Acupuncture for Stable Angina 'Too Good to Be True'

F. Perry Wilson, MD, MSCE


July 30, 2019

Welcome to Impact Factor, your weekly pinch of commentary on a new medical study. I'm Dr F. Perry Wilson.

This week, I need to tell you about a study with results that are simply unbelievable. And I don't mean that these results are amazing; I mean I don't believe them. But I hope that in discussing them, I'll share some insight on how to interpret all medical studies.

This was a randomized trial, appearing in JAMA Internal Medicine,[1] of acupuncture for the treatment of chronic stable angina in China. Around 400 patients with stable angina were recruited. At baseline, they were having an average of around 13 anginal attacks per month.

The participants were randomized into four groups, and here is where it gets interesting. First, there was a real acupuncture group, where two needles were placed on specific points that are supposed to lie on invisible energy lines — or meridians — that are beneficial for the heart, according to traditional Chinese medicine.


They called this the disease-affected-meridian group. Electric stimulus was applied to the needles as well. A second group got two electric needles applied to different acupuncture points — nonaffected meridians that are not supposed to be beneficial for the heart.


A third group got nonelectric needles applied to "sham" points — those that are not associated with an energy meridian. And a fourth group got waitlisted; they didn't get any acupuncture at all.

Anginal attacks decreased in all of the groups, but as you can see here, they decreased most in the disease-affected-meridian group, then the nonaffected-meridian group, then the sham group, then the waitlist group.


The explanation offered is that these significant differences demonstrate that, in fact, acupuncture works. It's not just the placebo effect. There is something special about sticking needles in these specific points that reduces angina more than sticking needles anywhere else. If true, this is a groundbreaking finding — an earth-shattering finding, really — that upends everything we think we understand about physiology.

And that I can't believe. I do not believe that there are hidden energy fields in our bodies that can be manipulated with needles. I don't believe that, because if there were, they could be directly detected by any variety of modern electromagnetic sensors.

So, how did these results come about? There are a couple of possibilities.

First, I may be wrong. My belief that undetectable energy paths do not flow through the body — founded on a mechanistic understanding of the universe — may be parochial, small-minded, pompous, and limited. I've been wrong before and I'll be wrong again. But there are other possibilities that we need to consider before we embrace a hypothesis that is inconsistent with observation.

The simplest explanation for these findings is that patients were aware of what group they were in. The study states that participants were blinded to the intervention, but it is quite possible that they could figure it out. Acupuncture is much more popular in China than in the US, and participants may have known that they were getting needles in heart-associated locations as opposed to non-heart-associated locations.

And, of course, the acupuncturist knew.

It is possible that — whether consciously or unconsciously — that information was conveyed to the participant. One easy way to test this is to ask the participants after the trial to guess what group they were in. As far as I can tell, this wasn't done.

Another possibility is that there was systematic error in measurement or recording of the data.

The point is that this study, and all studies, need to be interpreted not just by the P value that pops out at the end but by the likelihood of the underlying hypothesis. I am encouraged that the recent CMS proposal to fund acupuncture for patients requires enrollment in a randomized trial. I would encourage any such studies to robustly assess the quality of blinding.

If this trial had shown that needles in the disease-affected meridians and nonaffected meridians helped equally, better than nonelectric sham stimulation, I'd have less trouble believing it. But ironically, the significant differences between the disease-affected meridian group and the nonaffected meridian group in this trial actually undermine all of the results for me.

That's right. This study is too good to be true.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.