Skin Research and Technology 2009; 15: 218–223
Printed in Singapore  All rights reserved
doi: 10.1111/j.1600-0846.2009.00358.x
The evaluation of whitening efficacy of cosmetic products
using a human skin pigmentation spot model
Y. Tian1,2, T. Hoshino3, C. J. Chen2, Y. E4, S. Yabe3 and W. Liu1,2
1
China Medical University, Shenyang, China, 2Department of Dermatology, The General Hospital of Air Force, Beijing, China,
3
KOSÉ Corporation, Tokyo, Japan and 4VHBIOPHARMCOM, Shanghai, China
Background: To establish a pigmentation spot model on
human skin and to assess whitening efficacy for whitening
products by this established pigmentation spot model.
Methods: Twenty subjects between 20 and 45 years old
with skin phototype III or IV were selected. Three consecu-
tive daily UV exposures were performed on buttocks of
the subjects as follows: Day 1 5 1 minimal erythema dose
(MED), Day 2 5 0.5 MED and Day 3 5 0.5 MED. After
the first UV exposure, a selected whitening product was
applied to the subjects twice a day on UV exposure area.
The application of the whitening product to subjects on the
exposed areas was continued till Day 27. CM2500d chro-
mameter, Maxmeter MX18 and visual evaluation were used
to assess changes of skin color.
Results: A pigmentation spot model after UV exposure was
established. The measurement of the pigment spot showed
that L value declined abruptly at Day 3 and then slowly
reached to a lowest point at Day 6. L value of the pigment
spot almost remained at the same level until Day 20, there-
after increased slowly. The a value showed an abrupt
increase at Day 3 and slowly reached to a maximal level at
Day 6. The a value slowly declined toward its baseline level.
Likewise, the erythema index also increased significantly at
Day 3, and reached to a maximal level at Day 6 and then
slowly declined. However, L , a and erythema indices did
not return to their baseline levels during the 27-day period of
this study. On the other hand, b value started to increase
from Day 3 and such increase was observed continuously to
Day 27. Melanin index also showed a slow increase during
the first 3 days. It started to increase rapidly from Day 3 and
A for a fair skin is a common trend
PREFERENCE
among women in Asian countries. In China,
sales of whitening cosmetics are increasing, while
more and more patients with skin color disorders
demand treatments. Currently, whitening efficiency
evaluation for spot-removing and whitening pro-
ducts remains a difficult task. Good pigmentation
models are considered the basis for such kind of
whitening efficiency evaluation. Previous studies
showed that a single high-dosage UV exposure,
218
r 2009 John Wiley & Sons A/S
Skin Research and Technology
a to maximal level at Day 9 and maintain at a plateau till Day
27 (with an exception at Day 13). To assess the whitening
product by this pigmentation spot model, DL, Db , and DM
values were analyzed. It showed that absolute DL value and
Db value of whitening products were lower than those
values of the vehicle of the whitening product at each
checkpoint, while DM value of the whitening product was
lower only at Day 9 and Day 20, although no statistically
significant differences was found. The visual results also
strongly supported that the whitening product enhanced the
decrease of pigmentation.
Conclusion: This study showed that repeated UV exposure
was able to induce a long extensive period of pigment
formation. The resulted pigmentation spot was able to
maintain at an elevated level till Day 20. Clinical subjective
evaluations together with combined objective instrument
measurements were still important to assess whitening
and spot-removing ability of a material due to the instrument
limitation for color differentiations. This kind of pigmentation
spot model can be used to assess whitening efficacy
for whitening or spot-removing products. In addition, the
combinations of subjective and objective methods were able
to serve as advisable references to assess the whitening
efficacy of products.
Key words: whitening efficacy – pigmentation – human skin
– melanin index – erythema index – L a b
& 2009 John Wiley & Sons A/S
Accepted for publication 10 December 2008
usually at 2 minimal erythema dose (MEDs),
was irradiated on skin (1). In order to record and
assess changes of skin color, several methods (2–4)
were applied recently, such as visual evaluation,
reflectance sprectrophotometry, reflectance colori-
metry following the Commission Internationale de
l’Eclariage recommendations, and photograph.
Each of the above methods has its own limitation
and the result may only lead to a single inter-
pretation that may be somewhat arbitrarily.
 The evaluation of whitening efficacy of cosmetic products
The main purpose of our study was to establish
an in vivo pigmentation spot model with expo-
sure to repeated small dosage of UV radiation
and to evaluate a whitening product by this
pigmentation spot model. Visual evaluation and
two kinds of instrument measurement were in-
corporated to record the changes of skin color.
Material and Methods
Study subjects
Twenty Chinese females between ages of 20 and
45 with self-described skin phototype III or IV
were selected. Informed consents as well as
complete medical histories were obtained. None
of the subjects had any history of drug hyper-
sensitivity or abnormal reaction to sunlight. They
had not taken any medication for 4 weeks and
had avoided sun exposure on their buttocks for
3 months before the study. The subjects were
required and needed to avoid medical treatments
and no sun exposure on their buttocks during the
study too.
Test material
A kind of whitening product with 3% arbutin
was assessed in this study and a vehicle with-
out whitening-active ingredient was used as a
control.
Establishment of a human skin pigmentation spot
model
Radiation source and dosimetry
Multiport solar simulators GS-2004 (Aohua Co.,
Beijing, China) were used in the measurements.
Fig. 1. Relative transmission of the filters in the solar simulator. Schott WG 320/1 mm and UG 11/1 mm for minimal erythema dose measurement.
This solar simulator included a 450 W xenon
lamp and a dichroic mirror. Radiation was fil-
tered to deliver a continuous UV spectrum from
290 to 400 nm [solar-simulated radiation (SSR)]
by a WG 320 (Schott, Clichy, France), 1-mm-thick
short cut-off filter plus a UG 11, 1-mm-thick long
cut-off filter (Schott) (Fig. 1). The irradiances at
skin level were monitored with a calibrated
SUN5 digital spectroradiometer (National Insti-
tute of Measuring, Beijing, China).
UV irradiations
UV irradiations were performed using two con-
secutive processes. First, MED of individual sub-
jects was determined. On right side of buttock
skin, six series of 0.8 cm diameter round areas
were irradiated with UV doses ranging from
198.25 to 605.02 mJ/cm2 with increments of
15%. The MED was assessed 24 h after the ex-
posure and irradiation scheme on left side of
buttock was started. The irradiation scheme con-
sists of the following doses: Day 1 5 1 MED,
Day 2 5 0.5 MED, Day 3 5 0.5 MED. Three test
areas were involved: one was irradiation only
without any product application, and the other
two were applied with whitening product and
vehicle (twice a day) after each UV radiation,
respectively.
Evaluation methods
Objective evaluation: Two kinds of instruments,
CM2500d chromameter (Minolta Camera Co.,
Tokyo, Japan) and Mexameter MX18 (C-K Elec-
tronic, Cologne, Germany) were used to record
skin color with indices of L a b, melanin (M) and
219
Tian et al.

