Tai chi for fibromyalgia in the NEJM: A triumph of the Trojan horse

I tell ya, I go away for a few days and something always seems to happen that I’d be all over if I were at home and blogging normally. Either something major happens in the anti-vaccine movement or there’s a new study being touted by woos or womthing else big happens. In the old days, I’d try to cover it anyway, but lately I’ve learned just to let it go until I get back home. If I’m still interested in it, the end result will usually be better, and if I’m not still interested in it then it’s probably better that I never bothered writing about it anyway. This particular bit of blog material happened to appear in, of all places, in the New England Journal of Medicine, which a mere three and a half weeks ago published a credulous and one-sided article about acupuncture for low back pain.

Last week, as I was on my way to Chicago, it was tai chi.

Last Thursday, the NEJM published an article reporting a study of the use of tai chi in treating fibromyalgia. Entitled A Randomized Trial of Tai Chi for Fibromyalgia, the study comes out of the Tufts University School of Medicine and the Newton-Wellesley Hospital in Boston and was carried out by a team led by Chenchen Wang, MD, MPH. Not surprisingly, the study has gotten a lot of play in the media, for example, in this story in the L.A. Times, which is at least reasonably restrained, probably because it an AP wire story by Marilynn Marchione, who has written some excellent articles about “alternative” medicine before. Even the normally reliable GoozNews seems smitten with this study beyond what it rates, characterizing it as “rare victory for the National Institute of Health’s National Center for Complementary and Alternative Medicine and Sen. Tom Harkin (D-IA), who routinely comes under fire for pushing funding for these types of studies.” I’m less impressed. You’ll see what I mean in a few minutes, I hope.

First, however, let’s look at the study itself.

Fibromyalgia sucks. If there’s one condition or disease (or whatever you want to call it) that is pretty poorly understood by science thus far, it’s fibromyalgia. Indeed, like some diseases that are also not well understood, there has even been some debate whether or not fibromyalgia exists as a distinct disease, although that debate appears to have been largely settled in favor of fibromyalgia as a clinical entity. Unfortunately, the pathogenesis and pathobiology of fibromyalgia are not well understood, and there is little agreement on precise diagnostic criteria. The authors of this study describe it thusly:

Fibromyalgia is a common and complex clinical syndrome characterized by chronic and widespread musculoskeletal pain, fatigue, sleep disturbance, and physical and psychological impairment.

And an accompanying editorial by Gloria Y. Yeh, M.D., M.P.H., Ted J. Kaptchuk, and Robert H. Shmerling, M.D. describes fibromyalgia thusly:

Fibromyalgia is a common and poorly understood pain disorder that afflicts an estimated 200 million or more people worldwide.1 The lack of objective abnormalities detected on physical examination and standard blood and imaging tests has led many physicians to question the existence of this disorder.2 However, for those experiencing the pain and other associated symptoms (including fatigue, stiffness, and nonrestorative sleep), there is little doubt that the condition is real — and so is the need for relief. Studies over the past decade suggest that fibromyalgia may be due, at least in part, to an alteration in pain sensitivity in the central nervous system.3 Other potential mechanistic contributors include a genetic predisposition, emotional or physical stress, disordered sleep, and neurohormonal dysfunction.

These sorts of conditions are often the most frustrating of conditions to deal with, particularly for patients but also for doctors. After all, no one goes into medicine to tell patients that there’s not much we can do for them; yet that’s what doctors all too often end up having no choice but to tell fibromyalgia patients. Not surprisingly, patients are not satisfied with this. They’re suffering, and they want relief. Also not surprisingly they’re willing to try almost anything, including the rankest forms of quackery peddled by unscrupulous quacks.

