Article Text


Does the burning of moxa (Artemisia vulgaris) in traditional Chinese medicine constitute a health hazard?
  1. John Wheeler1,
  2. Belinda Coppock2,
  3. Cecil Chen3
  1. 1
    Secretary, British Acupuncture Council (BAcC), London, UK
  2. 2
    Member of BAcC executive committee and safe practice committee, London, UK
  3. 3
    Member of BAcC executive committee and chair of safe practice committee, London, UK
  1. Mr J Wheeler, Secretary, British Acupuncture Council, 63 Jeddo Road, London W12 9HQ, UK; john{at}


Background: Concerns have been expressed about potential toxicity of the smoke produced by the burning of moxa in traditional Chinese medicine. With the advent of strict anti-smoking legislation in the UK, it was decided to test the volatiles produced by moxibustion and compare them with current agreed safe exposure levels.

Method: Moxa, in the form of cigar shaped moxa “sticks” or “rolls”, was tested under International Organization for Standardization conditions in a tobacco testing laboratory, and the quantities of a number of pre-determined volatiles measured. The smoke tested was “sidestream smoke”, the smoke which arises from the burning tip of the moxa. The test results were then scaled up to reflect normal use and to provide direct comparisons with agreed national safety standards for both short- and long-term exposure levels.

Results: Levels of only two volatiles produced were equivalent or greater than the safe exposure levels, as was the carbon monoxide level reported, both as a consequence of using worst case assumptions for comparison. Under normal operating conditions neither volatile nor carbon monoxide would present a safety hazard. One group of chemicals tested, the aromatic amines, with known carcinogenic properties have no agreed safety levels. Results for these in the study compared favourably with background levels reported in urban environments.

Conclusion: There are no immediate concerns arising from the continued use of moxa as a therapeutic modality in traditional Chinese medicine. Further testing may be required to establish whether current recommendations for ventilation and cleansing of treatment room surfaces may need to be revised. Stronger recommendations may also be necessary on the inadvisability of using moxa on broken skin.

Statistics from

The introduction of new cigarette smoking legislation in the UK has raised safety concerns about the smoke generated by other, mainly non-industrial, processes. Industrial processes are relatively tightly controlled, with emissions subject to clearly defined safe or permissible occupational exposure limits such as those outlined in the UK Health and Safety Executive’s publication EH40 Occupational exposure limits 2007. The same testing has not usually extended to “social” smoke products. However, although the legislation was primarily aimed at tobacco smoking, usually defined as the transitive act of smoking a tobacco or tobacco-substitute product, questions have been raised about church incense and other social or non-industrial smoke-generating activities.

The British Acupuncture Council (BAcC) has been aware from patient feedback that some members of the public are concerned about the levels of smoke produced by the burning of moxa, a Chinese herb (Artemisia vulgaris) similar to mugwort. This is extensively used in Chinese medicine, where it is applied directly onto the skin, heaped on special needles or passed in cigar-shaped stick form over the skin. The burning of moxa generates heat which is used as a form of treatment. Although other heat devices are occasionally substituted (for example, infrared devices), moxa is widely believed within Chinese medicine to have specific energetic characteristics which make its use preferable. With over 15 000 practitioners in the UK potentially using moxa on a regular basis, it was considered important to establish what by-products arose from its use and whether these fell within safe exposure limits.

There is no currently available evidence to suggest that there are adverse effects from moxa smoke. Anecdotal evidence exists that practitioners and patients with respiratory disorders find its use problematic, but there is nothing to suggest that this is from the unique characteristics of moxa over and above the production of smoke as an irritant. Evidence does exist from several Japanese studies that moxa has possible cytotoxic effects.14

The benchmark levels for cigarette smoke ingredients were established by the use of standard equipment and protocols. It was decided to submit moxa to the identical testing regime at a laboratory which met these standards. Cigarette smoking generates two kinds of smoke: mainstream smoke and sidestream smoke. Mainstream smoke describes the smoke inhaled and then exhaled by the smoker, sidestream smoke is smoke generated at the burning tip of the cigarette in between puffs or while resting in an ashtray. An initial trial run established that testing mainstream smoke was not possible, since the moxa was packed too tightly within the roll to allow sufficient smoke to permeate the length of the stick for testing purposes. This sampling was also rejected on the grounds that moxa is never inhaled in this way. However, it did establish that only minute quantities of smoke were lost by being absorbed or filtered within the stick.

