Abstract
Objective Difficulty walking is a primary reason that individuals with knee osteoarthritis (OA) seek care. We examined the change in self-reported difficulty walking after participating in the Good Life With Osteoarthritis in Denmark (GLA:D) 8-week education and exercise program and assessed patient factors associated with improvement in difficulty walking.
Methods This was a registry-based cohort study of individuals in Denmark with knee OA who enrolled in GLA:D. Assessments were administered at baseline, program completion (~3 months), and 12 months. Our prespecified primary outcome was change in self-reported difficulty walking assessed using the EuroQol 5-dimension 5-level walking item. Exposures included sociodemographic factors, measures of OA illness severity, comorbidities, and psychological factors. In those with baseline moderate/severe difficulty walking, using multivariable regression analysis, we assessed the relationship between exposures of interest and improvement to no/slight difficulty walking.
Results We included 5262 participants. Of 2178 (41.4%) individuals with baseline moderate/severe difficulty walking, 51.4% and 58.3% reported no/slight difficulty walking at 3 and 12 months, respectively. Greater self-efficacy, younger age, female sex, lower BMI, less intense knee pain, and better function at baseline were associated with greater likelihood of improvement in difficulty walking, whereas severe difficulty walking at baseline and back pain intensity were associated with decreased likelihood of improvement.
Conclusion More than half of those with baseline difficulty walking experienced substantial improvement after completing GLA:D and this improvement was maintained at 12 months. Several patient factors were associated with the outcome, suggesting that some individuals may require additional support and extended treatment.
Self-reported difficulty walking, a complex construct that can be conceptualized as a relationship between a defined activity level and subjective perception of difficulty or effort,1 is a primary reason that individuals with knee osteoarthritis (OA) seek care, including joint replacement.2 Difficulty walking is also a precursor to disability and loss of independence3 associated with adverse health outcomes,4 including increased risk of cardiovascular events and all-cause death,5,6 and predicts future healthcare expenditure.3 Potential explanations include decreased physical activity and greater sedentary behavior, as people with knee OA often self-manage their OA-related symptoms by avoiding activities they find challenging,7 and associated decline in self-efficacy.8 Enabling people with knee OA to walk has the potential to improve pain and function,9 reduce presence and burden of comorbidities,10 improve independence,11 and augment overall quality of life (QOL).12 Therefore, addressing walking difficulty is a therapeutic priority with societal impact.
International guidelines strongly recommend exercise therapy and education as first-line treatments for knee OA.13 Good Life With Osteoarthritis in Denmark (GLA:D) is a structured group-based hip and knee OA treatment program that includes 2 to 3 patient education sessions and 12 supervised exercise therapy sessions delivered over 8 weeks by certified healthcare professionals14 and has been implemented worldwide.15 Cohort studies have shown that participation in GLA:D is associated with improvement in self-reported pain,15 physical function,15 QOL,15 and physical activity level,16 and that it is cost-effective from a healthcare payer perspective.17 It remains unclear to what degree participating in GLA:D can improve a person’s self-reported difficulties walking and whether there are differences in improvement based on patient-level factors.
Our objective was to determine the change in self-reported difficulty walking after participating in the GLA:D program, and to understand patient factors associated with improvement in difficulty walking.
METHODS
Study design and sample. This was a registry-based cohort study in Denmark using data from the GLA:D initiative. Individuals with knee OA who enrolled in GLA:D between July 2018 and May 2020 with complete follow-up data were included. We selected this period because of the availability of variables of interest. A detailed description of GLA:D has been previously published.14 In brief, it consists of a 2-day education course for therapists, and a treatment program consisting of 2-3 education sessions and 12 group exercise therapy sessions delivered twice weekly by a health professional; it is provided at no out-of-pocket costs to patients in Denmark. To be eligible for the program, participants needed to have joint problems consistent with knee and/or hip OA that have resulted in contact with the healthcare system and be able to communicate in Danish.
