Inclusion of the Marginalized: The Case of Sport Participation

Abstract

Sport plays an important role in society and is seen as a tool to promote social inclusion. However, sport involvement is socially stratified and people who are in marginalized populations (such as people with a disability, refugees, ethnic minorities) often do not participate and thus cannot experience the benefits of sport. This scoping review aims to add to the scientific understanding of why marginalized populations do or do not participate in sport by scoping, structuring, and synthesizing the research landscape on determinants influencing sport participation. A search was conducted in the interdisciplinary databases Science Direct and Web of Science. The initial search identified 829 potentially relevant studies, of which 38 studies were included in the analysis. This paper identifies research gaps and opportunities and employs a socioecological framework to present comprehensive overview of determinants of sport participation for people who are in marginalized populations to inform future research and interventions.

Keywords:

• Sport determinants
• marginalized populations
• inclusion
• literature review
• socioecological framework

1. Introduction

Sport plays an important role in current society (Woods & Butler, 2020) due to the diverse benefits that are associated with sport participation regarding a healthy human development and thus individual and social well-being (see e.g. Human Capital Model by Bailey et al., 2013, 2016). Focusing on social well-being, participation in sport has shown to be beneficial for social development regarding social capital, social cohesion, and social participation (Coalter, 2005; Collins & Haudenhuyse, 2015); also for people in marginalized populations (Marlier et al., 2015; Skinner et al., 2008) who are at the center of this review. Therefore, sport is seen as a tool to promote social inclusion (Council of European Union, 2010; Kelly, 2011; Morgan & Parker, 2017). Based on this, the promotion of sport participation is an explicit policy objective, also in the EU (Christiansen et al., 2014; Ekholm, 2016; World Health Organization, 2011) as already stated in the European Sport for All Charter in 1975 (Council of Europe, 1975). Thereby, a special policy focus lies on inclusion of marginalized populations in sport (Hoekman et al., 2017; Hoogendoorn & de Hollander, 2017; Ministry of Health & Welfare & Sports, 2006, 2012; Vandermeerschen, 2016) and on making sport accessible for everyone, removing barriers (Ministry of Health & Welfare & Sports, 2018).

However, current participation rates indicate that actual sport involvement is socially stratified (Moreno-Llamas et al., 2020; Mutz & Müller, 2021). While there is an overall increasing trend to be physically active (Duijvestijn et al., 2020; van Bottenburg et al., 2005) and 41% of the European population engages in sport at least once a week (European Commission Directorate-General for Education, Youth, Sport and Culture, 2014), the evidence for marginalized populations is different. Sport participation rates of, for example, people with a low income, low educational level, migration background, and individuals with a disability are much lower (Carroll et al., 2014; Duijvestijn et al., 2020; Hildebrandt et al., 2013; van Bottenburg et al., 2005). For instance, the prevalence rates in the Netherlands (Duijvestijn et al., 2020) indicate that only 37% of the seniors (65+ years) adhere to the physical activity guidelines in comparison to 50.1% of the adult population (18–64 years). Regarding education, the adherence level to the physical activity guidelines is 22.2% lower among low educated populations compared to high educated populations.

Due to the limited sport participation of marginalized populations, those who might need it the most often do not profit from the benefits of sport regarding individual and social well-being. The societal issue of socially stratified sport participation is further intensified by a rise in inequality and a growing gap in society regarding socioeconomic status, social inclusion, and subsequently also leisure and sport participation (Collins & Haudenhuyse, 2015; Mayer et al., 2018; Musterd et al., 2017). This social divide in sport participation is not likely to dissolve by itself (Taylor, 2001). Thus, current participation rates indicate that - even though European policy has been endeavoring to make sport accessible for everyone over decades - the issue of socially stratified sport involvement in our society could not be solved until today and will likely even be intensified in the future.

To help tackle this issue, research investigating factors influencing whether people in marginalized populations do or do not participate in sport is needed to inform policy and practice. However, even though this topic has been on research agendas for years, research outcomes have not been sufficient yet to enable a greater understanding of the determinants influencing sport participation of people in marginalized populations. This can partly be explained by a number of factors. First, research predominantly targets determinants of sport participation within the general population (European Commission Directorate-General for Education, Youth, Sport and Culture, 2014; Hovemann & Wicker, 2009; Trost et al., 2002) instead of different populations in society (Taylor, 2001). Even though the identified factors repeatedly point out the importance of aspects linked to marginalized populations, such as socioeconomic status (Biernat et al., 2020; Herazo-Beltrán et al., 2017), income, educational level, or migration background (European Commission Directorate-General for Education, Youth, Sport and Culture, 2014), people in these marginalized populations are often disregarded as a target group. Research on the context in which their sport participation is embedded is lacking (Taylor, 2001; Vandermeerschen & Scheerder, 2017). Yet, people in marginalized populations experience a variety of factors that may exclude them from participating in sport (Collins, 2004). Their experiences and needs often differ from people in nonmarginalized populations (see e.g. O’Driscoll et al., 2014; Taylor, 2001) and further can differ between marginalized populations (O’Driscoll et al., 2014). Therefore, it is crucial to consider the specific circumstances experienced by marginalized populations (O’Driscoll et al., 2014; Taylor, 2001). However, literature tends to homogenize marginalized populations (O’Driscoll et al., 2014). Vandermeerschen and Scheerder (2017) further claim that research most commonly focuses on only a few marginalized populations, such as people with a disease or ethnic minorities, while other populations stay neglected. Additionally, existing research mainly focusses on the influence of single aspects on sport participation instead of examining the broader sport participation context including the underlying levels of influence. Overall, the existing research is fragmented—focusing on a few marginalized populations, focusing on a few determinants, and often being small-scaled.

This is where the present scoping review is located. To help make sport inclusive, it is necessary to understand the greater context regarding determinants of sport participation for people in marginalized populations and to understand why people do or do not participate. Therefore, more integrated and holistic research is needed. To elevate and inform future research, it is thus decisive to review and bring together the fragmented existing research first in a systematic and exhaustive manner. Therefore, this paper scopes, structures, and synthesizes the research landscape on determinants of sport participation for people in marginalized populations published between 2016 and 2020. To meet this research aim, a scoping review approach allows to map existing literature addressing specific research topics - often emerging and complex topics - in order to synthesize existing research knowledge and to identify key concepts and gaps to inform future research as well as practice and policy (Colquhoun et al., 2014; Daudt et al., 2013). Following the scoping review framework by Arksey and O’Malley (2005), a broad scope was chosen to ensure wide and exhaustive coverage of the broad topic- acknowledging that the diversity of marginalized populations as well as the diversity of sport participation needs to be considered to make sport entirely inclusive for everyone.

Accordingly, the structure of the paper is as follows: first, a theoretical framework is presented including the applied socioecological framework and the definition of marginalized populations employed for this paper. Second, the study selection process is explained, and the selected studies are described. Third, an overview of the included studies and the identified determinants of sport participation for marginalized populations within a socioecological framework is given. Fourth, the findings of this scoping review are being discussed.

2. Theoretical Framework

2.1. Socioecological Framework

Socioecological models are increasingly implemented in research on sport participation. They are based on the central assumption that specific determinants influence health behavior on multiple interacting levels: the individual level (physical and psychological), the social level (interpersonal and cultural), the organizational level (routines and communication), the environment level (natural and built), as well as the political and societal level (laws and government; Bauman et al., 2012; Sallis et al., 2008). Research findings also confirm that diverse interacting socioecological levels influence health behaviors, as shown for health-seeking behavior (Chung et al., 2018), eating behavior (Brogan et al., 2021; Hermann et al., 2011), sexual behavior (R. Stevens et al., 2014), and physical activity (Biernat et al., 2020; Rawal et al., 2020).

