This study reports the outcomes of a secondary analysis from a C-RCT investigating the effects of adding self-monitoring of PA with a WAT on self-reported joint function and HRQoL in individuals with hip and knee OA. No statistically significant differences were observed between the groups in terms of changes during the study period for any of the outcomes. Both groups improved regarding pain and symptoms, but the changes were smaller than the minimal important change which previously has been established for KOOS subscale Pain (12.2 points) and HOOS subscale Pain (11.8 points) (35).
Impairments related to function and pain are common among people with hip and knee OA (4). This study investigates the interventions’ effect on joint related pain, symptoms, ADL, Sport/Rec. and QoL. As expected, participants had the lowest (worst) mean scores on the Sport/Rec. subscale and the highest (best) scores on the ADL subscale, reflecting more difficulties with activities such as jumping than rising from bed. These results are in line with a meta-analysis (focusing on knee OA only), reporting the highest scores on ADL and the lowest on Sport/Rec. (29). Compared to the meta-analysis, this study presents similar results on subscale Symptoms but better mean scores on the other four KOOS subscales (29).
The groups did not differ significantly in changes regarding outcomes throughout the measurement period. Additionally, subscales ADL, Sport/Rec. and QoL scores remained relatively stable during the measurements with, in general, no statistically significant differences according to the paired t-test. The PA levels among the participants also seems to have been relatively stable and high throughout the intervention period according to previously analysed WAT-data (intervention group only) and PROMS (21, 36). Both groups improved regarding subscale Pain during the study period, but the changes were smaller than the minimal important change identified for the HOOS/KOOS subscale Pain in a recent study (35). Rienstra et al (35) demonstrated that the HOOS/KOOS subscale Pain has adequate responsiveness in detecting minimal important change after joint arthroplasty but not after conservative treatment.
The average EQ-5D-3L index score for all participants in this study was 0.78, which is higher than other OA samples in Sweden reporting index scores between 0.65–0.67 (37, 38). Shalhoub et al. investigated factors related to HRQoL in people with OA (39) and found a negative association between OA pain and HRQoL. Furthermore, they also found that individuals with a higher educational level had higher HRQoL (39). In our study, most participants had post-secondary education, which might, in part, explain the relatively high EQ-5D-3L index and EQ-VAS scores. Participants mean scores on EQ-5D-3L and EQ-VAS remained stable throughout the 12-month study period, with no differences in changes between the groups. The change in EQ-5D-3L after participation in the SOASP has been investigated in previous research, where a slight increase from baseline to 3-month follow-up was reported (38). Nevertheless, that increase was only half of what is considered the minimal important difference for EQ-5D-3L index score in the knee OA population (score 0.15) (40). In our study, neither the SOASP alone nor the SOASP with the addition of self-monitoring PA with a WAT had any effect on HRQoL reaching the minimal important difference.
Both groups had slight improvements regarding joint-related pain and symptoms which may be related to their participation in the SOASP. In a previous study, reduced movement-related fear and increased OA self-efficacy were highlighted as mediators that improved pain and physical function in people with knee OA (41). In line with this, Åkesson et al. (42) showed that OA patients had an increase in empowerment (facilitating self-efficacy and autonomy) after participating in the SOASP. However, the absence of a control group receiving no treatment prevents us from drawing such conclusions. Natural fluctuations and regression to the mean are additional plausible explanations for the observed results (43).
The use of WATs in research has increased drastically during the last decade and PA-level is the most common outcome in research concerning OA and other populations (44). While insufficient PA can have negative health effects, excessive engagement in weight-bearing activities is not solely beneficial for people with OA (45, 46). Qualitative research on patients’ perspectives has, in fact, shown that WATs may be used as a guide to help in optimising PA to avoid a worsening of pain (47, 48). It is important to note that PA cannot cure OA or always alleviate symptoms but, for people with OA, being able to remain physically active has other important health-enhancing effects which should be emphasized.
The results in this study should be interpreted in light of some methodological limitations. While the randomised controlled trial design and the utilisation of well-established, reliable outcome measures are fundamental strengths, there are also several limitations that should be acknowledged. These limitations are akin to those observed in the preceding study within this project (21). The recruitment of participants through Facebook might have introduced a selection bias since it could lead to an overrepresentation of younger, white women (49). Participants were also highly physically active at baseline, about 40% already used a WAT and, as mentioned previously, their perceived joint function and HRQoL were higher than other OA cohorts (21, 36). Considering these factors, the possibilities for improvement during the study period was thus limited. Additionally, all participants were assigned the uniform step goal of 7,000 steps which might have been suitable for some but not all participants. In a qualitative study within this project, some participants said that the WAT facilitated and optimised PA while others said that they were unable to reach the step goal due to pain (47). They also expressed that the WAT used in this study were not able to adequately capture all types of PA, especially not activities such as bicycling or strength training, which was seen as a limitation. We suggest that future WAT- interventions for people with OA apply a more person-centered approach with individualised goals to help optimise PA level and type of activity. We also recommend the use of a WAT that can capture different types of physical activities since some OA patients might benefit from doing less weight-bearing activities such as bicycling or aquatic exercise.