Moderators of the effect of therapeutic exercise for knee and hip osteoarthritis: a systematic review and individual participant data meta-analysis

Key Points

1. Exercise for treatment of hip and knee osteoarthritis provided low to moderate improvements in pain and disability.
2. Treatment effects were higher for patients with higher levels of pain and disability.

BACKGROUND & OBJECTIVE

Randomized controlled trials have consistently shown that therapeutic exercise is beneficial for pain and physical function in patients with hip and knee osteoarthritis (OA). However, outcomes are variable, and many patients receive no benefit. Current research provides no good explanation for why exercise helps some patients and does not help others.

This study sought to shed light on this question by identifying individual patient-level moderators of the effect of therapeutic exercise on knee and hip OA.

Randomized controlled trials have consistently shown that therapeutic exercise is beneficial for pain and physical function in patients with hip and knee osteoarthritis.

Effect sizes for exercise for hip and knee osteoarthritis are small enough to ask questions about whether exercise provides a meaningful long-term clinical benefit to the average patient.

METHODS

• The authors identified 31 randomized controlled trials with individual patient data assessing the effect of exercise on people with knee or hip OA.

• These studies used 37 different exercise methods, including general aerobic, strengthening, and yoga.

• The studies also differed in the tracking of individual patient level data, which almost always included age, pain, disability level and BMI, and less commonly included self-efficacy, joint condition, physical activity, and education.

• A systematic review and meta-analysis was performed to identify potential moderators of the effects of exercise on pain and disability.

RESULTS

• Compared to non-exercise controls, exercise on average reduced pain 6.36 points, and improved function 4.4 points, both on a 100-point scale. Positive effects declined over timeand were larger in the short term (12 weeks) than the long term (12 months).

• The primary moderators for these effects were baseline pain and physical function - the patients with higher baseline levels of pain and disability experienced more benefit from exercise. This moderating effect was strongest at short term and less at long term.

• There was no evidence to suggest that age, BMI, physical activity, arthritis, self-efficacy, mental wellbeing, comorbidity, muscle strength (quadriceps), education, pain duration, or radiographic joint structure moderated the effect of exercise on pain and physical function in the short, medium, or long term.

• The authors concluded that the clinical benefits of exercise for hip and knee OA were of questionable clinical benefit, and more likely beneficial to patients with higher levels of pain and disability.

LIMITATIONS

• Some potential moderators were measured inconsistently, so there may have been insufficient data to detect their effect. Some potential moderators were not measured at all.

• The RCTs used different exercise methods, and the results may have been different for each method.

• The effect sizes in this study were smaller than previous studies, and this may have been due to the exclusion criteria used to select the studies.

CLINICAL IMPLICATIONS

Systematic reviews and meta-analyses of randomized controlled trials have consistently shown that therapeutic exercise is beneficial for pain and physical function in patients with hip and knee OA (1). A wide variety of exercise methods have been shown to be effective (2).

However, effect sizes tend to be small to moderate and decline over time (2). Further, there is considerable individual variability in patient response, with at least 50% of patients receiving no benefit (2).

Current research does not provide a good explanation for this variability, because post-hoc analyses have low statistical power to detect treatment- effect moderators (3).

This study sought to identify treatment moderators by examining individual patient-level data from 31 different RCTs examining the effect of exercise on hip and knee pain.

The main findings were that: (1) patients with higher levels of baseline pain and disability received more benefit; and (2) effect sizes are small enough to ask questions about whether exercise provides a meaningful long-term clinical benefit to the average patient.

+STUDY REFERENCE

Holden M, Hattle M, Runhaar J, Riley R, Healey E, Quicke J et al. (2023) Moderators of the effect of therapeutic exercise for knee and hip osteoarthritis: a systematic review and individual participant data meta-analysis, The Lancet Rheumatology,5(7), E386-E400.

SUPPORTING REFERENCE

1. Goh SL, Persson MSM, Stocks J, et al. Efficacy and potential 
determinants of exercise therapy in knee and hip osteoarthritis:
a systematic review and meta-analysis. Ann Phys Rehabil Med 2019; 62: 356–65.

2. Teirlinck CH, Verhagen AP, Reijneveld EAE, et al. Responders to exercise therapy in patients with osteoarthritis of the hip: a systematic review and meta-analysis. Int J Environ Res Public Health 2020;
17: 7380. 


3. Sun X, Ioannidis JP, Agoritsas T, Alba AC, Guyatt G. How to use a subgroup analysis: users’ guide to the medical literature. JAMA 2014; 311: 405–11.