The reporting of this study was guided by the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) checklist.
Assessments. Standardized questionnaires were administered at baseline, program completion (approximately 3 months), and 12 months.
Exposures. We selected exposures a priori based on variables collected within the GLA:D registry that the research team hypothesized could affect perceived improvement in difficulty walking. We drew on a biopsychosocial model18 of self-reported difficulty walking where biomedical, psychological, and social factors may all contribute importantly to whether a person with knee OA improves following completion of an education and exercise program.19 Exposures of interest included the following:
Sociodemographic factors: age, sex, and education (postsecondary education [yes/no])
OA illness severity: OA pain assessed on a visual analog scale (VAS) from 0 mm (no pain) to 100 mm (worst pain imaginable) and 12-item short form of the Knee Injury and Osteoarthritis Outcome Score (KOOS-12) physical function subscale (0 [low] to 100 [high])
Comorbidities: BMI, total number of nonmusculoskeletal comorbidities (0-12), back pain severity on an 11-point numerical rating scale (NRS; 0-10), and total number of hips/knees affected by OA
Psychological factors: 5-level EuroQol-5-Dimension questionnaire (EQ-5D-5L) symptoms of anxiety/depression item, Arthritis Self-Efficacy Scale (ASES; 10 [low] to 100 [high], for both pain and other symptoms subscales), and fear that physical activity will damage joints (yes/no).
Outcome. Our primary outcome was the EQ-5D-5L walking (mobility) item. Participants were asked to indicate the 1 response that best described their health that day: (1) I have no problems in walking about, (2) I have slight problems in walking about, (3) I have moderate problems in walking about, (4) I have severe problems in walking about, or (5) I am unable to walk about. We collapsed levels 4 and 5 because of low prevalence in level 5 (n = 2). Our prespecified primary endpoint was 3 months. Our secondary endpoint was 12 months.
Analysis. We compared characteristics of participants at each level of baseline difficulty walking by ANOVA, Kruskal-Wallis, or chi-square test as appropriate. We also compared baseline characteristics for those with and without missing data for our primary outcome. We assessed the proportion of participants within each level of difficulty walking (no problems, slight problems, moderate problems, and severe problems/unable to walk) at baseline, 3 months, and 12 months and compared these proportions using McNemar chi-squared test. This analysis was repeated after stratifying by age (< 65 vs ≥ 65 years), sex (female vs male), and baseline knee pain (VAS ≤ 60 vs > 60). We also calculated the magnitude of change in walking from baseline to 3 months as the proportions with change of −3, −2, −1, 0, 1, 2, 3 levels on the EQ-5D-5L walking item (with level 4 and 5 collapsed) for the overall sample and by the level of difficulty walking at baseline.
Finally, we assessed the proportion of those with moderate and severe problems/unable to walk at baseline who improved at 3 and 12 months (defined as no or slight problems walking [yes/no]). In this subgroup, we used a multivariable logistic regression model with participants nested within clinics to assess the relationship between exposures and improvement in difficulty walking at 3 months. Only variables with a variance inflation factor < 4 were included to avoid multicollinearity.
Analyses were conducted in R version 4.2.1 (R Foundation for Statistical Computing). Associations are presented as odds ratios (ORs) and 95% CIs. Statistical significance was set at a 2-sided P value of 0.05.
RESULTS
Study sample and characteristics. Of the 9913 participants with knee OA who registered for GLA:D between July 2018 to May 2020, we included 5262 participants with complete baseline, 3-month and 12-month data (complete case analysis, 53.1% completion). Mean age was 65.8 (SD 8.9) years, 68.4% were female, and baseline mean knee pain intensity was 45.8/100 (SD 21.8) mm. Full baseline participant characteristics are shown in Table 1. Baseline characteristics of participants with and without missing data for our primary outcome were similar for most variables (Supplementary Table S1, available with the online version of this article). Those with missing data were less likely to have a postsecondary education and had worse scores on measures of psychological wellbeing, self-efficacy, pain, physical function, and QOL; however, the differences were small and unlikely to be clinically important.20
Baseline characteristics by level of difficulty walking.
