Skip navigation

Acupuncture for Osteoarthritis

Clinical bottom line: There is no evidence that acupuncture is more effective than sham/placebo acupuncture for the relief of joint pain due to osteoarthritis (OA). Most of the existing trials have serious methodological flaws. The two most methodologically rigorous trials showed no difference between the effect of acupuncture and sham acupuncture for pain relief. The trials comparing acupuncture to active treatment controls were of insufficient methodological rigour to determine efficacy.

Acupuncture is amongst the most frequently used complementary therapy for OA. About half of all consultations with British acupuncturists are for OA conditions.

Systematic review

Ernst E. Acupuncture as a Symptomatic Treatment of Osteoarthritis. A Systematic Review. Scand J Rheumatol 1997; 26:444-7

Date review completed: Search completed 1996

Number of trials included: 13

Number of patients: A total of 437 in active and control groups, individual group sizes were not reported

Control group: sham acupuncture, TENS, physiotherapy, piroxicam

Main outcomes: Various pain outcomes; VAS, subjective reports, tenderness, range of motion and analgesic use

Inclusion criteria were controlled trials, patients with symptomatic OA of any joint, traditional Chinese and non-traditional acupuncture and an adequate control group (not defined further).

The author conducted a comprehensive search including the databases Medline and CISCOM (a database specialising in complementary medicine) and references of retrieved reports and also sent requests to experts in the field. The reviewer provided a descriptive summary of included trials as pooling of data was impossible due to their disparate nature.


Overall 13 clinical trials were included in the review, most used formula acupuncture (needling of a predefined set of points according to conventional diagnosis) as opposed to Chinese acupuncture (needling of individualised sets of points according to traditional Chinese diagnosis). The majority of the trials were of a poor quality with severe methodological flaws. These included lack of randomisation, blinding and statistical testing, small group sizes, and diagnostic criteria for OA and outcome measures were not well defined.

Only two trials were both randomised and double-blind providing the most valid data. Both trials compared formula acupuncture to sham acupuncture, one trial gave acupuncture once a week for eight weeks and the other gave acupuncture three times a week for three weeks. For each trial there was no difference between treatment groups for pain and tenderness outcomes. Although these are the highest quality trials in this review, they still have methodological flaws. Both trials were small with only 20 patients in each treatment group. It is not clear from the review when or how the outcomes were measured, whether the results were statistically tested, or what diagnostic criteria for OA was used.

Taking all 13 trials together, six trials showed greater pain relief in the acupuncture group than in the control group, seven trials showed no difference. The Table shows that the lower the methodological quality of the trial the more likely it was to show a positive effect.

Number of trials
Type of study
Control groups
Trial conclusion
CCT (not R, not B)
piroxicam (1)
physiotherapy (2)
1 positive, 1 negative
RCT (not blind)
no treatment (1)
sham TENS (1)
sham acupuncture (2)
1 positive, 1 no difference
TENS (1)
sham TENS (1)
no treatment (1)
sham acupuncture (1)
no difference
no difference
no difference
sham acupuncture
2 no difference

Abbreviations: CCT = controlled clinical trial; R = randomised; B = blind; SB = single-blind; DB = double-blind; TENS = transcutaneous electrical nerve stimulation

Adverse effects were not reported in the review

Further reading

Related topics

  • Identifier AT008 - 5858 ACUPUNCTURE FOR OSTEOARTHRITIS: Apr-2000