Content uploaded by Anthony D Ormerod
Author content
All content in this area was uploaded by Anthony D Ormerod
Content may be subject to copyright.
Randomized Trial of Aromatherapy
Successful Treatment for Alopecia Areata
Isabelle C. Hay, MRCP; Margaret Jamieson, SRN; Anthony D. Ormerod, FRCP
Objective: To investigate the efficacy of aromatherapy
in the treatment of patients with alopecia areata.
Design: A randomized, double-blind, controlled trial of
7 months’ duration, with follow-up at 3 and 7 months.
Setting: Dermatology outpatient department.
Participants: Eighty-six patients diagnosed as having
alopecia areata.
Intervention: Eighty-six patients were randomized into
2 groups. The active group massaged essential oils (thyme,
rosemary, lavender, and cedarwood) in a mixture of car-
rier oils (jojoba and grapeseed) into their scalp daily. The
control group used only carrier oils for their massage,
also daily.
Main Outcome Measures: Treatment success was
evaluated on sequential photographs by 2 dermatolo-
gists (I.C.H. and A.D.O.) independently. Similarly, the
degree of improvement was measured by 2 methods: a
6-point scale and computerized analysis of traced areas
of alopecia.
Results: Nineteen (44%) of 43 patients in the active
group showed improvement compared with 6 (15%) of
41 patients in the control group (P= .008). An alopecia
scale was applied by blinded observers on sequential
photographs and was shown to be reproducible with
good interobserver agreement (k= 0.84). The degree of
improvement on photographic assessment was signifi-
cant (P= .05). Demographic analysis showed that the 2
groups were well matched for prognostic factors.
Conclusions: The results show aromatherapy to be a
safe and effective treatment for alopecia areata. Treat-
ment with these essential oils was significantly more
effective than treatment with the carrier oil alone
(P= .008 for the primary outcome measure). We also
successfully applied an evidence-based method to an
alternative therapy.
Arch Dermatol. 1998;134:1349-1352
WITH THE recent re-
surgence of inter-
est in alternative
medicine, aroma-
therapy (aroma,
from the Greek meaning spice) has at-
tracted great public interest in its health-
promoting or medicinal properties. With
herbalism as its basis, aromatherapy in-
volves the use of essential oils and es-
sences derived from plants, flowers, and
wood resins, which are generally mas-
saged into the skin. These essential oils
have been used to complement tradi-
tional medicine with some benefit.1As with
other forms of alternative medicine, sci-
entific bases for the claims made are few,2,3
but in dermatologic studies, physiologi-
cal and psychological benefits were found
after treatment of psoriasis with aroma-
therapy.4Significant, consistent improve-
ment in behavior occurred in 1 of 4 pa-
tients with dementia5in 10 repeated
experiments. Aromatherapy is also use-
ful in hospice care.6
As examples of proven therapeutic
benefit, use of sandalwood oil has been
shown to significantly inhibit skin papil-
lomas in mice.7Tea tree oil is an effective
bacteriocide8and fungicide, especially for
Malassezia furfur.9Use of essential oils can
also alter the barrier functions of the skin10
and induce contact dermatitis.11
Cedarwood, lavender, thyme, and
rosemary oils have hair growth–
promoting properties. These oils have been
anecdotally used to treat alopecia for more
than 100 years. To date, there have been
no controlled trials to evaluate this treat-
ment. Our experience using aroma-
therapy provides anecdotal evidence that
several patients have had marked improve-
ment with this form of therapy. Our aim
in this study was to test the hypothesis that
pharmacologically active stimulants for
hair growth are present in these oils and
that use of these oils can be therapeutic
in patients with alopecia.
Alopecia areata is a common con-
dition affecting 1% of the Western
world. It can cause substantial social
and psychological distress and is often
highly detrimental to the patient’s well-
being and self-esteem. In 1 study,12
STUDY
From the Department of
Dermatology, Aberdeen Royal
Infirmary, Foresterhill,
Aberdeen, Scotland.
ARCH DERMATOL / VOL 134, NOV 1998
1349
©1998 American Medical Association. All rights reserved.
on October 19, 2010 www.archdermatol.comDownloaded from
patients with alopecia areata had an increased risk of
developing a psychiatric illness. An autoimmune
cause is widely accepted. However, it is thought that
stress can affect its course. Current conventional treat-
ments have limited success or unacceptable toxic
effects. The course of alopecia areata is unpredictable,
but factors affecting prognosis are well established.
A poor prognosis is indicated by the long duration
of the condition (.1 year), associated autoimmune
conditions (eg, thyroid disease), atopy, and family
history.13-16
RESULTS
Eighty-four patients entered the trial; 28 (68%) of the pa-
tients in the control group and 35 (81%) of the patients
in the active group completed the trial (Figure 2). The
patients were well matched for the important demo-
graphic indicators that might affect response to therapy
(Table 1). The distribution of patients by the 4-point
scale was similar in both groups.
