NY Times FAIL – Acupuncture

Today I found this article in NY times health – touting the benefits of acupuncture used for depression relief in pregnant women.  I have a few problems with the journalist, Shirley S. Wang, failing to remain skeptical in her piece.  But this comes as no surprise, so I won’t bother to point out obvious flaws in the media.

I will point out the flaws in the study.  Please find the original full text, here.  So after reading this paper I have come to a few conclusions.

1) Very small sample size of 150 patients, only 141 treated.

2) The study is blinded ONLY for the depression scoring, not administration of acupuncture.  In fact, their release form states “Participants in one of these two groups will receive acupuncture that focuses on depression symptoms and the other treatment will not.” Makes sense, getting a needle shoved in your skin is easy to notice.  So, these “randomized and blinded” participants knew what was coming and assumed that any poking was supposed to help their depression.  Furthermore, this (much better) study has shown that simulated acupuncture with toothpicks works just the same.  Careful, this is compelling evidence against the validity of acupuncture – not evidence for the usefulness of stimulated “pressure points”.  It is a placebo effect.

3) Study assumes validity of “depression specific” and “non-specific” acupuncture.  Meaning one method of pin sticking somehow cures depression over another.  What is this based on?  Oh wait, they say exactly the exhaustive scientific evidence right here: “patterns of disharmony according to the principles of traditional Chinese medicine”. I’m sorry, please, give me modern medicine over 2,000 year old mythical beliefs.  How was the average quality of life and life expectancy for an ancient chinese person?  Pretty damn horrible, and I’m sure the average life expectancy did not exceed 35 years.

4) Basing depression scores on one administration of the DSM-IV Hamilton Rating Scale.  Not being a psychiatrist I can’t speak to the efficacy of these tests.  But, I will go out on a limb and assume that a stronger baseline for depression should be established before comparing results.  The test may be accurate, but why not administer it more than once to reduce noise.

5) Selection of massage as a second control.  This is a bad attempt at token skepticism, they even state right off the bat “Massage was conceptualized as a control treatment because, although it improves mood immediately after a session, there is insufficient evidence to support its efficacy as a treatment for depression.” When I design a study I like to look at all known factors that I believe will disprove my hypothesis.  Choosing something that you already believe will fail only shows their hand of gross bias.

6) Failure to control for socioeconomic factors.  67% of the participants were white, the majority of which were well-educated.  They even go as far as to state in the discussion “Therefore, results might not generalize to specific minority groups that were underrepresented in our sample”. Is it just me, or does this statement negate the entirety of their research?  They are freely admitting that acupuncture might not work as well in other minority groups.  Why could this be?  The only logical and scientific answer is that a placebo effect differs across socioeconomic boundaries.  If, after all, acupuncture was a legitimate medical science, there would be an insignificant difference seen across physiologically identical organisms.

This study is appallingly bad science.  These researchers are beginning with the premise that acupuncture works, and searching for data to support their claims.  This is exactly opposite of how to conduct real science.  And, our health reporter at the NY Times didn’t even bat an eye.  Fail for you Mrs. Wang, and fail to the NY Times.