Australasian Science: Australia's authority on science since 1938

Evidence for Acupuncture: What Do Scientific Studies Show?

By Harriet Hall

Advocates of acupuncture claim that it has been proven effective by scientific studies. Critics claim that it is only a placebo. They can’t both be right.

It is often stated that acupuncture has been scientifically validated, but the truth is more complicated. Research on acupuncture is inherently problematic. The practice of acupuncture is not standardised, and some studies of “acupuncture” are actually of electro-acupuncture, ear acupuncture or other variants.

It’s impossible to do double-blind studies. The best studies use a retractable needle in a sheath so the patient can’t tell whether the skin has been penetrated or only touched by the needle.

It’s easy to find studies to support a belief in acupuncture, but it’s even easier to find studies showing it doesn’t work. The highest quality studies have shown that it doesn’t matter where you insert the needles or whether the skin is penetrated or merely touched with a needle or even a toothpick (

Two things that do matter are the patient’s beliefs and the acupuncturist’s behaviour.

Authors, editors and journalists can put a spin on study results to match their preconceived opinions. For example, in 2011 a major study in the British Journal of General Practice made conclusions that directly contradicted the data. It claimed there were benefits from acupuncture despite showing that needles are ineffective and that any placebo effects are trivial in size and of no useful benefit (

Acupuncture has been claimed to work for nearly every ailment in the alphabet, from allergies and asthma through menopause and multiple sclerosis to ulcers and whiplash. For most of these conditions, studies have found it is no more effective than a placebo. The most positive results have been for pain and for post-operative nausea and vomiting, but even here the evidence is weak and unconvincing.

A 2011 overview of systematic reviews of acupuncture for various pain conditions found only four pain conditions for which more than one review reached the same conclusion ( In three cases they agreed that acupuncture was ineffective, and in only one (neck pain) did they agree that it was effective.

There is a lot of noise in the data from these studies, and there doesn’t appear to be a signal mixed with the noise.

Even though it is pointless studying “how” something works before it has been established that it “does” work, researchers have tried to find a scientific basis for acupuncture. The gate control theory of pain was eventually rejected because it didn’t explain all the facts. While acupuncture releases natural painkillers (endorphins) in the brain, so too do placebos and other kinds of stimulation. For instance, pinching a mouse’s tail relieves pain better than acupuncture (

Many mechanisms have been postulated, involving things like neurovascular bundles, trigger points, connective tissue planes, electrical impedance and migration of nuclear tracers. The studies are flawed, inconclusive, contradictory, and have not been replicated.

We already have a good explanation of how acupuncture produces subjective effects through a combination of strong suggestion, Oriental mystique, impressive ritual, the beliefs of patient and practitioner, attention from a charismatic provider, long appointments, hands-on treatment, conditioned response and relaxation. The positive studies can be explained as false-

positives resulting from noise in the data, chance, bias, and flaws in research design. There are no specific objective effects; acupuncture has been called a theatrical placebo.

The World Health Organization (WHO) has endorsed acupuncture, but its assessment was based on outdated evidence. Cochrane reviews, The Center for Inquiry and The Medical Letter all dispute WHO’s conclusions.

When a treatment is studied for decades and the evidence remains weak and inconsistent, it becomes apparent that the treatment is not truly effective. We can confidently conclude that acupuncture “works” only as a theatrical placebo and produces no objective or specific effects.

Harriet Hall, MD, is a retired family physician. She writes the SkepDoc column in Skeptic magazine, is an editor of the Science-Based Medicine blog, and is co-author of Consumer Health: A Guide to Intelligent Decisions.