I know, I know, it’s hard to believe, given the sorts of posts I’ve done recently on acupuncture, but it’s true. Certainly, I didn’t believe the whole rigamarole about needles somehow “restoring the flow of qi” or anything like that, but I did wonder if maybe there was some physiologic mechanism at work behind acupuncture that produced real benefits in terms of pain relief above that of placebo. Sure, I may have dismissed homeopathy as the pure magical thinking that it was, but acupuncture I wasn’t so sure about.
Obviously, that’s changed.
The reason my opinion has changed and now I place acupuncture firmly in the “woo” category is that I’ve actually been reading the scientific literature on acupuncture over the last year or so. From such a reading of the literature, it has become very obvious to me that (1) the vast majority of research into acupuncture is shoddy in the extreme, with methodological problems that greatly increase the probability of false positive trials; (2) many investigators conflate electroacupuncture which is in reality nothing more than the “conventional” modality of transcutaneous electric nerve stimulation rebranded, with acupuncture itself (unless the ancient Chinese knew how to make electrical nerve stimulation devices, which I highly doubt); and (3) when trials are done with true sham acupuncture there is almost invariably no difference detected between the true acupuncture and the control group.
This time around, I’ve come across yet another acupuncture study that serves to demonstrate that acupuncture is nothing more than an elaborate placebo. The paper, published in the most recent issue of the Clinical Journal of Pain joins a long line of papers that show that, when the study is well-designed and includes true sham acupuncture, the results virtually invariably show acupuncture to be useless as a therapy. This particular study came out of a collaboration between Harvard Medical School, the University of Michigan School of Public health, and the Harvard School of Public Health and examined the effect of acupuncture on persistent arm pain due to repetitive stress injuires (RSIs).
The study design was fairly simple. Through June 2001 and April 2003, the investigators recruited men and women from the greater Boston and Cambridge areas through advertisements and referrals from health care professionals. Eligibility was determined through screening interviews by research assistants involved in the study. The only twist in this study is that apparently there were two studies (acupuncture versus sham acupuncture and acupuncture against medication), and participants were randomized to one of them and further randomized to either the sham acupuncture group or the “true” acupuncture group. Of course, one of the silly things about this or any acupuncture study is how the “true” acupuncture treatment is identified:
Acupuncture treatments focus on relaxing the muscles and opening channels to the circulation of what acupuncturists call ”qi.” Point selection in our study was based on the location of the pain, limitations to the range of arm motion, and local sensitivity to palpation. Local channel points proximal to the area of pain were combined with distal points that control the area of pain. Nonmeridian local ”trigger” points (”ah shi” points) were included. To incorporate the acupuncture theory of ”opening the gates” in pain conditions, all participants received needling on Liver 3 (a point near the big toe) on the contralateral side of the affected arm, which was paired with Large Intestine 4, a point on the hand of the affected arm.14 If both arms were affected, the latter point was needled bilaterally.
A consensus team of senior acupuncturists selected 20 allowable acupuncture points based on the acupuncture literature. We used a ”manualized” approach27 that allowed some flexibility to vary the location of points according to the specific location and nature of the pain, while providing standardization of treatment. Besides the required points, practitioners could select between 5 and 8 additional points at each session and could include local area points traditionally used to affect specific regions (ie, LI 5, P5, P6, P7, and TW5) and local and distal sensitive ”ah shi” points.
Clear enough to you? Personally, it’s clear as mud to me. However, since “certified” acupuncturists picked the needle points, I can only assume that they picked points that they thought would yield the best chance of a therapeutic result. The “sham” acupuncture was performed at the very same acupuncture points but using sham needles with blunt ends that retract back into the needle hub after hitting the skin. They have been validated before and look and feel like “real” acupuncture needles. The only weakness in this study was that, although patients and the research assistants who recorded the pain levels and measured range of motion were blinded to the experimental group, the actual acupuncturists were not. However, blinding was assessed by asking patients whether they thought they were in the placebo or true acupuncture group, and similar percentages thought that they were getting true acupuncture. Be that as it may, the acupuncture regimen used included eight treatments administered over four weeks.
So what were the results?
Here’s where things get amusing. Both treatment groups, “true” and sham acupuncture, experienced decreases in the intensity of arm pain, arm symptoms, and noted improvement in arm function. However, patients in the sham acupuncture group improved more than patients in the “true” acupuncture group in the intensity of arm pain and just as much in measures of arm function and grip strength. The difference between the two groups was not sustained at a followup visit one month after the treatment ended, although the improvement in both groups remained detectable compared to baseline. Indeed, arm pain and arm symptoms scores declined faster in the sham compared with the “true” acupuncture group.
In this study, which was the largest, best-designed trial thus far for acupuncture for arm pain due to RSI, sham acupuncture was better than “real” acupuncture!
Not that any of this keeps the authors from trying to explain how sham woo is better than the woo itself:
Reasons for the superiority of the sham acupuncture device during treatment are not clear. One possibility is that treatment effects were blunted in the true acupuncture group because of the higher rates of side effects, and especially mild pain during treatment. We speculate that this discomfort may have been due to some of the needle placements in the arm, in close proximity to the painful areas. Another possibility might be that the sham device may have conferred genuine treatment effects, because it was applied to real acupuncture points (an ”acupressure” effect). This seems less likely given the results of another RCT testing acupuncture in the treatment of fibromyalgia that found no significant differences between true acupuncture, noninsertive simulated acupuncture at true acupuncture points, and true acupuncture needling at nonacupoints.
Yeah, it’s a real bitch when your study shows about as conclusively as possible that acupuncture is nothing more than a placebo. Not that our intrepid investigators don’t try very hard to find something–anything– to salvage this study:
One limitation of our study may have been that 8 treatments over 4 weeks were insufficient to achieve maximum benefits from true acupuncture, because other studies have demonstrated the need for longer treatment periods to demonstrate statistical differences between the treatment and control groups. For example, in patients treated with acupuncture for osteoarthritis of the knee, Berman et al44 found no significant effects on pain scores until the 14th week of treatment, when pain had decreased by 40% in the true acupuncture group compared with 30% in the sham control. Even then, it is not clear whether the 10% difference would be clinically important.
Actually, it is, and it wouldn’t be.
The bottom line is that this study is yet another in an increasingly long line of studies that demonstrate that acupuncture is nothing more than an elaborate and fancy placebo. Personally, if we’re going to start using placebos to treat arm pain, I’d hope that we could find one that doesn’t necessarily involve sticking needles into one’s body to achieve its effects. Better yet would be to find and use therapies that actually produced a result greater than that of a placebo. Unfortunately, acupuncture isn’t one of those therapies.
Also unfortunately this study is yet another in a long line of negative studies funded by the National Center for Complementary and Alternative Medicine. Yes, indeed, it’s your tax dollars at work again.
Goldman, R.H., Stason, W.B., Park, S.K., Kim, R., Schnyer, R.N., Davis, R.B., Legedza, A.T., Kaptchuk, T.J. (2008). Acupuncture for Treatment of Persistent Arm Pain Due to Repetitive Use: A Randomized Controlled Clinical Trial. Clin J. Pain, 24(3), 211-218.