BJD

R EV IE W AR TI C LE British Journal of Dermatology

Hormonal changes in menopause: do they contribute to a

‘midlife hair crisis’ in women?
P. Mirmirani* à
*Department of Dermatology, The Permanente Medical Group, Vallejo, CA 94589, U.S.A.
Department of Dermatology, University of California, San Francisco, CA 94143, U.S.A.
àDepartment of Dermatology, Case Western Reserve University, Cleveland, OH 44106, U.S.A.

Summary

Correspondence Female pattern hair loss is a common problem affecting a large number of
Paradi Mirmirani. women worldwide but beset by a paucity of research. The study of androgens
E-mail: paradi.mirmirani@kp.org has hitherto dominated the field of hair biology but there is increasing scientific
and clinical data to suggest that nonandrogen signals can also affect the follicu-
Accepted for publication
23 August 2011
losebaceous unit, especially in women. The discovery of oestrogen receptor beta
has renewed and redefined prior concepts of oestrogen activity and signalling in
Funding sources hair biology. It is postulated that oestrogens modulate hair growth by their influ-
Financial support for publication of this ence on a number of other hormones, growth factors, transcription factors and
supplement was provided by Procter & Gamble. cytokines. The menopause is a period in which significant changes in oestrogen
levels are recorded, and this review discusses studies that help to clarify the link
Conflicts of interest
P.M. has received honoraria from Procter & between menopause and the perception of thinning hair. In a study of pre- and
Gamble for speaking at satellite symposia. postmenopausal women without alopecia, menopausal status significantly influ-
enced hair parameters, specifically hair growth rate, percentage anagen and hair
DOI 10.1111/j.1365-2133.2011.10629.x diameter distributions, most notably in the frontal scalp. Hair density decreased
with age, but was not correlated with menopausal status. Analyses of hair
amount using a model of hair density and hair diameters suggest that the impact
of changing hair parameters is most notable in the mid-forties for women.

In a typical case scenario, the archetypal patient is a 49-year- that the hormonal changes in menopause affect the hair folli-
old healthy woman who reports that in the past year, her hair cles? How much of the change is age related? What types of
is ‘just not the same’. It inadequately covers her scalp, does changes occur in the hair that can lead to a perception of hair
not achieve any length and the texture has changed. She is loss? Can these changes be quantified?
perimenopausal, with irregular periods for the last 2 years.
She has no significant family or personal history. All labora-
Perimenopause ⁄menopause
tory tests are normal. On examination, the patient is slender,
with clear skin, and with no other evidence of androgen Perimenopause is considered to be a transition phase prior to
excess. Her hair is fine textured, with some visibility of her menopause, in which menses become irregular, and hormonal
scalp and increased spacing between hairs on the vertex com- fluctuations occur. Perimenopause can commence as early as
pared with the occiput. Miniaturized hairs are noted. Her pull 35 years to the mid-forties.1 Menopause is defined either as
test is negative. A scalp biopsy shows a normal follicular the permanent cessation of menses, or the lack of menses for
count, preserved sebaceous glands and an increased percentage 12 consecutive months.2 The mean age women undergo men-
of miniaturized hairs. The patient is informed that she has opause is 51 years, and as a consequence women spend
female pattern hair loss (FPHL) (Fig. 1). approximately one-third of their life in the postmenopausal
The diagnosis of FPHL often unleashes a cascade of ques- period.2 One of the major changes during menopause is the
tions from patients and can be one of the more challenging virtual cessation of ovarian oestrogen production. The major
cases for a clinician in terms of providing answers, counselling source of oestrogen after the menopause is from conversion
and treatment options. The aim of this paper is to address of adrenal androgen to oestrogen by the enzyme aromatase in
some of these common questions and controversies: the peripheral tissues.
What is the aetiology of FPHL? Are there nonandrogen trig- While it has been recognized that oestrogen is an important
gers of hair thinning in women? Specifically, is there evidence modulator of hair growth, most evidently during pregnancy

 2011 The Author

BJD  2011 British Association of Dermatologists 2011 165 (Suppl. 3), pp 7–11 7
8 Hormonal changes in menopause, P. Mirmirani

is plausible that an intricate orchestration of these pathways

occurs in response to oestrogen. Further clarification and study
of oestrogen effects in different tissues, species and sexes is
ongoing.5,6,18