The improvement was statistically significant in all as-
sessments undertaken. The primary outcome measure of im-
provement vs no improvement showed improvement with
essential oils (P= .008, x2)(Table 2). The degree of im-
provement shown in the photographs was assessed by the
Mann-Whitney Utest and was significant (P= .05). The re-
sults of the alopecia scale, which scored the degree of im-
provement (from 1-6), are illustrated in Figure 3. The mea-
surement of traced areas, which could be performed in only
32 patients, showed a mean±SD reduction in area affected
of 103.9 ± 140.0 cm2compared with −1.8 ± 155.0 cm2in the
control group (Figure 4). This was significant, with P= .05
(Student ttest). A relative risk of 2.6 (95% confidence lim-
its, 1.2, 5.6) was calculated for the likelihood of improving
on the active therapy. Weighted kstatistic was 0.84
for agreement between scorers on the assessment of the pho-
tographic scale, showing good interrater correlation.
This indicates that this is a reproducible method of
assessment. One patient who received active treatment
and had an excellent response (ie, a score of 6 on our
scale) is pictured in Figure 5.
PATIENTS, MATERIALS,
AND METHODS
PATIENTS
Eighty-six patients diagnosed as having alopecia areata were
invited to take part in this randomized, controlled trial. These
patients were interviewed, and they completed a question-
naire. Patients with a medical history of hypertension, epi-
lepsy, or pregnancy were excluded. Two of 86 patients were
excluded because they had androgenic alopecia. Topical
medication and intralesional corticosteroid therapy for the
alopecia were discontinued before the trial.
Eighty-four patients were randomized into 2 groups
by the aromatherapist (M.J.). All patients signed a written
consent form before the trial.
The trained aromatherapist explained how to use the
oils and demonstrated the technique of scalp massage. The
oils were massaged into the scalp for a minimum of 2 min-
utes. A warm towel was then wrapped around the head to
aid absorption of the oils. Patients were advised to use this
technique every night.
There were 2 arms to the trial. The active group re-
ceived the essential oils: Thyme vulgaris (2 drops, 88 mg),
Lavandula agustifolia (3 drops, 108 mg), Rosmarinus offi-
cinalis (3 drops, 114 mg), and Cedrus atlantica (2 drops,
94 mg). These oils were mixed in a carrier oil, which was
a combination of jojoba, 3 mL, and grapeseed, 20 mL, oils.
The control group received the same carrier oils with-
out added essential oils, and the oils were identical except
in smell, which could not be mimicked in the control.
MEASUREMENT AND EVALUATION
Initial and 3- and 7-month assessments were made by 3
methods:
1. A 4-point scale, such as that described by MacDon-
ald Hull and Norris,17 was used to check that the severity was
similar in active and control groups, as follows: 1 indicates
vellus hair or no hair; 2, sparse pigmented or nonpigmented
terminal hair; 3, terminal regrowth with patches of alopecia
areata; and 4, terminal regrowth in all areas.
2. A standardized professional photographic assess-
ment of each volunteer was taken at the initial interview
and after 3 and 7 months. Changes in these photographic
assessments formed the primary outcome measure, with
improvement as the most important factor. These changes
were scored independently by 2 dermatologists (I.C.H. and
A.D.O.) who were unaware of the therapy administered.
Improvement in photographic assessment was graded us-
ing a numerical scale (Figure 1).
3. A further secondary outcome measure was per-
formed. A map was traced onto transparent film wherever
the alopecia occurred in patches. These tracings were then
transferred onto flat acetate sheets. A computerized image
analyzer was used to calculate the areas of alopecia at the
initial assessment and after 3 and 7 months.
STATISTICAL METHODS
We calculated that if improvement occurred in 20% more
patients with active treatment than in the control group it
would require 47 active and 47 control patients to detect
improvement at a 5% significance level with a power of
80%.18 Statistical analysis was performed on an intention-
to-treat basis. A pooled variance estimate Student ttest for
independent samples was used to test improvement in the
patients’ alopecia with the map-tracing method. The x2test
was used to detect improvement in the active and control
groups. A Mann-Whitney Utest, corrected for ties, was used
to assess the significance of the degree of improvement in
scored photographic assessment. The level of agreement
of the alopecia scoring scale between 2 assessors was ex-
amined using the Cohen weighted kstatistic.
The trial was approved by the Joint Ethical Commit-
tee for the Grampian Health Board and by the University
of Aberdeen, Aberdeen, Scotland.
