Moxibustion

Earliest textual evidence

The earliest surviving textual evidence usually cited comes from Han Tomb No. 3 at Mawangdui, Changsha, China, excavated in 1972 to 1974 and sealed in 168 BCE. Among the finds were two silk scrolls describing eleven channels and their associated symptoms, which a 1997 historical article suggests present a channel theory that predates the Nei Ching, the Canon of Medicine. A later philological study argues that in this material the channels were linked to moxibustion, and that moxibustion was then an experience-based technique not yet joined to channel theory.

A review by authors affiliated with a Kampo institute reports that the Mawangdui texts concern moxibustion and do not mention acupuncture, and reads this as a sign that moxibustion predates needling. It also dates references to moxa for cold conditions to the Huangdi Neijing in the 2nd century BCE. A Korean historical study places the appearance of needle acupuncture around 190 to 176 BCE. These are interpretations of a small number of manuscripts, and the sources reviewed here do not settle the sequence.

Later development and spread

The Kampo-affiliated review states that moxibustion reached Japan from the 6th century CE, that in Japan's Edo period (1603 to 1867) it was widely practised for pain management, health maintenance and disease prevention, and that publications documenting moxibustion techniques in China grew sharply during the Ming dynasty (1368 to 1644). These dates come from a single secondary source and should be checked against specialist historical scholarship.

Moxa, moxibustion and acumoxa

The Cochrane review of moxibustion for breech presentation defines the practice as burning a herb, Artemisia vulgaris or Artemisia argyi, close to the skin to induce a warming sensation. A review of adverse events describes it as stimulating acupuncture points with burning moxa wool. The Kampo-affiliated history traces the word moxa to the plant names Artemisia moxa or the Japanese mogusa, and notes that Chinese and Japanese join the words for acupuncture and moxibustion in a single term, from which the English hybrid acumoxa is drawn.

Direct and indirect forms

A systematic review of case reports separates direct moxibustion, in which a moxa cone is placed on the skin, from indirect moxibustion, in which the burning moxa is kept apart from the skin, for example by a layer of ginger or garlic. The Kampo history instead lists suspended sticks held 3 to 5 cm from the skin under non-scarring indirect methods, and describes two Japanese direct forms: okyu, using thread moxa in four sizes from string-like to rice-grain size, and a scarring form using larger cones with an ointment intended to encourage the discharge of pus. Classification therefore varies between sources, and a reader should check which technique a study or practitioner means.

Moxa cones and moxa sticks are both reported in trials. The Cochrane review describes sticks of about 1.5 cm diameter and 20 cm length or burning cones at the point Bladder 67, and the Kampo history says sticks are used mainly in China. Smokeless sticks made from carbonised material have also been tested against conventional smoke-producing sticks. The sources reviewed do not use 'moxa cone therapy' as a name for the whole practice; cones are one form of moxibustion.

Techniques in trials for breech presentation

In the 13 trials in the Cochrane review, moxibustion was applied with a stick or cone at Bladder 67 (Zhiyin), located at the tip of the fifth toe. Sessions lasted 14 to 30 minutes, most often 20, and frequency ranged from twice a day to twice a week over 7 to 14 days. Gestational age at recruitment ranged from 28 weeks to 36 weeks and 5 days. The trials took place in Italy (3), mainland China (2), France (2), Japan (2), Australia, Denmark, Spain and Switzerland. These are trial protocols and not a standard clinical schedule.

Home use and smoke control

In a Japanese pilot, women with a singleton breech presentation at 33 to 35 weeks applied moxibustion themselves at home for 20 minutes daily for 10 to 14 days, with the stick held in a stand 1.5 to 3 cm from the skin. One group used a conventional stick and another a carbonised stick in which smoke could be largely controlled. The authors cited earlier reports of nausea and throat problems from smoke as the reason for testing the smokeless version.

Factors that vary between sessions

The sources reviewed describe practice mainly through trial protocols and do not give a general account of assessment, session structure or practitioner roles in clinics. A review of case reports lists factors said to affect the safety of a session: the position, duration and distance between moxa and skin, the proficiency of the practitioner, the condition of the person treated, the presence of smoke and the treatment environment.

Traditional and practitioner framing

The Kampo-affiliated review reports moxa use for cold conditions in the Huangdi Neijing and, in Edo-period Japan, for pain management and for health maintenance and disease prevention. Its authors present moxibustion as a non-invasive treatment for painful syndromes including neuropathy and suggest that it may reduce neuroinflammation through the NF-kB and Nrf2 pathways. This is a hypothesis in a review written to make the case for the practice, and the sources reviewed here do not show that such mechanisms account for any clinical effect.

Breech presentation

Breech presentation at term can cause complications during birth and raises the chance of caesarean section. Moxibustion at Bladder 67 has been proposed as a way of changing breech to cephalic presentation. A Danish trial report notes that, in Chinese traditional medicine, stimulation of this point is recommended to favour cephalic version, by needle, ginger, pressure, laser or moxibustion.

