Cropwatch Newsletter 4 - Red Alert Issue


§5. Coumarin: The Real Story vs. the SCCP Opinion SCCP/0935/05.

What is it? Coumarin (2H-1-benzopyran-2-one) CAS No 91-64-5 when seen pure, is a crystalline white solid with a hay-like, sweet aromatic creamy odour with certain nutty shadings, much used in perfumes and in fragranced soaps and detergents. Coumarin is representative of the lactones (where a lactone is an ester group integrated into a carbon ring system).

Whats the problem? The FDA dubiously identified coumarin as a carcinogen in 1954. Subsequent studies initially upheld this opinion, but then disproved it. The net result is that because of the controversy, it is not used in foods (although it is famously present in many, including strawberries!). Coumarin has been approved for perfumery use, but was identified as a fragrance allergen by the SCCNFP/0017/98, although many perfumery professionals have refused to believe this (see below). Coumarin has been regulated within the 7th Amendment of the Cosmetics Directive (76/768/EEC) such that coumarin requires labelling if present at concentrations of ³10ppm in fragrances leave on products, or ³100 ppm in fragranced products washed off the skin. Floc’h et al. (2002) & Vocanson et al. (2006) have published data supporting their view that pure coumarin is not a sensitiser, but rather it is impurities that elicit any alleged reaction (see below), an opinion which anyway is widely accepted by industry but not by the SCCP. 

The metabolism of coumarin is different species to species. In humans coumarin is metabolized to umbelliferone (7-hydroxycoumarin) (Cohen 1979) whereas in the rat the major pathway is via 3-hydroxylation followed by ring opening and metabolism to o-hydroxyphenylacetic acid (Steensma 1994). The topic of metabolic differences between species has been revisited by Felter et al. (2006).

Coumarin-containing Natural Products.

Coumarin occurs widely in natural products, generally being liberated from the corresponding glycoside (melilotoside) on drying herb material. Dicoumarol, is a microbiological biotranformation product in spoiled Melilotus Clover, is responsible for fatalities by hemorrhaging in cattle. This is because dicoumarol interferes with vitamin K reductase in the liver and the liver is unable to reactivate vitamin K, which leads to a decrease in vitamin K-dependent clotting proteins. The study of this compound paved the way to the discovery of anti-coagulant drugs such as warfarin.


Coumarin occurs widely in natural products; the following natural aromatic materials are of note:


Some Natural Coumarin Sources                                                          


Anthoxanthum odoratum L.

Flouve oil

Both essential oil and absolute produced.

Cinnamomum cassia J. Presyl.

Cassia oil

Coumarin 4-11% (Burfield 1999). Eu Pharm V (2) allows 1.5 to 4.0% coumarin in cassia oil spec.

Dipteryx odorata (Aubl.) Wild. and  sometimes D. oppositifolia.

Tonka bean 

1-3% up to 10% coumarin in beans (Hagers Handbuch 1973); also dihydrocoumarin, o-coumaric acid etc (Ehlers et al. 1996). Absolute contains 65% coumarin.

Galium odoratum L.


Variable; coumarin content develops on drying herb. Used in alcoholic beverage flavourings.

Lavandula spp.

Lavender & Lavandin qualities.

Lavender absolute 8.0% coumarin; lavandin absolute 5.0% coumarin.

Lolium perenne L. & other spp. incl. Phleum pratense (Timothy grass), Poa pratensis L. (Meadow grass),

Cynosurus cristatus (Crested Dog’s-Tail), Anthoxylum odoratum L. and Melilotus spp.

Foin oil.

Essential oil and absolute produced. Oil contains 8% coumarin.

Carphephorus odoratissmus (J.F. Gemel) syn Liatris odoratissima Mich. syn. Trilisia odoratissima (J.F. Gmel.) Cass.

Deer tongue

1.6% coumarin. Ratio of coumarin:   dihydrocoumarin:  2,3 benzofuran in volatile fraction of extract  1:3:20  (Appleton & Enzell 1971).