The groups selected for testing were the most commonly found volatile organic compounds routinely assessed in cigarette analyses and/or chemicals:

  • Tar Tar, nicotine, carbon monoxide

  • Polycyclic aromatic hydrocarbons Benzo[a]pyrene, benzo[a]anthracene

  • Ammonia

  • Hydrogen cyanide

  • Poly aromatic amines 1 & 2-Aminonaphthalene, and 3 & 4-aminobiphenyl

  • Volatile organic compounds 1,3-Butadiene, benzene, acrylonitrile, toluene and styrene

  • Semi volatile compounds Pyridine and quinoline

  • Phenols Catechol, phenol, hydroquinone, resorcinol o- and m- & p-cresol

  • Carbonyls Formaldehyde, acetaldehyde, acetone, acrolein propionaldehyde, crotonaldehyde, butyraldehyde and methyl ethyl ketone.

There are a number of other chemicals and compounds assessed in cigarette testing such as the oxides of nitrogen, nitrosamines and metals such as mercury, nickel, lead, cadmium, chromium, arsenic and selenium. At this stage, however, it was decided to restrict the testing to those groups which represent the greatest known carcinogenic risk in cigarette smoke.56

In the UK there are two types of occupational exposure limits set by the Health and Safety Executive’s (HSE) regulations: maximum exposure limits (MEL) and occupational exposure standards. The main difference between the two limits is that an occupational exposure standard is set at a level at which there is no indication of a risk to health whereas for an MEL, a residual risk may exist and the level set takes socio-economic factors into account. Exposure limits of both kinds are set as the result of deliberation of expert committees and public consultation. MELs were chosen as the basis for comparison with the test results because most of the analytes for which the tests were conducted fall within this category by virtue of their potential toxicity.

For the purposes of analysing the significance of the results obtained from the testing of moxa, it was decided to use the MELs from table 1 of the Health and Safety Executive’s publication EH40 Occupational exposure limits 2007. Where there was no relevant information from the UK publication use was made of the USA Occupational Health and Safety Administration’s (OSHA) permissible exposure limits (PEL) data, Standard 1910.1000 Z-1 limits for air contaminants. Other sources were identified and used. These included the American Congress of Government Industrial Hygienists, the US National Institute for Occupational Safety and Health and the American Industrial Hygiene Association. Where a non-regulatory limit has been applied this usually means that the regulatory body has not yet set a limit, not that it does not regard a limit as necessary.

Table 1 A comparison of moxa smoke constituents with short-term occupational exposure limits set by the HSE and OHSA


Moxa is used in many different forms and with many other ingredients added to the basic material. It was important to select a form which would generate clear results. Only pure moxa was used; moxa production remains a cottage industry in China, and many varieties are adulterated, either intentionally or accidentally, with other herbal products. Of the pure forms available, it was decided to use the stick form or “moxa roll”. This is a tightly rolled cylinder of moxa with a paper casing, the size and shape of which were compatible with the standard cigar testing equipment. The brand selected, catalogue reference MX-09 from Oxford Medical Supplies in California, was chosen specifically because the paper coating had no ink printing, so there would be no hydrocarbon residues from the inks to skew the results. Certificates of authenticity were sought and received from the manufacturer in China which established that the product was 100% pure wild grown Folium Artemisiae Argyi with no chemical treatments used in its processing.