ARCH DERMATOL / VOL 134, NOV 1998
1350
©1998 American Medical Association. All rights reserved.
on October 19, 2010 www.archdermatol.comDownloaded from
COMMENT
The responses were variable but showed a clear and sta-
tistically significant advantage to treatment with this stan-
dardized regimen of aromatherapy. Although the tradi-
tion of aromatherapy is to combine several oils, it seems
likely that 1 of these agents has a stimulatory effect on hair
growth. One male patient also had severe androgenic alo-
pecia, which was not included in the assessment of effi-
cacy, and within this area there was some moderate re-
growth of hair and improvement in alopecia areata.
There was a higher dropout rate in the control group,
which could be explained by the fact that the volunteers be-
came discouraged with the 7-month protocol. The control
oil was not odorless because the carrier oils have some smell,
and patients did not know what aroma to expect. However,
they may have surmised that their treatment was inactive
and withdrew from the trial. The control group’s relative lack
of response again suggests a pharmacoactive property of the
topically applied therapy as opposed to an effect arising from
the comforting, relaxing effect of massage and of the appli-
cation procedure, which was the same for both groups.
Previous studies of alopecia therapies have used only
subjective scales of improvement, such as those de-
scribed by MacDonald Hull and Norris.17 We found these
scales less helpful because there are large intervals be-
tween the points in the alopecia scale. We validated the
method of sequential photography with a standardized
approach using a professional photographer’s studio. Im-
ages were judged by blinded observers and showed good
agreement. This proved to be the most blinded and un-
biased approach because there was no possibility of smell-
ing which treatment had been applied.
We encountered no significant adverse events from
this treatment, which makes the therapeutic ratio high
Scale Description Hair Regrowth, %
1 Worse NA
2 No Change NA
3 Slight but Definite Improvement 10-30
4 Marked Improvement 31-50
5 Very Good Improvement 51-80
6 Excellent Improvement 81-100
Figure 1.
The scale used in the photographic assessment to measure
changes in hair growth after intervention with active essential oils compared
with the control. NA indicates not applicable.
Registered Patients
(N
=
86)
Randomization
Not Randomized
(n
=
2)
Reason: Did Not Have
Alopecia Areata
Standard Intervention (n
=
41)
Did Not Receive (n
=
0)
Followed Up (n
=
28)
Withdrawn (n
=
13)
Completed Trial (n
=
28)
Withdrawn (n
=
8)
Completed Trial (n
=
35)
Followed Up (n
=
35)
Intervention (n
=
43)
Did Not Receive (n
=
0)
Control Active
Figure 2.
Flow diagram showing the randomization of patients to the active
and control groups.
No. of Patients
20
15
10
5
065
Active Treatment
Control
4321
Improvement Score
Figure 3.
Results of the photographic assessment of the degree of
improvement in alopecia on a 6-point scale (see Figure 1).
Reduction in Area Affected, cm2
300
200
100
0Active Group Control Group
Figure 4.
Results of computerized analysis of traced areas of alopecia in
active and control groups. Bars indicate SD.
Table 1. Demographic and Prognostic Factors
in the Active and Control Groups*
Active Group Control Group Total
Age, mean ± SD, y 38.9 ± 14.6 39.3 ± 13.6 . . .
Associated conditions
Thyroid disease 14.0 12.2 13.1
Atopy 18.6 17.1 17.9
Stress 48.8 53.7 51.2
Family history
Alopecia areata 18.6 17.1 17.9
Autoimmune thyroid disease 18.6 24.4 21.4
*
Values are expressed as percentages unless otherwise indicated. Ellipses
indicate not applicable.
Table 2. Contingency Table of Improvement With Each
Intervention: Active Essential Oils vs Control Carrier Oil Alone*
Improved Not Improved Total
Active 19 16 35
Control 6 22 28
Total 25 38 63
*P
= .008 (
x
2
).
ARCH DERMATOL / VOL 134, NOV 1998
1351
©1998 American Medical Association. All rights reserved.
on October 19, 2010 www.archdermatol.comDownloaded from
compared with other therapies, such as diphency-
prone,14,16 squaric acid dibutylester,19 and systemic or in-
tralesional corticosteroid injections.
Populations in other trials may differ in prognostic
factors. Shapiro et al20 and Gordon et al14 found a 38%
success rate in producing cosmetically acceptable re-
growth in patients with alopecia using diphencyprone.
A review of psoralen–UV-A therapy for alopecia areata
by Taylor and Hawk21 revealed that phototherapy was dis-
appointing, with minimal benefit. Therefore, this aro-
matherapy trial, with an improvement rate of 44%, is com-
parable to and possibly of more benefit than trials of
conventional therapies for alopecia areata. Compared with
these other treatments, its safety is also greater, offering
a better therapeutic ratio.
Accepted for publication April 16, 1998.
