When what an acupuncture study shows is much more interesting than what acupuncture believers think it shows

ResearchBlogging.orgOf all the “alternative” therapies out there, arguably the most studied is the modality known as acupuncture. Perhaps the reason is that, unlike homeopathy, which based on physics, chemistry, and biology alone is so implausible that, for it to “work,” huge swaths of well-established physics and chemistry would have to be shown to be not just wrong but extravagantly and outrageously wrong (making homeopathy far more akin to magic than science), or reiki, which, when you come right down to it, is nothing more than faith healing based on Eastern mysticism rather than Christianity, acupuncture actually involves doing something physical. Indeed, at the risk of annoying or boring long time readers, who have seen me write this many times before, even I used to think that there might be something to acupuncture. Certainly, I recognized that the concepts of meridians along which life energy (qi) flows and how sticking needles into just the right locations on these meridians somehow “unblocks” the flow of qi and thereby relieves pain and heals disease was prescientific nonsense on the order of thinking that evil spirits cause disease. Such ideas may have made sense hundreds of years ago, before scientists developed an understanding of how the body works, but now, in 2010, they are vestiges of a time when medicine really had close to no idea how the body works and when medicine was more often worse than the disease being treated.


Yet the fascination with acupuncture remains, so much so that an inordinate amount of research dollars are spent on studying it. Of course, as Steve Novella has pointed out, in general in medicine (at least these days), the trajectory of research is from bench research to animal models to small scale, less rigorous, pilot studies in humans to large scale, rigorously designed studies using many subjects. True, this order doesn’t always hold. For instance, if physicians make a compelling observation “at the bedside” of response to therapy or how a disease progresses, frequently, after making closer observations to confirm the initial observation, researchers will jump back to animal models and bench top research to try to figure out what’s going on. For such a progression to be useful, though, scientists have to be sure that the phenomenon in human patients under study actually exists. Unfortunately, such is not the case for acupuncture. As larger, more well designed studies using real placebo or sham acupuncture techniques, have increasingly shown that acupuncture does not function any better than placebo in human beings (and sometimes even worse), acupuncturists and acupuncture believers have been reversing the usual order of things, doing smaller studies and “pragmatic” (i.e., uncontrolled) clinical trials, where the placebo effect is not controlled for. Never mind that it doesn’t matter where the needles are placed (thus blowing the whole “meridian” idea out of the water) or even if the needles puncture the skin. Toothpicks work just as well as needles. Also never mind that the mythology of acupuncture as having been routinely practiced for over two thousand years (or, sometimes, four thousand years, is largely a creation of Chairman Mao, who elevated what was a marginal practice at the time to a modality that the state supported and promoted (1,2,3,4). Unfortunately, even the National Center for Complementary and Alternative Medicine (NCCAM) falls for this mythology.

Sometimes, acupuncture studies even end up in high impact journals like Nature Neuroscience. Of course, what is being studied is not really “acupuncture” per se, but rather sticking needles into either people or animals. Since a bunch of you have deluged my mail box with this particular study, I felt obligated to have a look at it. Before I get to the study itself, though, let’s take a look at the press release:

The research focuses on adenosine, a natural compound known for its role in regulating sleep, for its effects on the heart, and for its anti-inflammatory properties. But adenosine also acts as a natural painkiller, becoming active in the skin after an injury to inhibit nerve signals and ease pain in a way similar to lidocaine.

In the current study, scientists found that the chemical is also very active in deeper tissues affected by acupuncture. The Rochester researchers looked at the effects of acupuncture on the peripheral nervous system – the nerves in our body that aren’t part of the brain and spinal cord. The research complements a rich, established body of work showing that in the central nervous system, acupuncture creates signals that cause the brain to churn out natural pain-killing endorphins.

The new findings add to the scientific heft underlying acupuncture, said neuroscientist Maiken Nedergaard, M.D., D.M.Sc., who led the research. Her team is presenting the work this week at a scientific meeting, Purines 2010, in Barcelona, Spain.

