---
title: "Knee osteoarthritis"
url: "https://acuiq.com/wiki/evidence/knee-osteoarthritis"
updated: 2026-10-10
last_verified: 2026-10-09
license: CC-BY-4.0
license_url: "https://creativecommons.org/licenses/by/4.0/"
section: "evidence"
basis: "Modern research"
audience: "Anyone"
published: "2026-10-09"
reading_minutes: 7
sources: 9
---

# Knee osteoarthritis

_What the Cochrane review, later meta-analyses, two sham-controlled trials and the NICE and American College of Rheumatology guidelines report on acupuncture for knee osteoarthritis._

Two guideline bodies reviewed the trials of acupuncture for knee osteoarthritis within three years of each other and reached opposite recommendations. NICE, which writes clinical guidance for the NHS in England, [advised in 2022](https://www.nice.org.uk/guidance/ng226/chapter/Recommendations) that acupuncture and dry needling should not be offered for osteoarthritis. The American College of Rheumatology [conditionally recommended acupuncture in its 2019 guideline](https://doi.org/10.1002/art.41142) and kept that recommendation in its [2026 update](https://assets.contentstack.io/v3/assets/bltee37abb6b278ab2c/bltb3d12c34020da842/oa-guideline-summary-2026.pdf). The reviews and trials underneath both share one pattern: against no acupuncture, people report less knee pain; against sham acupuncture, the difference is small, and how small depends on which review or trial is read.

[Sham acupuncture](/wiki/research/sham-acupuncture) covers the controls that imitate treatment and why none of them is inert.

## The Cochrane review

The Cochrane review, [published in 2010](https://doi.org/10.1002/14651858.CD001977.pub2) by Manheimer and colleagues, covers osteoarthritis of the knee, hip and hand. It searched to December 2007 and included 16 trials with 3,498 people; 12 of the trials enrolled people with knee osteoarthritis only. It used the WOMAC index, the standard osteoarthritis questionnaire, whose pain scale runs from 0 to 20 and function scale from 0 to 68, and set in advance the smallest differences it would count as clinically relevant: 1.3 points for pain and 3.57 for function.

| Compared with | Time | Trials (participants) | Pain | Function |
| --- | --- | --- | --- | --- |
| Sham | Short term | 9 (1,835) | 0.9 points better on 0 to 20, below the threshold | 2.7 points better on 0 to 68, below the threshold |
| Sham | 6 months | 4 (1,399) | 0.4 points better, borderline significance | 1.2 points better, borderline significance |
| Waiting list | Short term | 4 (884) | 14.5 points better on 0 to 100, clinically relevant | 13.0 points better on 0 to 100, clinically relevant |

When the reviewers kept only the sham-controlled trials whose shams were most likely to keep patients unaware of their group, the pooled benefit shrank and was no longer statistically significant. Those were also the shams the reviewers judged most likely to have a physiological effect of their own. In head-to-head trials, acupuncture did better than supervised osteoarthritis education and than physician consultation, and about as well as home exercises with an advice leaflet or supervised exercise; added to an exercise-based physiotherapy programme, it gave no further improvement. Only 8 trials reported on safety, and incompletely. The authors concluded that the sham-controlled benefits were small and probably due at least partly to placebo effects from incomplete blinding, and that much of the larger benefit against a waiting list may come from expectation or placebo.

## Later reviews

A [2024 review in BMJ Evidence-Based Medicine](https://doi.org/10.1136/bmjebm-2023-112626) by Liu and colleagues searched English and Chinese databases to November 2023 and included 80 trials with 9,933 participants, nearly three times as many participants as the Cochrane review. It rated all its main findings very low certainty under GRADE, the standard scale from high to very low. On a 0 to 100 mm pain scale it estimated that acupuncture reduced pain at the end of treatment by about 18.5 mm more than sham, 21.5 mm more than anti-inflammatory drugs (NSAIDs), 25.3 mm more than usual care or a waiting list and 41.3 mm more than no intervention, with no difference from injections into the joint. Its network analysis, which compares treatments indirectly through shared controls, ranked electroacupuncture above manual acupuncture.

The individual patient data meta-analyses of the Acupuncture Trialists’ Collaboration, which reanalysed each participant’s raw data from high-quality trials, put the difference from sham for osteoarthritis at 0.16 standard deviations in [2012](https://doi.org/10.1001/archinternmed.2012.3654), and the difference from no acupuncture at 0.57, after setting aside an outlying group of trials. In that analysis’s worked example, 0.2 standard deviations corresponds to about 5 points on a 0 to 100 pain scale. [Chronic pain](/wiki/evidence/chronic-pain) covers those analyses and their 2018 update.

## Two trials

An Australian trial by Hinman and colleagues, [published in JAMA in 2014](https://doi.org/10.1001/jama.2014.12660), randomised 282 people aged 50 or over with chronic knee pain to needle acupuncture, laser acupuncture, sham laser or no acupuncture, delivered by family doctors trained in acupuncture for 12 weeks. Neither needle nor laser acupuncture beat sham laser on pain or function at 12 weeks: needle acupuncture differed from sham by 0.4 points on a 0 to 10 pain scale, against the 1.8 points the trial set as the smallest difference that matters. Against no acupuncture, needle acupuncture reduced pain by 1.1 points at 12 weeks, a difference gone at one year. The authors concluded their findings did not support acupuncture for these patients. [Laser acupuncture](/wiki/techniques/laser-acupuncture) covers the laser arm’s technique.