The socioecological approach (Bauman et al., 2012; Sallis et al., 2008) was further specifically adapted to the sport context by Eime et al. (2013). The Health through Sport Model (Eime et al., 2013) assumes, among other things, that sport participation, ranging from individual to team and from informal to organized forms, is influenced by determinants on the different socioecological levels; intrapersonal, interpersonal, organizational, environment, and policy level. Accordingly, individual and contextual characteristics can be perceived as barriers or enablers to sport participation by a person. Research correspondingly highlights that physical activity in adults is associated with factors originating from all socioecological levels (Herazo-Beltrán et al., 2017; Rech et al., 2018; Yarmohammadi et al., 2019). Research thus links sport participation to enabling or disabling factors across all socioecological levels and shows that marginalized populations experience a variety of general as well as specific determinants regarding sport participation (O’Driscoll et al., 2014). Sport participation behavior of an individual is accordingly embedded in its socioecological environment and is therefore influenced by a diversity of factors. To make research and interventions more effective it is accordingly crucial to be aware of all levels of influence (Sallis & Owen, 2015). This underlines the importance of understanding factors associated to sport participation in a more holistic way.

Therefore, socioecological frameworks are considered as valuable for contemporary sport research (Rowe et al., 2013). Acknowledging the impact of all socioecological levels, multilevel research on determinants explains behaviors, such as sport participation, more comprehensively than research focusing on specific levels (Sallis & Owen, 2015). Socioecological frameworks thereby further allow the integration of different theories in a comprehensive manner enabling to integrate existing research (Sallis & Owen, 2015).

Based on this, this scoping review applies a socioecological framework to comprehensively scope, organize, and integrate existing research. Thus, this paper structures the different types of factors identified in literature within a socioecological framework allocating the identified determinants of sport participation to the individual, social, organizational, environmental, and political and societal level.

2.2. Marginalized Populations

There is no unanimous definition of who belongs to a marginalized population. Gatzweiler et al. (2011, p.3) describe marginalization as the ‘position and condition of an individual or group at the margins of social, political, economic, ecological, and biophysical systems, that prevent them from access to resources, assets, services, restraining freedom of choice, and preventing the development of capabilities’. This reflects that marginalization is an umbrella term (Cheraghi-Sohi et al., 2020) for individuals or populations being excluded from participation in society. Various populations are considered to be marginalized, such as poor, elderly, and disabled persons, as well as sexual minorities, people with a migration background, or asylum seekers (Collins & Haudenhuyse, 2015). Also, marginalization can be temporarily, e.g. in case of short-term unemployment or divorce, or chronically in case of being old or permanently impaired.

This paper adopted a broad definition of people in marginalized populations by including populations that are commonly considered as socially marginalized to cover the diversity of such populations: people with a low socioeconomic status (SES), people with a disability, elderly, ethnic minorities, refugees, the LGBTQ + community, homeless people, people with substance abuse, and criminals.

3. Method

A scoping review is a research approach aiming to synthesize existing knowledge on a specific topic, especially for addressing upcoming and complex exploratory research questions (Colquhoun et al., 2014). Therefore, a scoping review is defined as a way to ‘map the literature on a particular topic or research area and provide an opportunity to identify key concepts, gaps in the research; and types and sources of evidence to inform practice, policymaking, and research’ (p.8; Daudt et al., 2013). The established scoping review framework by Arksey and O’Malley (2005) was employed as a guideline for the scoping review process. Thereby, the improvements suggested by Levac et al. (2010) were taken into consideration. The process included the following stages: identification of research question, identification of relevant studies, study selection, data extraction (charting the data), collation, summary, and report of results. An overview of every process stage is provided in the following.

3.1. Identification of Research Question

The objective of this study was to scope, structure, and synthesize the current state-of-the-art in research regarding determinants influencing sport participation of people in marginalized populations by establishing the scope of existing literature regarding a) descriptive markers (study characteristics including samples and research designs) and b) study content (including theoretical perspectives, sport types, and socioecological determinants of sport participation). Based on this, c) current trends in research, d) research gaps, and therefore e) research opportunities were thematized to inform future research on this topic. Following Arksey and O’Malley (2005), the research questions were defined using a wide approach to generate breadth of coverage and to allow identification of all relevant studies.

3.2. Identification of Relevant Studies

A literature search was conducted in the interdisciplinary databases Science Direct and Web of Science: Social Sciences Citation Index (SSCI). The electronic interdisciplinary databases were selected to ensure a wide coverage of literature. Relevant keywords covered terms related to sport, determinants (e.g. barriers and factors) and marginalized populations (e.g. minorities, poor, and homeless). The applied key terms were ((‘sport’ AND (‘barriers’ OR ‘constraints’ OR ‘challenges’ OR ‘factors’ OR ‘determinants’) AND

1. (‘minority’ OR ‘minorities’ OR ‘migrant’))

2. (‘refugee’ OR ‘reception’ OR ‘asylum’))

3. (‘poor’ OR ‘socioeconomic’ OR ‘income’ OR ‘unemployed’))

4. (‘disabled’ OR ‘disfigurement’ OR ‘ill’ OR ‘impairment’))

5. (‘prisoner’ OR ‘offender’ OR ‘felon’ OR ‘criminal’))

6. (‘substance abuse’ OR ‘addict’))

7. (‘gay’ OR ‘lesbian’ OR ‘bi-sexual’ OR ‘transgender’ OR ‘sexual minority’))

8. (‘homeless’ OR ‘shelter less’))

9. (‘elderly’ OR ‘older age’ OR ‘senior’)).

The title, abstract, and/or keywords in the database were searched for these key terms. The search was restricted to scientific journal articles written in English and published between January 2016 and December 2020. The search was restricted to literature from 2016 onwards in order to focus on the most recent research trends and developments. After removing duplicate entries, titles and abstracts of the identified articles were reviewed for eligibility based on the following inclusion criteria:

1. Including determinants or factors influencing sport participation or physical activity or both;

2. Marginalized population;

3. Age > 18 years*;

4. Scientific studies published between January 2016 and December 2020 in scientific journals;

5. Full-text available;

6. Published in English language.

*The age frame was chosen as this scoping review focuses on sport participation of adults.

3.3. Study Selection

The titles and abstracts of all articles were reviewed and screened for relevance to assure that they answered the research questions and that they met the inclusion criteria described above. Studies that did not meet the inclusion criteria and did not address our research topic were excluded.

3.4. Data Extraction

The following information was extracted and organized from each article: publication data (e.g. publication year, research country), target group (e.g. elderly, homeless), research design (quantitative, qualitative, mixed methods + cross-sectional, longitudinal), number of study participants, theoretical perspective (e.g. socioeconomic, psychological), sport type (e.g. soccer, cycling), sport entity (organized program versus not organized program), identified determinants of sport (e.g. social support, transport) and/or physical activity according to the associated socioecological level (e.g. individual, social). The selection of extracted data was based on the research questions.

3.5. Collation, Summary, and Report of Results

The extracted information was organized in a table and frequency and thematic analyses were conducted in line with Arksey and O’Malley (2005). The results were discussed with respect to the initial research questions. According to the scoping review framework (Arksey & O’Malley, 2005), the study qualities were not evaluated.