“Acupuncture has been a mainstay of medical treatment in certain parts of the world for 4,000 years, but because it has not been understood completely, many people have remained skeptical,” said Nedergaard, co-director of the University’s Center for Translational Neuromedicine, where the research was conducted.

Why does Nedergaard say “certain parts of the world,” instead of what she really means, namely China, from which the practice apparently spread? Who knows? As I will discuss in a moment, Nedergaard’s study is interesting, but it really doesn’t show that “acupuncture works” any more than it really shows compelling evidence for a specific mechanism behind acupuncture. Unfortunately, as is commonly the case, much of the press reporting this study earns an EPIC FAIL in falling for the spin being put on it. The Guardian, for instance, states:

The discovery challenges a widely held view among scientists that any benefit patients feel after having acupuncture is purely due to the placebo effect.

“The view that acupuncture does not have much benefit beyond the placebo effect has really hampered research into the technique,” said Maiken Nedergaard, a neuroscientist at the University of Rochester Medical Centre in New York, who led the study.

“Some people think any work in this area is junk research, but I think that’s wrong. I was really surprised at the arrogance of some of my colleagues. We can benefit from what has been learned over many thousands of years,” Nedergaard told the Guardian.

Ah, yes. The “arrogance” gambit, coupled with the “appeal to ancient wisdom” fallacy. Nice. If you want to irritate Orac, pull the “arrogance” gambit and then couple it with appeals to ancient wisdom. After all, the ancients were heavily into bleeding and purging, not to mention treating with toxic heavy metals. Does that mean such treatments were “wise” or efficacious. Meanwhile The Daily Mail entitled its report Let’s get straight to the point, acupuncture DOES ease pain, and The Telegraph entitled its report Acupuncture does work as it stimulates a natural pain killer, scientists find, while The Raw Story exulted, Researchers prove acupuncture’s effectiveness in pain therapy. Even Science Magazine was not immune. ScienceNow entitled its article on the study How Acupuncture Pierces Chronic Pain.

Perhaps the silliest and most credulous commentary on this study comes from Elizabeth Armstrong Moore on CNET, who even entitled her post Think Acupuncture’s a hoax? Think again (Scientific research shows natural healing compounds):

And yet another pin in the proverbial coffin for skeptics like myself: The researchers even found that in “adenosine receptor knock-out mice” not equipped with the adenosine receptor, acupuncture had no effect.

So what do revelations about a 4,000-year-old technique have to do with modern technology? The better we understand exactly how needles relieve pain, the more likely we are to invent modern acupuncture kits that are affordable, portable, and safe.

“Skeptic.” You keep using that word. I think it does not mean what you think it means, Ms. Armstrong, as you will soon see.

So, with all this hype going on, even during a holiday weekend in the U.S. and the U.K. (apparently Monday was some sort of Bank Holiday in the U.K.), what does the study actually show? Are the headlines of “acupuncture works” and “scientists discover how acupuncture works” justified? Well, not exactly. Despite all the spin, however, the study actually shows quite a bit of interesting biochemistry behind the pain response. Unfortunately, what it doesn’t show is that “acupuncture works,” despite all the whining about “arrogance” from the study’s lead investigator. All it shows are two things: (1) that a chemical called adenosine is released when needles are stuck into the skin of mice and twisted and (2) that adenosine decreases the pain response. These are actually very interesting findings. However, contrary to the way they are being spun, they in now way validate the belief system behind acupuncture or show that “acupuncture works.”

