In total 453 articles were found, of which 26 fulfilled the inclusion criteria of the review. The most common reasons for exclusion were not reporting on physical activity, exercise barriers, facilitators or benefits (see Fig. 1). An overview of the included studies relating to perceived barriers, facilitators and benefits of physical activity is presented in Tables 1, 2 and 3, respectively. The data presented in these tables reveal that studies vary with regard to the inclusion criteria of RA patients. For example, whereas some studies only report data from physician-diagnosed RA, others include those with a self-reported diagnosis of RA, or include patients with several types of arthritis, including RA. Additional analyses are reported for the ten studies that limited their recruitment to RA patients with a confirmed physician diagnosis of RA. Differences in sample size and assessment methods (quantitative or qualitative) are also evident from the tables. It is worth noting that, with one exception , all studies were conducted in Western European countries or the USA.
Perceived Barriers to Physical Activity and Exercise
Table 1 provides a detailed overview of the barriers that were identified by RA patients. Twenty-two studies were found, of which 12 used quantitative methodology [26, 27, 32–41], nine used qualitative methodology [24, 42–49] and one study reported on both quantitative and qualitative methods . Both qualitative and quantitative studies reported several barriers that are not specifically related to their disease, such as lack of time [34, 36, 42, 43, 47, 49, 51, 52] and the cost of exercise [39, 45, 47, 50]. In total, 15 (seven quantitative studies, eight qualitative studies) of the 22 studies that described barriers to physical activity and exercise reported at least one non-arthritis-specific barrier, similar to the barriers frequently reported by non-clinical populations. Studies that only included patients with a confirmed diagnosis of RA reported that lack of time was the most consistently reported barrier in both quantitative and qualitative approaches [33, 36, 42, 43, 47], followed by lack of motivation [37, 42, 48]. Even though the impact of these generic barriers should not be ignored, given the aims of the current review, the present results centre on the disease-specific barriers that are reported in patients with RA.
It is evident from both the quantitative and qualitative studies that pain (15 of 22 studies) and fatigue (12 of 22 studies) were the most commonly identified arthritis-specific barriers to participating in regular physical activity and exercise [26, 27, 33–43, 45–49]. Reduced mobility/functional ability (4 of 22 studies) and stiffness (2 of 22 studies) were other disease-related barriers mentioned as impeding physical activity participation [27, 35, 39, 41, 45, 46]. In addition to these physical barriers, which are reported in both quantitative and qualitative studies, qualitative studies also mentioned that a lack of provision of exercise programmes geared towards patients with arthritis [43, 45, 46] and a lack of knowledge about exercise regimens that are appropriate for patients with arthritis [24, 48] were perceived to negatively influence physical activity behaviour. This lack of knowledge regarding appropriate physical activity and exercise for RA has been related to fear of aggravating the disease or damaging joints [37, 43, 45, 46, 48]. Patients also felt that healthcare providers are unclear about the suitability of different types of exercise programmes for RA [26, 46, 48]. Similar results were found when analysing the studies that included only patients with a confirmed RA diagnosis. Pain was identified as a barrier by all eight studies [27, 33, 36, 37, 42, 43, 47, 48] and fatigue by seven of eight studies [27, 33, 36, 37, 43, 47, 48], with two qualitative studies reporting lack of advice from the healthcare provider as a perceived barrier to physical activity  and exercise .
