Knee bracing adds small short-term improvement
Adding a compartment-targeted knee brace plus adherence support modestly improved knee symptoms at 6 months compared with advice and home exercises alone.
*Randomized controlled trial; Level 1b (Oxford Centre for Evidence-Based Medicine).
Citation
Holden MA, Nicholls E, Abdali Z, et al. Provision of knee bracing for knee osteoarthritis (PROP OA): multicentre, parallel group, superiority, statistician blinded, randomised controlled trial. BMJ. 2026;392:e086005. doi:10.1136/bmj-2025-086005.
Background
Guidelines disagree on whether knee bracing helps knee osteoarthritis, partly because earlier trials were small and did not support long-term brace use. This trial tested compartment-targeted braces plus strategies to help people keep wearing them.
Patients
466 adults aged ≥45 years in England with symptomatic knee osteoarthritis and moderate-to-severe weight-bearing pain; excluded if recent knee injection/physiotherapy or current brace use, inflammatory arthritis, recent major knee surgery, or planned joint replacement within 6 months.
Intervention
Advice, written information, and home exercise instruction plus a fitted knee brace (type matched to knee compartment), a 2-week follow-up visit, brief adherence coaching, and text-message reminders for 6 months.
Control
Advice, written information, and home exercise instruction in a single visit.
Outcome
Primary: 0–100 composite patient-reported knee score at 6 months (higher is better). Secondary: pain measures and responder status.
Follow-up Period
12 months (primary endpoint: 6 months).
Results
| Outcome (6 months) |
Effect (brace group minus control) |
Practical meaning |
| Composite knee score, 0–100 (primary) |
+3.39 points (95% CI 0.96 to 5.82) |
Small average improvement |
| Pain during weight-bearing activity, 0–10 |
−0.80 (95% CI −1.15 to −0.44) |
Less pain with activity |
| Responder status (patient-relevant improvement) |
48% vs 33%; NNT ≈7 |
About 1 extra responder per 7 treated |
CI=confidence interval; NNT=number needed to treat.
Analyses followed an intention-to-treat approach (treatment-policy estimand). Minor expected harms occurred; skin irritation was most common (up to about 1 in 5).
A commonly used meaningful-change threshold for the primary score was 8 points; the average between-group difference did not reach this level.
Limitations
Benefits were small and may not be noticeable for many patients. Participants and clinicians could not be blinded; some control participants used braces and received adherence-style counseling, which could shrink differences. The study population was mostly White, limiting generalizability.
Funding
UK National Institute for Health and Care Research; Keele University; braces donated/discounted.
Clinical Application
Offer compartment-targeted bracing as an optional add-on to education and exercises, especially for patients willing to wear it regularly; counsel about modest benefits and skin irritation.