Let’s get to the science in more detail, although I hate to do it by pointing out the credulous and annoying opening paragraph of the paper that reviewers should have shot down in flames but didn’t:

Acupuncture is a procedure in which fine needles are inserted into an individual at discrete points and then manipulated, with the intent of relieving pain. Since its development in China around 2,000 B.C., acupuncture has become worldwide in its practice. Although Western medicine has treated acupuncture with considerable skepticism, a broader worldwide population has granted it acceptance. For instance, the World Health Organization endorses acupuncture for at least two dozen conditions and the US National Institutes of Health issued a consensus statement proposing acupuncture as a therapeutic intervention for complementary medicine. Perhaps most tellingly, the U.S. Internal Revenue Service approved acupuncture as a deductible medical expense in 1973.

Ugh. This is yet another example of an appeal to ancient wisdom, but this time it’s coupled with argumentum ad populum, better known as the appeal to popularity. I never thought I’d see the introduction to a scientific article appeal to, much less mention, the fact that the IRS allows a modality to be a deductible medical expense, but here it is. Where the hell were the reviewers? I could equally point out that the IRS allows medical deductions for the services of Christian Science prayer healers. Does that mean that Christian Science prayer is an effective treatment for anything? No, it does not. Add to that the whole false dichotomy between “Western” and “Eastern” medicine, a particularly odious and, yes, racist, construct in which the mythical “East” is represented as more “wholistic” and “spiritual” compared to the “reductionistic” and scientific brand of medicine, and it’s a perfect example of what Kimball Atwood would term the Weasel Words of Woo. Note how no evidence for the actual–oh, you know–efficacy of acupuncture is presented in the introduction. True, in the second paragraph, Nedergaard does mention the possibility of the release of opiod receptors, but there is zero discussion of the evidence for and against acupuncture, not even much of an acknowledgment other than that nasty “Western medicine” being so nastily “skeptical” of the practice, as though that were a bad thing.

The model used by Nedergaard is a model of inflammation that involves injecting complete Freund’s adjuvant (CFA) into the mice’s paws. As a result, the mice’s paws would become more sensitive to innocuous stimuli and the latency period for withdrawal to painful stimuli and heat would decrease; in other words, the mice’s paws would be more sensitive, and the mice would react more strongly and rapidly to the stimuli of heat or touching. This sensitivity peaked at day four or five and then decreased. As a preliminary experiment, the investigators noted that, after the insertion of acupuncture needles into the mouse limb at the “Zusanli point,” which is located near the knee a microdialysis probe inserted less than a millimeter away registered a spike in extracellular adenosine levels, as well as ATP (which is broken down to adenosine outside of the cells), ADP, and AMP, that peaked at around 30 minutes:

(Click for full image)

Having established that adenosine was increased within 30 minutes of an acupuncture stimulus, Nedergaard then injected an chemical that binds to the cell receptor activated by adenosine, the A1 receptor agonist, 2-chloro-N(6)-cyclopentyladenosine (CCPA). Injecting CCPA into the Zusanli point greatly improved touch sensitivity and in essence reversed the increased sensitivity to heat. So far, so good. Apparently in the mouse adenosine has a lot to do with modulating pain response in peripheral nerves. Even better, this effect was not observed in mice genetically engineered not to make the adenosine A1 receptor, known as A1 receptor knockout mice. Investigators also tested CCPA in a model of neuropathic pain (pain due to nerve dysfunction) and found it worked as well as it did in their model of inflammatory pain.

So what’s the link to acupuncture? Well, one thing that this paper failed to mention is that the Zusanli point is not actually a leg acupuncture point, although it is located on the leg, just inferior and lateral to the knee, it is described as the leg portion of the stomach meridian. According to traditional Chinese medicine, this point is also known as Stomach-36 and its current indications are:

The current standard indications for zusanli, as reviewed in Advanced Textbook of Traditional Chinese Medicine and Pharmacology (21) are: stomach ache, abdominal distention, vomiting, diarrhea, dysentery, indigestion, appendicitis, flaccidity and numbness of the lower limbs, edema, mastitis, mania, epilepsy, cough, vertigo, palpitation, and emaciation due to consumptive disease. This latter indication corresponds to the concept that needling this point can tonify the sea of qi and thereby help to stop the wasting disease and restore ones body weight and vitality.

