the physical, occupational, programs. Austr Rev Public Aff 10(1):23–40
Dooris M (2009) Holistic and sustainable health improvement: the
organisational, cultural and academic environments. This contribution of the settings-based approach to health promotion.
review has shown that academic-based interventions Perspect Public Health 129(1):29–36
enhancing the way students learn and academics teach are Dooris M, Wills J, Newton J (2014) Theorizing healthy settings: a
promising. However, the current body of evidence is scarce critical discussion with reference to Healthy Universities. Scand
J Public Health. 42(15 Suppl):7–16. doi:10.1177/
and more (high-quality) research is needed to recommend 1403494814544495
what are the best strategies. Evans GW (2003) The built environment and mental health. J Urban
Health Bull N Y Acad Med 80(4):536–555. doi:10.1093/jurban/
Acknowledgments The authors are grateful to the member of the jtg063
Mental Wellbeing Group at Healthy Sydney University, who over- Foster J, Allen W, Oprescu F, McAllister M (2014) Mytern: an
sighted this review during all the process, providing useful comments innovative approach to increase students’ achievement, sense of
and inputs. wellbeing and levels of resilience. JANZSSA 43:31–40
Friedman B, Freier NG, Kahn PH, Lin P, Sodeman R (2008) Office
Compliance with ethical standards window of the future?—field-based analyses of a new use of a
large display. Int J Hum-Comput Stud 66(6):452–465. doi:10.
Ethical standards This paper presents the results of a systematic 1016/j.ijhcs.2007.12.005
review of published studies. No ethical approval was required. Grawitch MJ, Trares S, Kohler JM (2007) Healthy workplace
practices and employee outcomes. Int J Stress Manage
Conflict of interest The authors declare that they have no conflict of 14(3):275–293. doi:10.1037/1072-5245.14.3.275
interest. Green LW, Glasgow RE (2006) Evaluating the relevance, general-
ization, and applicability of research: issues in external
validation and translation methodology. Eval Health Prof
29(1):126–153. doi:10.1177/0163278705284445
References Hamaideh SH (2011) Stressors and reactions to stressors among
university students. Int J Soc Psychiatry 57(1):69–80
Becker CM et al (2008) Making health happen on campus: a review Hassed C, de Lisle S, Sullivan G, Pier C (2009) Enhancing the health
of a required general education health course. J Gen Educ of medical students: outcomes of an integrated mindfulness and
57(2):67–74 lifestyle program. Adv Health Sci Educ Theory Pract
Bergen-Cico D, Possemato K, Cheon S (2013) Examining the efficacy 14(3):387–398
of a brief mindfulness-based stress reduction (brief MBSR) Jadad AR et al (1996) Assessing the quality of reports of randomized
program on psychological health. J Am Coll Health clinical trials: is blinding necessary? Control Clin Trials
61(6):348–360. doi:10.1080/07448481.2013.813853 17(1):1–12
Bloodgood RA, Short JG, Jackson JM, Martindale JR (2009) A Joffe P (2008) An empirically supported program to prevent suicide
change to pass/fail grading in the first two years at one medical in a college student population. Suicide Life Threat Behav
school results in improved psychological well-being. Acad Med 38(1):87–103
84(5):655–662. doi:10.1097/ACM.0b013e31819f6d78 Jones MC, Johnston DW (2006) Is the introduction of a student-
Bughi SA, Sumcad J, Bughu S (2006) Effect of brief behavioral centred, problem-based curriculum associated with improve-
intervention program in managing stress in medical students ments in student nurse well-being and performance? An

