The state of the art in European research on reducing social exclusion and stigma related to mental health: A systematic mapping of the literature

Abstract Stigma and social exclusion related to mental health are of substantial public health importance for Europe. As part of ROAMER (ROAdmap for MEntal health Research in Europe), we used systematic mapping techniques to describe the current state of research on stigma and social exclusion across Europe. Findings demonstrate growing interest in this field between 2007 and 2012. Most studies were descriptive (60%), focused on adults of working age (60%) and were performed in Northwest Europe—primarily in the UK (32%), Finland (8%), Sweden (8%) and Germany (7%). In terms of mental health characteristics, the largest proportion of studies investigated general mental health (20%), common mental disorders (16%), schizophrenia (16%) or depression (14%). There is a paucity of research looking at mechanisms to reduce stigma and promote social inclusion, or at factors that might promote resilience or protect against stigma/social exclusion across the life course. Evidence is also limited in relation to evaluations of interventions. Increasing incentives for cross-country research collaborations, especially with new EU Member States and collaboration across European professional organizations and disciplines, could improve understanding of the range of underpinning social and cultural factors which promote inclusion or contribute toward lower levels of stigma, especially during times of hardship.


Introduction
Stigma and social exclusion related to mental health are of substantial public health importance in Europe. Experience of mental health problems early in life can be associated with a trajectory of exclusion and disadvantage, for example, through reduced participation in higher education, exclusion from civil society (including functions such as democratic participation), increased risk of contact with criminal justice systems, victimization, less access to physical healthcare, poverty and homelessness and reduced life expectancy [16,21,26,35,42,47]. Stigma and exclusion of these kinds can directly reduce well-being and can also have significant consequences in terms of lower participation in healthcare, higher rates of mortality, higher levels of self-stigma, lower levels of empowerment and higher rates of unemployment [14,40,43]. Employment rates for people with mental health problems, for instance, are lower than those of the overall population and lower than those for people with physical health problems [25,38].
These issues may intensify at times of macro-economic adversity, as currently experienced in many European and other countries. Stigma and social exclusion of marginalized groups may increase during times of economic recession and can result in increased exclusion of people with mental health problems [10], for example in relation to employment [1,15]. Although the evidence is from outside Europe, other research has shown how relative incomerelated disadvantages, in relation to prevalence of mental health problems and to suicide, widen during a recession [24]. Given that recession and a poor economic outlook may be associated with a decline in population mental health and depletion of social resources associated with resilience [19,34,46], there is likely to be an increasing proportion of the population feeling the impact of the economic hardship. Thus, there is an urgent need for evidence to help us understand the mechanisms of and evidence-based solutions for decreasing stigma and social exclusion and to improving societal participation.
The importance of stigma and social exclusion was highlighted in a recent survey of 154 national associations/organizations of stakeholder groups in Europe, undertaken as part of the ROAMER (ROAdmap for MEntal health Research in Europe) project. Stigma and discrimination were considered priorities across all stakeholder groups (i.e., trainee mental health professionals, psychiatrists, other mental health professionals, service users and/or carers and trainees) [17]. The ROAMER project, which was funded under the European Commission Seventh Framework Research Programme (FP7), aims to develop a comprehensive and integrated mental health research agenda oriented to translational research and aligned with the policies of the new EU long-term research programme Horizon 2020 [23]. One part of the ROAMER project focused on social and economic issues in relation to mental health and well-being across Europe. The research underpinning this paper was designed to inform the ROAMER mental health roadmap by describing the most recent, state of the art research in relation to stigma and social exclusion. Understanding what research achievements have already been made would also help to identify existing research gaps, to understand weaknesses in the field, and hence, to ascertain which research needs might be prioritized for the future.
The overarching objective of this study was to map current research on stigma and social exclusion in relation to mental health and well-being across Europe. We had three specific aims: to identify where in Europe (i.e., in which countries) research on stigma and social exclusion in relation to mental health and wellbeing is conducted; to describe the type of research on stigma and social exclusion in relation to mental health in Europe (e.g., methodology, type of disorder, age groups); to investigate the ''main research'' topics in the areas of stigma and social exclusion in relation to mental health in Europe (e.g., focus of research).

