Skip navigation

Acupuncture for osteoarthritis

Clinical bottom line

There is no good evidence that acupuncture is any better than sham acupuncture.


YD Kwon et al. Acupuncture for peripheral joint osteoarthritis. Rheumatology 2006 45: 1331-1337.


There is much debate about whether acupuncture works. There is no substantial body of evidence that acupuncture is effective for anything. Despite this, acupuncture is much practiced, and even offered in the UK on the NHS.

Systematic review

A number of databases were searched for randomised trials of acupuncture in patients with peripheral joint osteoarthritis. Acupuncture had to be with needles, with or without electrical stimulation.


Statistical benefit with acupuncture was reported for 10 of 18 trials, including 7/10 higher quality trials, using all sorts of different control.

Manual acupuncture could be compared with sham in three trials, with about 330 patients, with a statistically better but trivial difference in favour of true acupuncture. The same comparison with electroacupuncture in two trials with about 260 patients again had statistical but not clinical significance.


In acupuncture trials the main concern is that of an appropriate control. The best control is sham acupuncture, where everything is the same except that the needles are inserted in places not supposed to be relevant to the pain being treated. According to acupuncture theory, sham acupuncture should be ineffective.

In fact, sham acupuncture is almost as effective as true acupuncture. This means that whatever is going on, the placing of the needles is irrelevant. Acupuncture doesn't need needles.