4. Results

4.1. Study Selection

A structured database search identified 829 potentially relevant abstracts. Additional potentially relevant articles were extracted via a manual search (n = 2). The set was refined by removing duplicate entries (n = 127). The titles and abstracts of the remaining references (n = 704) were then screened for relevance, leading to the exclusion of 610 articles. The full texts of the remaining 94 articles were then assessed for eligibility, subsequently excluding articles that did not meet the inclusion criteria. In total, 38 articles were included in this review. Figure 1 shows the sample development throughout the selection process.

Figure 1. Selection of studies: according to PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) flowchart.

An overview of the included studies is provided in Appendix 1 and the study characteristics as well as study content are further summarized in the following. The items in parentheses refer to the corresponding studies displayed in Appendix 1.

4.2. Study Characteristics

4.2.1. Target Group

Out of the 38 included studies, the largest category of studies (n = 12; items 5, 8, 9, 10, 11, 13, 14, 22, 26, 31, 33, and 34) targeted SES-related variables more generally covering income, education, and poverty. Eight studies included people with disabilities (items 2, 3, 4, 15, 25, 28, 32, and 35), six studies included elderly (items 12, 19, 27, 29, 37, and 38), five studies focused on ethnic minorities (items 7, 17, 20, 23, and 36), and two included refugees (items 1 and 21). Four studies investigated the LGBTQ + community (items 16, 18, 24, and 30). One study targeted a ‘hard-to-reach’ population by including homeless people as well as people reporting substance abuse (item 6). No studies on ex-prisoners or criminals were identified.

4.2.2. Research Design

The proportion of quantitative (n = 17; items 2, 3, 5, 8, 9, 11, 13, 14, 19, 23, 26, 27, 29, 31, 32, 33, and 37) and qualitative (n = 17; items 4, 6, 7, 10, 12, 15, 16, 17, 18, 20, 21, 24, 25, 28, 34, 35, and 36) study designs was equally distributed and four (n = 4; items 1, 22, 30, and 38) out of the 38 included studies applied a mixed methods approach. Quantitative research covered surveys (items 3, 5, 9, 11, 14, 19, 27, 31, 32, 33, and 37) including one panel study (item 13) as well as one cohort study (item 29), analysis of registration data (items 2 and 8), and two studies combining population census data and survey data (items 23 and 26). In qualitative research, semi-structured interviews (items 4, 10, 16, 17, 18, 20, 25, 34, and 35) and focus group discussions (items 7, 24, and 36) were used mostly. Combinations of qualitative methods were applied in five studies; interviews plus focus groups (items 15 and 28), interviews combined with observations and workshops (item 12), ethnographic and observational methods (item 6), and a case study including observations, interviews, follow-up interviews and focus groups (item 21). Studies applying mixed-methods covered surveys (items 30 and 38), a Delphi-study (item 1), and concept mapping (item 22).

As shown in Figure 2, however, the distribution of qualitative versus quantitative research designs differed greatly depending on the target group. Quantitative study designs were mainly implemented for SES-related variables (n = 9; items 5, 8, 9, 11, 13, 14, 26, 31, and 33) and for studies targeting elderly (n = 4; items 19, 27, 29, and 37). Research including people with a disability consisted of three quantitative (items 2, 3, and 32) and five qualitative (items 4, 15, 25, 28, and 35) studies. Studies on ethnic minorities, the LGBTQ + community, and ‘hard-to-reach’ populations, on the contrary, were almost exclusively based on qualitative research which is partly due to the absence of reliable sampling frames, with only one included quantitative study on ethnic minorities (item 23) and no quantitative research identified for the LGBTQ + community, and ‘hard-to-reach’ populations.

Figure 2. Number of publications by target group and research design.

The majority of studies was based on a cross-sectional design (n = 36) with only two identified longitudinal studies (items 21 and 29) including follow-ups after one year (item 21) and fourteen years (item 29).

4.2.3. Sample Sizes

The number of included study participants varied greatly between studies (see Appendix 1), ranging from n = 7 (item 21) to n = 488698 (item 8). Twenty studies reported at least 100 participants (n ≥ 100; items 2, 3, 5, 7, 8, 9, 11, 13, 14, 19, 23, 26, 27, 29, 30, 31, 32, 33, 37, and 38). These large-scale studies consisted primarily of research on SES (45%, n = 9; items 5, 8, 9, 11, 13, 14, 26, 31, and 33) and elderly (25%, n = 5; items 19, 27, 29, 37, and 38), followed by people with a disability (n = 3; items 2, 3, and 32), ethnic minorities (n = 2; items 7 and 23), and the LGBTQ + community (n = 1). Small-scale research (n ≤ 30), on the contrary, included people with a disability (n = 3; items 4, 25, and 35), refugees (n = 2; items 1 and 21), the LGBTQ + community (n = 2; items 16 and 24), and ethnic minorities (n = 2; items 17 and 36) as target groups. SES (16,67%, n = 2; items 22 and 34) and elderly (8,33%, n = 1; item 12) populations were presented to a smaller extend in small-scale compared to large-scale studies.

4.2.4. Theoretical Perspectives (See Appendix 1)

Seventeen out of 38 studies (n = 17; items 2, 3, 4, 5, 6, 7, 10, 11, 12, 16, 19, 22, 24, 27, 34, 35, and 37) did not base their research on a specific theoretical model. Regarding the 21 studies that did apply a theoretical model, the socioecological framework was most referenced (n = 7) across marginalized populations, referring to Bronfenbrenner’s ecological framework for human development (items 1, 14, 18, and 38), the four-level social ecological model by Vaseduvan et al. (item 32), or not further specifying the concrete socioecological model (items 26 and 8).

Sociological models (n = 4) were applied second most frequent and focused on the interrelation of social structures and sport participation among refugees, ethnic minorities, and elderly people. The four included studies covered Bourdieús concepts of habitus, taste, and practice (items 17, 20, and 21) as well as the social identity approach (item 29). Economic theories based on the household production theory and the economic theory of behavior by Becker were employed in three studies related to income and SES (n = 3; items 9, 13, and 31). Another three studies referred to frameworks specific for marginalized populations (n = 3); the critical disability theory (item 28), the concept of ableism (item 15), and the minority stress model (item 30). Two studies focused on environmental spatial perspectives for ethnic minorities and people with disabilities (n = 2) applying the framework of spatial deprivation (item 23) and the hypothetical theoretical accessibility model of physical environment of urban centers regarding accessibility for people with disabilities (item 25). One study was set up in the organizational context, referring to organizational capacity in addition to the theory on resource dependence, for explaining sport participation of people with a low SES (item 33). A further study was based on a psychological framework including motivational and health behavior theories regarding ethnic minorities (item 36).

4.2.5. Sport Type

Regarding the investigated sport types (Appendix 1), 66% of the included studies (n = 26; items 3, 5, 7, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 23, 24, 26, 27, 28, 29, 30, 32, 33, 34, 36, 37, and 38) did not specify the concrete type of sport or respectively physical activity within their research. Additionally, seven studies (items 1, 2, 4, 8, 10, 22, and 35) targeted ‘sport programs’, ‘sport clubs’ or ‘popular sports’ in general. Overall, five out of 38 studies (13%) had a focus on specified types of sport; two studies focused on sport centers (items 9 and 25), one study specifically investigated soccer (item 6), one was set up around a cycling course (item 21), and another study addressed a list of the most popular sports in a local area: soccer, swimming, dance, cycling, running, fitness, tennis, horse riding, winter sports, martial arts, volleyball, walking, and basketball (item 31).