Both quantitative and qualitative approaches were used to compare patients who participate in regular physical activity or exercise regularly and those who do not. These studies revealed no difference in perceived arthritis-specific barriers between the two groups in most [32, 37, 45, 46, 53], but not all [27, 44], studies. However, although the RA-related barriers appeared to be similar, qualitative studies showed a striking difference between the coping strategies between exercisers and non-exercisers. Whereas exercisers knew how to adjust their physical activity when experiencing a flare in disease activity or a high level of fatigue, those with insufficient levels of exercise were unable to do this [45, 46, 54]. Indeed, even when barriers were not different between exercisers and non-exercisers, self-efficacy for exercise was higher in those who exercise regularly [32, 55, 56]. Moreover, self-efficacy for exercise mediated the association between pain and exercise: pain was no longer associated with exercise when self-efficacy was taken into account . Finally, a quantitative study revealed that those RA patients who are more physically active also reported to have higher self-regulatory efficacy to overcome arthritis-related barriers to physical activity, while overall pain and number of flares were similar between patients with high levels of physical activity and those with low levels of physical activity . Thus, the majority of the studies suggested that exercising patients might not be different from inactive patients in terms of their perceived barriers, but exercisers are able to manage or overcome these barriers more effectively than inactive patients.
Perceived Facilitators for Physical Activity and Exercise
An overview of the facilitators for physical activity and exercise in patients with RA is shown in Table 2. In total, ten studies reported on facilitators, with the majority using qualitative methods [43–49, 58] and only two quantitative studies [39, 59]. Qualitative studies revealed that the most consistent RA-specific facilitating factor for regular physical activity and exercise was appropriate support from instructors and healthcare providers, which was reported in seven of nine studies [43–48, 58]. Similar findings were reported in the three studies that included only physician-diagnosed RA patients [43, 47, 48]. Social support or encouragement from family and friends (five of ten studies) also facilitated patients to participate in regular physical activity and exercise [43, 45, 46, 49, 58]. Indeed, those who currently exercise reported more support from family and friends than those who are inactive . In addition, the experienced or expected positive physical effects (e.g. reducing stiffness and increasing strength) as well as the psychological effects (e.g. happiness) were important facilitators for regular physical activity  and exercise [44, 59]. It is important to note that the most frequently reported facilitating factors were consistently linked to barriers to regular physical activity and exercise. For example, reducing stiffness was a facilitator for physical activity and exercise, whereas experiencing stiffness was also mentioned as a barrier. Similarly, support from a healthcare provider was mentioned as a facilitator, whereas lack of this support was reported as a barrier. An exception is social support from significant others, which was only occasionally mentioned as a barrier . Finally, it should also be acknowledged that social support was not a facilitating factor that is specific to patients with RA, but it is also often mentioned in other populations [60, 61].
Perceived Benefits of Physical Activity and Exercise
A variety of RA-specific and generic benefits of participating in regular physical activity and exercise have been reported, as presented in Table 3. Of the 11 studies, five applied quantitative methods [25–27, 35, 39] and six used qualitative methods [24, 43, 45, 46, 58, 62]. Reported benefits did not differ between the quantitative and the qualitative studies. Physical activity and exercise were perceived to be an important contributor to symptom management, as mentioned in eight of ten studies [24, 27, 35, 43, 45, 46, 58, 62], such as pain relief  or distraction from pain , improvements in joint function [27, 35, 45, 46] and increased energy . Together, these have a positive impact on daily tasks [35, 58]. These physiological benefits were also reported in studies only including patients with a physician diagnosis of RA [25, 27, 43, 62]. Feelings of independence and taking control were important perceived psychological benefits of physical activity and exercise [27, 35, 39, 45, 46, 58]. Similar to the experience of barriers, there did not seem to be a difference in perceived benefits between those who exercise and those who do not [53, 55], which is in line with the overall perception that RA patients are aware of the benefits of exercise in general and specifically for people with RA [23, 26, 48]. It is possible, though, that for inactive patients with RA, the perceived benefits are related to theoretical knowledge, whereas in those who are physically active the perceived benefits reflect the actual experience of such benefits. A recent study showed that, even though functional ability and social benefits of exercise were similar, those who regularly participated in exercise-related activities reported a broader range of physical and psychological benefits of regular exercise than those not regularly exercising .