Materials and methods
In order to provide a broad overview of the scientific literature we employed systematic literature mapping techniques as outlined by published guidelines [7,13]. We used a broad set of search terms for: mental illness and well-being (in the title or abstract); referring to social exclusion/inclusion and stigma (in the title or abstract) (see Appendix 1 for search terms used), and searched Pubmed and Psychinfo databases.
We focused on terms related to stigma, social exclusion and social inclusion to increase representation of research across different disciplines that may employ different terminology. The search terms were based on a bibliometric analysis used to assess mental health research funding in 2007 [6] and an earlier systematic mapping study of Social Exclusion and Mental Health Policy which covered the years 1948-2003 [12], both of which were supplemented with additional terms as recommended by scientific experts associated with the ROAMER project. To facilitate broad multidisciplinary thinking, the literature mapping strategy was presented to ROAMER scientific experts associated with the sub-project/work package devoted to social and economic issues as well as to all other ROAMER work packages, the scientific advisory board and the stakeholder advisory group. Recommendations from discussions at these meetings were incorporated into our mapping strategy.

Inclusion/exclusion criteria
As we aimed to identify the most recent ''state of the art'' research, studies had to be published in peer-reviewed journals between January 2007 and December 2012. This timeframe was established through consensus by the ROAMER consortium. The paper had to be an empirical paper and include an English language abstract. Review papers were excluded. The research must have been performed in a European Union (EU) country (including EU-28 countries and EU Candidate countries, i.e., Iceland, Montenegro, Serbia, The former Yugoslav Republic of Macedonia and Turkey) or the first author or corresponding author must have been based at a research institution in an EU country.
In regards to study outcomes, we reviewed papers using a twostage process. First, we screened abstracts to be considered for full review if the main focus was related to mental health problems, well-being, stigma or social exclusion/inclusion. Articles which simply performed an evaluation of psychological or pharmacological treatments were excluded unless the treatments included a social component or measured an outcome which was, for example, related to stigma or human rights, or which was closely related to social exclusion (such as social relations, social functioning, or employment). Articles that investigated outcomes solely in relation to e.g., clinical symptoms or continuity of care were excluded. Where an abstract was found to meet the initial criteria, we than reviewed the full paper and included for final selection any paper where the primary or secondary outcome or the main determinant focused on stigma or social exclusion/inclusion.

Selection of studies
Four reviewers (SEL, EC, ML, AF) participated in the selection. Each of them independently screened an initial set of 200 abstracts. Agreement for study exclusion between the authors ranged from 89-93%. Because of this high level of agreement, the remaining 4800 abstracts were divided between reviewers. Selected papers were then reviewed by the four reviewers (SEL, EC, ML, AF) independently to consider whether each retrieved paper met the inclusion criteria. Disparities in inclusion decisions were resolved through discussion. Reviewers did not contribute to inclusion decisions regarding studies in which they themselves were involved. Reference Manager was used to store all selected studies and extract relevant information into an SPSS database.

Results
The search on January 14, 2013 resulted in 11,078 articles from Medline and 16,047 articles from PsycINFO. When duplicates were removed, this resulted in 20,122 references. Although reviews themselves were not included, identified references were supplemented by recent systematic reviews of research related to social exclusion and mental health [9,11,37]. Of the total references, 5000 were randomly selected for review and data extraction. Three thousand eight hundred and eighty articles (78%) were excluded based on screening of the title and abstract. The remaining 1120 papers were reviewed using the full text version of the paper. Of those, 1023 were then excluded, as stigma/social exclusion was not considered to be the main focus of the paper. This resulted in 97 articles to be included in the systematic mapping (see Fig. 1) and the rest of this paper focuses on these articles.

