Physical activity and barriers and facilitators in patients with rheumatoid arthritis or spondyloarthritis: a cross-sectional study of 150 patients

Barriers and facilitators to physical activity in inammatory arthritis can be assessed through the Inammatory arthritis Facilitators and barriers questionnaire (IFAB) questionnaire. The objective was to measure the correlation between IFAB and self-reported physical activity levels. In this study, a majority (67%) of patients with inammatory arthritis reported a low to moderate level of physical activity and only 27% reached the recommended amount of physical activity by 7,000 steps per day. In this population of 150 patients with inammatory arthritis, a link was observed between a global score of barriers and facilitators and physical activity levels collected through IPAQ-S. This questionnaire could be a practical tool to use in clinical practice and in research to address perceived barriers and facilitators to physical activity in order to increase the physical activity levels of patients with inammatory arthritis.


Introduction
Physical inactivity has been identi ed as the fourth leading risk factor for global mortality around the world [1]. The positive effects of physical activity on health, wellness and reduced mortality are widely established and documented for all ages [2][3][4][5].
Patients with in ammatory arthritis (IA), such as axial spondyloarthritis (axSpA), rheumatoid arthritis (RA) or psoriatic arthritis (PsA) are more prone to physical inactivity than the general population [6,7]. Patients with IA derive speci c bene ts from regular physical activity [8][9][10]. In addition, patients with IA are at risk of other co-morbidities such as cardio-vascular diseases which can also be positively in uenced by physical activity [11,12]. Increasing physical activity in IA patients is a challenge [13]. Lifestyle changes should be addressed by a global approach taking into account behavioral barriers to increase chances of success [14].
Barriers and facilitators to physical activity are key elements to understand physical activity behavior in rheumatic diseases [15][16][17][18]. These elements can be classi ed as symptoms of the condition, social or physical environment of the person, and/or psychological status [19]. A Questionnaire for In ammatory arthritis patients assessing FAcilitators and Barriers to physical activity (IFAB) has been recently developed and validated in patients with IA [20]. The relevance of a score to assess barriers and facilitators would be increased if a link was shown with physical activity behaviors. Such a link would allow to consider alternative and enhanced approaches to physical activity assessment and interventions. Previous studies sought to determine which factors in uence physical activity levels in patients with IA. They showed a link between physical activity and general or arthritis-speci c barrier limitations but without considering a global score including barriers and facilitators [21][22][23].
The objective was to measure the correlation between barriers and facilitators, assessed through the IFAB questionnaire, and self-reported physical activity levels. We also explored other markers of physical activity, through stages of behavior change and steps per day indicated by smartphones.

Study design
Page 5/18 The ImBAIA study was an international, multicentric, cross-sectional study in a usual-care setting, performed in secondary and tertiary care hospitals in France (10 centres) and in Tunisia (one centre), between October 2019 June 2020 (ClinicalTrial NCT04426747) [24]. This study was approved by the ethics committee (CPP Sud-Est III, France, EudraCT 2019-A01413-54, methodology MR03 for noninterventional studies). All patients received at inclusion oral and written information, and oral consent was obtained. This report followed the STrengthening the Reporting of OBservational studies in Epidemiology (STROBE) statement [25]. Participants Inclusion criteria were: age above 18 years; de nite IA con rmed by the rheumatologist based on classi cation criteria: axSpA (referring to the Assessment of SpondyloArthritis international Society classi cation criteria)[26], RA (referring to the international classi cation criteria of RA) [27] or PsA (referring to the ClASsi cation of Psoriatic ARthritis (CASPAR) criteria)[28], with no restriction for comorbidities; ability to walk, having a smartphone compatible with apps that can track steps; ability to read and write in the language of the participating country.
Over the recruitment period, consecutive patients with de nite axSpA, RA or PsA who satis ed the inclusion criteria, seen in outpatient visits by one of the investigators, were invited to participate.

Outcomes collected
Barriers and facilitators to physical activity Barriers and facilitators to physical activity were measured through the IFAB questionnaire, validated in IA patients [20]. This questionnaire contains 10 items. The items are related to psychological status (N = 6), social support (N = 2), disease (N = 1) and environmental factors (N = 1). The total score ranges from − 70 to 70 with a higher score indicating a higher level of facilitators and/or a lower level of barriers. Results below − 5 were identi ed in the initial development as potentially justifying a targeted intervention [20].

