CLINICAL QUESTION
What exercise approach is best to lessen pain, increase function and gait performance, and improve quality of life in people with knee osteoarthritis?
BOTTOM LINE
Aerobic exercise produced the best outcomes for pain and function compared with other exercise approaches for people with knee osteoarthritis. The average effect is not large, which could mean that it will work for some people but not others (probably those who enjoy it and stick with it). (Level of Evidence = 1a–)
SYNOPSIS
The meta-analysis was conducted using the PRISMA process. The investigators systematically searched nine databases and three clinical trial registries, including the Cochrane Library, and identified 217 randomized controlled trials with 15,684 participants. The studies compared different exercise modalities in patients with knee osteoarthritis. The researchers used network analysis and employed a random effects model to compare the relative benefits of the treatments with one another. The risk of bias was high for 21% of the comparisons. For all interventions, on average, benefits were small and waned over 6 months. Aerobic exercise resulted in clinically relevant improvements in pain and function in the first 3 months and was better than other modalities (moderate certainty). Neuromotor exercises, such as dancing, walking heel-to-toe, and walking backward, produced a clinically relevant improvement in gait performance in the short term that did not last (moderate certainty). Mind-body exercises (eg, Tai Chi, Pilates) and flexibility exercises decreased pain outcomes in some patients in the short term. Strengthening did not seem to have a marked or consistent effect on outcomes. Study results were homogeneous and there was a low risk of publication bias.
Study Design: Meta-analysis (randomized controlled trials)
Funding Source: Government
Setting: Various (meta-analysis)
Reference: Yan L, Li D, Xing D, et al. Comparative efficacy and safety of exercise modalities in knee osteoarthritis: systematic review and network meta-analysis. BMJ. 2025;391:e085242.
Read the full article
Get immediate access, anytime, anywhere.
Choose a single article, issue, or full-access subscription.
Earn up to 14 CME credits per issue.