Geographic region
The majority of studies (in this case we use the term study to refer to each paper, even though some may have included multiple studies in one paper) took place in Northwestern Europe, primarily in the United Kingdom (32%), followed by Finland (8%) and Sweden (8%) and Germany (7%). When we adjusted for population size (see Fig. 2), however, Finland had the highest number of publications per capita (1.49 publications per million inhabitants), followed by Slovenia (1.43), Sweden (0.85), Denmark (0.54) and the UK (0.50). Only a small proportion of studies (8%) included a site outside of Europe.

Study characteristics
We investigated the number of published articles over each year and we found that the number of papers has grown over time (see Fig. 3) with the largest number of studies being published in 2012 (n = 22). Prior to that, there was a gradual increase in the number of publications leading up to 2012. A brief comparison with the total number of publications on mental health problems using MESH terms for mental disorders in Pubmed shows an increase in research interest for both mental disorders in general and the social exclusion or stigma associated with these disorders. It should however be noted that studies focusing on stigma or social exclusion still constitute only a very small proportion of the mental health research (0.035% in 2007 to 0.054% in 2012).
Most of the studies analyzed primary data collected for the purpose of the study (65%), while about one third (34%) used secondary data. In terms of study design, most studies employed a cross-sectional design (56%) and another one quarter (24%) used a longitudinal design, usually based on existing cohort or registry data. A small proportion of studies performed qualitative research (11%); although mixed-methods designs were also considered, we did not find any articles using this approach. The smallest category comprised studies that used a randomized controlled trial or a controlled pre/post study design (9%).

Study sample characteristics
Regarding sample composition, we characterized each study according to representation of gender, age group and mental health condition of study participants. The majority of studies included both (47%) or unspecified genders (47%), with only a small number focused on a specific gender (5% included females only). Regarding the age of study participants, the majority of studies focused on adults of working age (60%) with only a small number focused on children (6%), adolescents (1%), older adults (4%) or multiple age groups (other than children and adolescents) (5%). One quarter (25%) of studies did not specify the age group.
In terms of clinical or mental health-related characteristics, most studies referred to mental health broadly (20%) or common mental disorders without focusing on a specific mental illness (17%). Schizophrenia (16%) and depression (13%) were the only mental disorders for which there was a particular focus in the stigma and social exclusion literature (see Table 1).

Type of research performed
We characterized studies in terms of the type of study (i.e., what were the study aims) (see Table 2), study design and also in terms of the focus of the research. When categorizing the type of study, we used an adapted version of the Health Research Classification System of the UK Clinical Research Collaboration [22]. Most of the articles included in the mapping aimed at screening at-risk populations (40%) or were underpinning research (25%). A related finding is that the majority of studies (60%) were descriptive and just under one third (29%) aimed to identify mechanisms. Only a small number of studies (11%) looked at intervention development and/or evaluated interventions. In terms of the focus of the research (i.e., what are the main research topics in relation to research on stigma and social exclusion related to mental health and well-being in Europe), we identified seven main topics of study (see Fig. 4). These categories were developed inductively based on the broad categories reflected in the identified published studies. The most common focus of the research was on employment-related difficulties, and this represented 32% of studies. This was followed by research focused on stigma and self-stigma (26%), inequality (11%), social exclusion (10%), discrimination (9%), loneliness (6%) and quality of life (6%).
We also investigated the research topics by age group and found that the distribution of age groups varied by research topic (see Fig. 5). Though possibly self-evident, research pertaining to employment-related difficulties and to discrimination did not include any studies focused on children, adolescents or older adults. Research on loneliness was the category with the most research on older adults. Research on quality of life was the most equally distributed across age groups.
Finally, we examined whether published studies included a formal definition of the social issue being studied in order to better understand the theoretical underpinnings of the studies. There were 32% of the articles that included in the mapping provided a formal definition. Comparatively more studies focusing on stigma included a definition of the issue being considered. When explicitly defined, Goffman's seminal definition of stigma as ''an attribute that is deeply discrediting and that reduces the bearer from a whole and usual person to a tainted, discounted one'' [20] was often quoted. In recent years, Link, Phelan and colleagues have revisited this definition to reflect advances in stigma research. Their definition distinguishes between five main components of stigma (labeling, stereotyping, separation, status loss, and discrimination) [32,33] and was used in the majority of the studies that included a definition.