Clinical observations
Clinical observations have supported the hypothesis that oes-
trogen is an important regulator of hair follicle growth and
cycling. The use of topical oestrogens to treat androgenetic
alopecia and telogen effluvium has been proposed based on
small clinical studies and they are commercially available in
Europe, but their efficacy remains controversial.17,19 The use
of oral contraceptives, their cessation, or hormone replace-
Fig 1. A 49-year-old woman with a 2-year history of hair thinning
diagnosed with female pattern hair loss. ment therapy, have all been reported to cause a temporary
hair shedding or telogen effluvium.20 Although the proges-
terone component of oral contraceptives has been thought
and nursing, the molecular regulatory pathways have not been to be the culprit in such an effluvium, the oestrogen con-
well characterized. In contrast, the role of androgens on hair tent may also have an effect. Women with breast cancer
growth has dominated the field of hair biology. As our under- who take medications aimed at decreasing the effect of oes-
standing of the molecular and hormonal controls on the follic- trogens (selective oestrogen receptor modulators) or medica-
ulosebaceous unit has expanded, there has been renewed tions that block the peripheral production of oestrogen
interest in the role of oestrogens as well as androgens in mod- (aromatase inhibitors) commonly report hair loss. In preg-
ulating hair growth. nant women, there is an increase in hair growth rate, diam-
eter and the ratio of anagen ⁄telogen hairs.21–23 Although
higher oestrogen levels are likely to play a role in this alter-
Oestrogens and the hair follicle
ation of hair biology, it is difficult to attribute the phenom-
Oestrogen is synthesized in the ovary and in a number of enon to oestrogen alone because of the complex hormonal
peripheral tissues and acts via oestrogen receptors which changes seen in pregnancy including increases in progester-
belong to a superfamily of nuclear receptors. There are two one and prolactin.
oestrogen receptors, alpha (ER alpha) and beta (ER beta). The It has been suggested that the various clinical patterns of
relatively recent discovery of ER beta has broadened the range androgenetic alopecia in men and women may reflect quanti-
of potential oestrogenic target tissues and has also redefined tative differences in levels of androgen receptor and steroid-
prior concepts of oestrogen activity and signalling. In the converting enzymes in specific scalp regions at different
human hair follicle, immunohistochemical studies have shown ages.24 However, in recognition that there may be other non-
ER beta to be the predominant receptor.3,4 Similar to other androgen causes of hair thinning in women that have no
oestrogenic target tissues, the biological activity of oestrogen counterpart in men, the term FPHL has been coined to
in the hair follicle probably depends on a complex interplay encompass the clinical phenotype of hair loss in the central
of signals that may differ depending on the relative distribu- scalp region that may occur: (i) in genetically predisposed
tion and location of the two ERs, as well as the activity of the women due to androgens; (ii) due to nonhereditary androgen
peripheral converting enzyme, aromatase.5–7 excess; and (iii) due to menopausal hormonal changes.25,26
Several studies have demonstrated the influence of oestro- Although the aetiology of FPHL has not been fully eluci-
gen on the murine and other mammalian hair cycle; however, dated, understanding the many hormonal factors that are
it is clear that the distribution, expression and biological changing in concert during middle age and menopause may
activity of oestrogen receptors in murine models may be quite be helpful in this group. A study supporting the notion that
different than in humans.7–13 In a recent study, a topical anti- nonandrogen pathways may be involved in FPHL showed that
oestrogen that caused hair growth in a murine model was sebum secretion, a marker of end-organ androgen response,
ineffective in humans.14 In vitro studies have shown that organ was not elevated.27 The report of a woman with hypopituita-
culture of human scalp hair follicles exposed to estradiol rism and hair thinning in the absence of detectable androgen
results in decreased growth whereas cells of the dermal papilla levels also suggests that hair thinning is not exclusively andro-
responded with proliferation.15,16 Estradiol has also been gen dependent.28
noted to induce aromatase activity in human scalp follicles, Further complicating our understanding of the possible role
one possible mechanism by which it may exert biological of menopause on hair characteristics is whether age-related
activity.17 As hair growth is influenced by numerous hor- changes may coexist, overlap or supersede hormonal changes.
mones, growth factors, transcription factors and cytokines, Various reports have suggested that decreased hair density and
many of which are known to be modulated by oestrogens, it diameter occurs with advancing age in both sexes.29–33

 2011 The Author

BJD  2011 British Association of Dermatologists 2011 165 (Suppl. 3), pp 7–11
 Hormonal changes in menopause, P. Mirmirani 9

In summary, there is convincing evidence from clinical ob-

servations that menopause and specifically oestrogens have an
effect on hair biology. However, there is currently inadequate
information to attribute specific hair changes to hormonal
alterations seen in menopause. Despite the prevalence of FPHL,
there are still many shortcomings in our pathogenetic under-
standing of hair loss in women.