Melilotus alba Medik.

Bokhara Clover, or White Sweet Clover 

Less used than Common Melilot (q.v.)

Melilotus officinalis L. (Pallas)

Common Melilot or Yellow Sweet Clover

0.9% coumarin on dry weight basis. Wagner (1996) says 0.25-0.45% coumarin in herb, together with umbelliferone, scopolin etc.

Table 1: The occurrence of coumarin in some common herbs & natural products.


Coumarin also occurs in trace amounts in Narcissus spp, Artemisia vulgaris L., Michelia champaca L., Artemisia annua L., cinnamon bark & leaf oils (0.3% coumarin), spike lavender oil (0.2%) and many other natural products. Melilotus leaves from Melilotus officinalis L. have been used to flavour snuff & tobacco. Tonka bean absolute, deertongue absolute & melilotus absolute find some uses in perfumery, but woodruff absolute is no longer much used, apart from flavouring wines.  Coumarin derivatives such as the sweetly herbaceous 7–hydroxycoumarin (umbelliferone) also occur in natural products (e.g. in lavender absolute from Lavandula angustifolia) but derivatives like herniaren are banned IFRA

Synthetic Coumarin in Perfumery. Synthetic coumarin has a history of importance in perfumery, being the first synthetic (synthesised by W.H. Perkin 1868) to be used in a fragrance - Fougère Royale - and since that time coumarin has been fundamental to the fougère perfumery accord together with lavender and bergamot oils.

Coumarin in Flavourings. Use of coumarin and coumarin-containing herb extracts was common in earlier times. Walter (1916) relates the use of woodruff extract from the fresh flowering herb in the preparation of lemonade but remarks that it tends to be weak and prone to cause turbidity, and gives an alternative recipe for woodruff flavouring constructed from synthetic coumarin, alcohol and tonka bean tincture. Use of woodruff extract in cola, caramel, gooseberry and other flavourings is also detailed, and the use of tonka essence containing coumarin, vanillin etc is also outlined for flavouring of fondants.

Earlier reports of the toxicity and carcinogenicity of coumarin are now believed to be due to impurities, but coumarin is banned in foods in USA (21CFR 189.130), Japan, India, & the EC, and was banned in Germany from 1970 to 1991 (the ban is now replaced by a concentration limit) etc. Since many derivatives of coumarin are commercial poisons, e.g. warfarin, the well-known rat poison, it has been difficult to persuade people of coumarin’s safety. A detailed discussion of the beneficial uses of Melilotus extract & coumarin in phytotherapy (and there are many) & any remaining toxicological issues are given in a Meliltotus monograph by Mills and Bone (2000).


Coumarin as a Sensitiser (?).

Lake (1999) puts our maximum exposure to coumarin from fragrances in cosmetics to 0.04mg/Kg/day, with intake from the dietary sources contributing a further 0.02mg/Kg/day for a 60Kg person. An article by François Floc’h et al. (2002), of Rhodia Perfumery & Specialities, looked at pure coumarin applied in homogenous form to the skin of animals and humans, and concluded that coumarin is not a dermal allergen. Coumarin was one of the items you will remember cited in the SCCNFP position paper for Fragrance Allergy in Consumers (SCCNFP/0017/98 final Dec 1999) as being a skin sensitiser, this being the conclusion of previous COLIPA and RIFM opinions. Previous work by Malten KE et al. (1984), De Groot AC et al. (1988), Larsen W. et al. (1996), and Van Joost T et al. (1985) on coumarin was also reviewed by François Floc’h et al. who commented, amongst other things, on the lack of scientific rigor, and found no statements of the purity of the materials previously used, and who questioned the homogeneity and the stability of the coumarin in petrolatum suspension. Floc’h et al. further indicated the above work failed to distinguish allergy to coumarin and cross-reaction to allergens for which coumarin might be an indicator.