Discussion
Overall, this systematic mapping of research on stigma and social exclusion in relation to mental health and well-being across Europe shows growing interest in this field in recent years. The growth might not be specific for stigma and social exclusion research as there were similar findings in literature mapping related to public mental health and well-being [36], and also Leeuwen [30] found growth in publications in social psychology. The systematic mapping methodology allowed us to take stock of a representative proportion of available evidence and to better understand what is known about social and economic impacts of    social exclusion and stigma across the life course. Moreover, this strategy also allowed us to identify geographical disparity in research in Europe and gaps in the evidence base and how these might be filled [7,31].
In terms of geographical representation, the mapping suggests that publications in relation to stigma and social exclusion are unequally distributed across European countries with most publications being from Northwestern Europe and relatively few from countries that have joined the EU since 2004, with the notable exception of Slovenia. Although we identified publications from Central and Eastern Europe, there was a substantial disparity in publications per capita in Western Europe compared to Central/ Eastern Europe. It could be that this is linked to the very active role in health policy that Slovenia had during its 2008 EU presidency, including events such as the high level conference ''Together for Mental Health and Well-being''. This provides a potential example of how institutions such as the rotating EU presidency could allow smaller or newer member states to gain momentum in certain areas, thereby beginning to reduce geographical inequalities. The low proportion of publications from countries that have joined the EU since 2004 is, however, not a novel finding as previous studies have shown a similar distribution, at least in terms of a low proportion of research in Eastern Europe [8]. Although the rotating presidency may be a helpful influence, it is not likely to be practicable for all countries and there may be other EU funding incentives that could be tailored to develop research capacity as well as further training and collaborative opportunities. Interestingly, the geographical distribution of research corresponds to the geographical distribution of stigma and discrimination, as measured in recent publications in that there is greater stigma in places with less research on the topic [28,44]. Thus, it could be that recognition and prioritization of stigma and social exclusion at the policy level and among researchers are important across many countries.
In recent years European networks focused on social issues in relation to mental illness and well-being have formed to facilitate research in this area across Europe. These include: the Anti-Stigma Programme European Network (ASPEN, http://www. antistigma.eu/), the European Network for Mental Health Service Evaluation (ENMESH, http://www.enmesh.eu/), Adapting European Health Systems to Diversity (ADAPT, http://www. cost.eu/domains_actions/isch/Actions/IS1103) and the Mental Health: European Economic Network (MHEEN, http://www. mhe-sme.org/mheen.html). These networks generate opportunities to include a more diverse range of countries in research (all networks involved new member states), provide links to research which has not yet been published in peer-reviewed journals and may also help to ''institutionalize memory' as individuals (e.g., researchers, policy-makers and other officials) move on. Nevertheless, a gap may still exist in terms of leadership on publications and research focused in countries currently lacking a research base. This is important as greater geographic diversity could help us to better understand the mechanisms and consequences of stigma and social exclusion across different cultural, economic and legal systems. There were only a small number of studies identified which included countries outside of Europe; however, there are some good examples of broad international networks such as International Study of Discrimination Stigma Outcomes (INDIGO) which are led by European researchers and are publishing work in the field. [28,44].
