Moxibustion and cupping

What it is

Two treatments used alongside acupuncture in Chinese medicine, and independently in several other traditions.

Moxibustion (jiǔ) burns dried mugwort — Artemisia, prepared as "moxa" — on or near the skin to warm points and channels. Direct moxibustion places the cone on the skin, sometimes to the point of blistering or scarring; indirect methods hold a smouldering stick above the point, or place the cone on a slice of ginger, garlic or salt.

Cupping applies suction cups to the skin, traditionally by briefly heating the air inside a glass cup before placing it. Dry cupping is suction alone; wet cupping — Arabic ḥijāma — involves small skin incisions so that blood is drawn into the cup.

Where it came from

Moxibustion is at least as old as needling in China, and probably older: the Mǎwángduī manuscripts from 168 BCE describe channel therapy by cauterisation, with no needles and no points. It spread throughout East Asia, and in Japan direct moxibustion has a particularly long tradition.

Cupping is ancient and independently widespread. It appears in Egyptian medical papyri, in the Hippocratic corpus, in Islamic medicine — where ḥijāma is recommended in hadith and remains widely practised on religious grounds as well as therapeutic ones — and in European folk medicine well into the twentieth century.

It became briefly famous in the West at the 2016 Rio Olympics, when the circular bruises on Michael Phelps and other athletes drew wide attention.

How it is done

Moxibustion: the practitioner burns moxa at selected points, generally until the patient reports strong warmth. Sessions produce a characteristic and lingering smoke and smell.

Cupping: cups are applied for five to fifteen minutes, either stationary or slid across oiled skin. Wet cupping adds superficial incisions before reapplying suction to draw blood.

What we can and cannot say

We can say what the marks are. Cupping leaves circular discolouration because the suction ruptures small capillaries beneath the skin, producing a contained bruise. The colour is blood. The common explanation — that darker marks indicate more stagnation or more toxins being drawn out — has no basis; darker marks indicate more capillary rupture, which depends on suction strength, duration and the individual's skin and vasculature.

We can say the general evidence is weak. Reviews of cupping for pain and other conditions consistently find small, poorly blinded trials at high risk of bias, with results that do not hold up as quality improves. Blinding is genuinely difficult — the patient knows whether cups were applied — which limits what any trial can establish, but that is a reason for caution rather than for credit.

And we can say there is one interesting exception, which is worth knowing precisely because it is so specific. Moxibustion at the point BL67, on the outer edge of the little toe, is a traditional treatment for breech presentation in late pregnancy — the claim being that it encourages the baby to turn. This has been studied more seriously than most of the field, and Cochrane reviews have found some evidence that it may reduce the need for external cephalic version and possibly for caesarean section, while judging the quality of the evidence low and the results inconsistent. It is not established, and it is not nothing, and it remains the most intriguing loose end in this section.

Safety is a genuine issue here, more than with needling.

Moxibustion causes burns, including from direct methods where blistering is intended, and case reports of significant burns and scarring exist. The smoke contains particulates and is a respiratory irritant in enclosed spaces.

Wet cupping carries infection risk where blades or cups are not sterile, including bloodborne viruses, and repeated heavy bloodletting has caused iron-deficiency anaemia in documented cases. Fire cupping has caused burns. There are published case reports of serious harm, including skin infections and, rarely, worse.

We cannot say these practices do nothing — cupping produces real local effects on blood flow and tissue, and moxa produces real heat, and both may act on pain through mechanisms similar to those discussed under acupuncture. What can be said is that the traditional explanation for the marks is wrong, the evidence base is thin, and the risks are real enough that sterility and practitioner competence matter more than the theory.

Further reading

• Cochrane reviews on moxibustion for breech presentation, for the one genuinely open question.
• Case-report literature on cupping and moxibustion adverse events — sobering and short.