A multicentre trial in China by Tu and colleagues, [published in 2021](https://doi.org/10.1002/art.41584), tested a more intensive course: electroacupuncture, manual acupuncture or sham, three times a week for eight weeks, in 480 people with knee osteoarthritis. At week 8, 60.3% of the electroacupuncture group, 58.6% of the manual acupuncture group and 47.3% of the sham group reached a minimal clinically important improvement in both pain and function. Electroacupuncture differed significantly from sham and manual acupuncture did not, though both acupuncture groups did better than sham at weeks 16 and 26. [Electroacupuncture](/wiki/techniques/electroacupuncture) describes the method.

## What the guidelines say

NICE’s osteoarthritis guideline, NG226, published in October 2022, says in recommendation 1.3.8: do not offer acupuncture or dry needling to manage osteoarthritis. Its [rationale](https://www.nice.org.uk/guidance/ng226/chapter/Rationale-and-impact) states that the evidence, mostly for the knee, showed a lack of benefit and some evidence of harm, and that economic evidence found acupuncture not cost-effective. It found some benefit for electroacupuncture against sham, though not against acupuncture or no treatment, from very low quality evidence, and asked for further research on electroacupuncture rather than recommending it.

The American College of Rheumatology and the Arthritis Foundation, in their 2019 guideline by Kolasinski and colleagues, conditionally recommended acupuncture for osteoarthritis of the knee, hip and hand. Their text calls its efficacy a subject of controversy, attributes much of its benefit to a large contextual effect plus a small difference between true and sham acupuncture, and puts that difference at about the size of full-dose paracetamol against placebo; with minor risk of harm, the voting panel made a conditional recommendation. The College’s 2026 update, approved by its board in September 2026 and published as a summary ahead of the full paper, again conditionally recommends acupuncture for the knee, hip and hand, rating the certainty of evidence moderate for knee and hip and very low for hand. [Clinical guidelines](/wiki/evidence/clinical-guidelines) sets these positions beside other bodies’.

## What the reviews could not settle

The disagreement between the guidelines is about the sham comparison, and the sources do not resolve it. Cochrane found the sham-controlled benefit fell when blinding was best, Liu found a larger one from many more and mostly smaller trials at very low certainty, and Hinman and Tu found none and a modest one with different shams, doses and settings. Whether electroacupuncture or a more intensive course does better than ordinary manual acupuncture is suggested by Liu and Tu and remains, in NICE’s words, a question for research. No review established how long any effect lasts: Hinman’s gain over no acupuncture had gone at one year. The Cochrane review has not been updated since its 2007 search. Liu’s review drew heavily on Chinese databases, and [Where the trials come from](/wiki/research/where-the-trials-come-from) covers what the share of positive results by country shows. [Rheumatoid arthritis](/wiki/evidence/rheumatoid-arthritis) covers the trials in inflammatory joint disease.

AcuiQ’s [knee osteoarthritis page](/symptoms/osteoarthritis-knee) lists the protocols individual studies prescribed, each with its citation, and the [knee pain](/symptoms/knee-pain) and [knee stiffness](/symptoms/stiffness-knee) pages do the same for the terms those studies used.

### Sources

1. Manheimer E, Cheng K, Linde K, et al., Acupuncture for peripheral joint osteoarthritis, Cochrane Database of Systematic Reviews 2010 — https://doi.org/10.1002/14651858.CD001977.pub2
2. Liu CY, Duan YS, Zhou H, et al., Clinical effect and contributing factors of acupuncture for knee osteoarthritis: a systematic review and pairwise and exploratory network meta-analysis, BMJ Evidence-Based Medicine 2024 — https://doi.org/10.1136/bmjebm-2023-112626
3. Vickers AJ, Cronin AM, Maschino AC, et al., Acupuncture for chronic pain: individual patient data meta-analysis, Archives of Internal Medicine 2012 — https://doi.org/10.1001/archinternmed.2012.3654
4. Hinman RS, McCrory P, Pirotta M, et al., Acupuncture for chronic knee pain: a randomized clinical trial, JAMA 2014 — https://doi.org/10.1001/jama.2014.12660
5. Tu JF, Yang JW, Shi GX, et al., Efficacy of intensive acupuncture versus sham acupuncture in knee osteoarthritis: a randomized controlled trial, Arthritis & Rheumatology 2021 — https://doi.org/10.1002/art.41584
6. NICE, Osteoarthritis in over 16s: diagnosis and management (NG226), recommendations, published October 2022 — https://www.nice.org.uk/guidance/ng226/chapter/Recommendations
7. NICE, Osteoarthritis in over 16s (NG226): rationale and impact, published October 2022 — https://www.nice.org.uk/guidance/ng226/chapter/Rationale-and-impact
8. Kolasinski SL, Neogi T, Hochberg MC, et al., 2019 American College of Rheumatology/Arthritis Foundation guideline for the management of osteoarthritis of the hand, hip, and knee, Arthritis & Rheumatology 2020 — https://doi.org/10.1002/art.41142
9. American College of Rheumatology, 2026 update of the American College of Rheumatology recommendations for the management of osteoarthritis of the knee, hip, and hand: guideline summary, September 2026 — https://assets.contentstack.io/v3/assets/bltee37abb6b278ab2c/bltb3d12c34020da842/oa-guideline-summary-2026.pdf