erythema (E). The values of DLdayðnÞ , DMdayðnÞ and

DbdayðnÞ (DLdayðnÞ ¼ jLdayðnÞ  Ldayð0Þ j, DMdayðnÞ ¼
MdayðnÞ  Mdayð0Þ and DbdayðnÞ ¼ bdayðnÞ  bdayð0Þ )
were counted to assess changes of skin color at
different days. Measurements were made pre- and
post-UV radiations at successive time intervals of 3,
6, 9, 13, 20, and 27 Days. Three consecutive read-
ings were taken at each site and their mean values
were represented as study data. Study instruments
were calibrated before each measurement.
Subjective evaluation: Color assessment was
based on comparison between the whitening
product and vehicle to the sites of UV irradiation
only. Visual evaluation started from Day 13 be-
cause the earlier erythemous reation after UV
exposure made such evaluation difficult for the
changes of pigmentation. The scores were given
as 3, 2, 1, 0,  1,  2,  3 in representative to
remarkably effective, effective, somewhat effec-
tive, ineffective, somewhat reverse, reverse, and
remarkably reverse, respectively.

Statistical analysis
Software SPSS10.0 was used in the statistic calcu-
lation. t-test was adopted for comparison to the
differences among the whitening product, vehi-
cle, and UV irradiation only site. The pigmenta-
tion differences at different days were performed
by paired t-test. A P-value o0.05 was considered
significant.
Result
Establishment of a human skin pigmentation spot
model
After three consecutive exposures, erythema with
light swelling was observed, but no strong prur-
itus and soreness were complained by the study
subjects. Desquamation was not found in the test,
of which might be difficult to assess the changes
of pigmentation.
The irradiation only sites of each subject repre-
sented the formation of pigmentation without
disturbances of test products. After exposures,
L values (Fig. 2a) declined abruptly at Day 3 and
then slowly reached to lowest point at Day 6. It
then almost remained at the same level compared
with the lowest point (P40.05). From Day 20, the
L value increased significantly (Po0.05) com-
pared with the lowest point at Day 6, but had not
returned to its baseline completely at Day 27
(Po0.05).
Fig. 2. L (a), a (b), and b (c) values of the pigmentation spot model.
The asterisk indicates a statistically significant difference between the
baseline values and the measured values. And the double asterisk
indicates a statistically significant difference between the lowest point
and Day 27.
a values (Fig. 2b) showed an abrupt increase at
Day 3 and slowly reached to a maximal level at
Day 6. Then, a values slowly declined toward
the original level but not returned to its baseline
completely at Day 27 (Po0.05).
b values (Fig. 2c) started to incline gradually at
Day 3 and increased continuously to Day 27.
Compared with baseline, b values showed sig-
nificant increase from Day 13 (Po0.05).
Melanin index (Fig. 3a) also showed a slow
increase during the first 3 days. Then it started
rapidly from Day 3. Compared with the baseline,
melanin index showed significant increase to
maximal level at Day 9 (Po0.05) and maintain
at a plateau till Day 27 (with an exception at
Day 13).
Erythema index (Fig. 3b) increased signifi-
cantly at Day 3, and reached to maximal level at
Day 6 followed a slow decline (with an exception
at Day 13) toward the baseline till Day 27
(Po0.05).
Evaluation of the whitening product
Objective evaluation: To assess whitening efficacy
of the product, DL, Db , and DM values were used

