Exercise helps arthritis, by a modest amountStrong evidence
In a 2024 Cochrane review of 139 trials in knee osteoarthritis, exercise reduced pain by about 9 to 13 points out of 100 and improved function by about 11 to 13 points, compared with placebo or usual care. In a 2026 review of 18 hip studies, pain fell by about 7 points against no treatment. The reviewers judged these average effects small or of uncertain importance; NICE recommends exercise for everyone with osteoarthritis.
Sources 25, 26, 1
Knee osteoarthritis: better than nothing, about the same as other exerciseEmerging evidence
A 2025 review of 11 trials (476 people) found Pilates reduced knee pain more than no exercise (3 trials, 66 people, low certainty) but not more than conventional exercise (5 trials, 210 people, very low certainty). Another review of 8 trials found the same pattern, with signs that small studies with negative results may be missing. The trials were small and short, and few reported side effects.
Sources 9, 10, 27
Two network analyses ranked Pilates highlyEmerging evidence
One network meta-analysis found large effects of Pilates on knee pain and function against usual care and rated the evidence moderate certainty, but that rests on just 2 small Pilates trials. Another, of 29 trials, ranked Pilates first for knee function, with no significant effect on pain, at low or very low certainty.
Sources 28, 29
Hip, back and hand arthritis: little Pilates researchMixed evidence
We found no published trial of Pilates in hip osteoarthritis; one finished in June 2025 with no results posted. For back pain in general, a Cochrane review of 10 trials found Pilates eased pain and disability more than minimal treatment and about as much as other exercise, at low to moderate quality, and a review of exercise for worn spinal discs found no clear differences between types. We found no Pilates trials in spinal arthritis or hand arthritis.
Sources 30, 31, 32, 33, 34, 35
Ankylosing spondylitis: small trials, fairly consistent gainsEmerging evidence
In small trials of 20 to 96 people, Pilates improved disease activity and everyday function more than usual care or home exercise, and adding Pilates to aerobic training helped more than aerobic training alone. In a review of 15 studies of mind-body exercise, 4 of them Pilates, function and disease activity scores improved, at moderate to low certainty, with no serious side effects.
Sources 23, 36, 24, 37, 38, 39
Spinal mobility: the reviews disagreeMixed evidence
In a 2026 network meta-analysis of 59 trials, which we could read only as an abstract, Pilates improved disease activity, function and pain against usual care, but no type of exercise improved spinal mobility. Another network analysis ranked aerobic exercise plus Pilates first for disease activity and spinal mobility, and the review of mind-body exercise found a small gain in mobility. Whether Pilates beats other structured exercise isn't known.
Sources 40, 41, 39
Rheumatoid arthritis: too few trials to sayMixed evidence
There are only two small Pilates trials, of 30 people each: in one, Pilates, aerobic exercise and both together didn't differ; in the other, pain and knee strength improved more than in a control group. A network meta-analysis ranked Pilates first for pain, but on one of these trials, whose own Pilates group didn't improve significantly, so we don't rely on it. Regular exercise of any kind is the recommendation.
Sources 42, 43, 44, 3
Children with juvenile arthritisMixed evidence
In two small trials, of 50 and 40 children, Pilates on its own or added to physiotherapy improved quality of life, pain and function more than conventional exercise or physiotherapy alone.
Sources 45, 46
Side effects were rarely reportedMixed evidence
Only 2 of 11 knee trials reported side effects (none, or none serious), and reviewers noted that monitoring was limited. Exercise in general may bring more minor aches than usual care.
Sources 9, 10, 39, 25