There are so many things wrong with this study, I don’t know where to start.

First of all, there is no randomized control group, or a group that did a different exercise routine. The only thing they can concluded is that some exercise will help manage diabetes, which is kind of old news. They haven’t shown that Tai Chi is any better than any other exercise.

Even worse, instead of a control group, they compared people who stuck with the Tai Chi program to those who dropped out. The groups weren’t randomly assigned, and there was no controlling for other factors that could be related to why some people dropped out of the program. It even says in the results that the group that adhered to the Tai Chi routine also performed significantly more diabetic self-care activities. Perhaps that group is just better at sticking to a routine?

This is such a poorly designed study, I can’t believe anyone would publish it. It doesn’t tell us anything new about ways to manage diabetes. What surprises me is that the flaws are things that could have been easily remedied, but these researchers decided to go ahead with this terrible design anyway. Maybe because Tai Chi isn’t any better than any other exercise at helping to manage diabetes?

Alexis said: “Maybe because Tai Chi isn’t any better than any other exercise at helping to manage diabetes?”

Aw, does someone not like Tai Chi? Dr. Roberts is quoted as saying “Tai chi really has similar effects as other aerobic exercises on diabetic control,” not that Tai Chi was better than any other form of exercise for controlling diabetes. I can’t imaging my grandmother getting much aerobic exercise, but I can imagine her doing Tai Chi.

I don’t have anything against Tai Chi, and I certainly wouldn’t tell anyone not to do it. My beef is with the actual design of the study, as I detailed in my previous comment. There really aren’t any conclusions to be drawn here except that exercise (but not specifically Tai Chi) is good for diabetic people. This is not news.

Exercise being good for diabetes is old news, as you correctly pointed out. But Tai Chi is not exercise in the traditional sense, so the fact that it had a measurable impact on the physiological factors associated with diabetes is news.

OK, having read over the actual study, I understand your concern about the controls. However, it’s not like the authors of the paper are trying to hide these facts. The call the study “quasi-experimental” and clearly state in the discussion that the “findings of the present study should be interpreted with caution,” because of potential self-care differences between the two groups. It certainly looks to me like an honest presentation of some interesting results that is publication-worthy and may help spur future research efforts. The indication that Tai Chi, which is accessible to older individuals, has similar effects as other moderate forms of exercise on diabetes is good news.

I agree that it’s good for older people to have some low impact aerobic exercises, but this study doesn’t show that Tai Chi is any more effective than any other low impact exercise. My question is, why bother doing a “quasi-experimental” study at all? How difficult would it have been to add a control group and do a study that is actually empirical?

Again, the study doesn’t say Tai Chi is better than other low impact exercises, just that it is as effective. The conclusion in the abstract states that Tai Chi “count be an alternative exercise intervention to increase glucose control.” This is itself it worth publishing, especially since the paper’s introduction mentions that previous studies on the subject gave mixed results. What would you have expected the control group to show? Do an individual’s rate of decline in fasting blood glucose and hemoglobin A1c levels improve if they do nothing? I’m asking sincerely, since diabetes research is not my area of expertise. If it’s a well-known fact that an individual’s diabetic indicators do not improve significantly in the absence of any lifestyle changes, then a control group would not even be required, correct?

I understand you are considering the study, as you would being in the lab. Tai chi though improved lives of millions of people over 5,000 years. And the best proof for that point is that we are still doing it. I encourage you to try!

“the best proof for that point is that we are still doing it.”

That argument is bogus.

Astrology also has a pretty long history. As does chuild abuse.

Doesn’t mean that they have any proven scientific value.

Hi Bella. Yes, I agree that Alexis and I are both looking at it from the point of view of a scientist or a researcher. Dr. Roberts and I actually know each other because we both train under the same Tai Chi instructor. I do Tai Chi because it’s a wonderful practice that I enjoy immensely rather than for any empirically demonstrated scientific or medical benefits. There is value in researching it from an objective point of view, but all the research in the world will not actually give you the experience. I’m actually of to Tai Chi class right now!

True Michael, proof of scientific or medical value is obtained by doing objective research. However, Bella said that the continued practice of Tai Chi was proof that it improves lives, not that is has scientific value. One might even hypothesize that astrology improves the lives of those who believe in it.

The fact that people do something is not proof of anything.

Again, I’m not telling anyone not to practice Tai Chi. I’m just saying that all this study shows is that exercise helps people manage Diabetes. It doesn’t prove anything about Tai Chi at all. The researchers didn’t compare Tai Chi with anything, so you can’t even draw the conclusion that Tai Chi is as good as other exercises.

There should have been a group doing some other exercise routine for comparison. They should not have compared the group that stayed in the Tai Chi program with those who dropped out, as these groups are not even close to being randomly selected. If they wanted to compare Tai Chi to doing nothing, they could have had a randomized control group that does nothing.

These are all pretty simple fixes, and anyone who is conducting research should know this stuff. I understand that they’re calling this a “quasi-experimental” study. That, to me, is a red flag. “Quasi-experimental” studies are not scientific. They admit more research needs to be done, which is good. I hope they do it right next time.

Wikipedia has a page defining and discussing quasi-experimental design:

http://en.wikipedia.org/wiki/Quasi-experimental_design

Quasi-experimental design refers mainly to the lack of random assignment. Wikipedia defines it as a “scientific research method,” implying these studies can be scientific. The paper is a valid piece of work in its own right and hopefully will spur additional research with larger groups of individuals and more stringent controls.

Thank you very much for your comments. I teach Tai Chi, Martial arts, Chinese Yoga and several other disciplines that we use to improve and enrich peoples’ lives. I am quite familiar with research methods as I have a science degree from Russia and an MBA. However, our scientific knowledge is still too minimal to comprehend fully the arts that passed to us for millennia.

Several of my students were participating in the Tai Chi Study at UCLA and I will be the first to comment how incomplete was the study. Though, trying to always look on the positive side, I welcome any study that will bring the public towards this extremely beneficial practice. In fact, the evidence is so overwhelming in my experience, as a teacher, that I would recommend everyone to learn it or find a place to practice. The long term benefits of this practice will be only possible to trace as the practitioners would most likely outlive the test takers! Health care would have been much less costly, if more people practiced it.

[...] Futurity.org – BEST OF 2009: Tai chi lowers glucose Share and Enjoy: [...]