Acupuncture is one of the CAM modalities most widely accepted as scientifically proven to be effective, at least for pain and maybe nausea. Even providers skeptical of the mystical roots and language of the practice will often suggest that it might have some real benefit. Unfortunately, the bulk of the good quality clinical research in humans doesn’t support this notion. When compared with “fake” acupuncture (needles placed in non-traditional locations or depths, retractable needles that don’t actually pierce the skin, toothpicks twirled on the skin, and so on), “real” acupuncture generally gets the same results as the fake procedure; namely a small improvement in subjectively reported pain or nausea scores. There’s no question that sticking needles in people (or mice) has measurable effects on the body (releasing various chemicals, effecting pain receptor activity, and so on). This is a long way, however, from demonstrating that sticking needles in particular places and a particular way has meaningful clinical benefits (i.e., that “acupuncture works”).
The debate about the scientific evidence for acupuncture is muddled by the lack of a consistent definition for what acupuncture actually is. Many studies claiming to investigate acupuncture actually use “electroacupuncture” (a CAM pseudonym for what scientific medicine calls TENS, Transcutaneous Electrical Nerve Stimulation). This was the case for a recent set of studies in dogs, and it is also the case for a recent study in humans:
Suarez-Almazor, M., Looney, C., Liu, Y., Cox, V., Pietz, K., Marcus, D., & Street, R. (2010). A Randomized controlled trial of acupuncture for osteoarthritis of the knee: Effects of patient-provider communication Arthritis Care & Research
Orac at Respectful Insolence discusses the study in detail. In brief, it compared electroacupuncture with “fake” electroacupuncture (needles in different spots and different amount/duration of electrical stimulation) and with a no-treatment control, and it also compared these treatments between groups of patients given high expectations of success by the providers and those given neutral expectations. The results?
Not surprisingly, there was no difference between “real” and “fake” acupuncture. Both groups reported some improvements compared to the group that got no treatment, which is exactly what you’d expect if the “real” treatment was a placebo just like the “fake” treatment. What is cool about the study is that there were several measures that differed significantly between patients given high expectations and those given neutral expectations, regardless of which treatment they got. Placebo effects are well known to be greater when the fake treatment is presented with confidence by a supposedly knowledgeable professional. In this study the way the treatment was presented to the patients affected how much benefit they got from it and mattered more than which treatment they got, just as one would expect for a placebo therapy.
Of course, real therapies will also appear to be more effective for subjectively reported symptoms if the patient is given high expectations. However, since there was no difference between the effects of “real” and “fake” acupuncture, but there was a difference caused by the expectations the patients were given, the study is a nice illustration of both the fact that acupuncture is a placebo and that expectations are a key element in achieving placebo effects.
This is of particular concern to me as a veterinarian because I believe it is impossible to influence the expectations of my patients about the benefit of the treatments I give them. So unlike humans, they are unlikely to experience any benefit from placebo effects based on expectancy. Unfortunately, their owners are very likely to be influenced by a vet presents a therapy, which leads to a situation in which the client and the vet think the treatment is helping when in fact the patient feels no better. It is this placebo effect by proxy that I think keeps many ineffective CAM therapies alive and profitable in veterinary medicine, especially since the large, well-designed studies necessary to show the underlying reality about the treatments are seldom possible due to cost and practical constraints. We need to take not of such studies done in humans and recognize the implications they may or may not have for our field given the differences between humans and our patients.