220
 The evaluation of whitening efficacy of cosmetic products

Fig. 3. Melanin index (a) and erythema index (b) of the pigmentation
spot model. The asterisk indicates a statistically significant difference
between the baseline values and the measured values.

to record the changes of pigmentation. The abso-

lute DL values (Fig. 4a) of the whitening product
were lower than the DL values of vehicle at each
checkpoint from Day 3 to Day 27, but it did not
reach to a statistically significant level (P40.05).
Db values (Fig. 4b) of the whitening product
was lower than Db values of vehicle at each
checkpoint, although no statistically significant
difference was observed.
DM values (Fig. 4c) of the whitening product
were lower compared with vehicle only at Day 9
and Day 20, no significant differences were
observed.
Visual evaluation: (Fig. 5) Although the two
kinds of instruments of chromameter (Minolta)
and Mexameter MX18 (C-K Electronic) showed
no statistical significant differences on skin color
changes, our visual evaluation appeared to ob-
serve that the scores of the whitening product
were higher than the scores of vehicle at Day 13,
Day 20, and Day 27.
Discussion
There are many methods to assess the efficiency
of whitening or spot-removing active materials.
The inhibitory efficacy of tyrosinase, which is
known to be a key enzyme for melanin biosynth-
esis, is usually assessed in vitro (5, 6). But the
actual efficacy of whitening or spot-removing
products needs to be assessed in vivo with
Fig. 4. Evaluation of the whitening product by objective instrumental
measurement: DL (a), Db (b), and DM (c).
Fig. 5. Evaluation of the whitening product by visual evaluation.
specific and sensitive indices. In other studies, a
single high dose of UV radiation, usually at 2 or
2.5 MEDs, was exposed to skin (1). A single high
dose UV radiation may induce strong erythemal
response. When such erythema disappears, des-
quamations may occur. This desquamation phe-
nomenon will make it difficult to evaluate the
changes of skin color by subjective or objective
methods. Also it was said (7) that L value
reached to its lowest point at Day 7 after a single
exposure of 2 MEDs, and then L value started to
return to the baseline level. In our study, the L
value reached to the lowest point at Day 6 and
almost maintained at the same level until Day 20.
Meanwhile, melanin index showed a trend of