Analysis of hair growth parameters in women

Despite the absence of a known molecular pathogenesis of
FPHL, advances have been made in the clinical arena, bolster-
ing our understanding of the variations in hair growth
parameters. Specifically, refinement of computerized and com-
puter-aided techniques for measurement of hair growth
parameters has made such measurements more accurate and
reproducible, opening the door to novel types of analysis.
Two such techniques are the modified phototrichogram and
the optical fibre diameter analysis (OFDA). Briefly, the tradi-
tional phototrichogram technique involves photography of a

0.0
target scalp area in which the hair is clipped to approximately 0 50 100 150
1 mm in length. A repeat photograph of the target area taken Diameter (μm)

1–3 days later allows for identification of anagen hairs. Fur-

Fig 2. Optical fibre diameter analysis showing a bimodal distribution
ther enhancement of the phototrichogram technique has
of hair diameters.
included the use of digital luminescence microscopy and com-
puterized ‘blending’ in which the photographic images are
magnified, digitally captured and montaged, thus allowing for age.29,33 Therefore, it is likely that chronological age, but not
determination of length, diameter and growth rate, in add- menopausal status, dictates alterations in hair density. Hair
ition to density, percentage anagen and anagen ⁄telogen ratio. diameters as measured by OFDA were lower in the frontal
Improved speed and ease of measurement of hair diameters scalp in postmenopausal women; this finding was independent
has been made possible using OFDA, a technique that was ini- of age. Further analysis of this cohort as well as a larger
tially optimized and validated in the wool industry and has cohort of 177 women showed that hair diameters were not
now been adapted for research purposes in human hair. The normally distributed, and were typically bi- or trimodal (large,
OFDA instrument analyses electronic images of horizontally medium and small diameter).36 While the mean fibre diame-
cut and magnified hairs, which are distributed over a glass ter reported by OFDA is an accurate, reproducible and valu-
slide. Software analysis of these images derives large-volume, able tool for assessment of hair diameter, the diameter
rapid and highly accurate measurement of diameter of various distribution may report a more accurate and real-time assessment
longitudinal hair sections.34,35 In contrast, traditional hair of hair diameter and diameter changes. Using this new mea-
diameter measurement is too labour intensive to analyse more surement technology it was noted that there was a decrease in
than a few hairs from any individual sample. In addition to high diameter hairs in postmenopausal women. Future work
providing highly reliable and sensitive measurements, OFDA entails development of more quantitative tools for analysis of
has another advantage in generating data on the distribution OFDA distributions and their relevance to assessment of hair
of hairs as a function of diameter (Fig. 2). loss.
In a study of 44 pre- and postmenopausal women, without In order to give clinical relevance to the measurements of
overt hair loss, modified trichogram and OFDA measurements hair growth parameters, and to determine the variable impact
showed that there are significant differences in hair growth of each of these individual hair growth parameters on the per-
characteristics between pre- and postmenopausal women, spe- ception of hair loss, a new metric has been developed (Rob-
cifically hair growth rate, density, percentage anagen, hair bins C, Mirmirani P, Messenger AG, Birch MP, Youngquist RS,
diameters and hair diameter distributions.36 These changes Filloon T, Luo F, Dawson TL Jr. submitted). A convenient,
were more pronounced in the frontal compared with the descriptive and quantitative measure of hair ‘amount’ com-
occipital scalp. Whereas a decrease in hair growth rate and bines the technical measures of density and diameter. Hair
percentage anagen was associated with menopausal status, ‘amount’, or volume, is defined based on mass as simply the
scalp hair density decreased with age, and was not signifi- square of the average radius times the density. Relative scalp
cantly influenced by menopausal status. Prior studies of coverage is then defined as a function of the sum total of all
women without hair loss have also shown a wide, but normal hair cross-sectional areas in 1 cm2 of skin surface, expressed
distribution of scalp hair density that steadily declines with as lm2 hair surface lm)2 skin surface.

 2011 The Author

BJD  2011 British Association of Dermatologists 2011 165 (Suppl. 3), pp 7–11
10 Hormonal changes in menopause, P. Mirmirani

600 000 occurs in the mid-late forties. In other words, the com-
pounded effect of hair density and hair diameter changes may
500 000
lead to a heightened perception of decreased scalp coverage in
middle-aged women, leading to a ‘midlife hair crisis’ as high-
400 000
lighted in the case scenario.
Volume

300 000 Further study of hair changes in response to menopause

and age may provide an important opportunity for identifica-
200 000 tion of treatments targets and strategies that may significantly
benefit women.
100 000

0 What’s already known about this topic?

20 30 40 50 60
Age
• Androgens are known to have a role in the pathophysi-
Fig 3. Relative scalp coverage shows a more rapid decline in the mid- ology of female pattern hair loss, but other, nonandrogen
late forties in women. Age is given in years. signals may also cause hair miniaturization in women.