The mapping also identified a paucity of research on specific population sub-groups. Most research conducted focused on adults of working age. Previous reviews have also demonstrated a lack of research on stigma among children and adolescents [33]. A recent Cochrane review of mass media interventions identified none that were focused on children or adolescents [9]. This is important as early experiences of stigma and/or social exclusion among young people may lead to a cycle of social exclusion and disadvantage that is difficult to break later in life [29,41,48]. The mapping also highlighted gaps in relation to older people, a rapidly growing demographic group across Europe where loneliness has been identified as an important factor for mental health problems and decreased well-being [2,5,45]. Other research related to ROAMER has found a similar under-representation of older people in other related research areas, such as public mental health. There was also limited research in terms of other ethnic and religious minority groups including vulnerable population such as Roma, refugees or displaced people; and the relevance of this could be driven by the country in which the research was based. In terms of mental health conditions, most research focused on mental health generally or common mental disorders. Schizophrenia was an exception and was the mental health condition associated with the greatest number of publications. Although depression was also prominent in the literature, other disorders received little attention.
In terms of the type of research, overall the papers were characterized by descriptive research with few studies of interventions or mechanisms. Advances in the field of mental health stigma and social exclusion will require a move from descriptive approaches of the phenomenon to analyses of underlying social mechanisms, which could form the basis for future interventions to reduce stigma and promote inclusion.
Most research used primary data and employed cross-sectional designs. One implication is that further longitudinal research is needed to unravel the complex relationships between mental health problems, stigma and exclusion over the life course. The review shows that there is at the moment no conclusive evidence on causal links, which would provide a good basis for the design of interventions in that area. Study design varied by research focus, with for instance a larger proportion of intervention studies in the area of stigma. It is possible that opportunities for interrogation of existing cohort studies or registries may be considered further as administrative databases in health and social care exist in many European countries and offer possibilities for analysis of large longitudinal data sets. Further consideration of data linkage and the inclusion of validated stigma and or social exclusion-related questions [4] in cohort studies would also aid future research planning. This would also be important for the promotion and integration of multidisciplinary research on the determinants of stigma and social exclusion.
The mapping identified little published work on interventions. There is a growing number of national and regional initiatives in Europe for reducing stigma and promoting social inclusion [3,39]. However, despite the existence of interventions and social policy mechanisms that have been introduced to alleviate stigma and social exclusion across Europe, few of these initiatives are evaluated and much of the information is only available in local languages [3]. Moreover, evaluations of many existing interventions lack robust methods [9] that limits the conclusions that can be drawn about the most effective ways to reduce stigma and improve social inclusion of people with mental health problems.
Our review also noted that most published research did not include a formal definition of the terms used. Interventions which did refer to a theoretical model of stigma and cite a definition mainly referred to Goffman and or Link [20,32,33]. Although we have a long history around the development of social psychological and sociological understanding of stigma processes, there seems to be a disconnect between application of theory in relation to development and evaluation of interventions to reduce stigma and social exclusion. It may be that theoretical research could be further exploited in order to introduce a stronger theoretical base to intervention work. Additionally, clarifying the meaning and use of concepts in stigma and social exclusion research has practical implications for measurement of social outcomes in the area of mental health.

