Science in the Study of Taijiquan

Often, limiting ones study to specific factors that one is hoping to see can result in poor experimental design and therefore bias (skewed data). For example, if researchers are interested in Taijiquan for insomnia, then they will typically recruit study subjects already diagnosed with insomnia in order to see if Taijiquan can improve their condition. This leads to “selection bias” where the participating subjects who already have the condition are participating in something new that they may expect to help their condition. Their expectations, or the “alternative” nature of the therapy, or the reputation of Taijiquan’s benefits for health, or any number of other factors, can lead to the expectation of benefits. These expectations, if not satisfactorily controlled for, often produce unwarranted conclusions.

Subjective evaluations (e.g., participant questionnaires or self reporting) are often unreliable, and better studies try to include objective measures. MBIs can help participants feel better, but it is sometimes difficult to separate feelings from what is actually happening physiologically, especially when objective tests are not included. Analyzing feelings or opinions can easily misattribute improvements to what may instead be attributed to the placebo effect.

Unfortunately, the difficulty of designing appropriate rigorous controls often leaves researchers taking a simple approach and using inactive or wait list subjects as controls. But even gentle activities of many types (e.g., walking, cycling, stretching, etc., and even socializing and other interactions in group settings) can be found to benefit inactive people. Positive results would merely indicate that Taijiquan is better than inactivity! This is not particularly meaningful! Better are studies that use types of non-Taijiquan activities as controls while also including objective measurable criteria in their research designs.

To know if Taijiquan can be used therapeutically, researchers need to compare Taijiquan practice with alternative treatment options and show that Taijiquan’s benefits are similar to, or in some way better than, other options. This is often not done in preliminary experiments. If experiments show benefits from Taijiquan, but also show that alternative activities or treatments produce similar gains, then it may indicate that practitioners receive benefits, but that alternative approaches may be better as treatments.

Results from scientific studies on Taijiquan to date are rarely confirmed, primarily due to the limitations of the research protocols and methodology as well as the heterogeneity of the practices used in the studies. It is not uncommon for independent researchers to come to conflicting conclusions. One study does not provide conclusive evidence and must be tested independently in order to build support for any preliminary evidence. To date this has rarely been done. Different research protocols can result in differing conclusions, and one study is not enough.

It seems to me that lay readers of scientific research often tend to isolate things that look good to their particular interest/bias, and report an overblown significance for those results. Instead, one needs to evaluate the research methodology (a required component of serious research publications) and attempt to understand the limitations of the research. Many studies report statistical significance at a quite poor probability value, but lay readers put greater importance to the results than this confidence level warrants. Worse, some studies report improvements that are NOT SIGNIFICANT, but lay readers report those improvements as if they were important.

Reporting scientific results that indicate that Taijiquan can improve various conditions may make Taijiquan practitioners feel good about the art that they practice, especially for those whose practice emphasizes health. But reporting marginal improvements as if they were significant could be considered as being deceptive and self-serving. Many scientific studies really only show that short term practice of Taijiquan (or, more accurately, a Qigong derived from Taijiquan movements) improves the health of non-healthy or sedentary individuals, rather than showing long term benefits for healthy practitioners.

Preventative benefits for long-term practitioners have rarely been addressed. One can infer that benefits shown to be significant in scientific studies would also be beneficial in the long term, but only if those studies are performed on similarly active controls. If walking, for example, was similarly effective for the studied conditions, then that may be a better therapy than Taijiquan because people already know how to walk – there is no instruction needed, only motivation.

To seriously consider Taijiquan for therapies, studies need to be more rigorous than most current studies have been. Most studies also don’t examine the long term effects of serious in-depth practice of this art. We are left to wonder what, if any, potential health benefits Taijiquan practice confers are superior to other activities or treatments. Taijiquan includes physical and mental/cognitive factors as well as social and environmental influences, any of which could be responsible for observed therapeutic benefits.

For example, a relatively well designed study (Mortimer, et al., 2012, J. of Alzheimer’s Disease, 30: pp. 757-766) that had superior and unbiased randomized participant selection and several reasonable controls found that the brain volume for non-demented Chinese elders increased in the Taijiquan practice group (meeting 3x per week for 40 weeks with 20’ of warm-ups and balance training, 20’ Taijiquan, and 10’ cool-down exercises), but that a spirited debate group used as a control achieved similar benefit levels.

In this case, it appears that social interactions were as effective as was class instruction in Taijiquan. The researchers, however, reported that the debaters enjoyed it so much that they continued on their own for years after the study ended, indicating that debate groups may be a better therapy than Taijiquan.

Until preliminary research is confirmed by additional studies by independent groups, I would caution against making assumptions about the therapeutic benefits of Taijiquan based on the information that is currently available. Most of the research so far has mainly shown that sedentary or unhealthy individuals could benefit from practicing Qigong exercises that are based on Taijiquan. They mostly do not address long-term practice or benefits for otherwise healthy and active practitioners. They also rarely examine possible health benefits from either Taijiquan practiced for sport or for martial arts.

Conclusions that are unverified by independent research groups are too preliminary to rely on for therapeutic recommendations. We need to await additional robust studies before we can realistically recommend Taijiquan (or Qigong based on Taijiquan movements) as therapy. Robust studies use large numbers of randomized participants using several appropriate controls.