The challenge of exercise (non-)adherence: a scoping review of methods and techniques applied to improve adherence to physical activity and exercise in people with inflammatory arthritis

Abstract Objectives The aims were to explore the nature of methods/techniques applied to improve adherence to physical activity (PA) and exercise in people with inflammatory arthritis and to identify whether studies were theory based and/or used behaviour change techniques (BCTs). Methods Searches were undertaken of English language articles within four databases: Embase, Medline, PsycINFO and Cochrane. Articles were included if they assessed adherence to a PA and/or exercise intervention. A narrative synthesis of the findings is reported. Results Of 1909 studies screened, 18 studies met inclusion criteria. Adherence was most frequently included as a secondary outcome. Reporting of adherence measures was poor, in that 13 studies did not use a validated measure of adherence, with only three validated measures being identified. The majority of studies were not theory driven (n = 13), although the health belief model was the most used theoretical framework (n = 5). Only two studies mentioned both theory and BCTs. Four studies reported components that were mapped onto BCTs, with goal setting being the most prevalent. Conclusion This scoping review found that adherence to PA and/or exercise interventions was rarely the focus of research, despite its importance in maintaining health in people with inflammatory arthritis. Where research has been conducted in this area, serious shortcomings were revealed, in that psychological theory, evidence-based BCTs derived from theory and valid adherence measures were not used to inform intervention design and target adherence, meaning that interventions were suboptimal. These results suggest that there is considerable room for improvement and that more high-quality research is required to investigate determinants of adherence and develop impactful interventions.


Introduction
Physical activity (PA) and exercise (defined as planned, purposeful PA, designed to improve or maintain physical fitness) are key management strategies for people with inflammatory arthritis (including SpA, RA and PsA). People with inflammatory arthritis are advised to complete 150 min of moderateintensity PA per week, with strengthening and flexibility exercises twice a week [1][2][3], but adherence to this guidance is often low [4][5][6]. The World Health Organization (WHO) defines adherence as 'the extent to which a person's behaviour corresponds with agreed recommendations from a healthcare provider' [7]. However, this definition has been refined for exercise adherence by Frost et al. [8] as 'the extent to which individuals undertake a prescribed behaviour accurately and at the agreed frequency, intensity and duration'. Adherence to PA and exercise can be difficult to measure, and much of the evidence base does not assess adherence as a primary outcome, but only as a secondary outcome [9].
There are particular barriers to participation in PA and exercise for people with inflammatory arthritis, and many spend the majority of their time engaged in sedentary behaviour, meaning that non-adherence is a significant challenge [10][11][12]. Patients' perceptions of facilitators and barriers to PA and exercise need to be understood better, and interventions need to be more tailored to address individual determinants of this behaviour [13]. Recent research suggests barriers and facilitators in people with inflammatory arthritis are related to psychological status, social support, disease level and environmental factors [14]. However, a more in-depth understanding of adherence to PA and exercise is required, because there was poor adherence to this even among those who had high adherence to medication [15].
Several studies have been designed to address this problem using interventions such as exercise prescription, patient education and behavioural counselling. However, systematic reviews of interventions have revealed variable levels of success, with limited exploration of the methods/techniques used to assess adherence [16][17][18][19]. Furthermore, it is difficult to determine which aspects of these interventions were effective or how and why they might have worked, because most have not applied theory or tested fidelity [20]. Poor reporting of intervention design makes it difficult to draw conclusions about the effectiveness of theory or to assess whether the correct theory was chosen [21]. Few studies have described using evidence-based behaviour change techniques (BCTs) [22], and psychological theory has not been used to inform selection of behavioural change targets [23,24].
Many interventions to increase PA and exercise in people with inflammatory arthritis have demonstrated limited application of psychological theory and/or poor reporting, making it difficult to draw conclusions about the best strategies to use [20]. A scoping review, defined as 'a form of knowledge synthesis that addresses an exploratory research question aimed at mapping key concepts, types of evidence, and gaps in research related to a defined area or field by systematically searching, selecting and synthesizing existing knowledge' [25], is therefore appropriate for inflammatory arthritis, because this can examine how research is conducted on a certain topic or field, identify key characteristics or factors related to a concept and analyse knowledge gaps [26].
The objectives of this scoping review were to explore the nature of techniques/methods applied to improve adherence to PA and/or exercise in people with inflammatory arthritis and to identify whether studies were theory based and/or used BCTs.