Enter the SCCP. Again.

The SCCP Opinion on coumarin as a sensitiser SCCP/0935/05 (adopted 20th June 2006) considers whether coumarin of >99.99% purity had any sensitising properties (industry claims that it doesn’t), & if it doesn’t, whether the Opinion on Fragrance Allergy SCCNFP/0017/98 would need to be changed. The committee concluded that coumarin of 99.9% purity when patch tested at 2% would be able to elicit allergic contact reactions in humans.

Cropwatch Comments. Although the SCCP Opinion heavily criticises various published papers/abstracts/posters by Vocanson et al. (2006), Masamoto (2001), CIT (2001) & INSERM (2003/2004) on regarding alleged coumarin sensitisation on various grounds (i.e. is confusing, there is lack of evidence etc.), those very same remarks apply to their own Opinion. The SCCP document only looks half finished – sloppily, the Vocanson et al. paper is not fully referenced (full details are provided below) and the overall layout is very poor with lack of clear headings, so that it is not immediately apparent what study you are reading about (until the reader has gone over the paper several times). [The key to understanding the Opinion is that new evidence is considered under various headings: Patch testing, Animal data & LLNA (local lymph node assay) studies, with the identity of the study under consideration confusingly set out in normal type face towards the right hand margin at the bottom of the relevant text (instead of as a heading at the top)].  The discussion 3.3.14 needs rewriting with clear references to the work they are criticising.

The SCCP Opinion that coumarin of 99.9% purity when patch tested at 2% would be able to elicit allergic contact reactions in humans, seem to us to be largely based on the findings of the study by Vocanson et al. (2006), who claim that they found a reaction of only one subject in 512  hospital patients to pure coumarin (although, on a quick read-through, the SCCP Opinion seemingly only accounts for 510). Commercial samples of coumarin with coumarin derivatives as impurities were found to be weak or moderate sensitisers by the Vocanson team. The SCCP seems to have seized on this one reaction, and on the reaction of an individual positive from 101 patients positive to the fragrance mix (of which coumarin is not a component) as evidence that coumarin is a sensitiser. Crucially, the discrepancy between the Vocanson team’s finding of one positive and the SCCP’s reading of two positives is not explained, and presumably the SCCP did not bother to contact the authors for an explanation of why they had dismissed one of these positive reactions.

The SCCP wastes our time reporting on the evaluating an abstract by Masamoto (2001) and concludes, unsurprisingly, there is not enough evidence presented. The SCCP do not provide an explanation of why they were unable to obtain the full paper (maybe they are unable to cope with articles written in Japanese ?).

Cropwatch can only conclude the following:

1. That this SCCP Opinion SCCP/0935/05 only further establishes that the evidence for pure coumarin as a sensitiser is extremely weak. The SCCP defended their previous Opinion on coumarin only by nit-picking at the paper by Vocanson et al. (2006).

2. That criticism by the SCCP of the determinations of the purity of coumarin presented in the publications considered is a bit rich, considering Floc’h’s remarks (Floc’h 2002) that previous work up by Malten KE et al. (1984), De Groot AC et al. (1988), Larsen W. et al. (1996), and Van Joost T et al. (1985) had not paid any/sufficient attention to the issue. If the SCCNFP themselves had looked at the coumarin purity issue more closely in the first place, they would not have classified coumarin as a sensitiser in SCCNFP/0017/98 final Dec 1999.

3. We feel that the SCCP are now adopting a different set of evaluative criteria towards new submitted evidence on coumarin, in order to make judgements that support their previous Opinions – clearly they are being defensive rather than objective.

4. If the SCCP had contacted the authors of the papers on coumarin sensitisation that they were reviewing in SCCP/0935/05 for clarification/further information, it is probable that a different outcome would have resulted. The fact that this was not done has to be seen as having a political dimension. 

