Does acupuncture really work?

Dear Cecil: I agree with you that homeopathy is bunk, but what about acupuncture? Most of the commentary I’ve seen so far has been of the “maybe it works, maybe it doesn’t, more research is needed” variety. Come on, I’ve been reading about acupuncture for years — surely medical science has been able to form some tentative conclusions by now. Carol Fanchamps, Sturtavant, Wisconsin


Illustration by Slug Signorino

Cecil replies:

I’m tired of always pouring cold water on these things, so I’m not going to say acupuncture is silly. Who am I to make light of a therapy just because it uses the same technology as the voodoo doll? Getting miracle drugs from bread mold looked pretty stupid once, too. So I’ll just say this:

  1. People really, really want to believe acupuncture works.
  2. There’s pretty sparse evidence that it does.

The one place where acupuncture has been a big success is on the PR front. In 1996 the Food and Drug Administration declared that acupuncture needles were no longer experimental but would henceforth be considered standard medical equipment, along with syringes, trusses, and arch supports. In 1997 a conference sponsored by the National Institutes of Health concluded that acupuncture was sufficiently effective to warrant increased use in clinical practice.

But you don’t have to read much in the medical journals to think, these guys have got to be kidding. In July 1999 the British Medical Journal published an extensive review of Chinese research on “traditional Chinese medicine” (which also includes herbal and other techniques) and found numerous problems, including poor controls, inadequate protection against bias, etc.

The most revealing datum was a chart showing the results of 49 trials of acupuncture in the treatment of stroke. Normally in such a chart you’d see a bell-curve distribution, with a few data points at the far ends (indicating the treatment was either extremely effective or extremely ineffective) but most in the middle. In fact the chart shows half a bell — a few trials showed acupuncture was very effective, the largest number showed it was slightly effective, and almost none showed it was ineffective. Obvious conclusion: researchers in China only publish positive results.

Acupuncture enthusiasts say it will cure everything from cholera to overbite, but few of these claims can be taken seriously. Acupuncture is widely used to treat addiction, for example, but there’s little solid evidence it does any good.

It’s not even clear acupuncture is effective in treating pain, its most basic use. Acupuncture isn’t routinely used as an anesthesia substitute in China; reports to the contrary in the early 70s were based on observation of surgical patients who’d been selected for high pain tolerance and who in at least some cases were secretly given morphine. At best acupuncture can be said to alleviate rather than eliminate pain, and even then we don’t know whether it’s blocking the pain pathway or simply having a placebo effect.

Part of the problem with acupuncture is the dopiness of the underlying theory. Traditional Chinese medicine holds that disease is the result of imbalanced qi, or vital energy. Qi supposedly flows through the body in channels called “meridians,” with branches to the various organs. Acupuncture, along with breathing exercises, “moxibustion” (burning of certain herbs placed on the body), etc., restores the body’s yin/yang to equilibrium and ameliorates “channel obstruction,” “blood stasis,” organ “vacuity,” and so on.

Many acupuncturists — I’ve heard from a few — accept the qi business at face value. But there’s no scientific basis for it, and it’s of little use even in the practical sense of telling you where to place the needles. One study of recommended treatment for lower-back pain in 16 traditional texts found that fewer than 20 percent of the “acupoints” were recommended by more than half the texts.

Some of acupuncture’s defenders recognize its inadequacies. “Thirty years of active acupuncture research have failed to unequivocally demonstrate its clinical efficacy,” concedes a 1999 paper in the American Journal of Acupuncture. Shall we go back to Tylenol, then? Not at all. “Acupuncture and Chinese traditional medicine are based on a unique philosophical model, and the instruments of biomedical research may be inadequate and inappropriate,” the paper concludes. Translation: Don’t blame acupuncture, blame the test.

These guys can rationalize all they like, but I bet when they go in for a root canal the only needle they want to see is a shot of novocaine. For more, see Stephen Barrett’s review in Quackwatch, 01QuackeryRelatedTopics/acu.html.

Cecil Adams

Send questions to Cecil via