Barriers and facilitators of physical activity participation for young people and adults with childhood-onset physical disability: a mixed methods systematic review

Search strategy

A systematic search of the literature was conducted. Seven electronic databases (Embase, MEDLINE, PsychINFO, AMED, CINAHL, SPORTDiscus, and ERIC) were searched by one reviewer (GM) from inception to November 2019 to identify relevant original studies. The search strategy (Appendix S1, online supporting information) was designed to include key search terms relating to the research question: ‘physical disability’ (population), ‘attitudes’ or ‘barriers’ or ‘facilitators’ (phenomena of interest), and ‘physical activity’ (context) as well as relevant synonyms. Additional relevant studies were identified by manually checking reference lists of included studies and by citation tracking of included studies using Google Scholar. The search was limited to English language only as no funding was available for translation. No restrictions relating to participant age were used in the search strategy because of the lack of specificity within database options and our aim of capturing relevant literature relating to parents or other stakeholders. Instead, age restrictions were applied as an inclusion criterion.

Eligibility criteria

Original quantitative or qualitative studies, written in English, were included if they primarily investigated barriers and/or facilitators to physical activity, exercise, or physical recreation participation in young people or adults with childhood-onset physical disabilities. In addition to barriers and facilitators (and their synonyms), attitudes were included as a primary search construct because of their important influence on health behaviours such as physical activity. Studies were included if participants had: (1) a mean age of 15 years old or older, and (2) a primary diagnosis related to physical disability, where the disability was either congenital or acquired before the age of 15 years. Populations with associated or concurrent disabilities such as intellectual disability or vision impairment were included, providing the primary reported disability was physical. Studies that had investigated the experiences of parents or stakeholders such as carers, coaches, or policy makers were included, providing the population they reported on met the above participant criteria.

Studies were excluded if they did not include the population of interest, that is they: (1) investigated adult-onset progressive physical disabilities (e.g. multiple sclerosis, Parkinson disease), (2) did not define the age at onset for non-progressive conditions as less than 15 years (e.g. stroke, spinal cord injury), or (3) investigated populations with chronic, transient, or unstable medical conditions (e.g. cancer, heart failure). Studies that did not primarily investigate barriers, facilitators, or attitudes to physical activity participation (phenomena of interest) were also excluded; for example, studies that investigated associations between participant characteristics and physical activity, evaluations of interventions, or studies that explored physical activity experiences only. Studies were excluded if they primarily investigated non-physical activity related participation contexts such as sedentary activities (e.g. watching films), incidental daily activity (e.g. shopping), or other activities of daily living (e.g. cooking). Systematic and other literature reviews, conference presentations, and abstracts were excluded.

Eligibility criteria were applied to the titles and abstracts of the search yield independently by two reviewers (GM, and either NS or CW). When a decision on eligibility could not be made based on information in the title and abstract, a full text copy of that article was reviewed. Discrepancies between reviewers were discussed until consensus was reached. If consensus could not be reached, a third reviewer was consulted. Agreement between reviewers was measured using Cohen’s kappa statistic.⁴⁵

Assessment of methodological quality

Guided by the Cochrane recommended criteria for selecting a quality assessment tool⁴⁶ and the planned use of methodological quality within the data analysis, the McMaster Critical Review Forms for qualitative^{47, 48} and quantitative^{49, 50} studies were selected to assess methodological quality and relevance to the research question. The McMaster Critical Review Forms facilitate a narrative assessment of methodological quality³⁴ which can be used as a way of engaging with and better understanding the strengths and limitations of the primary studies.⁴⁶ These forms have previously been used to assess the quality of studies within mixed methods systematic reviews^{33, 34, 51-53} as they allow for the application of a rating criteria, such as that developed by Imms,⁵¹ to interpret methodological quality alongside processes such as data analysis. Qualitative studies were rated on the four widely accepted criteria of rigor: credibility, transferability, dependability, and confirmability. Quantitative studies were rated on three criteria: sample, measurement, and analysis. Using the rating criteria developed by Imms,⁵¹ each criterion was scored on a three-star system, where one star indicated no evidence of the study meeting the criterion, two stars indicated some evidence of the study meeting the criterion or unclear reporting, and three stars indicated evidence of the study meeting the criterion. Two reviewers (GM and CW) independently assessed methodological quality. Disagreements were discussed until consensus was reached. If consensus could not be reached a third reviewer (NS) made the final decision.