Difficulty walking. At baseline, the proportions reporting no, slight, moderate, and severe difficulty walking/unable to walk were 25%, 33.6%, 33.4% and 8%, respectively. At 3 months, there was statistically significant improvement in difficulty walking; this improvement was maintained at 12 months. The proportions reporting no, slight, moderate, and severe difficulty walking/unable to walk at 3 months were 40.1%, 32.8%, 21.6%, and 5.5%, respectively, and 44.6%, 30.2%, 19.1%, and 6.1% at 12 months, respectively (McNemar test, P < 0.001 at both 3 and 12 months).
The movement of participants across levels of difficulty walking from baseline to 12 months is shown in the Figure. Similar improvement in difficulty walking was observed in younger (≤ 65 years) and older (> 65 years) participants, in those who were male vs female, and those with higher (VAS > 60/100) vs lower (VAS ≤ 60/100) baseline knee pain intensity (Figure).
Change in difficulty walking after completion of GLA:D overall and stratified by variables of interest. GLA:D: Good Life With Osteoarthritis in Denmark; VAS: visual analog scale.
The magnitude of change in difficulty walking from baseline to 3 months is shown in Supplementary Table S2 (available with the online version of this article).
Factors associated with improvement in difficulty walking. In individuals who reported moderate/severe difficulty walking at baseline (41.4% of total sample; n = 2178), the proportion reporting no or slight problems walking were 51.4% at 3 months and 58.3% at 12 months. Greater self-efficacy (ASES other; OR 1.01 per unit, 95% CI 1.00-1.02), female sex (OR 1.55, 95% CI 1.27-1.89), and better knee function (KOOS-12 function at baseline; OR 1.01 per unit, 95% CI 1.00-1.02) were associated with greater likelihood of improvement in difficulty walking at 3 months. Older age (OR 0.98 per year, 95% CI 0.97-0.99), higher BMI (OR 0.97 per kg/m2, 95% CI 0.95-0.99), more intense knee pain (OR 0.99 per unit, 95% CI 0.98-1.00), and more intense back pain (OR 0.96 per unit, 95% CI 0.93-1.00) at baseline were associated with lower likelihood of improvement. Those who reported severe difficulty walking at baseline (compared to moderate; OR 0.50, 95% CI 0.39-0.64) were also less likely to report improvement. The full model is presented in Table 2.
Multivariable logistic regression in those with moderate and severe/unable to walk difficulty walking at baseline (n = 2178, 41.4% of sample) assessing baseline patient factors associated with no and slight difficulty walking at 3 months.
DISCUSSION
In this registry-based cohort study of individuals with knee OA who had participated in GLA:D, we found that a large proportion (41.4%) reported at least moderate difficulty walking at baseline. More than half of these individuals experienced substantial improvement in their difficulty walking following completion of GLA:D, and this improvement was maintained at 12 months. This study suggests that self-reported difficulty walking is not fixed and may be improved after participating in a structured, group-based, exercise therapy and education program. It also identified factors associated with improvement in difficulty walking, informing who is most likely to achieve their goals and who might need added supports. Given that the ability to walk is instrumental to functional independence, an outcome prioritized by patients, and a predictor of future cost, these results highlight the need to ensure those with knee OA and difficulty walking are identified and receive evidence-based OA treatment.
Knee OA is among the most important causes of difficulty walking in older adults.19 To combat the personal and societal impact of difficulty walking, OA treatment able to improve the ability to walk needs to be prioritized. In the current study, many individuals who completed a short (8-week) education and exercise therapy intervention perceived they had less difficulty walking following the program. The combination of both increased OA knowledge, confidence in disease self-management, reduced pain, and functional strength as a result of GLA:D may address key components required to improve walking. These results are from an observational cohort and we are therefore unable to make causal statements. A further understanding of the effects of such a program would have important implications for patients and health professionals, in addition to healthcare systems.
