Current evidence on exercise for osteoarthritis is largely inconclusive, showing negligible or short-term small effects that are similar to, or less effective than, other treatments, according to an overview of systematic reviews and randomized controlled trials published in RMD Open by Tim Schleimer and colleagues.
Exercise is recommended as a first-line therapy across osteoarthritis subtypes. However, the investigators wrote that their findings “question its universal promotion and highlight the need to revisit research priorities and clinical discussions around its worthwhileness.”
Study Details
Data were sourced from MEDLINE, Embase, Epistemonikos, PEDro, Cochrane, and trial registries. Searches for eligible systematic reviews covered records from inception through Oct. 31, 2025, and supplementary searches for randomized controlled trials extended through Nov. 8, 2025.
Reviews were eligible if they compared exercise with placebo, no intervention, or other treatments for pain and function in osteoarthritis. Additional trials were included to update areas where evidence was inconclusive.
A total of 5 reviews involving 100 unique randomized controlled trials across three treatment comparisons—exercise vs placebo (n = 1), no treatment/usual care (n = 3), and education (n = 1)—were included. These encompassed 8,631 patients with knee (n = 2), hip (n = 1), hand (n = 1), and mixed osteoarthritis (n = 1).
Researchers also included 28 supplementary trials (knee/hip: n = 23; hand: n = 3; ankle: n = 2) totaling 4,360 patients.
Two independent reviewers performed data extraction and risk-of-bias assessment. All outcomes were standardized to a 0–100 scale and reanalyzed using random-effects meta-analysis. Certainty of evidence was evaluated using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) framework.
Exercise Outcomes
The findings suggested small, short-term effects of exercise compared with placebo (mean difference = −10.8) and no treatment (mean difference = −12.4) for osteoarthritis pain, although the certainty of evidence was very low and the effects were smaller in larger or longer-term trials.
Moderate-certainty evidence indicated negligible effects for hip osteoarthritis (mean difference = −6.7) and small effects for hand osteoarthritis (mean difference = −10).
Evidence of varying certainty suggested that outcomes were comparable to those of education, manual therapy, analgesics, injections, and arthroscopy.
Single trials in selected populations found that exercise was less effective in the longer term than knee osteotomy (mean difference = 12.4) and joint replacement (knee: mean difference = 17.1; hip: mean difference = 24.2).
The investigators noted considerable uncertainty in the evidence base, including ongoing publication of small trials that may distort effect estimates, limited data on long-term durability, and limited evidence for less-studied osteoarthritis subtypes.
“Research should focus on priority questions with sound methodology to reliably inform policy decisions,” they wrote. Clinical practice, they added, should consider discussing with patients the worthwhileness of exercise relative to other treatment options, incorporating preferences, harms, and physician expertise.
The researchers reported no specific funding and no competing interests.
Source: RMD Open