4.3. Socioecological Determinants of Sport Participation

A key focus of this paper is to provide an extensive insight into socioecological factors being identified as decisive for sport participation of people in marginalized populations by current research. Therefore, the factors that were identified as relevant for sport participation of marginalized populations in the included studies were allocated to the associated socioecological levels—individual, social, organizational, environmental, political, and societal. The distribution of examined socioecological levels is summarized in Figure 3. An overview of the identified individual, social, organizational, environmental, as well as political and societal factors is outlined in the following. A socioecological framework including all identified determinants is further presented in Figure 4.

Figure 3. Number and proportion of investigated socioecological levels.

Note: The y-coordinate represents the percentage of included studies. The x-coordinate resembles the number of investigated levels.

Figure 4. Socioecological determinants of sport participation for people in marginalized populations.

4.3.1. Socioecological Level

As shown in Figure 3, presenting 42.1% (n = 16; items 3, 5, 9, 11, 13, 14, 22, 23, 26, 27, 29, 30, 31, 33, 34, and 37) of the included studies, most research on determinants influencing sport participation of people in marginalized populations focused solely on one socioecological factor level. Two (n = 9; items 2, 6, 8, 12, 19, 21, 24, 25, and 35) or three (n = 4; items 15, 17, 36, and 38) socioecological levels were further investigated by 34.21% of the studies. A more holistic analysis covering four (n = 7; items 1, 4, 7, 10, 16, 20, and 32) or all five (n = 2; items 18 and 28) socioecological levels were provided in 23.68% of the studies. Thereby, individual factors influencing sport participation were investigated most frequently (n = 29; items 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 26, 27, 28, 29, 31, 32, 36, 37, and 38) followed by organizational determinants (n = 17; items 1, 4, 10, 15, 16, 17, 18, 20, 21, 22, 24, 25, 28, 32, 33, 35, and 36). Environmental aspects were included in 15 (items 2, 6, 7, 8, 10, 11, 15, 16, 18, 19, 23, 25, 28, 32, and 38) studies. Political and societal (n = 11; items 1, 4, 7, 16, 18, 20, 24, 28, 30, 34, and 35) as well as social factors (n = 12; items 1, 4, 7, 10, 12, 17, 18, 20, 28, 32, 36, and 38) were the least examined socioecological levels. Thereby, studies covering one or two socioecological levels investigated individual factors most commonly (n = 16; items 2, 3, 5, 6, 8, 9, 12, 13, 14, 19, 21, 26, 27, 29, 31, and 37) especially focusing on the SES, whereas social factors (n = 1; item 12) were the least studied.

4.3.2. Individual Level

A high proportion of the determinants being identified as decisive for sport participation by people in marginalized populations were related to the individual level covering socio-demographics, life circumstances, as well as physical and psychological factors.

Regarding socio-demographics, factors being relevant for sport participation were mainly specific to certain marginalized populations. Gender and the associated gender habitus were identified as crucial for sport participation of female refugees (n = 2; items 1 and 21), whereas gender incongruence and a not binary system fitting trans*identity were decisive barriers for transgender people (n = 2; items 16 and 24). For female refugees and females with a migration background, sport participation capability was determined by requirements from their cultural background or religion, being in appropriate clothing, female-only environments, or less energy for sport during Ramadan. However, it was emphasized that neither cultural background nor religion were barriers themselves (n = 2; items 1 and 17). Language barriers were further mentioned as a hindering aspect for sport participation by female refugees (n = 1; item 1).

Yi et al. (2016; n = 1; item 38) stated the significance of educational background and knowledge in terms of sport participation for older adults. While the relevance of socio-demographic factors was mainly linked to certain marginalized populations, the socioeconomic status or financial situation were highlighted as determinants of sport participation frequently and independently of the marginalized population (n = 6; items 1, 6, 7, 28, 36, and 38).

A theme that recurred frequently across marginalized populations was associated with life circumstances: time and energy (n = 5; items 7, 17, 27, 28, and 36). Time and energy were substantially influenced by work (n = 4; items 7, 10, 17, and 36), as well as domestic or family responsibilities, and childcare (n = 4; items 7, 10, 17, and 36). Many people stated that it was challenging for them to establish a routine or schedule for sport due to their unstable or busy life situation (n = 3; items 6, 20, and 28).

The physical status became a determining factor for sport participation when it was either impaired, changing, or made people feeling out of place. Older adults reported that health, mobility, and physical functioning were decisive for their sport participation capability (n = 3; items 12, 27, and 38). People with disabilities further listed increased vulnerability to injury, the level and ranging severity of impairment (n = 2; items 2 and 28), as well as pain and fatigue (n = 1; item 32) as crucial factors. This was accompanied by concerns associated with possible physical limitations in combination with others’ ability to grasp and address their limitations (n = 1; item 28). Sport participation of transgender people, however, was influenced by their stage of transition and gender (in-)congruence, in addition to body appearance (n = 2; items 16 and 18). Looking specifically at the sporting ability, the general physical capital, including sport skills and experiences, was mentioned as a relevant factor for sport participation by people across marginalized populations (n = 4; items 4, 18, 21, and 38). Thereby, the influence of physical capital was especially relevant for females. Female refugees and women with a migration background further emphasized the role of the physical activity habitus formed by childhood socialization in their home country (n = 2; items 12 and 17).

Psychological factors, including attitudes and beliefs, were also regularly highlighted as determinants for sport participation by people in marginalized populations. Thereby, a main recurring theme was the motivation to do sport (n = 4; items 7, 12, 32, and 38). In addition, the general awareness about sport benefits and options was important for sport participation in general (n = 5; items 1, 7, 15, 27, and 28). People with disabilities further faced a variety of specific psychological factors and anxieties influencing their sport participation. For instance, they mentioned confidence as a factor in explaining own needs (n = 2; items 28 and 32), self-esteem or self-efficacy (n = 1; item 15), and enjoyment of sport (n = 1; item 15) as crucial for participating in sport. Fears of unknown places and transport routines (n = 1; item 15), feeling like a ‘burden’, being a ‘public spectaclé, or being patronized by instructors and anticipations of others’ intrusive questions (n = 1; item 28) were further perceived as barriers to sport participation by people with disabilities. Other marginalized populations named further psychological factors: Transgender people underlined the role of body (dis-)satisfaction and their anxiety of others’ reactions towards them (n = 1; item 16). People with a migration background reported the importance of personal preferences, e.g. for private facilities (n = 1; item 7).

4.3.3. Social Level

Social -or respectively interpersonal- factors being critical for sport participation were identified across marginalized populations. Thereby, a few themes emerged repeatedly and independent of the marginalized population: social support, safety, and role models. The most frequently mentioned social determinant included the support, motivation, encouragement, and (in-) activity of family, friends, the team, or even dogs, and thus highlighted the importance of the social network’s role in facilitating the sport participation of an individual in a marginalized population (n = 7; items 7, 12, 17, 18, 28, 32, and 38). Being accompanied by family, friends, or a team was further emphasized as a source of safety (n = 3; items 7, 18, and 28) during sport for either being secure in unsafe neighborhoods or feeling safe in unfamiliar situations. Whether people in marginalized populations participate in sport was further influenced by role models or mentors (n = 2; items 17 and 28). It was thereby stressed that ʿadequateʾ role models being from the same background and being relatable, respectively being everyday peers rather than ʿheroesʾ, were crucial for being a motivating example. Other identified social determinants were specific to marginalized populations: For the elderly and people with a migration background not being part of a social network or the loss of social contacts and activities were highlighted as important social barriers to sport participation, underpinning that these populations often suffer from social isolation (n = 3; 12, 17, and 38). On the other hand, refugees reported that group dynamics within the sport context, such as communication, language barriers, conflict, and intrapersonal openness to other cultures, affect the implementation of intercultural sport opportunities (n = 1; item 1).