Data extraction

The following data from included studies were extracted onto a standardized spreadsheet developed for the review: study design, country, year, participant age, sex, disability type, recruitment setting, method of data collection (e.g. questionnaire, interview, focus group), and reported attitudes, barriers, and facilitators. For qualitative studies, reported factors were extracted in the form of raw data (e.g. quotes) and reported themes and subthemes. For quantitative studies, all listed attitudes, barriers, or facilitators were extracted along with results (e.g. proportions, percentages). Data extraction was completed by one reviewer (GM) and checked by a second reviewer (CW). Descriptive statistics used to summarize data on participant age, sex, and diagnosis are reported in Table 1.

Data analysis

A convergent integrated analysis framework was employed.⁴² This framework is used when qualitative and quantitative data synthesis occurs concurrently, and is the predominant analysis framework used in mixed methods systematic reviews.^{42, 54} A convergent integrated approach involves integrating quantitative data with qualitative data through a process of data transformation where data are ‘transformed’ into the same format (i.e. ‘quantitizing’ or ‘qualitizing’).^{38, 42} In our review, data transformation involved ‘qualitizing’, where quantitative data were converted into ‘textual descriptions’ such as themes, categories, or narratives, to allow integration with qualitative data. ‘Qualitizing’ is preferred to ‘quantitizing’ as it is less error-prone than attributing numerical values to qualitative data.^{38, 42}

Qualitative synthesis was completed using a thematic synthesis approach.^{39, 42} This approach involved three stages: (1) thematic analysis, where integrated data were inductively coded into descriptive categories and themes, (2) thematic synthesis, where categories and themes were organized according to a conceptual framework, and (3) an interpretive synthesis of the thematic analysis and thematic synthesis results.

Thematic analysis was completed inductively, with extracted physical activity factors from all studies coded line by line into descriptive categories and themes based on participant perceptions of their influence on physical activity participation. A constant comparative approach was employed to compare and contrast emerging ideas across studies, with emphasis given to the studies of better methodological quality followed by frequency of occurrence.

After this, themes were organized according to the Capability, Opportunity, Motivation – Behaviour (COM-B) model, the framework selected for thematic synthesis (Fig. S1, online supporting information).^{55, 56} The COM-B model considers factors contributing to a health behaviour, in this case physical activity participation, according to the interaction between three psychological domains: capability (psychological and/or physical), motivation (reflective and/or automatic), and opportunity (physical and/or social).⁵⁶ While the use of this model as a framework for synthesis is an alternative to the frameworks proposed a priori, the analysis approach remained consistent with the thematic synthesis method proposed. The COM-B was selected as it recognizes that health behaviour is part of an interaction of systems and components that can be charted on to the behaviour change wheel which can aid clinicians, organizations, and policy makers to identify specific targets for interventions.^{56, 57} Thematic analysis and synthesis were completed by all researchers (GM, CW, NS), with codes and developed themes labelled, discussed, and critiqued through a series of critical discussion meetings.

Finally, an interpretive synthesis phase^{39, 40, 58, 59} was completed to explore relationships between themes that arose during thematic analysis and thematic synthesis phases. The purpose of this interpretive synthesis was to ‘go beyond’ the primary studies to develop a more conceptual understanding of the attitudes, barriers, and facilitators of physical activity participation within the context of the existing literature.⁴⁰ The findings of the interpretive synthesis phase of analysis are grounded in the evidence of the contributing studies but result from an interpretation of the available evidence as a whole. The results of the interpretive synthesis were developed by the research team, through several iterations, which enabled the revision, broadening, and confirmation of interpretive concepts that captured the scope of the descriptive themes.

Search results

The search strategy yielded 3953 articles, of which 19 studies (1034 participants, 417 females) were included (Fig. S2, online supporting information). There was ‘moderate’ agreement between reviewers (Kappa score⁴⁵ 0.73, 95% confidence interval 0.60–0.79). The included studies came from seven countries, and employed qualitative^{18, 65} and quantitative^{37, 66-73} designs (Table 1). Qualitative studies predominantly used semi-structured interviews (n=8) with one focus group⁶⁰ and one photovoice study.⁶⁴ The quantitative studies all used cross-sectional surveys (n=9).