Of course, tai chi is not a rank form of quackery pedaled by unscrupulous quacks. It’s a martial art that combines slow, deliberate, graceful exercises with deep breathing and relaxation. I was merely using the reference to quacks to explain why it’s not surprising that physicians might consider tai chi as an intervention in patients with fibromyalgia, given that the standard interventions recommended consist of exercise, sleep hygiene, and medications. All of these interventions have problems, particularly the pharmacological interventions, which in general aren’t all that effective. One might even say that demonstrating their efficacy to be greater than that of placebo interventions has been difficult. Although exercise is helpful in fibromyalgia, it is not known what kind, intensity, or combination of exercise modalities works, and, of the ones that may work, which one works the best. Tai chi, therefore, would appear to be an excellent candidate exercise regimen because of its gentleness, requirement for control, and relaxation. So testing tai chi in fibromyalgia is not an unreasonable hypothesis.

In this particular study, the design was fairly straightforward and is summarized below (click for a larger image):


Basically, the study examined 66 patients diagnosed with fibromyalgia by the time randomization was complete, of which 33 were assigned to the tai chi group and 33 to a sham group that underwent stretching exercises. This design had 78% power to detect a difference between means at a significance level of 5% with the use of a two-sided t-test.

The tai chi intervention was described thusly:

The tai chi intervention took place twice a week for 12 weeks, and each session lasted for 60 minutes. Classes were taught by a tai chi master with more than 20 years of teaching experience. In the first session, he explained the theory behind tai chi and its procedures and provided participants with printed materials on its principles and techniques. In subsequent sessions, participants practiced 10 forms from the classic Yang style of tai chi18 under his instruction. Each session included a warm-up and self-massage, followed by a review of principles, movements, breathing techniques, and relaxation in tai chi. Throughout the intervention period, participants were instructed to practice tai chi at home for at least 20 minutes each day. At the end of the 12-week intervention, participants were encouraged to maintain their tai chi practice, using an instructional DVD, up until the follow-up visit at 24 weeks.

The control intervention consisted of this:

Our wellness education and stretching program similarly included 60-minute sessions held twice a week for 12 weeks.19 At each session, a variety of health professionals provided a 40-minute didactic lesson on a topic relating to fibromyalgia, including the diagnostic criteria; coping strategies and problem-solving techniques; diet and nutrition; sleep disorders and fibromyalgia; pain management, therapies, and medications; physical and mental health; exercise; and wellness and lifestyle management.20 For the final 20 minutes of each class, participants practiced stretching exercises supervised by the research staff. Stretches involved the upper body, trunk, and lower body and were held for 15 to 20 seconds. Participants were instructed to practice stretching at home for 20 minutes a day.

One can certainly argue whether the sham control intervention was an appropriate sham control or not. It would depend to some extent what sort of stretching exercises were included, as well as other factors, such as the quality of the instructional material and the attention paid to the patients by the wellness education instructor. Also, one might imagine that the self-massage part of the tai chi intervention might have some benefit. It’s hard to say.

Patients were evaluated before starting the trial and then periodically for the severity of their fibromyalgia symptoms. The primary outcome measure was the change in the score on the Fibromyalgia Impact Questionnaire, a 100-point scale that estimates intensity of pain, physical functioning, fatigue, morning tiredness, stiffness, depression, anxiety, job difficulty, and overall well-being at 12 weeks. Secondary outcomes included the FIQ score at each week, as well as a number of other measures of fibromyalgia severity, chronic pain, and quality of life. Participants continued their regular medications and treatments, making note of any changes that occurred during the study period.

The results demonstrated improvement in FIQ scores in the tai chi group from 62.9±15.5 to 35.1±18.8, with an improvement in the control group from 68.0±11 to 58.6±17.6, with a highly statistically significant p-value. Investigators also noted improvements in other measures of pain and quality of life, all in favor of the tai chi group. More patients in the tai chi group had discontinued their medications to treat fibromyalgia by the end of the study period, but the difference was not statistically significant.

As you might expect, this study had a number of limitations. First, it isn’t very large. Another problem is that it looked at fairly short term outcomes. Fibromyalgia is a chronic condition for which 12 to 24 weeks do not represent a sufficiently long period of time to judge whether any potential responses will be durable. Previous studies have found that exercise can help fibromyalgia but that the effects are not always durable. This is particularly true given that fibromyalgia, like many chronic pain syndromes, is prone to placebo effects due to interventions, regardless of whether they’re effective or not. That’s one aspect of the condition that makes fibromyalgia so difficult to study.

