Androgenetic Alopecia in the Female

Treatment With 2% Topical Minoxidil Solution

Richard L. DeVillez, MD; James P. Jacobs, MD; Cheryl A. Szpunar, MPH; Michael L. Warner

Background: Women generally regard their hair loss
as socially unacceptable and go to great measures to
conceal their problem. In some cases, the negative
self-image brought about by hair loss may be the basis
of psychiatric illness. The purpose of this study was to
evaluate a 2% topical minoxidil solution (Rogaine/
Regaine, The Upjohn Co, Kalamazoo, Mich) for the
treatment of female androgenetic alopecia. A 32-week,
double-blind, placebo-controlled trial was conducted
in 11 US centers. Three hundred eight women with
androgenetic alopecia enrolled.
were Two hundred
fifty-six of these women completed the trial. A refined
photographic technique was used to objectively deter-
mine the number of nonvellus hairs regrown.
Results: After 32 weeks of treatment, the number of
nonvellus hairs in a 1-cm2 evaluation site was in-
creased by an average of 23 hairs in the 2% minoxidil
group and by an average of 11 hairs in the placebo
group. The 95% confidence interval for the difference
in mean hair count change between the treatment
groups was 5.9 to 17.5 hairs. The investigators deter-
mined that 13% in the minoxidil-treated group had
moderate growth and 50% had minimal growth. This
compared with 6% and 33%, respectively, in the
placebo-treated group. Similarly, 60% of the patients
in the 2% minoxidil group reported that they had new
hair growth (20% moderate, 40% minimal) compared
with 40% (7% moderate, 33% minimal) of the patients
in the placebo group. No evaluations of dense hair
growth were reported for either treatment group. No
clinically significant changes in vital signs were ob-
served and no serious or unexpected medical events
were reported.
Conclusions: Topical minoxidil was significantly more
effective than placebo in the treatment of female andro-
genetic alopecia.
(Arch Dermatol. 1994;130:303-307)

NDROGENETIC alopecia is Grade I as a perceptible thinning of the

From the Division of
Dermatology, University
of Texas, San Antonio
(Dr DeVillez), and the
Dermatology Division of
Upjohn Laboratories,
Kalamazoo, Mich
(Dr Jacobs, Ms Szpunar,
and Mr Warner).
the most common cause hair on the crown with preservation of
of hair loss in both men the frontal hair line, Grade II as a pro¬
and women. It affects nounced thinning of the hair on the
approximately one third crown, and Grade III as total baldness
of individuals of either sex who have a in the area seen in Grades I and II.
strong family history of hair loss.1 In fe¬ Since topical minoxidil solution
males, the process generally begins in is currently the
only approved therapy
the teens, 20s, or 30s, and is usually for the treatment of androgenetic
fully expressed by 40 years of age in alopecia,5"11 this study was conducted to
susceptible individuals.2 Female andro¬ assess its efficacyand safety for the
genetic alopecia begins as a diffuse thin¬ treatment of female androgenetic
ning of the hair on the frontoparietal alopecia.
area of the scalp
that becomes progres¬
sively thinner each year. One typical
feature is a more apparent central part,
although the frontal hairline is generally
See Patients and Methods
retained.3 Ludwig4 classified the degree on next page
of hair loss in women with androgenetic
alopecia into three grades. He defined

Downloaded From: http://archderm.jamanetwork.com/ by a Oakland University User on 06/07/2015

