Comparison of physical activity among different subsets of patients with knee or hip osteoarthritis and the general population

Hip and knee osteoarthritis (OA) are among the most common joint conditions, with treatment being predominantly symptomatic and focused on controlling pain, improving function, and health-related quality of life. In all stages of knee or hip OA, promotion of general physical activity (PA) is, in parallel with joint-specific exercises, considered to be a key component in the conservative management, including the trajectory after total joint arthroplasty (TJA) [1‐6]. PA is defined by the World Health Organization as “any bodily movement produced by skeletal muscles that requires energy expenditure” [7, 8]. According to this definition, PA is not only restricted to exercise but also comprises any activity in any domain of daily life, e.g. commuting, work activities, cycling, gardening, household activities and sports [9].

PA has multiple potential benefits in patients with knee or hip OA, as it has proven to play a role in improvement of pain, physical function, mobility, and weight management [1, 6, 10‐13]. In addition, PA is considered an important preventive measure for other chronic diseases (e.g. cardiovascular disease) associated with OA [11, 14‐16].

The proportion of OA patients meeting public health recommendations for health-enhancing PA varies largely in the literature, from 13 to 60% [14, 17‐19]. This variation could probably be due to heterogeneity in participants, settings, monitoring devices and methods across studies [17]. Currently, studies that compare the amount of PA between early and advanced stages of knee and or hip OA, including TJA, and the general population are scarce [17, 19, 20]. Regarding the comparison of PA in patients before and after TJA, a recent systematic review and meta-analysis concluded that PA does not change 6 months post-TJA compared with preoperative levels [20, 21]. Moreover, these studies concluded that PA after TJA was less compared to that of healthy controls [20‐22]. On the other hand, Meessen et al. found that patients following TJA were more physically active compared to the general population [23]. In none of the aforementioned studies, the nature of performed activities in different stages of the disease was presented.

It is conceivable that pain is a barrier for performing PA, and pain increases in more advanced stages of OA. Therefore, one would expect that PA decreases over the course of knee or hip OA [24, 25] and that the amount of PA of patients with OA is lower than that of the general population. On the other hand, we expect that patients after total joint arthroplasty (TJA) perform more PA than patients with advanced OA but without indication for surgery, since TJA results in less pain and improved function [14, 17].

Therefore, the aim of the present study was to deepen the insight in terms of amount and nature of self-reported PA levels in different stages of the disease (i.e. patients in primary and secondary care and post-TJA) and to compare PA characteristics of patients with knee or hip OA with those of the general Dutch population.

The aim of this study was to document the amount and nature of PA among patients with OA in different stages of their disease and to compare PA characteristics of patients with knee or hip OA with those of the general population. On average, patients reported to be physically active for at least 5.1 h per day, and to spend about one-third of PA in at least moderate-intensity PA. We found that OA patients after TJA are on average more physically active than patients in secondary care. No other relevant differences across different stages of osteoarthritis (OA) and the general population were found.

We found differences in the nature of PA that patients perform in different stages of OA and the general population: patients after TJA report more often low-impact activities (e.g. aerobic exercise (general exercises at moderate-intensity) and cycling) than other OA patients and the general population. On the other hand, swimming, also considered a low-impact activity, is most often reported by patients in secondary care. A possible explanation for the differences in the nature of PA is that high-impact sports such as running and contact sports are discouraged for patients after TJA [33‐35]. Our findings are in line with a meta-analysis concluding that patients after TJA return more often to low-impact activities than high-impact activities [36]. This may indicate that patients after TJA follow rehabilitation advice.

International guidelines recommend to perform at least 150 min per week of PA on at least moderate intensity for general health benefits [8, 37]. The WHO states that there is evidence for additional health benefits up to 300 min per week [8]. However, recommendations about the optimal amount and intensity of PA for OA patients are lacking. Apart from general health benefits, it is well known that increasing PA has positive effects on pain in OA patients [1, 10], suggesting a dose–response relationship. Currently, there is debate about the exact shape of the dose–response relationship between performed PA and health benefits [38‐40]. Some research groups assume a (curvi)linear relationship implying the more PA the better [37, 41, 42], whereas other research groups assume an optimal range beyond which health benefits may be partially lost [38‐40]. Epidemiologic studies on the optimal dose of PA are, however, mainly performed in cardiovascular patients and physically active volunteers rather than in OA patients. Therefore, future studies should assess the exact relation between PA and actual health benefits in OA patients to determine the optimal dose of PA.