123
Setting-based interventions to promote mental health at the university… 807

observational study of effect. Int J Nurs Stud 43(8):941–952. Reed DA et al (2011) Relationship of pass/fail grading and
doi:10.1016/j.ijnurstu.2005.10.013 curriculum structure with well-being among preclinical medical
Kahn-Marshall JL, Gallant MP (2012) Making healthy behaviors the students: a multi-institutional study. Acad Med
easy choice for employees: a review of the literature on 86(11):1367–1373. doi:10.1097/Acm.0b013e3182305d81
environmental and policy changes in worksite health promotion. Riley JB, McWilliams M (2012) Curriculum infusion: educating the
Health Educ Behav 39(6):752–776. doi:10.1177/ whole student and creating campus change-Georgetown Univer-
1090198111434153 sity (Washington DC) transforming undergraduate education:
Kinman G, Wray S (2014) Taking its toll: rising stress levels in theory that compels and practices that succeed. Rowman &
further education. In: UCU Stress Survey. http://www.ucu.org. Littlefield, Lanham, pp 319–323
uk/media/7264/UCU-stress-survey-2014/pdf/ucu_ Rohe DE, Barrier PA, Clark MM, Cook DA, Vickers KS, Decker PA
festressreport14.pdf. Accessed 26 June 2016 (2006) The benefits of pass-fail grading on stress, mood, and
Klainberg M, Ewing B, Ryan M (2010) Reducing stress on a college group cohesion in medical students. Mayo Clin Proc
campus. J N Y State Nurs Assoc 41(2):4–7 (quiz 18–20) 81(11):1443–1448
Kobau R et al (2011) Mental health promotion in public health: Shek DT (2012) Development of a positive youth development
perspectives and strategies from positive psychology. Am J subject in a university context in Hong Kong. Int J Disabil Hum
Public Health 101(8):e1–e9. doi:10.2105/AJPH.2010.300083 Dev 1(3):173–179
Lau S, Yang F (2009) Introducing healing gardens into a compact Shek DT (2013) Promotion of holistic development in university
university campus: design natural space to create healthy and students: a credit-bearing course on leadership and intrapersonal
sustainable campuses. Landsc Res 34(1):55–81. doi:10.1080/ development. Best Pract Ment Health Int J 9(1):48–61
01426390801981720 Silverman MM (2008) Campus Security Begins with Caring. Chron
Mallen C, Peat G, Croft P (2006) Quality assessment of observational High Educ 54(32):A51–A52
studies is not commonplace in systematic reviews. J Clin Epidem Slavin SJ, Schindler DL, Chibnall JT (2014) Medical student mental
59(8):765–769. doi:10.1016/j.jclinepi.2005.12.010 health 3.0: improving student wellness through curricular
McDuff SGR, McDuff D, Farace JA, Kelly CJ, Savoia MC, Mandel J changes. Acad Med 89(4):573–577
(2014) Evaluating a grading change at UCSD school of Stratford M (2012) Preventing suicide on campus may mean fences
medicine: pass/fail grading is associated with decreased perfor- and nets as well as counseling. Chron High Educ. http://
mance on preclinical exams but unchanged performance on chronicle.com/article/Preventing-Suicide-May-Mean/131291/.
USMLE step 1 scores. BMC Med Educ 14:127. doi:10.1186/ Accessed 26 June 2016
1472-6920-14-127 Tsouros AD, Dowding D, Thompson J, Dooris M (1998) Health
Moher D, Liberati A, Tetzlaff J, Altman DG, Group, P (2009) promoting universities: concept, experience and framework.
Preferred reporting items for systematic reviews and meta- World Health Organization, Switzerland. http://www.euro.who.
analyses: the PRISMA statement. J Clin Epidemiol int/__data/assets/pdf_file/0012/101640/E60163.pdf. Accessed 26
62:1006–1012 June 2016
Mykletun A, Harvey SB (2012) Prevention of mental disorders: a new Tucker P, Jeon-Slaughter H, Sener U, Arvidson M, Khalafian A
era for workplace mental health. Occup Environ Med (2015) Do medical student stress, health, or quality of life
69(12):868–869. doi:10.1136/oemed-2012-100846 foretell step 1 scores? A comparison of students in traditional
O’Reilly E, McNeill KG, Mavor KI, Anderson K (2014) Looking and revised preclinical curricula. Teach Learn Med 27(1):63–70.
beyond personal stressors: an examination of how academic doi:10.1080/10401334.2014.979178
stressors contribute to depression in Australian graduate medical Tupler LA, Hong JY, Gibori R, Blitchington TF, Krishnan KR (2015)
students. Teach Learn Med 26(1):56–63 Suicidal ideation and sex differences in relation to 18 major
Pignata S, Winefield AH (2013) Stress-reduction interventions in an psychiatric disorders in college and university students: anony-
Australian university: a case study. Stress and Health mous web-based assessment. J Nerv Ment Dis 203(4):269–278.
31(1):24–34. 10.1002/smi.2517 doi:10.1097/NMD.0000000000000277
Pignata S, Boyd C, Gillespie N, Provis C, Winefield AH (2014) Washburn CA, Mandrusiak M (2010) Campus suicide prevention and
Awareness of Stress-reduction interventions: the impact on intervention: putting best practice policy into action. Can J High
employees’ well-being and organizational attitudes. Stress Educ 40(1):101–119
Health. doi:10.1002/smi.2597 Whiteford HA, Degenhardt L, Rehm J, Baxter AJ, Ferrari AJ, Erskine
Reavley N, Jorm AF (2010) Prevention and early intervention to HE, Charlson F, Norman RE, Flaxman AD et al (2013) Global
improve mental health in higher education students: a review. burden of disease attributable to mental and substance use
Early Interv Psychiatry 4(2):132–142. doi:10.1111/j.1751-7893. disorders: findings from the Global Burden of Disease Study
2010.00167.x 2010. Lancet 382:1575–1586
Reavley NJ, McCann TV, Cvetkovski S, Jorm AF (2014a) A WHO (2004) Promoting mental Health: concepts, emerging evidence,
multifaceted intervention to improve mental health literacy in practice. Summary report. A report of the World health
employees of a multi-campus university: a cluster randomised Organization Department of Mental health and Substance Abuse
trial. J Public Health 13(1):25–39 in Collaboration with the Victorian Health Promotion Founda-
Reavley NJ, McCann TV, Cvetkovski S, Jorm AF (2014b) A tion and the University of Melbourne. World Health
multifaceted intervention to improve mental health literacy in organization, Geneva
students of a multicampus university: a cluster randomised trial.
Soc Psychiatry Psychiatr Epidemiol 49(10):1655–1666

123
Reproduced with permission of the copyright owner. Further reproduction prohibited without
permission.