 PATIENTS AND METHODS
STUDY DESIGN
A multicenter, double-blind, placebo-controlled trial was
conducted to compare 2% topical minoxidil solution (Ro-
gaine/Regaine, The Upjohn Co, Kalamazoo, Mich) (min¬
oxidil powder, propylene glycol, alcohol, and water) with
placebo (propylene glycol, alcohol, and water) for the treat¬
ment of androgenetic alopecia in the female. Patients ap¬
plied 1 mL of the solution to the scalp twice daily at ap¬
proximately 12-hour intervals for 32 weeks and were seen
in the clinic for efficacy and safety evaluations every 4 weeks
throughout the study.
PATIENTS
Patients were required to be female, between the ages of
18 and 45 years, with Ludwig's Grade I or II hair loss
and in good general health with no evidence of cardiac,
systemic, psychiatric, or scalp disease. Women previ¬
ously exposed to topical minoxidil solution were ineli¬
gible to participate in this study, as were women who
were pregnant, at risk of pregnancy, less than 12
months' post partum, or breast feeding. Patients were
also excluded from study enrollment if they had used
hair restorers or systemic drugs (eg, steroids, antihyper-
tensives, cytotoxic compounds, vasodilators, anticonvul-
sant drugs, ß-blockers, spironolactone, cimetidine, di-
azoxide, cyclosporin, ketoconazole, cyproterone acetate,
estrogens, or progestérones) within the previous 3
months. Informed consent was obtained from all pa¬
tients before study entry.
A total of 308 women were enrolled in the study
and randomized to receive either 2% minoxidil solution
or placebo. Two hundred fifty-six (83%) of these women
completed the 32 weeks of treatment and were consid¬
ered évaluable. Of these, 128 had applied 2% minoxidil
solution and 128 had applied placebo solution. Table I
shows the reasons why patients discontinued study par¬
ticipation, with voluntary withdrawal being the reason
given most frequently.
EFFICACY EVALUATION
Primary Efficacy Measure
Hair counts served as the primary efficacy measure and
were objectively determined by a method that combined
photography with computer-assisted image counting. A
1-cm2 evaluation site in the area of hair thinning was se¬
lected for each patient, and two opposing corners of the
square were permanently marked (using a template with
a square 1.2X1.2-cm opening) to ensure that the same
area wasclipped and photographed at each visit. The hair
Downloaded From: http://archderm.jamanetwork.com/ by a Oakland University User on 06/07/2015
in the outlined area was clipped to within 0.5 to 1.0 mm
of the scalp with an electric hair clipper or scissors as-
necessary, and a close-up photomacrograph was taken of
the outlined area. The film was then sent to The Upjohn
Co for uniform processing. The clipping and photo¬
graphic procedures were repeated at each 4-week clinic
visit. The photomacrographs were used to determine the
number of nonvellus hairs in the outlined area.1213
Secondary Efficacy Measures
At each clinic visit, both the investigator and the patient
subjectively assessed visible new hair growth since week
0 as none, minimal (definite growth but no substantial
covering of the thinning areas), moderate (new growth
partially covering thinning areas but less dense than in
nonthinning areas), or dense (full covering of thinning
areas and hair density similar to that in nonthinning ar¬
eas). A global photograph of the patient's scalp taken at
week 0 was made available for reference. Also at each
visit, the patient was asked to assess the degree of hair
shedding relative to that experienced before the start of
treatment as increased, decreased, or unchanged.
SAFETY EVALUATION
Safety assessments at each clinic visit included a history
of medical events and menstrual cycles; evaluation of
blood pressure, pulse rate, respiration rate, body weight,
and scalp; auscultation of the chest; and examination for
peripheral edema. At the initial screening of the patients
and at week 32 or whenever the patient discontinued the
study, the following evaluations were performed: elec-
trocardiography, serum chemistry studies, complete
blood cell count, platelet estimate, serum ferritin deter¬
mination, urinalysis, and endocrine function tests. A
pregnancy test (repeated throughout the study as
needed), chest roentgenography, and M-mode echocar-
diography were performed at the screening visit. Serum
samples for evaluation of minoxidil concentration were
obtained at screen, week 32, and any time systemic ef¬
fects of minoxidil were suspected.
STATISTICAL METHODS
The primary efficacy measure for this multicenter trial
was nonvellus hair count at week 32. This was examined
in terms of group means and mean change from pre-
treatment values. Treatment differences based on these
means were assessed by analysis of variance techniques
utilizing a full two-factor model with treatment and cen¬
ter as the factors. Categorical analysis techniques were
used to detect treatment differences in the secondary
efficacy measures, ie, investigator and patient
evaluations of hair growth and patient assessment of hair
shedding.
 Table 1. Reasons for Patient Discontinuation Table 2. Patient Demographics at Baseline

