The report below was in response to some dreadful quality research in 2007 claiming that
lavender and tea tree oil promoted the growth of breast tissue in boys. Members of this
same incompetent research team produced a similar report in 2018 (below) which was
published around the world by newcasters.

Call for Journal of New England Medicine to publish a retraction re the recent
article: Prepubertal Gynecomastia Linked to Lavender and Tea Tree Oil.
(published: New England Journal of Medicine 365 (5) pp 480-485 D. V.
Henley, Ph. D., Natasha Lipson, M. D., Kenneth S. Korach, Ph. D. and
Clifford A. Bloch, M. D.)

A critique after consultation with numerous research scientists by Christopher
Dean, Chairman of the Australian Tea Tree Oil Industry Technical and Safety

The recent reports alleging that Lavender and Tea Tree Oil may be causing
breast growth in very young boys has very little substance, is a product of poor
reasoning, and is cast into doubt on many grounds. The following article
challenges the poor conclusions and methodology, and calls on the New
England Journal of Medicine for a retraction of this data and the unwarranted
conclusions that give the impression that this has the endorsement reserved
for thorough, well reviewed science. This article focuses on the Tea Tree Oil
allegations in particular, which is the special knowledge area of the author.

Many researchers and scientist have looked at this article, and raised concern
and alarm at the poor methodology and conclusions which are certainly not
supported by science.

When such science is amplified by publication in a respected Journal, and the
media beats up the story, it has damaging consequences out of all proportion
to the facts. This article was uncritically reported around the world causing
alarm and commercial impacts and fear. Is this responsible?

One doctor in one town reports 3 cases in close succession and promotes as a
scientific conclusion that Tea Tree Oil is a causative agent – even though only
one of the 3 had any, and that a very tiny, exposure to Tea Tree Oil. There is
no good science to link this with the Gynecomastia, while there are dozens of
other plausible hypotheses that are not even considered. From this poor
reasoning this expert is able to make a claim and create a climate of fear that
is responsible for causing millions of consumers to be fearful of what they are
using and avoid an excellent proven therapeutic product which offers them
major benefit. Bad science outcome indeed. The Tea tree industry around the
world has a long history of documenting all adverse events reported – and two
of the largest companies selling retail products, with global sales of over 150
million units of tea tree products over three decades, have never had a single
instance of this bizarre side effect reported.

Let’s review the facts:
This article does not provide the type of scientific reasoning which one might
expect from a respected medical journal. It suggests an association but not a
logical connection between prepubertal Gynecomastia and lavender and tea
tree oils. The application of lavender oil and of tea tree oil in the case
studies is purely anecdotal. In only one case was tea tree oil involved
even anecdotally.
This does not seem to fit with the title of the article or with
the unsupported conclusion in the summary.

This paper consists of two separate parts which have no scientific
Part one describes three clinical case studies of prepubertal
Gynecomastia. Part two describes a cell culture assay of the estrogenic activity
of lavender oil and tea tree oil. In the summary we find the statement “We
conclude that repeated topical exposure to lavender and tea tree oils probably
caused prepubertal Gynecomastia in these boys.” No scientific basis for this
conclusion is stated
nor can it be found in the article.

A scientific basis would require that dosages and routes of administration could
be related to each other quantitatively. The cell culture studies show
quantitative dosages for lavender oil and for tea tree oil. There are no
quantitative dosages in the clinical case studies. The route of administration is
defined in the cell culture study. It is by direct application to unprotected cell
culture medium. The routes of administration for the case studies and even the
materials applied in the three case studies are different.

They are only by applied by topical administration. In two of the case studies
only lavender oil was applied. In one case a product containing unstated
amounts of lavender oil and tea tree oil was applied via shampoo and hair gel.
In only one of the cases was any amount of any oil product applied via a leave-
on skin topical treatment. In the other two cases only a rinse-off product was
used. The quantities of oils used in the cell culture are vastly greater than
would be possible to achieve by normal cosmetic application of products yet
there is no acknowledgement of this.

Skin Penetration studies show that only 3 components of this complex mixture
penetrate the skin when applied topically, and further that evaporation
removes over 90%. 1 Even if the results seen in the in vitro MCF-7 test is
correct (which it may not be), it is unlikely that a TTO mixture is able to cause
oestrogen receptor activation in the body since the TTO complex is altered
following breakdown/metabolism on the skin. Skin penetration studies for Tea
Tree Oil conducted at the University of Queensland by Dr Sheree Cross
(unpublished) have clearly shown that only extremely small amounts of 3 of
the over 100 components found in TTO have been found to penetrate the
surface of the skin so that any oestrogen receptor activity by TTO in vitro is not
relevant to topical application of TTO products. This may well be true with
Lavender as well. There is nothing in the literature that indicates that these
components (terpinen-4-ol, alpha-terpineol and 1,8-cineole) have estrogenic
receptor activity in vitro or in vivo.