Methods
The PRISMA Extension for Scoping Reviews (Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA)-ScR) [27,28] was followed and reported accordingly (see Supplementary Table S1, available at Rheumatology Advances in Practice online). A protocol for this scoping review has not been published because it was not eligible for registration on Prospero.

Search strategy
Search terms included adapted MeSH, keyword and wild card terms located in the title or abstract that reflected disease and outcome (e.g. adherence/compliance to physical activity/exercise) taken from two previous systematic reviews on a similar topic of interest [19,29]

Eligibility criteria
Articles were included if they assessed adherence to a PA and/ or exercise intervention. A full list of inclusion and exclusion criteria for study inclusion is shown in Table 1 using the population, intervention, comparison, outcome(s) and study design framework.

Data collection and analysis
Selection of studies Primary screening was undertaken by the first coder (H.C.), with a random sample of 10% of studies cross-checked by a second coder (M.S.) at the screening stage, which resulted in a 0.90 kappa level of agreement (strong) between the two coders (3) because there were only four discrepancies. Raters discussed discrepancies and reached agreement on the final studies included for the review.

Data extraction and analysis
With the use of a study-specific data extraction table, information about each study (e.g. author, year of publication, country, study design), patient population, description of the intervention, details of adherence assessment (e.g. adherence measurement, validation and study outcome type) and involvement of theory or BCTs were extracted by the first coder (H.C.). A random sample of 10% of studies were extracted by a second coder (M.S.). Only published data have been extracted, with no further data requests or confirmation from study authors undertaken.
A narrative synthesis is presented to describe the methodology used to assess adherence to PA and/or exercise interventions with descriptive statistics [31,32]. Depending on the studies meeting inclusion, a mapping of the theory used and BCTs was also undertaken [22]. Given that the required data concerned methodology and reporting, there was no differentiation in how the data from either qualitative and quantitative studies were dealt with. A quality assessment or critical appraisal was not conducted, because the aim of this scoping review was to examine how research is conducted on adherence to PA and/or exercise interventions, to identify key characteristics or factors related to adherence assessment and to analyse knowledge gaps [31].

Study selection
Combined searches yielded a total of 1909 citations, of which 1676 remained after removal of duplicates, with 23 studies meeting inclusion at the title/abstract screening stage ( Fig. 1). At full-text screening and data extraction, 18 achieved final inclusion, with five studies being excluded for the following reasons: intervention not clearly described (n ¼ 1); no intervention, and adherence was not measured (n ¼ 1); full text not accessible (n ¼ 1); intervention did not involve PA/exercise (n ¼ 1); and duplicate study/sample (n ¼ 1).

Study characteristics
As shown in Table 2, the majority of included studies investigated RA (n ¼ 12, 66.7%), mainly from Europe and UK (n ¼ 13, 72.2%), with all studies published between 1999 and 2022. A variety of study designs were used, but the most common were randomized controlled trials or cross-over trials (n ¼ 4, 22.2%). The sample size ranged from 14 to 328, with a median size of 42.5, with most studies (n ¼ 14, 77.8%) representing a female majority of participants and with average age ranging from 21.54 to 63.6 years.

