The revised COLIPA in vitro UVA method

D. Moyal*, V. Alard

, C. Bertin

, F. Boyer

, M.W. Brown

, L. Kolbe**, P. Matts

and M. Pissavini

*LOreal Research & Innovation, 25-29 Quai Aulagnier, 92665, Asnie`res Sur Seine, France,

LVMH Recherche, Branche Parfums et Cosmetique,
185, Avenue de Verdun, 45804, St Jean de Baye, France,

Johnson & Johnson Sante Beaute France, 1 Rue Camille Desmoulins, 92787, Issy-les-
Moulineaux, France,

Pierre Fabre Dermo Cosmetique, 2 Rue Viguerie, 31025, Toulouse, France,

Alliance Boots Limited, Nottingham NG2 3AA,
U.K.,
**
Beiersdorf AG, Forschung und Entwicklung, Troplowitzstrae 17, D22529, Hamburg, Germany,

Procter & Gamble, London Innovation
Centre, Egham, Surrey TW20 9NW, UK and

Coty-Lancaster, International Research & Development Center, 2 rue de la Lujerneta, 98000, Monaco
Received 25 June 2012, Accepted 30 July 2012
Keywords: COLIPA, in vitro, protection, UVA
Synopsis
A multicentred study derived from the COLIPA in vitro UVA
method was performed to assess the inuence of test conditions on
UVA protection factor (UVAPF) values in terms of amplitude, repro-
ducibility between laboratories and correlation with in vivo UVA
results. Eight products with a range of in vivo UVAPF from three to
29 were used. Two different types of plates, namely high-roughness
(5 lm) and low-roughness (2 lm) plates, were used with a differ-
ent application rate for each (1.3 mg cm
2
and 0.75 mg cm
2
respectively). The UVR dose applied to both plate types followed the
same principle as the original test (1.2 J. cm
2
9 UVAPF0).
Strong, signicant correlations between in vitro and in vivo UVAPF
values were observed for both plate types (Pearson correla-
tion > 0.9, P 0.01). The correlation and slope obtained with
the low-roughness plates conrmed the previous results obtained
by COLIPA. Across all laboratories, higher UVAPF values were
obtained on the high-roughness plates (P < 0.01). Reproducibility
of UVAPF values between laboratories was comparable between
the two plate roughness values (low roughness, COV = 8%; high
roughness, COV = 12%). Considering the in vitro/in vivo compari-
sons, a regression slope of 0.83 was observed for the low-roughness
plates, in comparison with a value of 1.05 for the high-roughness
plates. The accuracy of the method was improved, therefore, with
the use of the high-roughness plates. With a constraint to recom-
mend the use of only one plate type in the COLIPA UVA in vitro
Test, the high-roughness plate was selected on an on-going basis to
limit variability of results and to provide better accuracy with in
vivo data.
Re sume
Une etude multicentrique, basee sur la methode UVA in vitro du
COLIPA a ete realisee pour evaluer linuence des conditions dessai
sur les valeurs du facteur de protection UVA (UVAPF) en termes
damplitude, de reproductibilite entre les laboratoires et de correla-
tion avec les resultats UVA in vivo. 8 produits couvrant une gamme
de valeurs in vivo UVAPF de 3 a` 29 ont ete utilises. Deux differents
types de plaques, de rugosite elevee (5 m) et de faible rugosite
(2 m) ont ete utilises, avec des taux dapplication differents pour
chacune.(1.3 mg/cm et 0.75 mg/cm, respectivement). La dose
UVR appliquee aux deux types de plaques a suivi le meme principe
que lepreuve initiale (1.2 J/cm x UVAPF0). Des correlations fortes
et signicatives entre les valeurs UVAPF in vitro et in vivo dans les
valeurs UVAPF ont ete observees pour les deux types de plaques
(Correlation Pearson > 0.9, P 0.01). La correlation et la pente
obtenue avec les plaques a` faible rugosite ont conrme les resultats
anterieurs obtenus par le COLIPA. Sur lensemble des laboratoires,
des valeurs UVAPF plus elevees ont ete obtenues sur les plaques de
rugosite elevee (P < 0.01). La reproductibilite des valeurs UVAPF
entre les laboratoires a ete comparable quant aux valeurs de rugo-
site des plaques (faible rugosite, COV = 8%; forte rugosite, COV =
12%). Considerant la comparaison in vitro/in vivo, une pente de
regression de 0.83 a ete observee pour les plaques de rugosite
faible, comparee a` une valeur de 1.05 pour les plaques de rugosite
elevee. La precision de la methode a donc ete amelioree avec lutil-
isation des plaques de rugosite elevee. Avec la contrainte de recom-