More importantly, one has to remember that this study was not double-blind. The authors justify this lack of blinding by arguing that there is no accepted and validated sham tai chi intervention. Fair enough. It took many years before scientists studying acupuncture managed to develop and validate various forms of sham needles that successfully blinded both patient and practitioner as to whether they were undergoing “true” or sham acupuncture. The authors also point out that investigators told study subjects that tai chi was being tested. Rather, they told subjects that they were testing the effects of two different exercise training programs, one of which was combined with education. Fair enough as well, but whether this “deemphasis” of tai chi worked or not to minimize expectations and placebo effects is highly arguable. Do the investigators really think that most of the tai chi group didn’t realize they were doing some form of “Eastern” martial art? Yeh et al sum up the difficulty in coming up with an adequate sham intervention for tai chi very well:

The authors state that they tried to minimize any a priori differences between expectations for tai chi and the control intervention, which consisted of stretching and health education, and they report that expectations in the two groups were similar at baseline. However, it seems likely that when a persuasive and enthusiastic teacher of tai chi first explained its potential benefits to the class, expectations in this group were heightened. The authors dutifully suggest that a sham tai chi intervention would have been desirable as a control. Ideally, a placebo control matches all aspects of the therapeutic intervention except for the “active” element of that intervention. But what is the active element of a complex, multicomponent therapy such as tai chi?11 Is it rhythmic exercise, deliberate and deep breathing, contemplative concentration, group support, relaxing imagery, a charismatic teacher, or some synergistic combination of these elements? If so, would the matched control include awkward movements, halted breathing, participant isolation, unpleasant imagery, or a tepid teacher? Would the resulting sham intervention be credible, valid, or even genuinely inactive?

These are all excellent questions. Unfortunately, Yeh et al then conclude that a “quixotic” search for the “ideal sham” is a waste of time, which is rather annoying because such a search doesn’t have to be quixotic. All you have to do is to break down tai chi into its key elements: the slow, graceful stretching exercises, the breathing exercises, the relaxation exercises. On the other hand, maybe Yeh et al have a point–except that they didn’t go far enough. I’ll show you what I mean.

Let’s for the moment and for the sake of argument accept the findings of Wang et al. Let’s say that tai chi is the greatest thing since sliced bread and that it slaps down fibromyalgia symptoms better than anything we’ve yet come up with. Let’s assume all of those things, just for the moment and then think about it. What thought comes to mind to you? I know what thought comes to mind to me:

Why on earth is this result “provocative” (as Yeh et al describe it) or even the least bit surprising? Why on earth is it in the least bit “alternative”? Stripped to its essence and particularly stripped of its woo elements about qi, all tai chi is is exercise and relaxation, and we already know that exercise is useful for fibromyalgia! The only question is what kind, intensity, and regimen does the most good, and this study answers that question to the extent that it tells us that an exercise regimen resembling tai chi is pretty good. Excellent! We can build on that! But why is this finding “alternative”? It shouldn’t be. Why was this study funded by the National Center for Complementary and Alternative Medicine (NCCAM) and why is it being touted by NCCAM (and GoozNews) as a “success”? It’s not “alternative.”

As I’ve said time and time again, exercise and relaxation interventions have been completely co-opted by the CAM movement as “alternative” when there is no a priori reason that they should be considered anything other than science-based interventions. Just because they are difficult to study doesn’t make them any less science-based. However, CAM supporters know that co-opting less controversial modalities with some degree of prior plausibility that have no real reason not to fall under the mantle of science-based medicine allows them to slip the woo in alongside it. The Trojan horse is diet, exercise, and natural products derived from plants (i.e., herbal medicine). The Greeks jumping out of the belly of the horse is the woo, including acupuncture, “energy medicine” (including the “qi” aspect of tai chi), and even homeopathy.

Can anyone tell me again why this study is in any way “alternative”?

Why are we seeing videos like this one, for example, from the Mayo Clinic?