221
Tian et al.
continuous increase, while b value increased
from Day 3 until the end of the test at Day 27.
As a result, repeated exposure of low-dose UV
radiation could induce pigmentation formation
with longer time period compared with single UV
exposure. And the extensive length of pigmenta-
tion formation is helpful to evaluate whitening
efficacy.
In this study, L value, decreased abruptly at
Day 3, and kept decreasing steadily to the lowest
point at Day 6. Park et al. (7) explained this
phenomenon as persistent pigmentation dose
(PPD) by UVA component within the radiation
source. However, the abrupt decrease in L value,
of which the definition is the measure of the
lightness of an object regardless to its chromacity,
should be considered as an integrative response
to the erythematous reaction, but not merely a
change in melanin pigment. This phenomenon
can be explained simply by the actual reaction
observed during this period that it is not easy for
the naked eyes to differentiate the pigmentation
accurately due to the prominent erythematous
changes on skin. This was also supported by the
changes of a , b , M and E indices. Both a value,
the degree of erythema, and E index, represent-
ing the presence of hemoglobin (8–12), abruptly
increased at Day 3 and reached a maximal at
Day 6. It showed that the two values were
positively correlated. Both melanin index and b
value persistently increased from Day 3, support-
ing the two values are positively correlated.
However, S. H. Lim (13) thought that melanin
index correlated with L value not b value. This
might be due to the differences of UV exposure
methods.
Based on our pigmentation spot model, a
whitening product with 3% arbutin was assessed.
Arbutin is a commonly used whitening ingredi-
ent, which serves as a control in whitening study.
Many in vitro studies (14, 15) showed that arbutin
had inhibition effect on tyrosinase activity in a
dose-dependent manner. Cytotoxicity of arbutin
also occurred with the increase of concentration.
Considering to arbutin cytotoxicity and skin
absorption issues, the actual arbutin concentra-
tion of 3% used in our study might limit its
whitening effect in vivo compared with the vitro
study results. As a result, the whitening product
with 3% arbutin in our study showed no statis-
tical significant differences of DL, Db , and DM,
but showed the trends of whitening effect. We
believe that we need to combine both instrument
222
measurements and visual evaluations in actual
in vivo study. This is because the instruments may
have their limit to record skin color changes. For
an establishment of a model system, the instru-
ment measurement results should always meet
clinical results to avoid bias and incorrect con-
clusions, so the clinical visual evaluation is im-
portant too. In this study, visual evaluation
results showed whitening effect of the whitening
product. Combined with the subjective results,
we believe that the whitening product with 3%
arbutin has a certain level of whitening effect,
eventhough no statistical significance was ob-
tained due to the reasons discussed above. It
becomes important to select other whitening
reagents to further evaluate this pigmentation
spot model.
This study was valuable in the establishment of
pigmentation spot model with extensive time
period for pigment formation. Also Mexameter
(C-K Electronic) and chromameter (Minolta)
were compared with some observed correlations.
Acknowledgement
This work was supported by KOSÉ Corporation
in 2007.
References
1. Takiwaki H, Shirai S, Kohno H, Soh H, Arase S. The
degree of UVB-induced erythema and pigmentation
correlate linearly and are reduced in a parallel manner
by topical anti-inflammatory agents. J Invest Dermatol
1994; 103: 642–646.
2. Piérard GE. EEMCO guidance for the assessment of skin
colour. J Eur Acad Dermatol Venereol 1998; 10: 1–11.
3. Park BS, Youn JIL. Topographic measurement of skin
color by narrow-band reflectance spectrophotometer
and minimal erythema dose(MED) in Koreans. Skin
Res Technol 1998; 4: 14–17.
4. Kawada A, Kametama H, Asai M et al. A new approach
to the evaluation of whitening effect of a cosmetic using
computer analysis of video-captured image. J Dermatol
Sci 2002; 29: 10–18.
5. You-Jung KIM. Antimelanogenic and antioxidant prop-
erties of gallic acid. Biol Pharm Bull 2007; 30: 1052–1055.
6. Szu-Hsiu LIU, I-Horng PAN, I-Ming CHU. Inhibitory
effect of p-hydroxybenzyl alcohol on tyrosinase activity
and melanogenesis. Biol Pharm Bull 2007; 30: 1135–1139.
7. Park SB, Suh DH, Youn JI. A long-term time course of
colorimetric evaluation of ultraviolet light-induced skin
reactions. Clin Exp Dermatol 1999; 24: 315–320.
8. Seitz JC, Whitmore CG. Measurement of erythema and
tanning responses in human skin using a tri-stimulus
colorimeter. Dermatologica 1988; 177: 70–75.
9. Takiwaki H, Overgaard L, Serup J. Comparison of nar-
row-band reflectance spectrophotometric and tristimulus
 The evaluation of whitening efficacy of cosmetic products

colorimetric measurements of skin color. Skin Pharma- 14. Sugimoto K, Nishimuo T, Nomura K e t al. Inhibitory
col 1994; 7: 217–225. effects of alpha-arbutin on melanin synthesis in cultured
10. Fullerton A, Fischer T, Lahti A et al. Guidelines for human melanomn cells and a three-dimensional human
measurement of skin colour and erythema. Contact skin model. Biol Pharm Bull 2004; 27: 510–514.
Dermatitis 1996; 31: 1–10. 15. Chakraborty AK, Funasaka Y, Komoto M, Ichihashi M.
11. Liu Wei MD, Wang Xuemin MD, Lai Wei MD et al. Skin Effect of arbutin on melanogenic proteins in human
color measurement in Chinese female population analy- melanocytes. Pigment Cell Res 1998; 11: 206–212.
sis of 407 cases from 4 major cities of China. Int
J Dermatol 2007; 46: 835–839. Address:
12. Liun W, Lain W, Wang XM et al. Skin phototyping in a Liu Wei
chinese female population: analysis of four hundred and Department of Dermatology
four cases from four major cities of China. Photoderma- The General Hospital of Air Force
tol Photoimmunol Photomed. 2006; 22: 184–188. PLA No. 30, Fucheng Road
13. Lim SH, Kim SM, Lee YW et al. Change of biophysical Beijing 100036
properties of the skin caused by ultraviolet radiation- China
induced photodamage in Koreans. Skin Res Technol Tel/Fax:186 10 6848 5082
2008; 14: 93–102. e-mail: lwei5811@yahoo.com.cn

223