The sidestream smoke was generated under conditions specified by the International Organization for Standardization smoking regime for cigars. The moxa roll was held in horizontal position in a smoking machine port and the sidestream smoke generated by burning the stick for five minutes was collected on a 44 mm Cambridge filter pad using a fish-tail chimney mounted above the burning stick end. Each test was repeated three times. The tests for each of the groups of volatiles listed above required specific variations on this general procedure. Each is outlined in the full laboratory report available on the British Acupuncture Council website (

In order to draw comparison with MEL, it was decided to adjust the results as follows:

(a) Short-term exposure limits are usually calculated in mg/m3 per 15 minute exposure. It was decided to allow for the unlikely circumstance that the patient and practitioner occupied a cube with 1.5 m sides (3.375 m3) into which moxa was burned without any ventilation or diffusion of smoking by-products. Accordingly, the figures from the laboratory reports for compounds generated by moxa burned in one minute were scaled up to produce a direct comparison with agreed MELs and PELs.

(b) Long-term exposure limits, generally referred to as a time-weighted average, are also usually expressed in mg/m3. Again, it was decided to make an assumption of a significant but possible case of the burning of moxa for 15 minutes per hour in an eight hour day in an average-size treatment room of 3 m×2.5 m×2 m, or 15 m3, with equal diffusion throughout the day and no effective means of ventilation.

The figures have also in some cases been re-calculated and shown in parts per million (ppm) for comparisons, where the safety data are only available in this form. The data were assembled primarily from the HSE and OSHA documents mentioned previously, but also made use of other industry sources in which limits were variously expressed in mg/m3 and ppm. The other source most frequently cited was the American Congress of Government Industrial Hygienists but there are also figures from the US National Institute for Occupational Safety and Health and the American Industrial Hygiene Association. Tables 1 and 2 show the source for each figure MEL/PEL quoted in both short- and long-term exposure limits, respectively.

Table 2 A comparison of moxa smoke constituents with long-term occupational exposure limits set by the HSE and OSHA


Tables 1 and 2 show the results produced for the duration of the International Standards Organisation test cycle in the form of quantities of analyte generated per minute, and then show comparisons with existing occupational exposure limits and permissible exposure limits. Even with concentrations of an order of magnitude greater than would normally be achieved with normal dispersal and ventilation, few of the analytes reached levels at which there would be significant concerns.

Table 3 shows a comparison between the amounts of aromatic amines generated by the tests and figures from the Tobacco Manufacturers Association UK Benchmark Report Study 2003: Smoke constituents study annex part 12. This study shows the amounts of the same four aromatic amines produced by burning a cigarette. The Tobacco Manufacturers Association tests were based on average weights per cigarette of between 0.6 and 0.8 mg oven dry weight of tobacco. This contrasts with the amount of weight lost through combustion of moxa over a five-minute period reported at 663 mg.

Table 3 Tobacco Manufacturers Association benchmark values for concentration of aromatic amines in sidestream tobacco smoke in ng (from National Institute for Occupational Safety and Health Pocket guide to chemical hazards)7


Of the various groups of volatiles tested, only two (acrolein and hydroquinone) generated results which exceeded the current long-term MEL set by the HSE (although test results for hydroquinone fall within the MEL set by OSHA, which is four times greater than the HSE level). Given that the test figures were extrapolated to levels of concentrations unlikely to be found with normal rates of dispersal and diffusion in a treatment setting, the results were not considered to be significant. The same may be stated for the carbon monoxide figures which are over 12 times higher for MELs (420 mg/m3 against 35 mg/m3) for long-term exposure. As well as the unlikely high extrapolations used for comparison, it is also suspected that the testing method, which allows the moxa stick to smoulder without movement, might well have caused incomplete combustion around the tip producing more carbon monoxide than carbon dioxide. However, since the concentrations would certainly fall below the MEL with normal dispersal and ventilation, this was not considered worthy of further examination.

The one group of chemicals of potential concern are the aromatic amines. The test results show these only to have been present at minute levels, but there are no agreed safety limits for any of the four analytes. These all appear on the OSHA’s list of 13 known carcinogens, OSHA Standard No. 1910.1003, for which the only current recommendations are that exposure should be maintained at the lowest possible levels. Their use in industrial processes has been largely superseded by the use of alternatives, there being a strongly suggested causal connection between aromatic amines and bladder cancer.8