To illustrate the uniformity of indications amongst the Chinese authorities, the following were listed in Chinese Acupuncture and Moxibustion (22, 23), with slight differences on translation between the original Chinese and later Western publications: gastric pain, hiccup, abdominal distention, vomiting, diarrhea, dysentery, emaciation due to general deficiency, constipation, mastitis, intestinal abscess (acute appendicitis), numbness (motor impairment) and pain of the lower extremities, edema (beriberi), manic depressive psychosis.

Thus, according to acupuncturists, most associated with stomach and abdominal problems far more than lower extremity pain. Strange that the investigators didn’t mention that. Be that as it may, here’s what the investigators found:

  • In normal mice of adenosine, acupuncture reduced discomfort by two-thirds.
  • In A1 receptor knockout mice, acupuncture had no effect on the reactions of the mice to the stimuli of touch or heat.
  • During and after an acupuncture treatment, adenosine levels in the tissues near the needles was 24 times greater than before the treatment.
  • Deoxycoformycin, a drug that inhibits the removal of adenosine by the tissues, increased the length of time that the adenosine remained in the tissues (surprise! surprise! given its known mechanism of action) but also appeared to increase the length of time that acupuncture treatment was effective.
  • In mice who had acupuncture but in which the needle wasn’t rotated every five mintues, acupuncture had no effect.

So what does this all mean? First of all, this study is actually quite interesting for its implications for adenosine as a mediator of both inflammatory and neuropathic pain. It was cleverly done and is pretty clear in supporting the conclusion that mimicking the action of adenosine or somehow increasing its local concentration around a nerve might be a good strategy for relieving pain in humans. But does it actually support the efficacy of acupuncture, as Nedergaard claims.

Not so fast.

All this study shows isthat sticking needles in mice causes adenosine production and that that adenosine can blunt the pain response in nerves by binding to the A1 receptor. That’s all well and good, but it doesn’t validate acupuncture. The only thing in common with acupuncture is the needle sticking part, and the investigators might as well conclude that this study validates ear piercing for pain relief. (Egads! Battlefield acupuncture strikes back!) So, it’s quite possible that needles twisted in the area near a nerve might release a flood of adenosine that might bind to A1 receptors in nearby neurons and blunt the pain sensation. No “meridians” or qi needed to explain that. Moreover, this study notwithstanding, Nedergaard seems at a loss to explain how her results might be reconciled with numerous studies in humans that show clearly that (1) it does not matter where you stick the needles and (2) it doesn’t even matter if the needles are stuck through the skin. As I’ve pointed out before, just twisting the end of a toothpick against the skin produces the same effect as acupuncture. She does, however, give it the old college try to explain this, although she does so using the hated term “allopathic” to describe “Western” medicine:

One may speculate that other non-allopathic treatments of chronic pain, such as chiropractic manipulations and massage, modalities that involve the mechanical manipulation of joints and muscles, might also be associated with an efflux of cytosolic ATP that is sufficient to elevate extracellular adenosine. As in acupuncture, adenosine may accumulate during these treatments and dampen pain in part by the activation of A1 receptors on sensory afferents of ascending nerve tracks. Notably, needle penetration has been reported to not confer an analgesic advantage over nonpenetrating (placebo) needle application, as opposed to our observations (Supplementary Figs. 2 and 3) and those of others. However, it is possible that ATP release from keratinocytes in response to mechanical stimulation of the skin results in an accumulation of adenosine that transiently reduces pain, as A1 receptors are probably expressed by nociceptive axon terminal in epidermis. In fact, vibratory stimulation applied to the skin depressed the activity of nociceptive neurons in the lower lumbar segments of cats by release of adenosine. However, this effect differs from the anti-nociceptive effect of acupuncture, which does not depend on the afferent innervation of the skin. Acupuncture is typically applied to deep tissue, including muscle and connective tissue, and acupoints may better overlap with their proximity to ascending nerve tracks than to the density of cutaneous afferents.

