Monday, August 23, 2010

Yoga Might Not Have An Effect On Practitioner's Moods, But This Study Sure Affected My Mood

"New Study Proves Nothing!!!"  Hmm, I guess I can see why they didn't go with that for a headline, but they very well could have.  According to a study in the Journal of Alternative and Complimentary Medicine or NAMBLA, yoga has "mood altering effects."

I'm not going to trash CAM in general, or the Journal of Alternative and Complimentary Medicine (well aside from that one cheap shot at the beginning).  Instead, I'm just going to arm you with some of the flaws of this study so you'll be prepared when Hoda and Kathy bring it up during hour 4 of the Today Show tomorrow.

No Blinding - When you hear "no blinding," your inner Babel Fish should throw up in his mouth a little bit, then translate that as "useless data."  Especially in a study like this, where the primary indicator of "mood" was self reported.

Ns of Less than 20 - It is perfectly acceptable to use a low number of participants in a pilot study, funds are sparse, and when you are simply trying to figure out which direction research should go in, an N of 19 for the experimental group and 15 for the control group is fine.  That being said, you know what Pilot studies are designed to prove . . . NOTHING.  Even rigorous, well designed studies don't prove anything in and of themselves, they have to be balanced against the state of the scientific consensus.  Small pilot studies are simply a tool for determining which areas of research may yield results, nothing more nothing less.

Self Reporting - Other than a test for GABA levels, self reported anxiety and mood scales were used to determine the effects of Yoga.  Self reporting doesn't by itself invalidate a study, but what it does do is raise a red flag in terms of placebo effects.  The combination of self reporting and no blinding makes this little more than hearsay as there would have been powerful social pressures towards validating the researcher's biases.

GABA testing -  The tests for GABA levels were only performed 3 times, once for a baseline, and twice 12 weeks later before and after the participants participated in their respective exercises.  If you were trying to design a study so that by chance there would be some significant outcome, that is exactly how you would design it; non repeated testing of a small group.  In addition, walking (used for the control) and yoga are very different types of exercises.  Why wasn't non-yoga based stretching in a quiet room used for the control?  My guess is (intentional or not) to help ensure there would be a significant difference between the control and experimental conditions in the data.  From what I understand, resting GABA levels are typically higher (someone please correct me if I'm mistaken), it's hard to imagine that slipped the minds of these researchers.  

No comments:

Post a Comment