The paper discusses three case studies - only Patient 2 was exposed to TTO
and only in a styling gel & shampoo. From the usage of styling gel, the
possibility of skin absorption is very low (as the gel is applied to the hair, not
scalp). The composition of the actual shampoo that was used is known - it is a
well formulated product of less than 1% TTO (not a commercial secret - it is
easy to determine by solvent extraction & GC of T4-Ol). The expected
deposition rate of TTO from a normal surfactant based shampoo like that used
by Patient 2 is very low - shampoos are, after all, designed to remove
hydrophobic materials, not deposit them - so the likelihood of skin absorption
of tea tree oil resulting from use of this shampoo is very low. Additionally,
Patients 1 & 3 were exposed only to lavender oil, Patient 2 to lavender & TTO.
Alternatively, all three cases may be due to some other material in these boys
environment. The fraternal twin of one of the boys apparently using the same
materials was not affected. Note that all 3 affected boys lived in the Denver
area, yet no other environmental or health factors were considered.

When they received this information from Bloch, Henley and Kenneth Korach,
both researchers at the National Institute of Environmental Health Sciences,
performed test tube experiments of the effects of lavender oil on breast cancer
cells. They also decided to test tea tree oil because of Dr. Bloch’s request. They
observed that both oils exhibited “estrogen-like” qualities on the cells. At the
annual meeting of the National Endocrine Society held in Boston in June 2006,
Henley reported the results of the research, which was subsequently published
on February 1, 2007. What Henley’s report failed to mention is that there are
literally thousands of harmless natural oils and other natural plant substances
that exhibit similar “estrogen-like” qualities when applied directly to a cell
culture. Just a few common examples of products that have similar effects as
essential oils in similar tests are: soy, hops, garbanzo beans, red clover, lentils,
flaxseed, sunflower seeds, alfalfa sprouts, liquorice, and ginseng. Were these
boys screened for liquorice consumption or garbanzo beans or one of these
other hundred of suspect substances each?

Dr. Henley told a representative from Melaleuca Inc, a USA corporation selling
Tea Tree oil products, that while he was being interviewed by reporters about
the report, he had the definite impression that they were trying to get him to
say that lavender oil and tea tree oil cause Gynecomastia so that they could
publish a headline that these products should not be used. Scaring consumers
about dangers of ‘safe’ products sells papers (and gives exposure to scientific
Journals in the face of their credibility). Mainstream Channel 10 news in
Australia carried a 3 minute prime time segment on this article warning
mothers of the risk of using tea tree oil products. Henley told Melaleuca Inc.
that he was concerned about how the stories had come out as they just took
portions of what he said instead of publishing everything he said. Henley
emphasized that the research does not conclude that either lavender oil or tea
tree oil are the direct cause of the Gynecomastia in the young boys – but that
there “may” be a correlation. He pointed out that the only common ingredient
among all of the products used by the patients was lavender oil and that only
one boy had used a product that contained both lavender oil and tea tree oil.
In his report Henley cautioned patients of prepubertal Gynecomastia to avoid
repeat exposure to these essential oils, but in the phone interview he said
there is not nearly enough evidence to indicate that people should stop using
products with lavender oil or tea tree oil, even young boys.

It seems very odd to us that tea tree oil was even mentioned in this story. It
appears that lavender oil is the only common substance identified as being
used by the three boys in question. It appears that the only reason that tea
tree oil was mentioned in the story was because the source of lavender for one
of the three boys was a Tea Tree Hair Gel and Shampoo. There does not appear
to be any evidence whatsoever that the symptoms of that one boy had
anything to do with Tea Tree Oil. Industry records certainly support this
conclusion. Over the past 21 years the two leading companies supplying tea
tree products, Melaleuca Inc and TP Health Ltd, have sold over 150 million
bottles of product containing Tea Tree Oil. Both companies maintain meticulous
adverse event reporting records. At the time of writing there has never been a
single case of prepubertal Gynecomastia reported to either company anywhere
in the world in all those years.

Additional poor methodology and erroneous conclusions are to be found in the
in-vitro study – the second part of this flawed research. This is a unique
protocol with no relationship to any other body of work. Apart from flawed
technique, there is the absurd conclusion that to achieve equivalence in the
human clinical situation would require 40 bottles of shampoo per dose for a
20Kg child (see below). No responsible scientist should draw conclusions on
such grossly distorted comparisons.