mander lutilisation dun seul type de plaque dans la methode de
test UVA in vitro du COLIPA la plaque de forte rugosite a ete selec-
tionnee an de limiter la variabilite des resultats et de fournir une
meilleure concordance avec des donnees in vivo.
Introduction
In recent years, there have been concerted efforts around the globe
to incorporate UVA protection into commercial sunscreens and to
develop methods to measure the photoprotection afforded by these
products. The COLIPA (European Cosmetics Trade Association) in
vitro Sun Protection Methods Task Force was given the remit to
develop an in vitro measure of protection from UVA wavelengths,
correlated with an in vivo measure of the same. The result was the
COLIPA technical guideline published in 2007, Method for the in
vitro determination of UVA protection provided by sunscreen
products [1].
The test is based on the measurement of UV radiation (UVR)
transmittance through a thin lm of sunscreen sample spread on
a UVR-transparent roughened substrate, before and after expo-
sure to a controlled dose of UVR from a dened source of solar-
simulated radiation. By convoluting the ensuing transmission
data with the action spectrum for in vivo Persistent Pigment
Darkening (PPD) and the spectral irradiance received from the
Correspondence: P. Matts, Procter & Gamble, London Innovation
Centre, Egham, Surrey, TW20 9NW, U.K. Tel.: +44 1784 474454; fax:
+44 1784 474508; e-mail: matts.pj@pg.com
2012 Society of Cosmetic Scientists and the Societe Francaise de Cosmetologie 35
International Journal of Cosmetic Science, 2013, 35, 3540 doi: 10.1111/j.1468-2494.2012.00748.x
UVA source, an in vitro UVA protection factor (UVAPF) is pro-
vided, which is correlated with its corresponding in vivo PPD
value [2, 3].
In a previous publication [4], we described and summarized the
methods and results from the two ring studies critical in the devel-
opment of the published COLIPA in vitro UVA Test Method [1].
Eight laboratories tested a total of 13 sunscreens using this method
and established a very good correlation (r
2
= 0.83; slope = 0.84,
P < 0.0001) between resulting in vitro UVAPF values and
corresponding values derived from the in vivo PPD method. We
demonstrate through these data that this new method can be used
to provide a reliable in vitro metric to describe and label UVA ef-
cacy in sunscreen products, in line with the EU Commission recom-
mendation 2006/247/EC [5].
Substrate roughness and the corresponding dose of sunscreen
applied are two critical parameters in determining reproducibility of
results and correlation with in vivo data. In the 2007 COLIPA in
vitro UVA guidelines, a dose rate of 0.75 mg cm
2
and a polym-
ethylmethacrylate (PMMA) plate with a roughness of S
a
= 2 lm
were dened. In continuing optimization of the method, the COLI-
PA in vitro Methods Task Force investigated the use of plates with
a higher roughness and this present article summarizes the results
from a new multicentre ring study comparing the 2007 conditions
[1] and new proposed conditions.
Materials and methods
Materials
UV spectrophotometer
Six test laboratories used a variety of different UV spectrophotome-
ters to conduct the transmission measurements, but it was ensured
that they all conformed to minimum requirements, as outlined in
the COLIPA in vitro UVA Method [1,6,7].
All UV spectrophotometers used, spanned the primary wave-
band of interest, 290400 nm. The wavelength accuracy of the
devices was within 1 nm, as checked using a mercury spectral
standard lamp or a xenon lamp with a specially doped lter.
Device detectors were capable of collecting both the direct and
diffuse portions of UVR transmitted through the roughened
PMMA substrate, either with or without applied sunscreen. The
dynamic range of the device detectors was at least 2.2 absor-
bance units. The maximum measured absorbance was <90% of
the dynamic range of the device used.
Lamp sources used by the spectrophotometers emitted continu-
ous radiation with no peaks within the 290400 nm waveband
and irradiance was low (<0.2 J cm
2
per measurement cycle), so
that photostability of the sunscreen did not become a factor during
spectrophotometric measurement.
Monitoring of the UV spectrophotometer