Within the transgender community, anxiety associated with the social interaction aspect of sport was mentioned as a barrier including fears about team reactions on body changes and feeling uncomfortable with strangers for female transgender persons especially (n = 1; item 18). People with disabilities further reported that over-concern for their safety by family or instructor kept them from being engaged in sport, as well as exaggerated expectations in terms of their physical sporting ability by others (n = 2; items 4 and 28). However, for people with a low socioeconomic status the social responsibility (n = 1; item 10) that they would take away the spot from someone else who would like to participate if they would miss a sport class and the obligation to their mentor was seen as a facilitator of sport participation.

4.3.4. Organizational Level

Determinants of sport participation being associated with structures and processes of sport organizations were reported repeatedly in the included studies and were found across marginalized populations. Thereby, prices recurred as an important determinant (n = 3; items 15, 32, and 36), followed by organizational capacity (n = 1; item 1), and an understaffed workforce resulting in waiting lists for participation (n = 1; item 4). One study (n = 1; item 1) further showed the systemic challenges in joining sport organizations that refugees face. Their uncertain situation can cause difficulties in being integrated into regular club operations, their membership processing, and the required flexibility of organizations as the number of arriving refugees and the associated demand fluctuates. A number of studies further highlighted the influence of the binary-gendered system within organizations on sport participation of transgender people, being expressed in binary divided facilities or teams and no flexible or adapted regulations for transgender individuals (n = 3; items 16, 18, and 24). People with disabilities also underlined the role of unwelcoming standard fitness practices, such as exercises they could not join due to physical limitations, and an inclusive sporting culture with an understanding of providing accessible sport opportunities regarding their sport participation (n = 2; items 4 and 28). Úbeda-Colomer et al. (2019, n = 1; item 32) specifically addressed the need for adapted programs for people with disabilities. This is in line with the frequently mentioned factor of need orientation which entails that sport programs fit the needs, motivations, and resources of the target group (n = 2; items 4 and 21). In detail, women with a refugee or migration background preferred women-only courses (n = 2; items 17 and 21) and transgender individuals underlined the factor of trans-only environments or teams in terms of their sport participation (n = 1; item 16). Individuals with a migration background further addressed the need to offer free, flexible, informal, and intercultural sport opportunities (n = 1; item 21). People with disabilities and refugees further addressed the advertisement and promotion strategies of sport organizations.

To increase the awareness of ‘hard-to-reach’ populations more effective information and communication on sport opportunities was demanded (n = 3; items 4, 15, and 32), including tailored messaging (n = 2; items 1 and 28) and adequate advertisement which takes into account that the internet might be challenging for some people (n = 1; item 15). Furthermore, internal as well as external communication and cooperation to inform about opportunities and to stay in touch was mentioned as decisive for sport participation by people who attended sport programs for asylum seekers and refugees (n = 2; items 1 and 21). The importance of attentive professional staff within the sport organization for guidance, support, and motivation was highlighted repeatedly by people in marginalized populations. It was emphasized that all employees (i.e. coaching, voluntary, as well as administration staff) need to be trained adequately to offer inclusion. People with disabilities specifically reported the relevance of the staff being aware of their special needs and having the competencies and training needed to address these special needs (n = 7; items 1, 4, 22, 25, 28, 32, and 36). Additionally, people in marginalized populations appreciated it if part of the staff also came from a marginalized background to understand their situation better. Accordingly, Brown and Pappous (2018; n = 1; item 4) reported the need for disabled experts or instructors with disabilities and refugees underlined that they would be encouraged to participate in sport if a part of the staff would also have refugee background (n = 1; item 21).

4.3.5. Environmental Level

Environmental factors influencing sport participation, ranging from neighborhood characteristics over transport to architectural features of sport facilities, were mentioned frequently across studies and marginalized populations. They concern the accessibility of sport facilities and the opportunities to participate.

In terms of neighborhood characteristics, the importance of local sport facility provision, including green spaces (n = 6; items 1, 2, 6, 7, 26, and 38) and the safety of an area (n = 2; items 7 and 38) were highlighted as critical factors for sport participation commonly and independently of the marginalized population. The condition of routes to sport facilities, sidewalks and streets, further specifically influenced the capability to participate in sport of mobility impaired people (n = 1; item 32).

Transport being decisive in terms of sport participation was furthermore mentioned very frequently by all populations referring to limited mobility, financial aspects, and, specifically for visually and mobility-impaired people (n = 1; item 28), physical barriers of means of transport (n = 4; items 1, 6, 7, and 15).

However, architectural or built features of sport facilities were solely relevant for people with disabilities and the LGBTQ + community. Accessibility of sport facilities, e.g. in terms of parking spaces, entrances, elevators, equipment, changing rooms, showers, and toilets, was crucial for sport participation of people with disabilities (n = 2; items 25 and 28). Transgender people, on the other hand, emphasized the importance of changing rooms, showers, and toilets offering privacy and not being binary gendered (n = 2; items 16 and 18).

4.3.6. Political & Societal Level

The political and societal level covered a wide range of determinants of sport participation for people in marginalized populations covering societal attitudes, politics, and laws, but also media, economy, and the healthcare system.

For refugees and people with disabilities, societal attitudes played a decisive role in terms of sport participation. In a study by Anderson et al. (n = 1; item 1), the importance of an inclusive social mindset as well as the societal awareness and appreciation in terms of refugee health and integration were highlighted as important factors along with a mutual understanding of religion and culture. Socio-cultural norms of the home country were further named in relation to forming a physical activity habitus of especially female refugees (n = 1; item 21). Regarding people with disabilities, it was additionally stated that societal stereotypes about people with disabilities, public education about disability and sport, and linguistic terms disconnecting between disability and fitness played a role in sport participation (n = 1; item 28). Specifically for people with disabilities, the media’s influence on society’s perceptions of disability, e.g. via coverage of the Para Games, was seen as an important factor in fostering sustainable sport participation (n = 1; item 4).

A large proportion of the identified determinants were further linked to political and legal regulations. These included bureaucracy (n = 1; item 1), security (n = 1; item 7), or obtaining assistance (n = 1; item 28). The main topics in this context were funding as well as austerity and the associated restrictions in the provision and the time-consuming efforts to secure funding of sport programs (n = 4; items 1, 4, 7, and 28). These factors were decisive regarding whether programs were offered and whether regulations made accession complicated or simplified. In addition, Vandermeerschen and Scheerder (n = 1; item 34) thoroughly investigated the impact of differing local social politics in municipalities concerning sport provision for people with a low SES. It was shown that solely financial measures are often taken instead of increasing awareness and expertise about social inclusion mechanisms and reaching people with a low SES who often are not even considered as a target population. Furthermore, intersectoral communication and cooperation with other municipalities, the social sector, and sport organizations were considered as important for increasing sport participation. Within the LGBTQ + community, the legal binary gendered regulations and the juridical gender recognition process impacted the sport participation of Transgender people including their capability to be in a team or competition (n = 2; items 18 and 24).

Sport participation was further assumed to be impacted by the availability of appropriate clothing for females with a migration background and transgender people (n = 3; items 16, 17, and 18) and the availability as well as affordability of adapted equipment for people with disabilities (n = 1; item 28).

Furthermore, the role of health professionals in promoting and supporting sport participation based on the needed expertise was emphasized by people with disabilities as well as by transgender people (n = 2; items 18 and 28).