Physical activity behavior
Different indicators of physical activity behaviors were used in this study.
Self-reported levels of physical activity were collected through the International Physical Activity Questionnaire Short form (IPAQ-S) and were analysed though the number of metabolic equivalent of task (MET) minutes per week (energy expended while performing various activities throughout the whole week) [29]. The level of physical activity was categorised as low, moderate and high following the IPAQ scoring protocol [30].
A simple question developed by the authors, assessing the feeling of doing enough activity was also used ("Do you think you do enough physical activity?") and was scored on a 0 to 10 numeric scale. Involvement in an active lifestyle was assessed through a questionnaire, the stage of behavior change regarding active lifestyle [31]. Active lifestyle is de ned in this questionnaire as 150 minutes of moderate activity or 75 minutes of intense activity per week, according to the recommendation of the World Health Organisation [1]. Stage of behavior change ranges 1 to 5, from precontemplation '(I do not engage in regular physical activity and do not intend to in the next 6 months') to maintenance ('I engage in regular physical activity and have been doing so for more than 6 months') [31].
The level of physical activity was also collected through mean daily steps per month on the last four full weeks through smartphone Apps, either installed by default (such as Health on iPhone or Samsung Health on Samsung) or installed by the patient (such as Runtastic or Fitbit) [32,33]. The results were selfreported into the patient CRF. The threshold of 7,000 steps per day has been used as a recommended level of physical activity for patients with chronic disabilities [34]. The following classi cation was used to interpret steps per day: <5000 = sedentary, 5000-7499 = low active, 7500-9999 = somewhat active, 10 000-12 499 = active and ≥ 12 500 = highly active [35].

General data collected
Other variables collected were socio-demographic data, as well as information about the underlying condition (type of IA, year of diagnosis of IA), current treatment and comorbidities, collected using the Functional Comorbidity Index (0 = no comorbidity; to 18) [36]. In patients with RA: the Disease Activity Score 28 (DAS28) was collected using the last available data [37]. In patients with axSpA, disease activity was measured through the Bath Ankylosing Spondylitis Disease Activity Index (BASDAI) and through the Disease Activity Index for Psoriatic Arthritis (DAPSA) for PsA patients [38]. Function was measured via the modi ed Health Assessment Questionnaire (mHAQ). Physician global assessment was assessed on a numeric scale (0-10).

Statistical analyses
Sample size calculation to demonstrate a link between the IFAB questionnaire score and IPAQ-S with a relative risk of 0.5, 144 patients were needed (with α 0.5 and β 0.20).
Descriptive statistics relied on mean (standard deviation, SD) and median values. The link between the IFAB and physical activity, using primarily the IPAQ-S, then exploring stage of behavior change and steps by smartphone, was tested using Spearman correlation. Both total score and individual items of the IFAB questionnaire were analysed.
Three multivariable linear regressions were performed, using as dependent variable the IFAB, and as explanatory variables, rst IPAQ-S (MET-minutes per week) and the stage of behavior change. Mean daily steps per month was not used as explanatory variable because not linked with IFAB score in univariate analysis. Other variables were included in the models based on statistical association in univariate analysis (p < 0.20) and on their clinical relevance: these were: age, gender, disease duration, physician global assessment and number of comorbidities. For each analysis, assumptions regarding linearity, homoscedasticity, and normality were checked. The -level of signi cance was set at 0.05.
The coherence of the 3 physical activity behaviors measures was assessed by the correlation or t-test between the IPAQ-S score and mean daily steps per month, stage of behavior and binary feeling of enough activity.
Statistical analyses were performed using R version 3.5.1. There was no imputation of missing data. However, for mean daily steps per month on the smartphone, a correction was applied for outliers.

Barriers and facilitators
The mean score of the IFAB questionnaire was 6.0 (SD 19.2), median was 4 (Fig. 1). The two most frequent barriers or facilitators were item 1 (presence or absence of symptoms, n = 111, 74%) and item 9 (knowledge that physical activity is good for my health, n = 110, 73%). A total of 39 (26%) patients scored less than − 5 and thus could justify a targeted intervention. The IFAB score was slightly negatively associated with physical function; and furthermore, with disease activity only for patients with psoriatic arthritis (online supplementary table 1). Multivariate analyses showed correlation between the IFAB score and number of comorbidities (estimate 3.6, p < 0.05), and physician global assessment (estimate − 2.0, p < 0.03).

Physical activity behaviors
Self-reported physical activity through IPAQ-S was moderate (although in the high range of the moderate category; i.e., between 600 and 3000): the mean MET-minutes per week was 2,837 (2,668), median 1,784 ( Fig. 2). Level of physical activity was high for 38% of the participants, moderate for 46% and low for 16%. In all, 56 (37%) patients reported having the feeling of doing enough activity, and 82 (54%) reported following the WHO recommendations for physical activity (54%) (stage of behavior: action and maintenance).
Physical activity was low when collected by smartphone through mean daily steps per month: mean 5,600 (SD 3,797), median 4578 with 27% walking over 7,000 steps per day (Fig. 2).
Link between the IFAB questionnaire and physical activity behaviors Univariate associations: The global score of the IFAB questionnaire was linked with 3 of the 4 parameters of physical activity: IPAQ-S (rho 0.28, p < 0.001), stage of behavior change (rho 0.35, p < 0.001) and the IFAB score was signi cantly higher in patients who felt su ciently active: 13.9 (SD 16.2) vs 1.3 (SD 19.4), p < 0.0001) ( Table 2). However, there was no correlation with steps on smartphones ( Table 2).

