A Look at Moxibustion

by chelseajin

I first heard about moxibustion from an attending physician at the Mobile Health Clinic. To recap, it’s a traditional Chinese medicine technique where you put incense of Mugwort (Artemesia Vulgaris) as close as you can to a specific point on the fifth toe. Doing this treatment for several weeks is supposed to help turn a fetus within the womb from a breech position to cephalic (head down). I actually like drinking mugwort tea, especially for menstrual cramps. I feel that it does help, but that could also be because my mother told me it does before she introduced me to it.

Even with my Korean background, I was skeptical. I considered doing my own chart review as a small project, but I don’t think there are enough patients to do it. (A thought for another time). Even with remote access to libraries and databases from my City College ID, I get limited access to journals. I did manage to print out 2 articles, A randomised controlled trial of moxibustion for breech presentation by Francesco Cardini et al. and Effectiveness of a acupuncture-type interventions versus expectant management to correct breech presentation: a systematic review by van den Berg et al.

Side note: scientists and doctors they may be, but their grammar sure needs work. I realize this is coming from someone who consistently ends sentences with prepositions.

I was attracted to the Cardini article since it was a randomized control trial, which is supposed to be a very strong study. An RCT is a “real” experiment, where you ideally have 2 same/similar groups, one undergoes a placebo or standard-of-care treatment while the other has the experimental. You can see how this would be more reliable than just picking out one or two people and making them try something new with no one to compare to.

The Bad

Summary: Cardini looked at six obstetric departments in Italy and finally had 65 women use moxa and 58 on observation only. All the women were pregnant with their first babies (primagravid).

Results: Inconclusive.

Cephalic presentations at delivery were 34/65 (52%) in the moxa group and 29/57 (51%) in the control group.

Limitations: (I’m going to note that there were glaring issues with the design)

1. small sample size –> low Power

2. Low external validity

3. poor planning

a. no treatment oversight –> bad compliance

1. Cardini mentions that their yardstick of measurement was 74 people in each group, which would be enough people to provide results showing a 27% difference in treatment vs none. This number wasn’t met. So they are less than 90% confident that their results are even significant. (We now have inconclusive results which might not even matter because the researchers are uncertain. The standard is demonstrating 90-95% confidence)

2. The only admirable part of this study is that the researchers managed to create a homogenous group of women. Of course this also narrows the relevance of the results to this particular population (non-Chinese primagravid women at 32-33+ 3days gestational age with fetus in breech position) aka, it has low external validity (not generalizable results).

3. A major reason why their numbers fell so low is because these women were trained to apply the treatment to themselves at home. So now we wonder:

A) Did they learn how to do it correctly?

B) Are they even doing it at home?

It’s mentioned that several women couldn’t bear the smell of the herbs and so dropped out. It’s possible that the ones who did comply didn’t do it as often as they were told to because they didn’t like the odor. Which brings me to the other point, how did the researchers think they would get reliable results for this type of treatment without actually overseeing the treatments take place? Yes, it can be argued that it’s costly and almost impractical to call these women in everyday/week to watch them do this, but I know from my experience with moxibustion and general public opinion that this is a foreign treatment that smells bad. The natural reaction is to refuse to do it. I’m glad that they at least acknowledge this, but it’s a real shame that they didn’t think of it sooner because it makes discussing the article almost useless.

The Good

The next paper is meta-analysis, meaning it takes a survey of several independent studies and pools those numbers together to find a result. This is the most powerful type of study, but it is subject to some weakness because it depends on finding well-planned studies to combine. It is also vulnerable to publication bias, because articles that have certain results are more likely to be published.

Summary: Three authors independently selected articles to use. They came up with 9 total studies, 6 were RCTs and 3 were cohort studies.

Results: “Acupuncture-type interventions” (acupuncture, moxa, electro-acupuncture) on BL 67 do correct breech position compared to expectant management.

from the van den Berg paper

Interestingly, BL 67 stimulation only works while the fetus is alive, so it’s not a simple reflex reaction.

People have suggested that stimulation of BL 67 increases cortico-adrenal secretion, placental estrogens, and changes in prostoglandin levels, leagind to raised basal tone of the uterus and enchanced movement of the fetus, making version more likely.


1. Publication Bias

2. Relatively small population

3. No standard method for moxa

4. Cultural differences (Chinese vs. Italian, etc)

5. Couldn’t adjust for parity.

6. No placebo.

7. Other characteristics might affect the outcome.

1. There were no unpublished studies found. Actually, I felt that the studies had some significant differences in the population used, but the researchers tried to adjust for this with a “random-effects model”.

2. Ranging from 23-130 people in each study.

3. The treatment is variable, and there’s no agreement on the best regimen. The most common method according to this paper was 1x/day for 15-20 min for as long as 1-2 weeks.

4. 3 of the papers were from Caucasians. 4 were Chinese. 2 Japanese. Considering that the treatment comes from traditional Chinese Medicine, Chinese people may be more bias towards believing it would work (more relevant because there was no blinding)..

5. Multiparous people with breech babies might be more likely to have their babies turn because their uterus have stretched more.

6. There were no placebo treatments used in the control groups because sham/fake treatments would be easily discovered. This is problematic because there’s no way to blind the subjects/researcher to which treatment they’re getting.

7. Age, condition, female babies, fetal motor activities, room temp can all affect the outcome, but there haven’t been studies looking at them.


I’m a little cautious of the results because the Chinese studies are the ones that show a really significant treatment effect, but there are obvious biases that can be figured in. I’d actually have to go hunt down those articles and read them too. The most worrying aspects of the meta-analysis is that they weren’t able to find studies using placebos and that their population was not well-sampled. Since moxa doesn’t have any harmful side effects, I wouldn’t mind asking people to try it out, but I’m not sure if I would completely believe its purported effects. The 2nd study really looks at stimulation of BL 67, not moxabustion, we’d have to consider all traditional Chinese medicine techniques similar. I thought I might come up to a different opinion after reading up, but what I’d probably do as a provider is follow the standard procedure for managing breech position, then for a few weeks before having to revert to manual version, have the patient this try out using moxa/acupunture/electro-acupuncture.


A randomised controlled trial of moxibustion for breech presentation

Effectiveness of a acupuncture-type interventions versus expectant management to correct breech presentation: a systematic review