5. Discussion

This scoping review scopes, structures, and synthesizes the current state-of-the-art in research regarding determinants influencing sport participation of people in marginalized populations, including people with a low socioeconomic status, people with a disability, elderly, ethnic minorities, refugees, the LGBTQ + community, homeless people, people with substance abuse, and criminals. Via reviewing the literature published in the period 2016–2020, this paper identified research gaps and opportunities through reviewing the applied research methods and provided a socioecological framework on determinants of sport participation for people in marginalized populations.

Thereby, this review identified a number of patterns, gaps, and opportunities regarding current research approaches and study content. Even though various populations are considered marginalized by society (Collins & Haudenhuyse, 2015), a large proportion of the included studies on marginalized populations focused solely on a few populations; people with a low SES, people with disabilities, and the elderly. This focus on specific marginalized populations was also claimed by Vandermeerschen and Scheerder (2017). Current politic agendas might provide a possible reason why research focuses on specific populations. Looking at the social inclusion policies and activities of the European Commission as an example, there is a strong emphasis on social inclusion regarding people in poverty being socially excluded, equality and accessibility for people with a disability, and active ageing strategies for the elderly (European Commission, Directorate-General for Employment, Social Affairs and Inclusion, n.d.). These mentioned populations correspond to the marginalized populations that were studied the most in this review. Accordingly, research might especially focus on topics that are relevant in current politics to help feed and drive politics with empirical evidence. Thereby, specific topics or target populations might be more present in research than others that are rather being neglected.

Additionally, even though sport participation of people with a low SES and the elderly is researched regularly and often in a large-scale manner, the research set up of the included studies was often restricted to quantitatively investigate the link of the SES and age with sport participation as general and isolated factors. Though, it was shown that sport participation of the elderly depends on various determinants besides age itself (Jenkin et al., 2017). In addition, sport initiatives targeting people with a low SES that solely rely on financial support (Vandermeerschen & Scheerder, 2017) are often not sufficient as this group is marginalized and disadvantaged in various ways (Mowat, 2015; Saatcioglu & Corus, 2014) and not only financially. By only including SES and age as determinants and not investigating associated factors that may influence sport participation more contextually, the specific circumstances that prevent people with low SES and elderly from taking part in sport cannot be fully understood and thus not be targeted. Research on other marginalized populations included in this research, on the contrary, was predominantly based on cross-sectional, small-scale, and qualitative research designs. This unilateralism of research designs in terms of sport participation of marginalized populations limits the overall range of insights, conclusions, and generalizability of outcomes drawn from research. Furthermore, a large proportion of the studies on sport participation of marginalized populations did neither refer to a theoretical framework nor specified the type of sport or physical activity, although determinants might be related to types of sport and therefore might differ, further limiting the validity of research outcomes.

A focus of this review was furthermore to provide an extensive overview of determinants identified as relevant for sport participation of people in marginalized populations within a socioecological framework. In line with Collins (2004) who argued that marginalized populations are very diverse and thus experience a variety of personal, mediating, and structural factors excluding them from participating in sport, this review identified a variety of general factors being associated with sport participation independently of the marginalized populations as well as factors being highly specific to marginalized populations. These combinations of determinants formed unique profiles regarding sport participation. Interestingly, the most commonly identified determinants were mainly independent of the marginalized population and covered all socioecological levels; SES, time, energy, work, domestic responsibilities, physical capital, and motivation as individual factors, social support, safety, and social networks as social factors, pricing as organizational determinant, transport and local sport facility provision regarding environmental influences, and funding of sporting opportunities on the political and societal level. These determinants and levels correspond to factors being identified as decisive for sport participation in the general population as well (Biernat et al., 2020; Giles-Corti & Donovan, 2002; Rech et al., 2018). Accordingly, there appear to be general individual, social, organizational, environmental, as well as political and societal aspects being, independently of the population, crucial for sport participation. Research, currently predominantly targeting determinants of sport participation within the general population (European Commission, Directorate-General for Education, Youth, Sport and Culture, 2014; Hovemann & Wicker, 2009; Trost et al., 2002), could therefore provide interesting insights into general factors, also for marginalized populations, which further could be targeted to facilitate sport participation. However, marginalized populations experience a variety of general but also very specific determinants considering their sport participation, making their sport experiences unique and their needs different from nonmarginalized populations (see e.g. O’Driscoll et al., 2014; Taylor, 2001). This is also reflected in the still lower sport participation rates of people in marginalized populations (Carroll et al., 2014; Duijvestijn et al., 2020; Hildebrandt et al., 2013; van Bottenburg et al., 2005) showing that current approaches are not sufficient to facilitate their partaking in sport. The conducted scoping review highlighted this by revealing that general determinants of sport participation come along with a variety of very specific factors which, to some extent, differ greatly between marginalized populations. Accordingly, many factors play a role regarding sport participation; while some aspects are relevant for all, the impact of other factors depends on the type of marginalization resulting in variable impact of the socioecological levels pending from the marginalized population. This underlines the necessity to look at marginalized populations specifically. Rawal et al. (2020) also stressed the need to get a more complex understanding of positive as well as negative factors regarding sport participation of marginalized populations to enable a more critical insight into research gaps. Thus, it is crucial to consider the specific circumstances and the complex combination of determinants that marginalized populations encounter regarding sport participation in order to understand why people in marginalized populations do or do not take part and to be able to facilitate their sport participation (O’Driscoll et al., 2014; Taylor, 2001). Sport participation in general and sport participation of marginalized populations, as shown in this review, is embedded in a complex interplay of individual, social, organizational, environmental, and political and societal factors (Bakhtar et al., 2019; Biernat et al., 2020; Rawal et al., 2020).

Therefore, a multilevel framework considering all socioecological levels enables a more comprehensive understanding of the factors that influence sport participation of people in marginalized populations as marginalized populations are often being marginalized at many levels (Saatcioglu & Corus, 2014: Sallis & Owen, 2015). Applying a socioecological framework further allows to scope and integrate existing fragmented research into a comprehensive whole as conducted in this scoping review. This is due to the fact that (socio-)ecological frameworks are considered as valuable contemporary meta-frameworks that allow the coherent organization and integration of different theories while not replacing these but placing them into a broader context (Sallis & Owen, 2015).

Research on marginalized populations, as indicated by the included studies, however, focusses on only a few determinant levels and does not pay sufficient attention to the social as well as political and societal level as the majority of included studies investigated individual factors. This review identified a number of social aspects, including social support, safety, social networks, and a variety of specific factors at the political and societal level, including funding, which are decisive in terms of sport participation of marginalized populations. It was also shown by other research that social aspects, especially social support (see amongst others Allender et al., 2006; Bakhtar et al., 2019; Wendel-Vos et al., 2007), as well as well thought out sport promoting policies and practices of governments (Ball et al., 2015) do have a substantial influence on sport participation. These outcomes highlight the potential of investigating and addressing these determinants in terms of facilitating sport participation of marginalized populations.

A further challenge originating in 2019 that needs to be mentioned here with regard to the inclusion of marginalized populations in sport nowadays is the Covid-19 pandemic. Though not yet being covered in the studies in this review as they were being conducted in the time just before the Covid-19 outbreak, the effects of the worldwide pandemic will most likely have a substantial impact on factors influencing sport participation of marginalized populations. The link of social inequality and Covid-19, including more negative pandemic effects for low educated people (Ohlbrecht & Jellen, 2021) and the deepening of social inequality (Grodach, 2020), have the potential to widen the social inequality gap. This must also be considered regarding sport participation along with the potential that the experienced barriers to sport participation will be even amplified. Accordingly, the impact of the Covid-19 pandemic needs to be considered in future research as it most likely even amplifies the marginalization of certain populations in terms of sport participation making the topic even more relevant in the near future.