5. The matter of whether it should be required by law that coumarin – a weak sensitiser at best - should need labelling as required under the 7th Amendment to the Cosmetics Directive, now needs investigation by Judicial Review.  Further, the assumption that natural botanical products which contain coumarin are sensitising also needs reviewing; deertongue incoloure (which has a high coumarin content) was after all, previously reported non-sensitising by RIFM (Opdyke D.L.J. (1976) Food Chem Toxicol 14 (Suppl.) 743).


Appleton R.A. & Enzell C.R. (1971) “Triterpenoids & Aromatic Components of Deer Tongue Leaf” Phytochemistry  10, 447-449.


CIT (2001): CIT/Study No. 21214 TSS/Rhodiascent TM Coumarine/Rhodia Services – RSP 13 Dec 2001


Cohen A.J. (1979) “Critical Review of the toxicology of coumarin with special reference to interspecies differences in metabolism and hepatoxic response & their significance  to man” Food Cosmet. Toxicol. 17, 277-289.


Clarke G.S. (1995) “Coumarin” Perf & Flav. 20, (Nov/Dec 1995) 23-34.


de Groot, A.C. et al. (1988) “Allergens in Cosmetics” Arch. Dermatol. 124, 1525-1529.


Ehlers D. et al. (1996) “Reducing the Coumarin Content of Tonka Bean Extracts Using Supercritical CO2.” Int J. Food Sc & Techn 31, 93-95.


Felter S.P., Vassallo J.D., Carlton B.D., Daston G.P. (2006). “A Safety Assessment of Coumarin taking into account Species-Specificity of Toxio-Kinetics”. Food & Chem. Toxicol. 44, 462-475. 


François Floc’h (2002) “Coumarin in Plants and Fruits: Implications in Perfumery.” Perf. & Flav. 27 (Mar/Apr 2002), 32-36.


INSERM U503/Société Rhodia Services. “Evaluation du potential de sensibilisation cutanée des coumarines à l’aide du Local Lymph Node Assay murin.” 10th Dec 2003.


INSERM U503/Société Rhodia Services. “Evaluation du potential de sensibilisation cutanée des coumarines à l’aide du LLN A.” 26 mai 2004.


Lake B.G. (1999) “Coumarin Metabolism, Toxicity & Carcinogenicity: Relevance for Human Risk Assessement.” Food & Chemical Toxicology 37(4), 423-453. 


Larsen W. et al. (1996) “Fragrance contact dermatitis. A worldwide multi-centre investigation (part 1).” Am. J. of Contact Dermatitis 7, 77-83.


Malten K.E. et al. (1984) “Reactions in Selected Patients to 22 fragrance materials” Contact Dermatitis 11, 1-10.


Masamoto Y. (2001) “Sensitisation & Cross-Reaction of Simple Couymarins.” Yagugaku Zasshi, 121, 97-103.


Mills S. & Bone K. (2000) Melilotus monograph in Principles & Practice of Phytotherapy Churchill Livngstone 2000 pp 463-489.


Steensma A., Beamand D.G., Walters D.G. et al. (1994) “Metabolism of Comarin and 7-ethoxycoumatrin by rat, mouse, guinea pig, Cynomolgus monkey & human precusion-cut liver slices” Xenobiotica 24, 893-907.


Van Joost T et al. (1985) “Simultaneous allergy to perfume ingredients.” Contact Dermatitis 1é, 115-116 (1985)


Vocanson M., Goujon C., Chabeau G., Castelain M., Valeyrie M., Floc’h F., Maliverney C., Gard A. & Nicolas J.F.  (2006) “The Skin Allergenic Properties of Chemicals may depend on Contaminants” Int Arch Allergy Immunol 140, 231-238 


Wagner H., & Bladt S. (1996) Plant Drug Analysis: a Thin-Layer Chromatolography Atlas 2nd edn. Springer-Verlag, Berlin.


Walter E. (1916) Manual for the Essence Industry  John Wiley.


Back                                                                                                                                                                                                                                                            Next