No. (%) 2% Minoxidil Placebo

I-1 (n=157) (n=151)
2% Minoxidil Placebo I- 1 1
Reasons (n=157) (n=151) Characteristic* Mean SD Mean SD
Voluntary withdrawal* 18(11.5) 17(11.3) Age, y 33.6 6.67 34.4 6.32
Lack of efficacy 0 (0.0) 0 (0.0) Weight, kg 70.1 18.73 69.8 17.97
Local Irritation 1 (0.1) 1 (0.1) Height, cm 163.3 5.89 163.6 6.73
Pregnancy 2(1.2) 0(0.0) Race, %
Other health problems 6 (3.8) 6 (3.3) White 91 91
Use of excluded medication 0 (0) 1 (0.1) Other 9 9
Total 27(16.6) 25(14.8) Medical/surgical history, %
Cardiopulmonary disease 2.6 0.7
'"Includes patients who voluntarily discontinued medication, moved from Other disease(s)t 17.8 21.2
area, were unavailable for follow-up, were unable to keep appointments, etc.
Prior surgery 66.2 63.6
Regular use of medications 19.8 21.2
RESULTS Menstrual and reproductive
historyi
PATIENT CHARACTERISTICS Age at onset, y 12.5 1.29 12.5 1.36
Cycle Interval, d 28.7 2.55 28.5 2.31
AT BASELINE
Duration, d 5.0 1.17 4.9 1.25
No. of pregnancies 1.8 1.64 1.5 1.47
Patients ranged in age from 17 to 46 years at study entry,
No. of live births 1.3 1.44 1.1 1.21
wi th an overall mean age of 34 years. (One 17-year-old pa¬ 0.4 0.83 0.73
No. of abortions 0.3
tient was enrolled in the study; however, she had turned No. of living children 1.3 1.40 1.2 1.22
18 by the time she was randomized to study medication, Hair loss characteristics
thus satisfying the age requirements of the study.) Com¬ Duration, y 9.5 6.67 9.0 6.68
parability of the treatment groups with respect to age; Age at onset, y 24.1 7.26 25.4 7.14
weight; height; race; medical, surgical, menstrual, and re¬ Degree of thinning, %
productive histories; and hair loss characteristics was as¬ Grade I 48 53
sessed, and the results are summarized in Table 2. No Grade II 52 47
significant difference between the treatment groups was *?>.07 for all characteristics.
detected for any of the characteristics listed. Hair loss sta¬ f Includes patients with gastrointestinal, hepatic, genitourinary, and
tus was evaluated for each patient at study entry and de¬ neurologic diseases.
scribed in terms of duration of hair loss, age at onset of XMenopausal symptoms were reported at study entry by two patients in
the minoxidil group and two in the placebo group.
hair loss, and degree of thinning. None of these loss attrib¬
utes differed significantly between the treatment groups.
The comparability of the treatment groups with re¬
Table 3. Nonvellus Hair Counts
spect to hair thinning status was further examined in terms
of nonvellus hair counts at week 0 (means and standard 2% Minoxidil Placebo
errors are shown in Table 3). Patients randomized to Week ; n r
the 2% minoxidil group averaged 140 nonvellus hairs in of Visit No. Mean SEM No. Mean SEM
the evaluation area (range, 59 to 317 hairs) compared 0 126 140.4 3.89 124 138.6 3.33
with a mean of 139 nonvellus hairs in the placebo group 4 122 144.0 3.92 122 142.4 3.44
8 125 156.7 4.10 124 144.5 3.43
(range, 70 to 262 hairs). These means were not signifi¬ 12 121 4.40 121 147.8
165.9 3.75
cantly different (P=.720), and the patient population over¬ 16 126 4.23 126 147.0 3.55
all had an average of 140 nonvellus hairs in the 1-cm2 166.8
20 120 165.7 4.29 123 150.4 3.49
evaluation area at baseline.
24 125 162.6 4.35 125 148.5 3.52
28 119 167.0 4.58 121 154.3 3.87
EFFICACY
32 128 163.1 4.69 128 148.7 3.55

Mean nonvellus hair counts determined at 4-week

were
intervals for each group (Table 3), and the
treatment
change in count is shown in Figure 1. At the week 8 study. At the final visit,
the mean hair count in the 2%
visit, there was a pronounced difference in mean non¬ minoxidil group (163 hairs) was significantly greater
vellus hair count between the treatment groups, and that (P=.02) than that in the placebo group (149 hairs). The
difference continued throughout the remainder of the patients in the 2% minoxidil group had a mean increase