We could not confirm our first hypothesis that the duration of PA decreases over the course of OA. Subgroups in the present study were based on the assumption that primary care patients (recruited trough searching GP electronic patients records and advertisements in local newspapers) and secondary care patients (recruited after referral by a rheumatologist to self-management programme of a hospital) reflect increasing rates of severity, rather than upon well-accepted criteria (i.e. Kellgren and Lawrence classification or joint space width). Our results are in line with a meta-analysis, showing no clear differences in PA between mild or moderate and severe OA [17]. A possible explanation could be that patients with more severe knee or hip OA substitute activities so that PA can be performed despite their complaints. However, our findings are contradictory with results of a longitudinal study in over 1200 OA patients showing that the amount of PA decreases with 11% over the course of 4 years [24]. The observed decrease in PA over time in the latter study could be explained by the influence of ageing; higher age is associated with lower PA levels [20, 43]. Future studies should focus on unravelling the influence of ageing and severity of the disease on PA in OA patients.

Additionally, we could not confirm our hypothesis that patients in primary and secondary care were less active than the general population. Our results are in line with a recent study showing comparable levels of objectively measured activity of at least moderate intensity for patients with knee OA and the general population [44]. On the other hand, our findings are inconsistent with a study concluding that patients with end-stage OA were less active than controls [45]. However, it is likely that in the latter study more severely affected patients were included. Our findings suggest that having OA in the early and more advanced stages does not impact PA levels, and that patients with OA manage to stay as active as the general population. However, prospective longitudinal research is needed to study the impact of symptoms related to OA on actual PA levels.

This study confirms our second hypothesis that patients who underwent TJA are more physically active than patients in secondary care. Regarding the effect of TJA on PA, recent studies showed no differences in preoperative PA levels compared to 6, 12 and 24 months after TJA, as well as compared to matched controls [20, 21, 46]. There are several possible explanations that patients following TJA spent more time on PA in our study. First, TJA results in less pain and improved function [14, 17]. Second, in our sample, 50% of patients received physical therapy for more than 3 months after TJA [28]. Therefore, these patients could be better instructed and motivated to perform PA. Another explanation could be that relatively active patients with end-stage OA receive TJA, and will therefore resume their old PA level easily.

This is the first study that compares the amount of PA in different stages of OA and the general population. This study also reports activities that are actually performed in different stages of OA and in the general population. However, data must be interpreted with caution, since this study has some potential limitations. Compared to the literature, we studied a relatively physically active cohort of patients [17, 43, 47], however, comparison with studies utilizing objective measures of PA (e.g. accelerometers) should be done with caution. Despite the regular use of self-administered measures, questionnaires tend to overestimate PA [30]. The SQUASH takes walking and cycling into account in three different modules; commuting, recreational and as a sports activity, and overlap in reporting these activities cannot be excluded. In particular, overestimation of the amount of PA spent on cycling, a common daily activity in the Netherlands, is likely. On the other hand, all datasets in this study used the same questionnaire assessing PA, i.e. the SQUASH, and, in our view, our results regarding differences between groups are valid. A note of caution is due here since we compared baseline data of four separate, previously published studies of different clinical settings in different time periods. Due to the heterogeneity of the subjects included in the separate studies, it is possible that there is some overlap in patients’ characteristics among different subsets of OA patients. The subgroups comprising OA patients in different stages of disease were relatively small, and this could result in lack of power. Although we adjusted for age, sex and BMI, we cannot rule out possible confounding of other factors such as education, profession, and comorbidities not included in our analysis. However, sample sizes were appropriate to detect differences of 10% or higher.

In conclusion, this study showed that patients with different subsets of OA and the general population spend on average considerable time on PA and that at least one-third of time active was spent on PA of at least moderate intensity. On the one hand, we found no major differences in the amount of PA among OA patients in primary, secondary care and the general population, and on the other hand we found substantial differences in the amount of PA between patients in secondary care and post-TJA patients. Although we found that PA levels for patients in different subsets of OA and the general population were comparable, it is well known that PA has beneficial effects on OA symptoms. Thus, continued efforts are needed to enhance PA in patients in different stages of OA. More research is needed to assess the exact relation between PA (in terms of duration and intensity) and actual health benefits in patients with OA over time.