This is pure speculation without any compelling evidence. It would also have been easy enough to test. Contrary to her implication, Supplemental Figures 2 and 3 do not refute the results showing that non-penetrating acupuncture works as well as penetrating acupuncture. In Figure S2, all that is shown is that using acupuncture at the Zusanli point on the other leg doesn’t affect the reaction of the leg being tested. All Figure S3 shows is that failing to rotate the needle results in loss of the analgesic effect. Neither refute findings in humans that non-penetrating acupuncture “works” just as well as penetrating acupuncture, where the needles are inserted to the “correct” points. Absent evidence from the current study showing that stimulating the skin results in the release of adenosine and subsequent blunting of the pain response in the extremities of these mice, Nedergaard would have been better off leaving this paragraph out of the paper. Finally, remember: This paper says absolutely nothing at all about all the other myriad claimed benefits of acupuncture in fertility treatments, headache, or hot flashes after menopause, to name a few.

Finally, there are two huge problems with this paper. Here’s the second biggest. Mice are much, much smaller than humans. This Zusanli point in a mouse is going to be within a couple of millimeters of the sciatic nerve. In the human, it’s going to be centimeters away. In the mouse, the size of the needle relative to the size of the leg and distance from the sciatic nerve, as well as the nerve’s branches, the tibial and peroneal nerves, is going to be very close. The tissue damage, virtually no matter where the needle is stuck, is going to be close to these nerves. In humans, this is unlikely to be the case, particularly since the nerves are much further beneath the skin than they are in mice. Thus, there is no good reason to think that these results will necessarily translate easily to humans. Moreover, contrary to the claim that “merdians” map to ascending nerve tracts is stretching it a bit, if you look at these maps. For instance, the kidney, stomach, and spleen acupuncture points line up somewhat with nerves at certain points in the body but are nowhere near ascending nerves in other parts. However, this works in mice, size differences alone will make it hard to replicate in humans.

What really bugs me about this article, though, is that it’s cool science. These guys have found something interesting that may even have a potential clinical application. For instance, local injection of A1 receptor agonists works the same as the “acupuncture.” Adding compounds that slow the removal of adenosine fromt the tissues improves the efficacy of the adenosine released into the tissues by minor trauma. Scientists can work with that. Scientists could take this observation and use it as a justification to work on better, more specific, and longer acting A1 agonists. Perhaps they could even develop oral drugs that are broken down into adenosine or A1 receptor agonists in the peripheral tissues. If this paper’s conclusions regarding the importance of adenosine in pain signaling are correct, these would represent stragies that could very well work and very well improve pain control. One could even envision implantable pellets that could be placed in wounds or near relevant nerves to release A1 receptor agonists right where they’re needed over a long period of time. Unfortunately, Nedergaard and her team are too enamored of the woo that is acupuncture to emphasize the true significance and potential usefulness of their findings.

Instead of working to turn these observations into usable therapies, no doubt Nedergaard and her collaborators will use this preliminary data to apply to NCCAM for a grant to study adenosine in acupuncture further, rather than pursuing this observation in a manner far more likely to lead to a clinical benefit in human beings. So much the pity. Woo poisons the real science it touches.


Goldman, N., Chen, M., Fujita, T., Xu, Q., Peng, W., Liu, W., Jensen, T., Pei, Y., Wang, F., Han, X., Chen, J., Schnermann, J., Takano, T., Bekar, L., Tieu, K., & Nedergaard, M. (2010). Adenosine A1 receptors mediate local anti-nociceptive effects of acupuncture Nature Neuroscience DOI: 10.1038/nn.2562


  1. Acupuncture Works, Say Scientists
  2. A biological basis for acupuncture, or more evidence for a placebo effect?
  3. Why was a study on ‘acupuncture’ reported so badly?