Downloaded From: http://archderm.jamanetwork.com/ by a Oakland University User on 06/07/2015

 30 n

16 20 24 28 32

Week of Visit

Figure 1. Change in nonvellus hair count from week 0 (mean and standard
error of mean).

of 23 nonvellus hairs in the evaluation area from week 0

compared with a mean increase of 11 hairs for patients
in the placebo group (P=.0004). The 95% confidence in¬
terval for the difference in mean hair count change be¬
tween the treatment groups was 5.9 to 17.5 hairs. The
investigators and patients were asked to provide sepa¬
rate ratings of overall new hair growth since the start of
the study, using the response categories of none, mini¬
mal, moderate, and dense. At week 32, the investigators
reported that 63% of the patients in the 2% minoxidil group
showed minimal or moderate new hair growth (50% mini¬
mum, 13% moderate) compared with 39% of the pa¬
tients in the placebo group (33% minimum, 6% moder¬
ate) (P<.0005). At week 32, 60% of the patients in the
2% minoxidil group reported that they had new hair
growth (40% minimal, 20% moderate) compared with
40% (33% minimal, 7% moderate) of the patients in the
placebo group (P=.002). There were no evaluations of
dense hair growth in either treatment group at the con¬
clusion of the study.
Patients were asked to rate their degree of hair shed¬
ding. A difference between the two treatment groups be¬
gan to emerge after the second month of treatment and
appeared consistent throughout the remainder of the
study. Over half (55%) of the patients in the minoxidil
group reported that they had a decreased shedding com¬
pared with 46% in the placebo group, but this difference
was not significant (P=.169). Figure 2. Top, Clinical photograph of patient at week 0. Bottom, Clinical
photograph of same patient at week 32.
SAFETY
One patient using 2% minoxidil solution discontinued
Topically applied 2% minoxidil solution was safe and well the study because of scalp folliculitis, probably related
tolerated. No clinically significant changes in laboratory to the minoxidil solution.
values or vital signs (blood pressure, pulse rate, respira¬
tion rate, and body weight) were observed during the COMMENT
study. The frequency of abnormalities in electrocardio¬
grams; hématologie, biochemical, and urinalysis find¬ Since the pattern of androgenetic alopecia in the female
ings; and endocrine results was similar for the two groups. is different from that observed in the male, there was a

Downloaded From: http://archderm.jamanetwork.com/ by a Oakland University User on 06/07/2015

