Walking, cycling, and swimming have long been recommended for people with knee osteoarthritis, but until recently, no large-scale analysis had determined whether aerobic exercise actually outperforms other types. A network meta-analysis published in The BMJ now provides the most comprehensive answer yet: across 217 randomized controlled trials involving 15,684 participants, aerobic exercise was the only modality that showed benefit across every measured outcome, including pain, physical function, gait speed, and quality of life.
No other category of exercise, not strengthening, flexibility, mind-body, neuromotor, or mixed programs, matched that consistency. Some improved pain but not function. Others helped gait speed without meaningfully changing quality of life. Aerobic activity was the sole modality that delivered across the board.
What the review measured and found
The BMJ analysis used network meta-analysis methods to compare six categories of exercise against control groups and against one another. Researchers pooled data from trials conducted across multiple countries, with most interventions lasting roughly eight to twelve weeks. The statistical approach allowed comparisons between exercise types even when they had not been tested head-to-head in the same trial, a method that strengthens conclusions when direct comparisons are scarce.
A clinical commentary published by NEJM, reviewing the BMJ analysis, described aerobic exercise as the only modality demonstrating benefit across all outcomes at the evaluated timepoints. The commentary highlighted walking, bicycling, and swimming as practical examples, all low-cost, widely accessible, and requiring no specialized equipment or clinical supervision. (The full commentary may require a subscription to access.)
These results align with existing international guidance. The Osteoarthritis Research Society International (OARSI) already positions exercise as a first-line non-surgical intervention for knee osteoarthritis in its 2019 guidelines, stratifying recommendations by patient phenotype and comorbidities. What the BMJ review adds is specificity: among exercise types, aerobic activity rises to the top.
Where the evidence has limits
The finding is strong, but it is not bulletproof. A 2025 overview of systematic reviews published in RMD Open raises broader concerns about the exercise research literature, arguing that exercise effects in osteoarthritis may be small and transient. That analysis flags publication bias and small-study effects as persistent methodological problems in the field. These concerns do not directly contradict the BMJ findings, but they do raise a fair question: do the benefits observed in controlled, supervised trials hold up in everyday life over months and years?
Most of the trials pooled in the BMJ review measured outcomes over weeks, not the longer timeframes that matter most to people managing a chronic, progressive condition. Durability of benefit remains a genuine gap in the evidence.
Subgroup data is also thin. The OARSI guidelines acknowledge that patient responses to exercise vary by body mass index, disease severity, and coexisting conditions like diabetes or cardiovascular disease. The BMJ review does not break down aerobic exercise benefits along these lines. For a clinician trying to advise a specific patient, someone with advanced joint damage or severe obesity, for instance, the broad conclusion that “aerobic is best” may not capture important individual variation.
Cost-effectiveness data presents another blind spot. Exercise is generally cheaper than drugs or surgery, but no economic analysis accompanies the BMJ review, and direct cost comparisons between exercise modalities remain scarce. Health systems looking to structure and reimburse exercise programs at scale have limited evidence to guide those decisions.
The review was preregistered on the PROSPERO database (CRD42023469762), which documented objectives, eligibility criteria, and analysis plans before publication, a transparency measure that strengthens confidence in the methodology.
What this means for people with knee osteoarthritis
For people living with knee osteoarthritis, the practical message is unusually direct. Walking, cycling, or swimming is the single exercise approach with evidence of benefit across pain, function, gait, and quality of life. Starting at a low, tolerable intensity and gradually increasing duration is consistent with both the trial evidence and current guideline recommendations.
Because most of the studied programs were supervised or structured, people who can access physical therapy, community exercise classes, or clinician guidance may find it easier to start and stick with an aerobic routine. But the activities themselves, particularly walking, require nothing more than a pair of shoes and a safe place to move.
The evidence does not mean other forms of exercise are useless. Strengthening, flexibility, mind-body, and neuromotor programs all showed at least some benefits in specific domains. Someone with balance problems might reasonably prioritize neuromotor training alongside walking. A person with marked muscle weakness may still benefit from targeted strengthening even if it does not match aerobic exercise across every outcome. Combining modalities based on preference, access, or coexisting conditions remains a sound approach.
Why aerobic exercise stands apart in the data
What distinguishes the BMJ review from earlier work is not just its size but its specificity. Previous meta-analyses grouped exercise broadly or combined knee and hip osteoarthritis data. This review isolated knee osteoarthritis, compared six distinct exercise categories using network meta-analysis, and evaluated multiple timepoints. The result was a clearer ranking than the field has previously produced.
That ranking carries a two-part message for clinicians and patients. First, aerobic exercise deserves to be presented as the default, first-line exercise option for knee osteoarthritis, given its consistent performance across key outcomes in the largest body of randomized evidence assembled to date. Second, exercise prescriptions should remain individualized and responsive to patient experience, with room to adjust modality, intensity, and frequency over time.
As longer-term and more finely stratified studies emerge, these recommendations may sharpen further. For now, regular aerobic activity stands as the most evidence-backed starting point for managing knee osteoarthritis symptoms without surgery, and the gap between it and every other exercise type is wider than many clinicians may have assumed.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