UV spectrophotometers were tested at regular intervals (on a
monthly basis) by the measurement of reference samples. A twofold
test was recommended:
(a) Monitoring the instruments response and dynamic range with
standard PMMA plates (see Appendix IIA in the published CO-
LIPA in vitro UVA Method [1,6,7]).
(b) Checking wavelength accuracy with an approved standard
material (e.g. holmium perchlorate, as recommended in Appen-
dix IIB of the published COLIPA in vitro UVA Method [1,6,7]).
Source for SSR irradiation
The spectral irradiance of the articial UVR source (at the sample
plane) that was used for irradiation was as similar as possible to
the irradiance at ground level under a standard zenith sun as
dened by COLIPA (2006) [8] or in DIN67501 (1999) [9]. Target
UV irradiance therefore was set within the following acceptance
limits (measured at the sample plane):
Total UV irradiance (290400 nm) 50140 W m
2
Irradiance ratio of UVA (320400 nm) to UVB (290320 nm) 822
All laboratories used an Atlas Suntest
TM
insolator (Atlas Material
Testing Technology GmbH, Linsengericht, Germany), type CPS/CPS
+ or XLS/XLS+, tted with the UV short cut-off lter and IR
dichroic mirror with a temperature during irradiation of samples
maintained below 40C.
Monitoring of the SSR source
The emission of the SSR source was checked by a suitably qualied
expert using a calibrated spectroradiometer for compliance with the
given acceptance limits. On an on-going basis, the SSR source
emission was also monitored using a radiometer. All spectroradi-
ometers and radiometers were calibrated according to COLIPA
recommendations in the Guideline Monitoring of UV light sources
(2007) [10]. The calibration of the radiometers provides a
coefcient of correction between radiometry and spectroradiometry
that helps ensure that all laboratories are applying the same UVR
dose.
Substrates
Each test sunscreen was applied to two types of polymethylmethac-
rylate (PMMA) plate:
(a) The rst type of plate was that described in the COLIPA UVA in
vitro Method published in 2007 [1] and revised in 2009 [6].
The average roughness value (S
a
) of these sand-blasted PMMA
plates (supplied by Schonberg GmbH, Hamburg) was approxi-
mately 2 lm (S
a
dened by EUR 15178 EN [11]) with dimen-
sions of 5 9 5 cm (surface 25 cm
2
).
(b) The second type of plate used in the study (PMMA plates HD6
from Helioscreen, France; dimensions 4.7 cm 9 4.7 cm and
surface area of 22.1 cm
2
) had a moulded surface topography,
providing both lower intra- and inter-batch variation in surface
roughness [12]. The average Ra value was determined as
4.85 lm (referred to as 5-lm plates from hereon).
Reference plate
Reference plates were produced by spreading a few microlitres of
glycerine on the roughened side of the plate, using just enough to
thinly coat the entire plate surface (approximately 15 lL for both
types of plate).
Amount of applied product
An application rate of 0.75 mg cm
2
was used on the 2-lm plates
(total amount 18.75 mg), and an application rate of 1.3 mg cm
2
was used for the 5-lm plates (total amount 28.7 mg). The
1.3 mg cm
2
dose was chosen to ensure a sufcient lm thickness
on the higher roughness 5-lm plates [13].
2012 Society of Cosmetic Scientists and the Societe Francaise de Cosmetologie
International Journal of Cosmetic Science, 35, 3540 36
The Revised COLIPA in vitro UVA method D. Moyal et al.
Sample application
Product was applied to the roughened side of PMMA plates as a
large number of small droplets of approximate equal volume,
distributed evenly over the whole surface of the plate. Positive-
displacement automatic pipettes were found ideal for this purpose.
To check for the correct application rate, pipettes and/or plates
were weighed before and after dispensing the product.
After application and check weighing, the sunscreen product
was spread immediately over the whole surface using light strokes
with a naked ngertip (no nger-cot) pre-saturated with the
product. Spreading was completed in two phases: (a) the product
was rst distributed over the entire plate using light pressure, in
<30 s and (b) the distributed sample was then rubbed into the
rough surface using stronger pressure over a period of 2030 s.
Treated samples were then allowed to equilibrate in the dark, at