This study provides insights into patient factors that are associated with walking improvement. Those with moderate/severe walking difficulty who were younger, with less pain and better physical function, and with lower BMI were more likely to improve at 3 months. These individuals may have found it easier to engage with and progress with the program in the relatively short time frame. Our data suggest individuals with some physical and psychological factors (eg, worse baseline OA symptoms, comorbid musculoskeletal disease [eg, back pain], obesity, and lower self-efficacy) may require additional support, such as additional supervised sessions or treatment, to see sufficient improvement. Given the heterogeneity of people with knee OA, future research should consider ways to individually adapt GLA:D and similar programs to help ensure a greater number of individuals can improve their OA symptoms, including walking ability.
This study has several strengths, including a large sample size afforded by the use of registry data that captures community-dwelling individuals with knee OA. Further, it begins to bridge the gap in literature around difficulty walking, an outcome important to patients.2
There are several limitations of this study. This was an observational study and effects compared to a control group should be confirmed in a clinical trial, since regression to the mean21 and contextual factors20 could account for improvements in OA symptoms, including difficulty walking. These results should be viewed as hypothesis-generating. Many participants did not have complete follow-up data and were excluded from our analyses; however, consistent with prior studies using data from the GLA:D registry,22 we found only small differences in characteristics between those with and without complete data. Given the time frame from which we included participants, some study participants may have been affected by lockdowns relating to the coronavirus disease 2019 (COVID-19) pandemic. GLA:D continued during the COVID-19 pandemic, and participant characteristics and pain and physical function outcomes for people with knee OA were similar during the pandemic compared to before and after.23 If there was any effect, we believe pandemic restrictions would bias our results toward the null. Those who seek out exercise therapy may share characteristics that make them more likely to improve compared to all individuals with OA. Finally, these data are drawn from within a single healthcare system. However, GLA:D has been implemented worldwide and prior studies have shown similar responses for pain, physical function, and QOL across multiple countries.15 Therefore, our findings are likely to be widely generalizable.
In conclusion, we found difficulty walking to be frequently reported by individuals with knee OA enrolled in the GLA:D program, and most improved by the completion of the program, with effects that extended 12 months after completion. We found that older individuals and those with greater disease burden were less likely to improve. Future research should evaluate whether added individualized support could increase the proportion of those who experience improvement in difficulty walking.
ACKNOWLEDGMENT
We wish to thank the GLA:D clinicians and participants for contributing to the data collection.
Footnotes
The initiation of GLA:D was partly funded by the Danish Physiotherapy Association’s fund for research, education, and practice development; the Danish Rheumatism Association; and the Physiotherapy Practice Foundation. These institutions had no role in the study design, collection, analysis and interpretation of data, in the writing of the manuscript, or in the decision to submit the manuscript for publication.
LKK reports support from Canadian Institutes of Health Research Doctoral Award, and grant funding from Physician Services Incorporated Resident Grant. JJY is a member of the leadership team (no financial interest) for the GLA:D International Network, a nonprofit organization aimed at improving access to guideline recommended treatments for osteoarthritis (OA) and low back pain. STS and EMR are the founders of GLA:D. GLA:D is a not-for-profit initiative hosted at University of Southern Denmark aimed at implementing clinical guidelines for OA in clinical practice. STS is associate editor of the Journal of Orthopaedic and Sports Physical Therapy and has received personal fees from Munksgaard, TrustMe-Ed, and Nestlé Health Science outside the submitted work. The remaining authors report no conflicts of interest relevant to this article.
- Accepted for publication May 30, 2024.
- Copyright © 2024 by the Journal of Rheumatology