Conditions studied in reviews

Systematic reviews have examined moxibustion for pain of several kinds, including osteoarthritis, scleroma and herpes zoster pain, chronic low back pain, and cancer pain when added to drug treatment. These are conditions that have been studied, and the reviews summarised below do not support transferring a result from one of them to another.

Breech presentation: the Cochrane review

The updated Cochrane review (searches to November 2021) included 13 randomised or quasi-randomised trials with 2,181 women with a singleton breech presentation. Against usual care alone, moxibustion plus usual care probably reduced non-cephalic presentation at birth (7 trials, 1,152 women; risk ratio 0.87, 95% confidence interval 0.78 to 0.99; moderate certainty) and probably made little or no difference to caesarean section (6 trials, 1,030 women; 0.94, 0.83 to 1.05; moderate certainty). The effect on the need for external cephalic version was very uncertain (4 trials; 0.62, 0.32 to 1.21; low certainty).

In one trial of 272 women with a sham moxibustion comparator, moxibustion probably reduced non-cephalic presentation at birth (0.74, 0.58 to 0.95; moderate certainty) with little or no difference in caesarean section. Most trials reported adverse events inadequately, and one trial stopped early was at high risk of bias.

Other breech studies and a guideline

A Danish randomised trial of 200 women at week 33 compared daily moxibustion for two weeks with no moxibustion and found no significant difference in breech position afterwards (risk ratio 1.05, 0.8 to 1.38). Whether it was included in the review's pooled analyses was not checked. A non-randomised Japanese pilot with 60 women found more cephalic presentations after smokeless moxibustion than in controls at the end of treatment (60% against 25%; risk ratio 2.40, 1.04 to 5.56), but no significant difference at birth.

A 2019 French clinical guideline (CNGOF), reviewing literature from 1980 to 2019, states that acupuncture, moxibustion and postural methods are not effective in reducing breech presentation at birth and does not recommend them, advising that women be told external cephalic version should be attempted. This differs from the later Cochrane review, which pooled more trials.

Pain

A 2010 systematic review of moxibustion for pain searched 14 databases and found four randomised trials meeting its criteria, two of which compared indirect moxibustion with drug therapy in osteoarthritis (200 participants; risk ratio 1.11, 1.02 to 1.21). Most other trials were at high risk of bias, and the authors concluded that effectiveness could not yet be determined.

A 2020 review of chronic low back pain (10 trials, 987 patients) found lower pain scores than with Western medicine or acupuncture but not than with core stability training, and rated the evidence very low to low in certainty; the authors said firm conclusions were difficult. A 2023 review of moxibustion added to drug treatment for cancer pain (10 trials, 999 cases) reported better pain relief and lower pain scores than drugs alone and called for higher-quality trials. Its abstract describes no sham comparator and reports no certainty rating.

Gaps in the sources reviewed

A PubMed search for Cochrane reviews with moxibustion in the title returned three records, consistent with versions of the breech review. None addressed pain or other conditions, and other databases were not searched.

Comparators and certainty

Most evidence compares moxibustion, added to usual care or to drug treatment, with usual care or drugs alone, and a comparison with no moxibustion cannot separate the heat treatment from attention or expectation. The Cochrane review noted that blinding participants is difficult with this intervention, and the pain reviews rated certainty very low to low or reported high risk of bias. Findings should be read as preliminary for pain.

Presentation at birth and birth outcomes

In the Cochrane review a lower chance of non-cephalic presentation at birth did not translate into a measurable difference in caesarean section, and effects on need for external cephalic version, premature rupture of membranes and low cord blood pH were very uncertain because of few data. A guideline reached a more negative conclusion on a different search and appraisal, so the breech evidence is not uniformly read as supportive.

Adverse events in breech trials

The Cochrane review states that adverse events were inadequately reported in most trials. In the one study with usable data, 27 of 65 women given moxibustion reported events such as nausea, unpleasant odour, abdominal pain and uterine contractions, against none of 57 controls (very low certainty). The review also lists increased fetal movements, headache and burns from holding the stick too close to the skin. A small Japanese pilot recorded no serious adverse events.

Burns, allergy, infection and smoke

A review of case reports found 64 cases in 24 articles from six countries, mostly from China. Burns were most frequent (43 cases), especially in newborns and people with impaired sensation, followed by allergic reactions (7), infections (6), two cases of fetal harm and one death; one allergy followed inhaling smoke in a closed room. An earlier review found rubefaction, blistering, itching, smoke discomfort and headaches in trials and premature birth and rupture of membranes in pregnancy. Incidence cannot be estimated from case reports. A PubMed search for studies of moxa smoke and indoor air pollutants found none measuring exposure.

Suitability, training and care boundaries

Pregnancy, infancy and reduced skin sensation are the groups in which the reviews report the more serious events. Breech presentation at term is assessed within obstetric care, and the trials tested moxibustion as an addition to usual care and not a replacement. No statutory regulation or training standard was found in the sources reviewed, and professional registers were not searched. The authors of the case-report review advise improving practitioner skill and controlling time, distance, dose and protection.