Based on the insights of this scoping review, a number of implications for research as well as society can be derived. First, future research should overall widen their research focus not only looking at sport participation in general but also regarding sport participation of marginalized populations. Including diverse target populations, such as people with low socioeconomic status, people with disabilities, the elderly, ethnic minorities, refugees, the LGBTQ + community, homeless people, people with substance abuse, and criminals, the experiences these populations make regarding sport participation should be investigated nuanced and specific to their population. Therefore, applying a socioecological framework might be beneficial to investigate and understand the complex patterns and levels of sport participation determinants. Thereby, a focus on social as well as political and societal level determinants would add to currently still lacking research insights. Further points of interest for future research might be to investigate the role of multiple marginalization, as people often belong to more than just one marginalized population, and to quantify the relative level of impact each determinant and associated socioecological level has regarding sport participation of marginalized populations. Overall, the use of more diverse research approaches (quantitative, qualitative, mixed-methods) as well as longitudinal designs would add to current research and empirical knowledge. Following up on the advanced empirical knowledge on determinants influencing sport participation of marginalized populations, policies and initiatives aiming at facilitating sport participation of these populations should further overcome mainstream one-size-fits-all approaches towards tailor-made nuanced approaches. These adapted approaches might help to make sport accessible for everyone by taking the specific circumstances that people experience regarding sport participation into account supporting social inclusion of marginalized populations in the long-term.

A number of limitations need to be mentioned regarding this scoping review. One of them concerns the use of only two databases in the literature search, Science Direct and Web of Science: Social Sciences Citation Index (SSCI), which might limit the scope of included studies in this review. However, as these are large interdisciplinary databases, the range of identified papers is most likely still appropriate. A second limitation is due to the focus on current research trends of this review as the search was limited to papers being published between 2016–2020. Accordingly, future research might also look at developments in research over a longer period. A third limitation is given due to the decision to include a variety of marginalized populations and socioecological levels, which might lead to very broad outcomes. However, this decision was made to comprehensively scope the entire research landscape on this broad topic acknowledging that the diversity of marginalized populations as well as the diversity of sport participation needs to be considered to understand the context of sport participation for people in marginalized populations and to make sport inclusive for everyone. Furthermore, the disbalance of existing research regarding the focus on only a few marginalized populations makes it less reasonable to focus a scoping review on a specific marginalized population as it leads to neglect of the same (very) marginalized populations again, and the greatly lacking specification of sport types is not well-suited to review a specific sport type as not enough literature is provided. The broad scope, therefore, allows to point out future research directions. Future research thus might focus on specific marginalized populations and specific sport types to enhance the in-depth understanding of their sport participation.

Despite this limitation, this scoping review provides a better understanding of the current state-of-the-art in research on determinants influencing sport participation of people in marginalized populations and thereby reveals possible directions for future research. By using a socioecological framework to allocate all identified determinants to the respective socioecological levels, this paper further provides an extensive overview on the diverse factors influencing sport participation of marginalized populations. The applied socioecological framework and the associated differentiation of individual and contextual levels of sport determinants enable a comprehensive understanding of the complex interplay of overarching and specific factors influencing sport participation of marginalized populations and thereby may point out concrete aspects and starting points for future research and interventions to help make sport participation inclusive for everyone.

Acknowledgments

This work was supported by ZonMw under Grant 546003003.

Disclosure Statement

No potential conflict of interest has to be reported.

Additional information

Notes on contributors

Susen Lange

Susen Lange (M.Sc., TUD Dresden University of Technology, Germany) is a PhD researcher in the Department of Human Geography and Spatial Planning at Utrecht University and the Department of Industrial Design at the University of Technology Eindhoven (The Netherlands). She holds a master’s degree in psychology: Cognitive-Affective Neuroscience and her research focuses on sport participation of people in marginalized situations.

Gideon Bolt

Gideon Bolt (PhD, Utrecht University, The Netherlands) is an associate professor of urban geography at the Department of Human Geography and Spatial Planning in Utrecht (The Netherlands). He is one of the coordinators of the ENHR (European Network of Housing Researchers) working group Housing and Minority Ethnic Groups and is an editor of the Journal of Housing and the Built Environment. His publications focus on the link between residential segregation and integration, bridging and bonding social capital, policies aimed at reducing segregation, as well as the issue of social cohesion in a society that is increasingly diversifying.

Steven Vos

Steven Vos (PhD, KU Leuven, Belgium) is a full professor at the Department of Industrial Design at the University of Technology Eindhoven, and head of the research group ‘Move to Be’ at Fontys University of Applied Sciences. His main area of work is in profiling people during physical activity and sports participation in different settings and designing tailored services and products to support physical activity, sports participation, and vitality.

Beate Völker

Beate Völker (PhD, Utrecht University, The Netherlands) is a sociologist and director of the Netherlands Institute for the Study of Crime and Law Enforcement (NSCR). Her research focusses on social networks and social capital, how social relations come about in various meeting places or settings and what kind of resources they contain. She published a.o. in Social Networks, Network Science, Sociological Science and Nature Sustainability.

Appendix

Appendix 1. Overview of included articles.