In a large comprehensive dataset of 1099 caucasian women What does this study add?
(aged 18–66 years) with self-perceived hair loss and 315 cau-
• The hormonal changes of menopause lead to decreased
casian women (aged 17–86 years) with no complaints of hair
hair diameter, growth rate and percentage anagen. Hair
loss scalp hair diameter and density measurements were analy-
density is affected by chronological age. The compounded
sed and hair ‘amount’ was calculated. It was noted that parietal
effect of these changes may lead to a heightened percep-
scalp hair diameter increased until 40–45 years and then
tion of decreased scalp coverage in middle-aged women.
decreased with increasing rate. Hair density peaked between
20 and 30 years of age and then steadily decreased. Hair
amount and relative scalp coverage increased only slightly until
the early forties and then decreased significantly (Fig. 3). These
References
findings suggest that increasing hair diameter minimizes the
perception of decreasing hair density until the mid-late forties 1 Burger HG, Dudley EC, Robertson DM et al. Hormonal changes in
when the combined decreases in density and diameter impact the menopause transition. Recent Prog Horm Res 2002; 57:257–75.
2 Rymer J, Morris EP. Extracts from ‘Clinical evidence’: menopausal
perception of ‘amount’. Thus, determination of hair amount
symptoms. BMJ 2000; 321:1516–19.
may provide a useful tool to understand the contributions of 3 Thornton MJ, Taylor AH, Mulligan K et al. Oestrogen receptor beta
hair density and diameter on women’s perception of hair loss. is the predominant oestrogen receptor in human scalp skin. Exp
The age at which these changes occur supports the notion that Dermatol 2003; 12:181–90.
hormonal influences, which may be starting at the perimeno- 4 Thornton MJ, Taylor AH, Mulligan K et al. The distribution of
pausal transition stage, are impacting hair parameters. estrogen receptor beta is distinct to that of estrogen receptor alpha
and the androgen receptor in human skin and the pilosebaceous
unit. J Investig Dermatol Symp Proc 2003; 8:100–3.
Conclusions 5 Ohnemus U, Uenalan M, Inzunza J et al. The hair follicle as an
estrogen target and source. Endocr Rev 2006; 27:677–706.
Androgens are known to cause hair miniaturization in FPHL; 6 Conrad F, Paus R. Estrogens and the hair follicle. J Dtsch Dermatol Ges
however, other, nonandrogen signals may also influence the 2004; 2:412–23.
hair. Our understanding of the effects of oestrogen on the hair 7 Ohnemus U, Uenalan M, Conrad F et al. Hair cycle control by
makes this hormone the leading target for ongoing study in estrogens: catagen induction via estrogen receptor (ER)-alpha is
FPHL. Until the complex molecular effects of oestrogen on checked by ER beta signaling. Endocrinology 2005; 146:1214–25.
8 Stenn KS, Paus R, Filippi M. Failure of topical estrogen receptor ag-
peripheral tissues are fully understood, progress can be made in
onists and antagonists to alter murine hair follicle cycling. J Invest
identifying and describing relevant clinical observations. Specif- Dermatol 1998; 110:95.
ically, refinement of computerized and computer-aided tech- 9 Smart RC, Oh HS. On the effect of estrogen receptor agonists and
niques such as the modified phototrichogram and OFDA have antagonists on the mouse hair follicle cycle. J Invest Dermatol 1998;
allowed for reliable measurements of hair growth parameters. 111:175.
There are significant differences in hair growth characteris- 10 Ohnemus U, Unalan M, Handjiski B et al. Topical estrogen acceler-
tics between pre- and postmenopausal women, specifically ates hair regrowth in mice after chemotherapy-induced alopecia by
favoring the dystrophic catagen response pathway to damage.
hair growth rate, percentage anagen, hair diameters and hair
J Invest Dermatol 2004; 122:7–13.
diameter distributions. Hair density is correlated with chrono- 11 Oh HS, Smart RC. An estrogen receptor pathway regulates the telo-
logical age, as opposed to hormonal status. Combination of gen–anagen hair follicle transition and influences epidermal cell
hair diameter and density into a single metric of hair ‘amount’ proliferation. Proc Natl Acad Sci USA 1996; 93:12525–30.
suggests a dramatic impact on the perception of hair loss that

 2011 The Author

BJD  2011 British Association of Dermatologists 2011 165 (Suppl. 3), pp 7–11
 Hormonal changes in menopause, P. Mirmirani 11

