Lavender oil does not mimic estrogen nor does it enhance the body’s own estrogens. It is therefore not a ‘hormone disruptor’, cannot cause breast growth in young boys (or girls of any age), and is safe to use by anyone at risk for estrogen-dependent cancer. The lack of estrogenic action is the conclusion of a new report, which used a novel form of ‘uterotrophic’ assay.
This measures the effect of a test substance on the uterus of immature or estrogen-deprived female rats over three days. Any estrogenic action causes a rapid and measurable increase in uterine weight. The assay has been in use since the 1930s, was adopted by the OECD in 2007, and is now regarded as the “benchmark animal assay for estrogenic effects” (Politano 2013).
A 2007 report by Henley et al found that both lavender and tea tree oils had a weak in vitro estrogenic action. Lavender was suggested as the cause of three cases of prepubertal gynecomastia (breast growth) in boys, and tea tree was suggested in one case. The report was subsequently criticized on a number of grounds. For example it was pointed out that there was no evidence of either essential oil causing the gynecomastia in any of the four cases, and that in vitro estrogenic findings frequently do not extrapolate to a similar action in warm bodies (see Rebuttals, below).
Since 2007 there have been many warnings about lavender and tea tree oils that relate back to this study. Others, including myself, have suggested that such cautions are unnecessary and premature. The new research, in which considerable quantities of lavender oil were used, found that there was no increase in uterine weight, and so no estrogenic action.
The novel aspect of the uterotrophic assay was that the test substance – lavender oil – was applied to the skin, while the most common method is subcutaneous injection. This was changed in order to mimic the circumstances of the Henley et al 2007 case reports, and it also mimics the use of lavender oil in fragrances and personal care products. Lavender oil was used in two concentrations, 4% and 20% in corn oil. According to Politano et al 2013, these concentrations are respectively more than 6,000 and 30,000 times greater than a conservative estimate of human skin exposure from multiple cosmetic products containing lavender oil. They are also 5,000 and 1,000,000 times greater than the estimated exposure to lavender oil experienced by the Henley et al boys.
These calculations may seem exaggerated, but the massive differences are because in the uterotrophic assay, the lavender oil was applied in Hilltop Chambers patches, which do not allow any evaporation, or loss of essential oil by means other than dermal absorption. Examples of the quantities of fragrant substance absorbed from personal care products are shown below. The amount of fragrance absorbed from a shampoo, for instance, is 200 times less than the amount applied to the head. If amounts of dermally-absorbed lavender oil that are at least 5,000 times greater than any normal human exposure are not estrogenic, then we can be confident that this particular safety issue is not a concern.
Safety in pregnancy is a broader question than hormonal action alone, since affects on the fetus, or even miscarriage, are theoretically possible with any substance. The question of uterine stimulation and miscarriage was thoroughly investigated here, and there is clearly no risk. As for fetal effects we know that linalool, a major constituent of lavender oil, is not fetotoxic to pregnant rats at an oral dose 1 g/kg/day for 16 days (equivalent to a human oral dose of 60 g, or 2 oz per day, and a total dose of 32 oz). The chances of lavender oil being fetotoxic in the amounts used in personal care products are negligible, considering the small amounts both applied to, and subsequently absorbed by human skin (Cadby et al 2002).
The new research findings represent a major development in our knowledge of lavender oil safety, since the possibility of estrogenic action now looks remote. While no single test should be taken as absolutely conclusive, we can expect the volume of noise about lavender being estrogenic to diminish considerably.
Dean CJ 2007 Prepubertal gynecomastia linked to lavender and tea tree oils. New England Journal of Medicine 356:2543-2544
Kalyan S 2007 Prepubertal gynecomastia linked to lavender and tea tree oils. New England Journal of Medicine 356:2542-2544
Kemper KJ, Romm AJ, Gardiner P 2007 Prepubertal gynecomastia linked to lavender and tea tree oils. New England Journal of Medicine 356:2541-2542
Kurtz JL 2007 Prepubertal gynecomastia linked to lavender and tea tree oils. New England Journal of Medicine 356:2542-2543
Lawrence BM 2007 Estrogenic activity in lavender and tea tree oils: Part I. Perfumer & Flavorist 32:20-25
Lawrence BM 2007 Estrogenic activity in lavender and tea tree oils: Part II. Perfumer Flavorist 32:14-20
Tisserand R Tea tree and lavender not linked to gynecomastia
Cadby P A, Troy W R, Vey M G 2002 Consumer exposure to fragrance ingredients: providing estimates for safety evaluation. Regulatory Toxicology & Pharmacology 36: 246-252
Henley DV, Lipson N, Korach KS, Bloch CA 2007 Prebubertal gynecomastia linked to lavender and tea tree oils. New England Journal of Medicine 365(5): 479-485
Politano VT, Lewis EM, Hoberman AM et al 2008 Evaluation of the developmental toxicity of linalool in rats. International Journal of Toxicology 27:183-188
Politano VT, McGinty D, Lewis EM et al 2013 Uterotrophic assay of percutaneous lavender oil in immature female rats. International Journal of Toxicology Open access article