 need to develop an objective, unbiased, reproducible meth¬
odology to determine hair counts in women. The pho¬
tographic and computer imaging methods used to de¬
termine nonvellus hair counts in this multicenter trial
yielded accurate and objective counts that were devoid
of center-to-center variations in counting technique. The
methods are reproducible and provide a permanent re¬
cord of treatment effects.
The use of 2% minoxidil solution in this 32-week
study provided some hair regrowth for approximately 60%
of the female patients with androgenetic alopecia. The
mean increase of 23 nonvellus hairs in the 1-cm2 target
site in those treated with topical minoxidil solution was
significantly greater than the response of patients using
the placebo solution.
The increase in the hair count between weeks 0 and
4 is noted as an early response in both treatment groups.
Although the reason for the early effect is not known, it
may be related to scalp care, vehicle effect, or a mechani¬
cal influence from the massage associated with the ap¬
plication of the study medication. The week 32 increase
of 11 hairs in the placebo group indicated that a placebo
effect did occur during this study.
Increased hair growth as it relates to normal scalp
coverage provides a measure for the cosmetic benefit de¬
rived from the treatment. The frontoparietal scalp of
women with no history of hair loss contains a mean of
211 (range, 104 to 318) nonvellus hairs in a 1-cm2 site.13
Using this number (211 ) as a measure of what would be
considered "normal" scalp hair count, one can infer the
percentage of loss of coverage in females with androge¬
netic alopecia. For example, at week 0, the estimated hair
loss in the minoxidil group was 71 hairs (211 140) and
—
the decrease from normal coverage was 34%; after 32 weeks
of treatment, the mean change in hair count was in¬
creased by 23 hairs (163 nonvellus hairs). This change,
equivalent to a 11% increase in scalp coverage and a 32%
regrowth of hairs lost as a result of androgenetic alope¬
cia, represents a desirable cosmetic response derived from
using topical minoxidil solution (Figure 2). The esti¬
mated hair loss and decrease from normal coverage in
the placebo group (71 hairs and 34%) were not unlike
those seen at baseline in the minoxidil group. However,
this group realized only a 4% increase of scalp hair cov¬
erage and a 13% regrowth of lost hairs.
No serious or unexpected medical events were re¬
ported during this study.
Downloaded From: http://archderm.jamanetwork.com/ by a Oakland University User on 06/07/2015
Accepted for publication April 6, 1993.
We would like to thank foan P. Olree for her assis¬
tance in preparing the manuscript.
Principal investigators were Wilma Bergfeld, MD, Cleve¬
land (Ohio) Clinic; Charles N. Ellis, MD, University of
Michigan Medical Center, Ann Arbor; Virginia C. Fiedler,
MD, University of Illinois Medical Center, Chicago; Ruth
Freinkel, MD, Northwestern University, Chicago, 111; Elise
A. Olsen, MD, Duke University Medical Center, Durham,
NC; Vera Price, MD, Kaiser Permanente Medical Center,
San Francisco, Calif; Robert Rietschel, MD, Ochsner Clinic,
New Orleans, La; Janet Roberts, MD, private practice, Port¬
land, Ore; Ronald Savin, MD, Adult and Adolescent Der¬
matology, PC, private practice, New Haven, Conn; Jerome
Shupack, MD, New York (NY) University; David Whit¬
ing, MD, Baylor Hair Research and Treatment Center, Bay¬
lor University, Dallas, Tex.
Reprint requests to The Upjohn Co, 7000 Portage Rd,
Kalamazoo, Ml 49001 (Dr Jacobs).
REFERENCES
1. Bergfeld WF. Etiology and diagnosis of androgenetic alopecia. Clin Dermatol.
1988;6:102-107.
2. Price VH. Androgenetic alopecia and hair growth promotion state of the art:
present and future. Clin Dermatol. 1988;6:218-227.
3. DeVillez RL, Dunn J. Female androgenic alopecia: the 3\g=a\,17\g=b\-androstanediol
glucuronide/sex hormone binding globulin ratio as a possible marker for fe-
male pattern baldness. Arch Dermatol 1986;122:1011-1015.
4. Ludwig E. Classification of the types of androgenetic alopecia (common bald-
ness) occurring in the female sex. Br J Dermatol. 1977;97:247-254.
5. Civatte J, Degreef H, Dockx P, et al. Topical 2% minoxidil solution in male
pattern alopecia: the initial European experience. Int J Dermatol. 1988;27
(suppl):424-429.
6. DeVillez RL. Topical minoxidil for androgenetic alopecia: optimizing the chance
for success by appropriate patient selection. Dermatologica. 1987;175(suppl
2):50-53.
7. Dutr\l=e'\e-MeulenbergROGM, Neiboer C, Koedijk FHJ, Stolz E. Treatment of male
pattern alopecia using topical minoxidil in the Netherlands. Int J Dermatol. 1988;
27(suppl):435-440.
8. Roenigk HH, Pepper E, Kuruvilla S. Topical minoxidil therapy for hereditary
male pattern alopecia. Cutis. 1987;39:337-342.
9. Olsen EA. Treatment of androgenetic alopecia with topical minoxidil. Resid Staff
Physician. 1989;35:53-69.
10. Katz HI. Topical minoxidil: review of efficacy and safety. Cutis. 1989;43:94-98.
11. Kulick MI. Topical minoxidil: its use in treatment of male pattern baldness. Ann
Plast Surg. 1988;21:273-275.
12. Chambers JH. Photographic methods for evaluating hair regrowth. Presented
at the 25th annual meeting of the Drug Information Association; June 20,1989;
Boston, Mass.
13. Jacobs JP, DeVillez RL. Androgenetic alopecia. Arch Dermatol.1990;26:1371-1372.