12 Chanda S, Robinette CL, Couse JF et al. Am J Physiol Endocrinol Metab 24 Sawaya ME, Price VH. Different levels of 5alpha-reductase type I
2000; 278:E202–10. and II, aromatase, and androgen receptor in hair follicles of
13 Moverare S, Lindberg MK, Faergemann J et al. Estrogen receptor women and men with androgenetic alopecia. J Invest Dermatol 1997;
alpha, but not estrogen receptor beta, is involved in the regulation 109:296–300.
of the hair follicle cycling as well as the thickness of epidermis in 25 Olsen EA. Female pattern hair loss. J Am Acad Dermatol 2001;
male mice. J Invest Dermatol 2002; 119:1053–8. 45:S70–80.
14 Gassmueller J, Hoffmann R, Webster A. Topical fulvestrant solution 26 Olsen EA, Hordinsky M, Roberts JL et al. Female pattern hair loss.
has no effect on male and postmenopausal female androgenetic J Am Acad Dermatol 2002; 47:795.
alopecia: results from two randomized, proof-of-concept studies. 27 Birch MP, Lashen H, Agarwal S et al. Female pattern hair loss,
Br J Dermatol 2008; 158:109–15. sebum excretion and the end-organ response to androgens. Br J
15 Kondo S, Hozumi Y, Aso K. Organ culture of human scalp hair Dermatol 2006; 154:85–9.
follicles: effect of testosterone and oestrogen on hair growth. Arch 28 Orme S, Cullen DR, Messenger AG. Diffuse female hair loss: are
Dermatol Res 1990; 282:442–5. androgens necessary? Br J Dermatol 1999; 141:521–3.
16 Conrad F, Ohnemus U, Bodo E et al. Substantial sex-dependent differ- 29 Birch MP, Messenger JF, Messenger AG. Hair density, hair diameter
ences in the response of human scalp hair follicles to estrogen stimu- and the prevalence of female pattern hair loss. Br J Dermatol 2001;
lation in vitro advocate gender-tailored management of female versus 144:297–304.
male pattern balding. J Investig Dermatol Symp Proc 2005; 10:243–6. 30 Courtois M, Loussouarn G, Hourseau C et al. Ageing and hair
17 Hoffmann R, Niiyama S, Huth A. et al. 17alpha-estradiol induces cycles. Br J Dermatol 1995; 132:86–93.
aromatase activity in intact human anagen hair follicles ex vivo. Exp 31 Kligman AM. The comparative histopathology of male-pattern
Dermatol 2002; 11:376–80. baldness and senescent baldness. Clin Dermatol 1988; 6:108–18.
18 Thornton MJ. Oestrogen functions in skin and skin appendages. 32 Loussouarn G, El Rawadi C, Genain G. Diversity of hair growth
Expert Opin Ther Targets 2005; 9:617–29. profiles. Int J Dermatol 2005; 44 (Suppl. 1):6–9.
19 Orfanos CE, Vogels L. [Local therapy of androgenetic alopecia with 33 Tajima M, Hamada C, Arai T et al. Characteristic features of Japa-
17 alpha-estradiol. A controlled, randomized double-blind study]. nese women’s hair with aging and with progressing hair loss. J Der-
Dermatologica 1980; 161:124–32. matol Sci 2007; 45:93–103.
20 Bergfeld WF. Telogen effluvium. In: Hair and Scalp Diseases (McMich- 34 Qi K, Lupton CJ, Pfeiffer FA et al. Evaluation of the optical fibre
ael AJ, Hordinsky MK, eds). New York: Informa Healthcare, 2008; diameter analyser (OFDA) for measuring fiber diameter parameters
125. of sheep and goats. J Anim Sci 1994; 72:1675–9.
21 Nissimov J, Elchalal U. Scalp hair diameter increases during preg- 35 Lupton CJ, McColl A, Stobart RH. Fiber characteristics of the Hua-
nancy. Clin Exp Dermatol 2003; 28:525–30. caya alpaca. Small Rumin Res 2006; 64:211–24.
22 Lynfield YL. Effect of pregnancy on the human hair cycle. J Invest 36 Mirmirani P, Luo F, Youngquist RS et al.. Hair growth parameters
Dermatol 1960; 35:323–7. in pre- and post-menopausal women. In: Aging Hair (Trüeb RM,
23 Muallem MM, Rubeiz NG. Physiological and biological skin Tobin DJ, eds). Heidelberg: Springer-Verlag, 2010; 49–60.
changes in pregnancy. Clin Dermatol 2006; 24:80–3.

 2011 The Author

BJD  2011 British Association of Dermatologists 2011 165 (Suppl. 3), pp 7–11