Characteristics of included studies

Ten studies included participants with various physical disabilities,^{21, 71} eight studies, accounting for 484 participants, included participants with cerebral palsy only,^{18, 23, 37, 62, 64, 66, 67, 73} and one study included 33 participants with muscular dystrophy only.⁷² The over representation of participants with cerebral palsy is not unexpected considering it is one of the most common motor disorders of childhood.⁷⁴ Based on the available data from 13 studies, the overall mean age of participants was 23 years.^{21-23, 69, 70, 72, 73} All studies, with one exception,⁶² included male and female participants. Fourteen studies included only young people or adults with physical disabilities,^{21-23, 73} with five studies including other stakeholders such as parents,^{18, 64} paid carers,⁶⁷ and recreation staff.^{62, 65} Participants were recruited from the general community,^{18, 23, 37, 62, 64, 67, 73} schools,^{70, 71} colleges,^{21, 22} sporting groups,^{61, 63, 65, 68} clinical settings,^{60, 69, 72} and elite level sport.⁶⁶ Eleven studies investigated both barriers and facilitators to physical activity.^{18, 71, 73} Four studies investigated barriers only^{37, 67, 68, 72} and one study investigated facilitators only.⁶² Five studies included data relating to attitudes towards physical activity.^{61, 66, 67, 69, 71}

Assessment of methodological quality

Five of the 19 studies (four qualitative,^{21, 22, 62, 63} one quantitative⁷¹) achieved the maximum quality rating on all assessment criteria. Emphasis was given to data reported in these studies followed by the frequency of reporting.

The overall quality of the 10 qualitative studies was higher than the nine quantitative studies. All except one qualitative study⁶¹ demonstrated evidence of trustworthiness through use of triangulation, member checking, detailed participant and setting descriptions, rich data reporting, clear methods, and the use of reflection and peer review. Seven quantitative studies^{37, 66, 67, 70-73} reported the sample population in sufficient detail; however, power calculations or sample size justification were rarely reported. Validity and reliability for outcome measures was reported in only four quantitative studies.^{66, 71-73}

Thematic analysis (stage 1 analysis)

An overarching theme of ‘finding the right balance’ emerged. Physical activity participation was perceived as ‘the right fit’ if predominantly enabling factors were experienced, or ‘all too hard’ if predominantly barriers were experienced. Inductive thematic coding of extracted data resulted in six subthemes, each with contrasting enabling or inhibitory aspects, as perceived by the participants (Fig. 1; Table S1, online supporting information). Data relating to attitudes were found to relate to other enabling or inhibiting factors and were therefore incorporated into the relevant subthemes. The interaction between subthemes, in addition to the concept of ‘finding the right balance’, lent itself to a scale metaphor where the themes within the capability, opportunity, and motivation categories combine to ‘tip the scale’ towards being physically active or not (Fig. 1).

Thematic synthesis according to COM-B framework (stage 2 analysis)

The six subthemes identified (Table S1) were organized according to the COM-B framework as follows.

Interpretive synthesis (stage 3 analysis)

An interpretive synthesis identified three critical elements were essential to finding the right balance for physical activity participation: social connectedness, social support, and the physical environment. These elements permeated the narratives of the included studies and ultimately shaped physical activity behaviour (Fig. 1). In particular, having social support and positive social experiences could outweigh other barriers that participants experienced.

Social support, which encapsulates attitudes, beliefs, and behaviours of all relevant stakeholders (family, peers, sport and recreation staff, programmes, organizations, and policymakers), strongly contributed to capability, opportunity, and motivation of young people and adults with physical disability to be active. The influence of social support extended to beliefs about benefits, fears, experiences, and prioritization of physical activity, and to the participants’ sense of self. In particular, family support facilitated physical activity, and this continued into adulthood. Universally, participants reported a lack of support for funding and programme availability.

Experiences of social connectedness acted to magnify enabling or inhibitory aspects of the emergent themes. The experience of positive social connections through inclusive behaviours of peers or staff alone could overcome other recognized barriers to participation. Positive social connections and a sense of belonging was considered to allay fears related to being physically active and acted as a strong motivator to participate and prioritize physical activity. Conversely, negative social interactions that contributed to feeling self-conscious or discomfort functioned as barriers to participation, despite other capability or opportunity facilitators being present.

The physical environment predominantly impacted opportunity factors. Inadequate community facilities, equipment, and transport were almost exclusively reported as barriers to participation. This in turn had a negative impact on capability and motivation, contributing to concerns and apprehensions about being physically active, and deepening feelings of isolation in physical activity settings.

Conclusion

We found physical activity participation for young people and adults with childhood-onset physical disabilities was primarily influenced by the social and physical environment. Positive social connection, availability of social support, and an appropriate physical environment were essential elements to ‘finding the right balance’ to be physically active. Interventions focusing on social and physical environmental modifications such as addressing societal attitudes, facilitating social connections, and improving physical accessibility should be the focus of future intervention studies for this population.