Synthesis of results from individual studies
Details of interventions, adherence assessments, theory and BCTs for each included study are presented in Table 3. A wide range of 14 interventions were described across the 18 studies, with most following physiotherapist-or occupational therapist-supervised moderate-to high-intensity exercise programmes (n ¼ 5, 35.7%) or using predominantly educationbased programmes (n ¼ 4, 28.6%). Only one intervention was delivered exclusively online [32], with other interventions incorporating dance-based exercise [43] and Nordic walking [45] as PA and/or exercise.
Adherence was most frequently included as a secondary outcome (n ¼ 7, 38.9%), with only five studies reporting it as their primary outcome (27.8%). Most studies used the term adherence, with three studies using the term compliance (16.6%). Reporting of adherence measures was poor, with most studies not using a validated measure of adherence (n ¼ 13, 72.2%), typically using a study-specific measure (n ¼ 9, 69.2%) or simply presenting a descriptive statistic (usually a percentage or frequency) of those completing the intervention/course (n ¼ 4, 30.8%), often only in the Discussion. The validated measures of adherence were as follows: joint protection behaviour assessment [38][39][40]; using ecological momentary assessment to capture frequency data alongside group attendance [46]; and a definition of adherence to standard exercise therapy stated as exercising for 30 min per day and performing back exercise on 5 days per week based on previous literature [47].
The majority of studies were not theory driven (n ¼ 13, 72.2%); however, of those five that mentioned theory, all used the health belief model as the theoretical framework, either explicitly mentioned, alluded to or in conjunction with self-efficacy theory. Only two studies mentioned both theory and BCTs (10.5%) [39,44]. Four studies in total (21%) mentioned components that can be mapped onto BCTs [33,39,40,44], although they did not use the terminology of BCTs, ranging from 1 to 13 BCTs (median ¼ 4.5) across studies, with goal setting the most common (n ¼ 3, 75%). All types of clinician-guided or self-directed behaviour change interventions (defined as coordinated sets of activities designed to change specified behaviour patterns [30], which may or may not include recognized behaviour change techniques) Treatment groups must have received intervention content that has the aim of changing participant behaviour (e.g. any effort by the health-care professional or researchers to change, or support change, of a behaviour); these might include, but were not limited to, goal-setting activities or behaviour monitoring Intervention descriptions must include a specific, measurable prescription of physical activity or exercise (i.e. a set of planned, structured and repetitive movements to be followed for the duration of the intervention) No physical activity or exercise intervention clearly described

Comparison
Not applicable: studies with or without comparison groups included Outcomes Self-reported measure of adherence to physical activity and/or exercise at the end of the intervention Outcomes could be reported as exercise diaries, questionnaires, levels of physical activity by any validated measure (e.g. monitoring device, i.e. step-count, accelerometer) Other types of adherence, such as medication adherence Adherence not explicitly reported

Study design
Randomized controlled trials, quasi-experimental trials, prospective cohort studies, retrospective cohort analyses and before-after trials that reported baseline and follow-up measurements of adherence to physical activity and/or exercise or physical activity/exercise levels in at least two groups, including qualitative Laboratory studies using animal models or cells Conference abstracts The challenge of exercise (non-)adherence It is important to note that three studies/interventions were exemplified across seven papers [34, 35, 38-40, 44, 50], with some papers representing follow-up (both qualitatively and quantitatively). This is particularly significant given that Hammond et al. [38][39][40] assessed adherence in a more theory-driven and rigorous way, with the use of BCTs to inform their intervention, compared with the rest of the literature. However, the level of detail in reporting was inconsistent across publications for the same/similar studies, particularly regarding theory usage, reporting of BCTs and interventions.

Summary of evidence
Eighteen studies met inclusion criteria and had data extracted and analysed as part of this scoping review. A narrative synthesis was completed to describe the methodology used to assess adherence to PA and/or exercise interventions using descriptive statistics. Interestingly, adherence was reported as a primary outcome in only five studies and was most frequently included as a secondary outcome. Although the term adherence was used most commonly, reporting of adherence measures was poor, with most studies not using validated measures of adherence. In addition, many studies did not underpin interventions with theory, and the five that did all used the health belief model, sometimes in combination with selfefficacy theory. Four studies mentioned components that could be mapped onto BCTs [33,39,40,44], although they did not always apply proper BCT terminology, ranging from 1 to 13 BCTs across studies. Goal setting was the most commonly used BCT, but only one paper [39] specified that goalsetting behaviour was used, whereas two others [33,44] did not specify whether goal-setting behaviour or outcome was used. However, there was inconsistent reporting of this across publications, even within the same or similar studies, particularly regarding theory usage and reporting of BCTs and intervention components.
The scoping review reported similar shortcomings to Fenton et al. [20], because limited application of to draw conclusions about the best strategies to use to increase adherence to PA and exercise in people with inflammatory arthritis. However, this scoping review has added to the existing literature and advanced understanding by finding that psychological theory and evidence-based BCTs derived from theory have not been engaged to inform intervention design and target adherence. Furthermore, the included research did not distinguish between the initiation and maintenance of PA or exercise, which might be influenced by different determinants, and importantly, most studies did not use a validated measure of adherence.