Strengths and limitations
This systematic mapping review covered a broad range of scientific papers and investigated a wide range of outcomes. Nevertheless, our review was limited in that it only included papers with English language abstracts and we did not perform an assessment of the quality of the studies. Covering such a broad area of research resulted in a very large number of abstracts to be retrieved and thus we were only able to review a random subsample of publications, even though this was still quite a large sample of 5000 papers. As we did not review all studies, it is possible that there may be some countries with a small number of studies which were missed from our review and thus, we lacked the statistical power to test and examine some more finely grained subquestions. Additionally, as we were focused on mapping scientific publications, we did not include grey literature or doctoral theses; the inclusion of these sources may have generated different patterns of activity as certain countries or disciplines may have different traditions in relation to the publication of research findings in books or monographs. Using a random sample of a large number of studies, however, should have led to a representative set of papers in terms of the study characteristics and geographic patterns.

Conclusion
The current social and economic circumstances in Europe highlight the social challenges of improving population mental health and well-being during times of hardship, particularly as these are periods during which the impact of stigma and social exclusion may be exacerbated. Our findings suggest that there has been a growth in research on stigma and social exclusion across Europe in recent years, which has helped decision-makers and citizens more generally gain a better understanding of the characteristics of individuals and populations which are associated with greater or less stigma and social inclusion. More research is needed on the mechanisms of reducing stigma and social exclusion, in addition to factors which might promote resilience or protect against stigma and social exclusion across the life course. In relation to mechanisms, we need greater connection between application of theory on development and evaluation of intervention, in addition to more robust evaluation of interventions to understand the most effective ways of promoting inclusion and reducing stigma across individuals and populations. Currently, research around stigma and social exclusion is unequally distributed across Europe and little research exists to contextualize these issues globally. Increasing incentives for cross-country research collaborations, especially with new EU Member States and with collaborations outside Europe, could help improve understanding of the diverse range of underpinning social and cultural factors which promote inclusion or contribute toward lower levels of stigma, especially during times of hardship. Additionally, collaboration across various European professional organizations and across a range of scientific disciplines could facilitate more diverse networks, and potentially promote creative approaches and more translation of research theory into practice. Europe is uniquely placed and may have a comparative advantage to research in comparison to other parts of the world as the heterogeneity in economic circumstances, population characteristics, and social-welfare/protection systems allows for interesting comparisons across countries. Incorporation of stigma measures in cohort or panel studies could help elucidate mechanisms and explore associations, while also facilitating natural experiments as a basis for evaluating policies or interventions which are introduced unevenly either over time [27] or across different cultural or other contexts. Proxy measures of social inclusion and exclusion are available in administrative databases, and measures to promote access and analyses of ''big data'' need to be fostered. Finally, cross-country collaborations and interdisciplinary collaborations should create opportunities for research which makes use of the rich theoretical frameworks that have been developed over time and evaluates them in real world settings. Mental illness and mental health and well-being search terms (title or abstract) ((''drug abuse'' OR ''drug addict'' OR ''drug addicts'' OR ''drug addiction'' OR ''drug addicted'' OR ''drug dependent*'' OR ''drug dependence*'' OR ''drug withdrawal'' OR ''drug abuse'') OR (''addictive disease*'' OR ''addictive disorder*'') OR (''alcoholic patient*'' OR ''alcoholic subject*'' OR alcoholism OR ''alcohol dependent*'' OR ''alcohol dependence*'' OR ''fetal alcohol*'' OR ''prenatal alcohol*'' OR ''chronic ethanol*'' OR ''chronic* alcohol*'' OR ''alcohol withdrawal'' OR ''ethanol withdrawal'') OR (''caffeine dependent*'' OR ''caffeine dependence'' OR ''caffeine addiction'' OR (caffeine AND addict*) OR ''caffeine withdrawal'') OR (((cocaine OR heroin OR cannabis OR mdma OR ecstasy OR morphine*) AND (abuse OR depend* OR dependent* OR dependence* OR addict* OR addicts OR addicted OR addiction* OR withdrawal)) OR methadone) OR (addiction OR addictive OR ''substance abuse'' OR ''withdrawal syndrome'' OR psychoactive*) OR ((schizophrenia OR schizophrenic) OR Schizotyp* OR ((Delusional OR paranoid) AND disorder*) OR hallucination* OR Psychotic OR Schizoaffective OR psychosis) OR (((manic OR bipolar OR mood) AND disorder*) OR (depressive AND (disorder* OR episode*)) OR ''depressive symptom*'' OR hypomania OR mania* OR ((major OR psychotic OR disorder*) AND depression) OR ''suicide attempt*'' OR suicidal* OR cyclothymia OR Dysthymia) OR (((anxiety OR panic OR ''Obsessivecompulsive'' OR adjustment OR conversion OR dissociative OR Somatoform OR Somatization OR neurotic) AND disorder*) OR (''hypochondriasis*'' OR ''body dysmorphic disorder*'' OR ''pain disorder*'') OR agoraphobia OR ''social phobia*'' OR ''Posttraumatic stress'' OR ''stress disorder*'') OR (''Eating disorder*'' OR ''Anorexia nervosa'' OR ''Bulimia nervosa'' OR ''sleep disturbance'' OR (sexual AND (disorder* OR dysfunction)) OR ((postnatal OR postpartum) AND depression) OR ((antidepressant* OR laxative* OR analgesic* OR psychotropic* OR vitamin* OR steroids