The Soropotomists International of Aberdeen, Scot-
land, provided financial support for the trial.
We thank Jill Mollison, BSc, for statistical advice and
the members of the Aberdeen Alopecia Self Help Group for
their participation.
Corresponding author: Isabelle C. Hay, MRCP, De-
partment of Dermatology, Ward 48, Aberdeen Royal Infir-
mary, Foresterhill, Aberdeen AB25 2ZN, Scotland (e-mail:
ad.ormerod@abdn.ac.uk).
REFERENCES
1. Whitmore SM, Leake NB. Complementary therapies: an adjunct to traditional thera-
pies.
Nurse Pract.
1996;21:10-13.
2. Vickers A. Yes, but how do we know it’s true? knowledge claims in massage
and aromatherapy.
Complement Ther Nurs Midwifery.
1997;3:63-65.
3. Martin GN. Olfactory remediation: current evidence and possible applications.
Soc Sci Med.
1996;43:63-70.
4. Walsh D. Using aromatherapy in the management of psoriasis.
Nurs Stand.
1996;
11:53-56.
5. Brooker DJ, Snape M, Johnson E, Ward D, Payne M. Single case evaluation of
the effects of aromatherapy and massage on disturbed behaviour in severe
dementia.
Br J Clin Psychol.
1997;36:287-296.
6. Welsh C. Touch with oils: a pertinent part of holistic hospice care.
Am J Hosp
Palliat Care.
1997;14:42-44.
7. Dwivedi C, Abu-Ghazaleh A. Chemopreventive effects of sandalwood oil on skin
papillomas in mice.
Eur J Cancer Prev.
1997;6:399-401.
8. Hammer KA, Carson CF, Riley TV. Susceptibility of transient and commensal skin
flora to the essential oil of
Melaleuca alternifolia
(tea tree oil).
Am J Infect Con-
trol.
1996;24:186-189.
9. Nenoff P, Haustein UF, Brandt W. Antifungal activity of the essential oil of
Melaleuca alternifolia
(tea tree oil) against pathogenic fungi in vitro.
Skin Phar-
macol.
1996;9:388-394.
10. Yamane MA, Williams AC, Barry BW. Terpene penetration enhancers in propyl-
ene glycol/water co-solvent systems: effectiveness and mechanism of action.
J Pharm Pharmacol.
1995;47:978-989.
11. Weiss RR, James WD. Allergic contact dermatitis from aromatherapy.
Am J Con-
tact Dermat.
1997;8:250-251.
12. Koo YM, William VR, Shellow MD, Hallman CP, Edwards JE. Alopecia areata and
increased prevalence of psychiatric disorders.
Int J Dermatol.
1994;33:849-850.
13. van der Steen PHM, van Baar HMJ, Happle R, Boezeman JMB, Perret MJ. Prog-
nostic factors in the treatment of alopecia areata with diphencyprone.
Am Acad
Dermatol.
1991;24:227-230.
14. Gordon PM, Aldridge RD, McVittie E, Hunter JAA. Topical diphencyprone for alo-
pecia areata: evaluation of 48 cases after 30 months’ follow up.
Br J Dermatol.
1996;134:869-871.
15. De Prost Y, Paquez F, Touraine R. Dinitrochlorobenzene treatment of alopecia
areata.
Arch Dermatol.
1982;118:542-545.
16. Weise K, Kretzschmar L, John SM, Hamm H. Topical immunotherapy in alope-
cia areata: anamnestic and clinical criteria of prognostic significance.
Dermatol-
ogy.
1996;192:129-133.
17. MacDonald Hull S, Norris JF. Diphencyprone in the treatment of long standing
alopecia areata.
Br J Dermatol.
1988;119:367-374.
18. Pocock SJ, Hughes MD. Estimation issues in clinical trials and overviews.
Stat
Med.
1990;9:657-671.
19. Frattasio A, Germino M, Cargnello S, Patrone P. Side-effects during treatment
with SADBE.
Contact Dermatitis.
1997;36:118-119.
20. Shapiro J, Tan J, Ho V, Tron V. Treatment of severe alopecia areata with topical
diphenylcyclopropenone and 5% minoxidil: a clinical and immunopathologic evalu-
ation.
J Invest Dermatol.
1995;104(suppl 5):36S.
21. Taylor CR, Hawk JLM. PUVA treatment of alopecia areata parrialis, totalis and
universalis: audit of ten years’ experience at St. John’s Institute of Dermatology.
Br J Dermatol.
1995;133:914-918.
Figure 5.
A patient showing an excellent response (6 on the scoring scale) to essential oil therapy.
ARCH DERMATOL / VOL 134, NOV 1998
1352
©1998 American Medical Association. All rights reserved.
on October 19, 2010 www.archdermatol.comDownloaded from