Implications of the scoping review
There are implications of this review for both researchers and clinicians. It is clear that researchers need to design interventions that are theory based, then to identify the specific BCTs impacting on PA-or exercise-adherent behaviour in order to change this behaviour, if they are to be effective [51]. There is also potential to improve future studies by using valid and reliable measures of adherence as primary or secondary outcomes; for example, the exercise adherence rating scale (EARS) [24,52]. This measure has been widely validated, used across different populations and translated into several languages [53][54][55]. Clinicians could be trained to use effective BCTs [56] and could also use brief measures, such as the EARS, to improve and assess the success of their treatment, because adherence to PA and exercise is an important issue for people with inflammatory arthritis [11,15].

Strengths and limitations
An important strength of this scoping review was that it followed the PRISMA Extension for Scoping Reviews (PRISMA-ScR) process [27,28]. In addition, to check for accuracy, a random sample of 10% of studies was extracted by a second coder (M.S.). One of the limitations of this scoping review is that no quality appraisal of studies was completed. However, this was not relevant to the aims of this scoping review, which was designed purely to examine how research was conducted on adherence to PA and/or exercise interventions, to identify key characteristics or factors related to adherence assessment and to analyse knowledge gaps [26]. It is also possible that some relevant research was not assessed, because only English language papers and published data were included, with no further data requests or confirmation from study authors undertaken. A further limitation is that a librarian or information specialist was not consulted when developing the search strategy, and therefore some key terms might have been missed, although this is unlikely given the authors' experiential knowledge. The challenge of exercise (non-)adherence  The JP group education programme consisted of four weekly 2-h sessions, plus an optional home visit within 2 weeks of the end of the programme. It was led by an experienced rheumatology occupational therapist covering RA, drug treatments, diet, exercise, pain management, relaxation and joint protection Joint protection behaviour assessment: performances of 20 tasks when making a hot drink and snack meal were assessed as incorrect, partly correct or correct joint protection methods, with scores converted to percentages. A higher score indicates increased adherence Yes Primary outcome Instruction on how to perform the behaviour, demonstration of the behaviour, feedback on behaviour and problem-solving, behaviour practice/rehearsal, habit formation, information about health consequences No [41] The active group received a multidisciplinary education programme, including training in home-based exercises and guidelines for leisure physical activity. The control group received a booklet added to usual medical care Compliance with home-based exercises was defined as a practice rate 30% of the prescribed training. Compliance with leisure physical activity was defined as 20% increase in Baecke questionnaire score. Additional assessments involved possible predictors of compliance and changes with regard to the compliance Exercise compliance assessment at a given visit. The compliance rate for home-based exercise was measured as described [33]. The mean weekly practice was calculated as the proportion of self-reported mean weekly number of exercises to total number of exercises included in the home-based programme. To be compliant, each participant had to have a compliance rate 30%, meaning at least a daily mean practice of a set of three different exercises whatever the exercises performed and have disrupted training for <1 month before the 6-month follow-up visit and <2 months before the 12-month follow-up visit Leisure physical activity compliance was measured by comparing the baseline and follow-up (6-or 12month) level of leisure physical activity as assessed by the Baecke questionnaire. Given that identification of a minimal clinically important difference is lacking for the Baecke score, we decided that compliant participants had to have increased their score by 20% over that at baseline. The challenge of exercise (non-)adherence

Conclusions
This scoping review found that adherence to PA and/or exercise interventions was rarely the focus of research studies, despite the importance of PA and/or exercise to maintaining health in people with inflammatory arthritis. Where research has been conducted in this area, serious shortcomings were revealed, because in many studies psychological theory and evidence-based BCTs derived from theory were not used to inform intervention design and target adherence, meaning that interventions were suboptimal. Furthermore, reporting of intervention components and choice of adherence measures was poor, with most studies not using validated measures of adherence. These results suggest that there is considerable room for improvement in this area and that more high-quality research is required to investigate the determinants of PA and/ or exercise adherence and develop targeted interventions to enhance it in people living with inflammatory arthritis. Researchers and clinicians should use valid and reliable measures and carry out theory-informed research that targets adherence accurately using appropriate BCTs, in order to improve the outcome and provide better support for people living with inflammatory arthritis.

Supplementary data
Supplementary data are available at Rheumatology Advances in Practice online.

Data availability statement
The secondary data generated that support the findings of this study are available from the corresponding author upon reasonable request.

Funding
No specific funding was received from any bodies in the public, commercial or not-for-profit sectors to carry out the work described in this article.
Disclosure statement: The authors have declared no conflicts of interest.