However, two studies undertaken in Italy appear to show that the levels of aromatic amines in indoor air and outdoor air are significantly higher than the levels recorded in our testing of moxa. Luceri et al found that a group of 16 such chemicals appeared at levels of between 20 000 and 30 000 ng per cigarette in sidestream smoke, and were present at significant levels in rooms adjacent to where smoke had first been produced.9 Palmiotto et al demonstrated that for a range of ten aromatic amines levels of 3 ng/m3 (in Siena) and 104 ng/m3 (Brindisi) were found, with most samples of outdoor air showing aromatic amine levels lower than 40 ng/m3.10 Although these readings are for a wider range of aromatic amines, they provide useful comparison with the figures detected in the tests of moxa, where the highest figure recorded expressed as a long-term time-weighted average was only 0.8 ng/m3.

Two factors arose in the preparation of the analysis which warrant further study. First, although moxa was tested in stick form and the volatile by-products of combustion tested and quantified, moxa is also used directly on the skin by some practitioners and when used in this way leaves a slight oily residue. Korinth et al report a number of cases where dermal absorption of aromatic amines has been linked to cancers.11 The exposure of most patients to aromatic amines through direct moxa use will be infinitesimal in comparison to some of the industrial processes for which these trials were carried out, but the report did indicate that a damaged epidermal barrier significantly increased the existing risk of absorption. Professional acupuncture bodies may have to assess whether their guidelines on moxa use may have to be adjusted to take this into account, and several already forbid the use of moxa on broken skin.

Wider research into the dispersal of potentially harmful concentrations of volatiles by improved ventilation may need to take into account the conclusions of Kotzias et al into the effect of ventilation on the reduction of air pollution from environmental tobacco smoke.12 It has been argued that potential risks could be reduced by better ventilation. Kotzias et al found, however, that many of the more dangerous volatiles stuck to internal surfaces rather than remaining in suspension in the air, accounting for the smell of smoke lingering in a room for many days after it has first been produced. The authors conclude that “diffusion of the emitted compounds (sidestream smoke and burning products) is relatively slow, so dilution via mixing with new incoming fresh air is not very effective as a control measure”.12

Furthermore, Kotzias et al conclude that wind-tunnel like rates of dilution ventilation would be necessary to achieve the kinds of reduction in pollutants to equivalent levels in ambient air.12 This may also have to be factored into any new recommendations to practitioners, since it may be more important to clean or cover surfaces in the room regularly than to insist on investment in air conditioning or ventilation plant. A great deal depends, however, on the rate of oxidation of some of the volatiles involved, many of which have a relatively short life span in their more dangerous forms. This in turn may be affected by the nitrogen content and combustion temperature of the moxa; one study in America pointed to a direct correlation of both of these factors to the amount of aromatic amines produced by cigarette smoke.13


In summary, the results of testing moxa for safe use are encouraging, and the by-products of combustion do not appear to be generated in sufficient quantities to cause concerns about exceeding the agreed industrial safety standards. Where some analytes approach current safe exposure limits in the tests and extrapolated results, this can be explained by the assumptions in the extrapolations from the test results which reflect highly improbable concentrations of smoke.

Summary points

  • Smoke from mouldering moxa may be hazardous to health

  • This study analysed moxa smoke for concentrations of known harmful constituents

  • Concentrations of three substances could cause concern under abnormal operating conditions

  • Practitioners should consider ventilation of clinic rooms, and the use of moxa on broken skin


The authors would like to thank Ulrike Wirth (BAcC Safe Practice Officer), Mark Bovey (Acupuncture Research Resource Centre), Dr Mike Cummings (BMAS), and Dr Helen Taylor and Dr Tatiana Humphreys (Arista Laboratories) for their help and advice in the preparation of this paper. The British Acupuncture Council, which funded the study, would also like to thank Arista Laboratories, Richmond, Virginia for their help and advice.


View Abstract


  • Competing interests: The study was funded by the Executive Committee of the British Acupuncture Council in response to questions posed by patients. The analyses were conducted by an independent laboratory. The authors, two of whom (BC and CC) are members of the Executive Committee, converted the data, located the published maximum exposure limits for comparison and interpreted the findings. The BAcC Executive Committee itself played no role in the interpretation of data or preparation of this report.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.