The procedure for dosing the essential oils into the cell culture medium is not
fully described in the article. It is stated that the oils were dissolved in
dimethylsulfoxide before addition to the cell culture in the text. In the figures it
is stated that the solvent control was ethanol. It is not clear which of these
solvents was actually used. Nor is it clear whether the same amount of solvent
was added to the cell culture at the different dose levels. Dismissing this
discrepancy there is a technical problem with the cell culture studies in that the
quantities of oils added to the cell cultures exceed the solubility of the oils in
water and presumably also the solubility in the culture media. For example the
addition of 0.01% of lavender oil corresponds to 100 mg/l of oil. This is a
considerable quantity of free, insoluble oil. Thus the distribution of the
components of the oils in the cell culture is not controlled and free droplets of
oil would be floating around at the higher concentrations. The potential effects
of this on the assays are unclear. It could certainly either add to the variability
of the assay or possibly lead to unpredictable results. The authors note that
lavender oil was cytotoxic at levels above 0.025% oil. They do not state
whether they tested for long term cytotoxicity at the lower levels. By
comparison the addition of nanomolar amounts of estradiol (0.28 micrograms
per litre) would not lead to apparent physical separation.

If we attempt to equate the level of 0.01% lavender oil used in cell culture to a
human dosage the quantity would be 0.01% of the body weight or a systemic
dose of 100 mg/ kg. The quantity of topically applied lotion or cream required
to achieve this type of a systemic dose would, of course, be enormous. Even if
the lotion contained the improbably high amount of 10% lavender oil and even
if the transdermal delivery were as improbably high as 10% a 20 Kg child
would require a single dose of 200 g of lotion.

In the case of a rinse off product such as a shampoo or bar soap the quantities
become quite ridiculous. The maximum amount of tea tree oil which is
practically incorporated into a shampoo or bar soap is about 1%. Further since
these are administered as rinse off products the potential for transdermal
transport is much less than for a lotion or cream. Let us estimate it with the
improbably high figure of 1%. To achieve the 2.0 g systemic dosage required
for a 20 Kg child via 1% transdermal delivery of a 1% product we would
require an application of 20 Kg of product. Attempt to visualize this process.
This would correspond to about 80 bath sized soap bars or 40 bottles of

To summarize:
In only one of the case studies does the product claim to contain tea tree oil
even qualitatively. The conclusion that Tea Tree is a causative agent with only
one case study is preposterous. It ignores numerous other hypotheses.

There is no discernable dosage of lavender oil or tea tree oil in the three case

The authors make no attempt to relate the dosages of lavender oil and tea tree
oil in the case studies to the dosages used in the cell culture experiments.
The authors do not account for the improbability of transdermal delivery of oil
components from a lavender oil lotion (Case 1) or from a lavender oil and tea
tree oil shampoo (Case 2) or from a lavender oil soap bar (Case 3) Nor do they
note that what transmits is very different from the whole oil mixture applied
directly to a cell, and nor do they consider other materials in these products.
The actual components of lavender oil and of tea tree oil are almost totally
chemically distinct from each other; it is unlikely that they would have similar
effects. The paper claims the opposite.

The quantities of lavender oil and tea tree oil which elicited a response in the
cell culture studies are vastly greater than the quantity of estradiol in the
positive control treatment. The difference is a factor of 100,000 to 1,000,000.
Thus if the studies do show any estrogenic activity it is at a level one hundred
thousand to one million times less than that of estradiol, the positive control.
The amounts of lavender oil and tea tree oil which elicited a positive response
are stated as concentrations as 0.01% to 0.025%. At these nominal
concentrations these water insoluble oils will physically separate from water
and presumably from the cell culture medium. Thus there will be droplets of
pure oil floating around in the culture. It is incorrect to refer to these amounts
as concentrations due to this fact.

The authors do not state exactly how the oils were added to the cell cultures.
In the text they state that they were dissolved in dimethylsulfoxide before
addition to the cell culture. In the figures they state that the solvent was
ethanol. It is not clear which was used or whether the amounts of solvent
added to the different concentrations were the same. In either case however
one would expect physical separation of oil droplets at concentrations as high
as 0.005% and above.

The authors make no reference to these huge differences in dosage and in
potency or the physical difficulties in the cell culture experiments cited above.
There is no supporting field data from decades of actual exposure that
supports this conclusion.

In short the authors ignore numerous alternate possibilities and draw an
unsound conclusion from non significant and very noisy data.

This publication is, to say the least, unscientific. The conclusion stated in the
summary is not supported by the cell culture studies. The authors show no
curiosity at all about the enormous difficulties in attempting to connect the cell
culture studies with the case studies scientifically. It is disappointing to see the
New England Journal Of medicine publishing such work uncritically, allowing
such material to damage its own reputation and to create unwarranted alarm
and commercial damage around the world. A retraction is warranted.

1. Cross S. and Roberts M. (2005). In-vitro human epidermal membrane
penetration of tea tree oil components from pure oil and a 20% formulation. A
report to RIRDC (Australian Rural Industry Research and Development

Source: Copyright: ATTIA Ltd
More posts
Rosemary oil (Tunisian) - growth and production