This study evaluated the amount of physical activity, according to intensity levels, in moderate-to-severe knee OA patients with and without DM, to confirm whether DM affects physical activity in this patient population. After adjusting for age, sex, and physique, the presence of DM and walking mobility (TUG test) affected step count and LPA time. MVPA time was not affected by the presence of DM or walking mobility but was impacted by contralateral knee-extension strength.
The prevalence of DM among knee OA patients was reported to be 14.4% in a meta-analysis by Louati et al., although the prevalence varied from 5.2–38.0% in the studies (since 2010) included in their analysis . The prevalence in the present study was 25.1%. This was higher than the reported value of 14.4% but was within the range of the studies reported since 2010, which indicates that the participants in our study were not a unique population and that the results can be generalized.
Our results showed that, in addition to the presence of DM, TUG time affected step count and LPA time. TUG time is an appropriate clinical measure of functional mobility and balance . Patients with balance-related disabilities most often exhibit impairments in standing balance. About half of these patients have slow walking speeds or are unable to perform the chair stand test . In arthritis patients, which includes the majority of OA patients, mobility and gait limitations are factors that can lead to insufficient activity . The above suggests that the decrease in mobility and balance as evaluated by the TUG test affected the step count and LPA time detrimentally.
Our results did not support the hypothesis that DM would have an effect on MVPA time. Rather, we observed that contralateral knee-extension strength influenced MVPA time. In previous research on the effect of knee-extension strength on MVPA time, increased MVPA time was associated with higher knee-extension strength in middle-aged women, but LPA time was not associated with knee-extension strength . Aoyagi et al.  reported that walking speed and knee-extension strength exhibit significant positive correlations with both daily step count and MVPA time in older people. Later research found that lower-limb muscle mass was more closely associated with MVPA time than with the daily step count . Furthermore, the participants in the present study had lower knee-extension strength on the affected side than on the contralateral side (Table 1), suggesting that activities focused on the contralateral side of the knee joint with fewer knee symptoms could help the activity of these patients. The above indicates that, in patients with moderate-to-severe knee OA for which TKA is indicated, MVPA cannot be maintained because of reduced contralateral knee-extension strength, rather than because of DM, as only contralateral knee-extension strength was associated with MVPA time.
The comparison between the groups with and without DM showed no significant differences in severity (KL grades) or age, and among the knee functions, only ROM was significantly more restricted in the DM group than in the non-DM group. The main factors affecting joint mobility in DM are increased stiffness of joint capsules, ligaments, and tendons, for which the underlying mechanism involves non-enzymatic glycosylation of collagen, accompanied by formation of advanced glycated end products. Reduced contractility of the muscle fibers is also an important factor. This is mediated by a mechanism that involves increased collagen content in the muscles , which limits knee ROM. However, Miner et al.  reported that the satisfaction and quality of life of patients who underwent TKA did not correlate with knee ROM, and Robertson et al.  reported that knee ROM at 1‒10 years after TKA was worse in patients with DM, even though the Knee Society Scores did not change. This may explain why knee ROM did not affect physical activity in the hierarchical multiple regression analysis in the present study.
Neither the DM group nor the non-DM group in the present study met the levels of physical activity recommended by international guidelines . Almost none of the participants were spending any time engaged in MVPA. Furthermore, LPA time was 32.9% lower and MVPA time was 44.4% lower in patients with DM than in those without DM. When a person's activity is limited to inside the home, they engage in about 4000 steps/day and less than 5 minutes/day of MVPA [44–46]. In the present study, the daily average step counts were 4,656 steps in patients without DM and 3,122 steps in those with DM, and MVPA time was 4.5 minutes/day in patients without DM and 2.5 minutes/day in those with DM. Both groups had less than 5 minutes/day of MVPA, and the figure was particularly low for patients with DM. The finding that the DM groups also took fewer than 4000 steps/day, suggests that knee OA patients with DM, in particular, are active only at home and highlights their high levels of inactivity. According to the American College of Sports Medicine, regular exercise can make people more active and extend life expectancy by controlling the onset and progression of chronic diseases and disabilities . The above indicates that increasing physical activity is important not only for reducing medical costs by delaying TKA, but also for lowering the overall burden of disease .
Regardless of the presence or absence of comorbidities, the main non-surgical approaches for treating knee OA are exercise therapies, such as muscle strength training, aerobic exercise, and balance training; the use of walking aids; and educating patients on healthy lifestyle habits, such as maintaining regular physical activity [49, 50]. The results of the present study also suggested that patient education and exercise therapies to strengthen the knee-extension muscles and improve balance are necessary to increase physical activity. Although knee OA patients with DM may have the ability to engage in LPA, the results of the present study showed that they have fewer step counts and shorter LPA time. Therefore, education on the necessity of exercise is particularly important for patients with moderate-to-severe knee OA who also have DM, as a way of encouraging them to increase their step counts and LPA time.
In addition, because the TUG times affects the step count and LPA time, physiotherapy programs focused on balance training and using walking aids for improving mobility could be useful. Furthermore, it has been reported that the TUG time was shortened when knee OA patients with DM performed dynamic resistance exercises for 12 weeks , implying that dynamic resistance exercises would also be useful for these patients.
Moreover, the present study found that MVPA time was affected by contralateral knee-extension strength, which is believed to represent the effects of functional limitations resulting from knee OA. Therefore, a physiotherapy program focused on muscle strength training for contralateral knee-extension strength could be useful for increasing MVPA.
A strength of this study is that the physical activity of patients with moderate-to-severe knee OA, with and without DM, was objectively measured using an accelerometer, which likely led to more accurate measurements of activity intensity than would have been possible by using a questionnaire, as was used in most previous studies that have examined physical activity. To the best of our knowledge, these has been no previous detailed description of physical activity in patients with moderate-to-severe knee OA that took DM into account. This study had several limitations that should be considered when interpreting the findings. First, accelerometers cannot track certain types of physical activity (for example, water-related activities and cycling) and cannot distinguish posture. Second, some participants had undergone TKA on the contralateral side. Third, because control of DM is one of the criteria for TKA, patients with severe DM, such as those with HbA1c levels > 8% and those being treated with insulin injections, were not included among the participants. Fourth, in addition to age, sex, and physical factors, physical activity has been reported to be affected by psychological factors, such as self-efficacy, psychosocial factors, and environmental factors [31, 45, 52]. As such, these factors may have contributed to the low adjusted R2 values in the hierarchical multiple regression analysis. Additionally, this was a cross-sectional study and thus could not identify causal relationships. In the future, we will endeavor to conduct a longitudinal study to identify and better understand determining factors that affect physical activity in patients with moderate-to-severe knee OA.