No	Reference	Research Country	Marginalized Population	Sports Type	Sports Entity	Socioecological Determinant Level	Theoretical Perspective	Research Design	N	Title
1	Anderson et al. (2019)	Germany Netherlands	Refugee	N.s.	Prog/ orga	Individual Social Organisational Political & Societal	Socioecological	Mixed methods Cross-sectional	17	Managerial perceptions of factors affecting the design and delivery of sport for health programs for refugee populations
2	Blauwet et al., (2017)	US	Disabled	N.s.	Prog/ orga	Individual Environmental	N.s.	Quantitative Cross-sectional	134	Functional and Environmental Factors Are Associated With Sustained Participation in Adaptive Sports
3	Borland et al., (2020)	Australia	Disabled	N.s.	N.s.	Individual	N.s.	Quantitative Cross-sectional	305	Participation in sport and physical activity in adults with intellectual disabilities
4	Brown and Pappous (2018)	UK	Disabled	N.s.	Prog/ orga	Individual Social Organisational Political & Societal	N.s.	Qualitative Cross-sectional	30	The Legacy Element. … It Just Felt More Woolly: Exploring the Reasons for the Decline in People With Disabilities’ Sport Participation in England 5 Years After the London 2012 Paralympic Games
5	Chikafu and Chimbari (2020)	South Africa	SES	N.s.	N.s.	Individual	N.s.	Quantitative Cross-sectional	392	Levels and Correlates of Physical Activity in Rural Ingwavuma Community, uMkhanyakude District, KwaZulu-Natal, South Africa
6	Curran et al., (2016)	UK	Homeless &/or substance abuse	Soccer	Prog/ orga	Individual Environmental	N.s.	Qualitative Cross-sectional	34	The challenge and impact of engaging hard-to-reach populations in regular physical activity and health behaviours: an examination of an English Premier League ‘Football in the Community’ men’s health programme
7	Payán et al., (2019)	USA	Ethnic minority	N.s.	N.s.	Individual Social Environmental Political & Societal	N.s.	Qualitative Cross-sectional	111	Intrapersonal and Environmental Barriers to Physical Activity Among Blacks and Latinos
8	Eime et al. (2017)	Australia	SES	Team sports	Prog/ orga	Individual Environmental	Ecological	Quantitative Cross-sectional	488698	The relationship of sport participation to provision of sports facilities and socioeconomic status: a geographical analysis
9	Elasri-Ejjaberi et al., (2020)	Spain	SES	Sports centre	Prog/ orga	Individual	Economic	Quantitative Cross-sectional	7299	An Analysis of the Determinants of Sport Expenditure in Sports Centers in Spain
10	Garner-Purkis et al., (2020)	UK	SES	N.s.	Prog/ orga	Individual Social Organisational Environmental	N.s.	Qualitative Cross-sectional	35	A community-based, sport-led programme to increase physical activity in an area of deprivation: a qualitative case study
11	Garrett et al., (2020)	UK	SES	N.s.	N.s.	Environmental	N.s.	Quantitative Cross-sectional	18447	Urban nature and physical activity: Investigating associations using self-reported and accelerometer data and the role of household income
12	Guell et al., (2018)	UK	Elderly	N.s.	N.s.	Individual Social	N.s.	Qualitative Cross-sectional	27	Towards co-designing active ageing strategies: A qualitative study to develop a meaningful physical activity typology for later life
13	Hallmann et al., (2017)	Germany	SES	N.s.	N.s.	Individual	Economic	Quantitative Cross-sectional	22361	Leisure participation: modelling the decision to engage in sports and culture
14	Hoekman et al., (2016)	Netherlands	SES	N.s.	N.s.	Individual	Socioecological	Quantitative Cross-sectional	17910	Sport participation and the social and physical environment: explaining differences between urban and rural areas in the Netherlands
15	Ives et al., (2019)	UK	Disabled	N.s.	N.s.	Individual Organisational Environmental	Marginalised situation specific	Qualitative Cross-sectional	32	´I’ll always find a perfectly justified reason for not doing it’: Challenges for disability sport and physical activity in the United Kingdom
16	Jones et al., (2017)	UK	LGBTQ+	N.s.	N.s.	Individual Organisational Environmental Political & Societal	N.s.	Qualitative Cross-sectional	14	Barriers and facilitators of physical activity and sport participation among young transgender adults who are medically transitioning
17	Lenneis and Pfister (2017)	Denmark	Ethnic minority	N.s.	N.s.	Individual Social Organisational	Sociological	Qualitative Cross-sectional	26	When girls have no opportunities and women have neither time nor energy: the participation of Muslim female cleaners in recreational physical activity
18	López-Cañada et al., (2019)	Spain	LGBTQ+	N.s.	N.s.	Individual Social Organisational Environmental Political & Societal	Socioecological	Qualitative Cross-sectional	43	Socio-ecological analysis of trans people’s participation in physical activity and sport
19	Machón et al., (2020)	Spain	Elderly	N.s.	N.s.	Individual Environmental	N.s.	Quantitative Cross-sectional	634	Proximity to Facilities and Its Association with the Health-Related Habits of Functionally Independent Older Adults
20	McGovern, (2020)	US	Ethnic minority	N.s.	N.s.	Individual Social Organisational Political & Societal	Sociological	Qualitative Cross-sectional	31	The intersection of class, race, gender and generation in shaping Latinas’ sport experiences
21	Mohammadi, (2019)	Germany	Refugee	Cycling course	Prog/ orga	Individual Organisational	Sociological	Qualitative Longitudinal	7	Social inclusion of newly arrived female asylum seekers and refugees through a community sport initiativ---=p[ ;.e: the case of Bike Bridge
22	Mulderij et al., (2020)	Netherlands	SES	N.s.	Prog/ orga	Organisational	N.s.	Mixed methods Cross-sectional	19	Effective elements of care-physical activity initiatives for adults with a low socioeconomic status: A concept mapping study with health promotion experts
23	Ouyang et al., (2017)	China	Ethnic minority	N.s.	N.s.	Environmental	Environmental spatial	Quantitative Cross-sectional	313	Spatial deprivation of urban public services in migrant enclaves under the context of a rapidly urbanizing China: An evaluation based on suburban Shanghai
24	Phipps, (2019)	UK	LGBTQ+	N.s.	N.s.	Organisational Political & Societal	N.s.	Qualitative Cross-sectional	8	Thinking beyond the binary: Barriers to trans* participation in university sport
25	Reklaitiene et al., (2016)	Lithaunia	Disabled	Sports centre	Prog/ orga	Organisational Environmental	Environmental spatial	Qualitative Cross-sectional	28	The accessibility for people with disabilities - New challenge and possibilities for fitness and recreation services development
26	Robinson et al., (2016)	US	SES	N.s.	N.s.	Individual	Socioecological	Quantitative Cross-sectional	5236	The impact of neighborhood on physical activity in the Jackson Heart Study
27	Rowiński et al., (2017)	Poland	Elderly	N.s.	N.s.	Individual	N.s.	Quantitative Cross-sectional	3743	Participation in leisure activities and tourism among older people with and without disabilities in Poland
28	Saxton, (2018)	US	Disabled	N.s.	N.s.	Individual Social Organisational Environmental Political & Societal	Marginalised situation specific	Qualitative Cross-sectional	67	Hard bodies: exploring historical and cultural factors in disabled people’s participation in exercise; applying critical disability theory
29	Stevens & Cruwys (2020)	UK	Elderly	N.s.	N.s.	Individual	Sociological	Quantitative Longitudinal	3896	Membership in Sport or Exercise Groups Predicts Sustained Physical Activity and Longevity in Older Adults Compared to Physically Active Matched Controls
30	Symons et al., (2017)	Australia	LGBTQ+	N.s.	N.s.	Political & Societal	Marginalised situation specific	Mixed methods Cross-sectional	294	The impacts of discriminatory experiences on lesbian, gay and bisexual people in sport
31	Thibaut et al., (2017)	Belgium	SES	13 popular sports types	N.s.	Individual	Economic	Quantitative Cross-sectional	3005	Time and money expenditure in sports participation: The role of income in consuming the most practiced sports activities in Flanders
32	Úbeda-Colomer et al. (2019)	Spain	Disabled	N.s.	N.s.	Individual Social Organisational Environmental	Socioecological	Quantitative Cross-sectional	1219	Barriers to physical activity in university students with disabilities: Differences by sociodemographic variables
33	Vandermeerschen et al. (2017)	Belgium	SES	N.s.	N.s.	Organisational	Organisational	Quantitative Cross-sectional	580	Sports, Poverty and the Role of the Voluntary Sector. Exploring and Explaining Nonprofit Sport Clubs´ Efforts to Facilitate Participation of Socially Disadvantaged People
34	Vandermeerschen and Scheerder (2017)	Belgium	SES	N.s.	N.s.	Political & Societal	N.s.	Qualitative Cross-sectional	20	Sport managers’ perspectives on poverty and sport: The role of local sport authorities
35	Walker and Hayton (2017)	UK	Disabled	N.s.	Prog/ orga	Organisational Political & Societal	N.s.	Qualitative Cross-sectional	15	Navigating austerity: balancing "desirability with viability’ in a third sector disability sports organisation
36	Warbrick et al. (2016)	New Zealand	Ethnic minority	N.s.	N.s.	Individual Social Organisational	Psychological	Qualitative Cross-sectional	18	Provider, father, and bro - Sedentary Maori men and their thoughts on physical activity
37	Yamakita et al., (2020)	Japan	Elderly	N.s.	N.s.	Individual	N.s.	Quantitative Cross-sectional	23200	Association between childhood socioeconomic position and sports group participation among Japanese older adults: A cross-sectional study from the JAGES 2010 survey
38	Yi et al., (2016)	China	Elderly	N.s.	N.s.	Individual Social Environmental	Socioecological	Mixed methods Cross-sectional	1580	Associations between individual and environmental factors and habitual physical activity among older Chinese adults: A social–ecological perspective