Discussion
In this population, we observed a link between a global score of barriers and facilitators and physical activity levels collected through IPAQ-S. A correlation was also observed in stage of behavior regarding active lifestyle and patients reported less barriers and/or more facilitators when feeling enough active. This study failed to show a correlation between barriers and facilitators and mean daily steps per month collected through patients' smartphones. This questionnaire could be a practical tool to use in clinical practice and in research. Addressing perceived barriers and facilitators to physical activity by using the IFAB questionnaire may therefore be key to increasing the physical activity levels of IA patients.
This study has strength and weaknesses. We observed a correlation of barriers and facilitators with selfreported physical activity but not with mean daily steps per month. This could be explained the fact that a wide variety of physical activities are not covered by steps per day through smartphone, such as swimming or arm movement [39]. A substantial part of the physical activity can be related to domestic activity such as gardening or cleaning. These activities are generally not well captured by the smartphone as steps, leading to an unrepresentative measure of physical activity [40]. In this study, only 61% of patients included had analysable data. This exclusion rate is due to participant recruitment methods related with the di culty of using the electronic form. The demographics of the study sample, such as being mainly White, middle class women, limits the generalizability of the ndings. This study included IA patients through 3 different conditions (axSpA, RA, PsA). These 3 conditions are the most prevalent in ammatory joint and spine diseases. They share common characteristics such as pain and fatigue, swelling in the joints or axial stiffness, systemic manifestations and can potentially lead to structural changes in joint or spine with loss of function [41,42]. This study used a validated questionnaire providing a global score.
In our study, we found that barriers that were most correlated with all parameters of self-reported physical activity were items related to physical condition (i.e., symptoms). This correlation was also observed in a cross-sectional study of Suh 2019 including 245 RA patients, where the 18-item Barriers to Health Activities Scale (BHAS) was compared with physical activity self-reported through IPAQ [43], and in Freid et al in 2020 including 108 IA patients [44]. c In this study, a majority of patients with in ammatory arthritis reported a moderate to low level of physical activity (62%) and only 37% patients reported having the feeling of doing enough activity. The mean daily steps per month collected with apps was low (5,600; SD 3,797) with 27% walking over 7,000 steps per day [34,35]. This means that only a minority of people with IA are undertaking the recommended amount of physical activity (7,000 steps per day) and are at risk of complications.
The low level of physical activity observed in this study is coherent with other studies[6, 45,46]. This is a major health concern as physical activity is associated with a decrease of cardiovascular risk, a decrease of disease activity and an increase of physical function [8][9][10][11].
A quarter of patients (26%) scored less than − 5 on the IFAB questionnaire (the lower 25% of the group) and could justify targeted intervention. The rst most frequent barriers or facilitators was the presence or absence of symptoms. Indeed, symptoms such as pain, fatigue and stiffness are highly prevalent in IA patients and lead to disability. Controlling symptoms might be a relevant strategy to enable regular physical activity. The second most reported determinant of physical activity was the knowledge that physical activity is good for health. This item was correlated with stage of behavior change and the feeling of being enough active (but not with IPAQ-S and mean daily steps per week) and could be an easy target to modify through patient education. This underlines the importance of physical activity education and its health bene ts.
Similar results were observed in other studies. Brittain et al in 2011 examined in 248 women with arthritis the link between barrier categories and participation in moderate physical activity [22]. They concluded that both arthritis-speci c and general barrier limitations were the strongest predictors of self-reported moderate activity. This study did not consider the implication of facilitators in participation to physical activity and did not use a global score. This link between arthritis-speci c, general barrier and selfreported moderate activity (GPAQ questionnaire) was also observed in a cross-sectional study of 96 RA patient [21].
In 2002, Bell et al studied in a cross-sectional study of 137 patients with in ammatory joint disease the link between physical activity objectively measured with thigh worn physical activity monitor and exercise believe questionnaire [23]. They observed that attending an exercise facility in the community and low role limitations due to physical health predicted low physical activity. Following the physical activity guideline was linked with low role limitations due to emotional problems, higher physical tness and healthier exercise attitudes and beliefs.
In conclusion, perceived barriers and facilitators to physical activity were correlated with physical activity, indicating that targeting patients with high barriers and low facilitators to physical activity could be an effective option to improve physical activity levels. Addressing perceived barriers and facilitators to physical activity by using the IFAB questionnaire may be key to increasing the physical activity levels of IA patients. Availability of data and material: The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.
Code availability: The codes used in R software during the current study are available from the corresponding author on reasonable request.