Barriers and Benefits Related to Physical Activity Behaviour
Even though barriers and benefits to physical activity and exercise in patients with RA are well-described, less is known about associations between physical activity behaviour and perceived barriers/benefits or the confidence to overcome these barriers (i.e. barrier efficacy). Perceived barriers were predictive of levels of physical activity or exercise in some [40, 41], but not in all, studies . Care should be taken when interpreting and comparing these results, as different approaches have been used to quantify barriers in the literature. Whereas some studies evaluated barriers in terms of identification as well as the perceived impact of the barrier (i.e. how limiting is this barrier) [40, 41], others only measured the presence of a barrier . Interestingly, levels of physical activity were associated with barriers when perceived impact as well as presence were taken into account [40, 41], but the presence of a barrier itself was not associated with exercise behaviour . Therefore, it seems that barrier efficacy (i.e. the confidence to overcome a barrier) is a key aspect when exploring physical activity and the obstacles to regular engagement in physical activity. This is in agreement with the quantitative and qualitative studies comparing barriers between those who exercise regularly and those who do not reported above. The perceived barriers are similar between these groups of patients. Nevertheless, those who exercise have developed methods to overcome the indicated challenges. In other words, even though the barriers still exist in exercising patients, the impact of the barriers on physical activity and exercise behaviour is substantially reduced [32, 37, 45, 53, 54]. It should be acknowledged that these studies were not restricted only to RA patients; therefore, these findings need to be confirmed specifically in patients with a physician diagnosis of RA.
Perceptions of the benefits of physical activity have been shown to be positively related to participation in physical activity or exercise in most [34, 53, 63], but not all, cross-sectional studies . In addition, lack of perceived benefits of regular physical activity was associated with physical inactivity  and, perhaps unsurprisingly, patients who complied with home exercise regimens reported more perceived benefits of exercise than those who did not comply with the exercise regimens . However, it is worth noting that adherence to an exercise programme was not predicted on the basis of perceived exercise benefits prior to programme onset [55, 67] or self-reported physical activity post-intervention . Given that the patients included in these studies were all about to start a physical activity or exercise intervention, they are likely to rate the benefits of exercise higher than the general RA population. However, this suggestion remains speculative, as a direct comparison between the perceived benefits in those about to start an exercise intervention and those of the general RA population is not possible due to the different methods used to quantify the benefits of exercise in existing studies. It should also be noted that different methods have been used to define and quantify physical activity and exercise (e.g. semi-structured interviews, questionnaires), which can influence the findings. For example, Greene and colleagues  made a distinction between leisure physical activity and household physical activity. Outcome expectations were associated with household physical activity; however, this association was not apparent for leisure physical activity or total physical activity . In an observational longitudinal study, it was specifically leisure time physical activity and not work-related physical activity that was associated with improvements in functional ability in people with arthritis . Therefore, the research to date suggests that modalities of physical activity are differentially related to (perceived) benefits of physical activity and perhaps exercise. This premise warrants further examination in patients with arthritis. As before, only a few studies have restricted their inclusion criteria to RA patients with a confirmed diagnosis; therefore, further studies are needed to explore these associations in this particular population.
Little is known about the interactive or additive effects of barriers or benefits in predicting physical activity or exercise. Multivariate path analyses revealed that only perceived benefits were associated with physical activity participation, with perceived barriers and health status not linked to exercise after controlling for potential modifying factors such as age, education, pain and disease duration . Two further studies have explored the interactive effects of individual barriers in predicting exercise. Fatigue influenced the association between a combined measure of generic, non-arthritis-specific benefits and barriers with exercise participation. In the presence of high levels of fatigue, other barriers and benefits were not related to exercise, whereas when the levels of fatigue were low, generic barriers and benefits were associated with exercise . Similarly, Der Ananian et al.  reported that when taking physical limitations into account, pain was no longer related to exercise levels, providing evidence for physical limitations as a mediating factor in the associations between exercise and pain. Thus, the existing evidence indicates that relationships between barriers and exercise behaviour are complex. Therefore, when examining predictors of exercise behaviour, the interaction between individual perceived barriers and/or benefits should be taken into account.