ambient temperature, for at least 15 min to help facilitate forma-
tion of a standard stabilized product lm.
Transmission measurements through product-treated plates
The UV transmission (monochromatic absorbance data over 290
400 nm with 1 nm steps) of at least three PMMA plates, treated
with product in the manner described above, was measured using
a calibrated UV spectrophotometer. Depending on the spectropho-
tometer type, each plate was measured either on a large spot area
or at a number of different sites to ensure that an area of at least
2.0 cm
2
was measured in total. Care was taken to ensure that
exactly the same sites on each plate were measured before and
after irradiation, to help reduce variability. Each laboratory applied
each test product to at least three plates of each type.
Exposure to SSR
Care was taken to ensure that samples were not exposed to temper-
atures of >40C during irradiation. In the Atlas Suntest
TM
insola-
tors, for example, air-conditioning units or water-cooled trays were
employed for this purpose. PMMA plates were also xed in place
using suitable means (e.g. the use of a template with wells in the
surface to accommodate the plates) and placed against a non-UV
reective background to minimize reection of UVR back through
the sample.
The irradiation dose Dx used for exposure of the samples was cal-
culated using the same principle for the two different plate types:
Dx = UVAPF
0
9 1.2 J cm
2
Summary of the different procedure steps
After standardized sunscreen lm application on roughened PMMA
plates (Calculations are detailed in Appendix)
(i) transmission measurement with a UV spectrophotometer
(ii) iterative adjustment of the absorbance values by coefcient C
such that in vitro SPF now equals in vivo SPF
(iii) calculation of UVAPF
0
from the corrected spectra
(iv) determination of the irradiation dose D;
D = UVAPF
0
9 1.2 J cm
2
(v) irradiation of the sunscreen samples with dose D
(vi) calculation of UVAPF from the absorption spectra after irradiation
(vii) calculation of critical wavelength value after irradiation
Calculations are detailed in Appendix.
Products tested
Eight products were tested in total, of which seven were marketed
products of different types with labelled SPF values from 6 to 50+.
Mean in vitro UVAPF values for each of these products were deter-
mined on 10 subjects. Measured values ranged from 2.9 to 29.4,
as shown in Table I.
Reference sunscreen S2 (proposed in the COLIPA UVA in vitro
Method issued in 2011 [7]) was also tested. The SPF in vivo value used
for calculations was 16. The in vivo UVAPF acceptance range for the
reference formulation S2 was from 10.7 to 14.7 (mean value, 12.7).
Statistical analysis
Normality and homogeneity of variance were checked using a
ShapiroWilk test. Intra-laboratory comparisons between the two
plate types were performed using a Students t-test (with
signicance set at P < 0.05). For each product, inter-laboratory
comparisons between the two types of plates were performed using
a Students t-test (with signicance set at P < 0.05). The relation-
ship of in vitro and in vivo UVAPF values obtained for the eight
products was modelled using simple linear regression for both plate
types, to yield slope, signicance and Pearson correlation coefcient
metrics [correlation is signicant at the 0.01 level (2-tailed)].
Results
Mean UVAPF
0
, UVAPF and critical wavelength results are summa-
rized in Table II for the 2-lm plate and in Table III for the 5-lm
plate.
Table I Labelled SPF, in vivo UVA protection factor values, mean and 95%
condence intervals on 10 subjects, types of products
Products Labelled SPF In vivo UVAPF Mean (95% CI) Type of product
UVA1 SPF50+ 29.4 (26.532.3) Gel cream
UVA2 SPF50+ 25.4 (20.530.3) Lotion
UVA3 SPF50+ 21.4 (18.024.8) Cream
UVA4 SPF30 11.9 (10.813.0) Fluid lotion
UVA5 SPF6 2.9 (2.43.4) Lotion
UVA6 SPF10 4.9 (4.05.8) Oil
UVA7 SPF30 18.5 (15.521.5) Lotion
S2 SPF16 12.7 (10.714.7)
*
Lotion
*Acceptance range of in vivo UVAPF values for S2 proposed reference formulation.
Table II UVAPF
0
, UVAPF and critical wavelength results on the 2-lm
plates, mean values obtained across six laboratories, standard deviation (SD)
and % variance between laboratories
Product
UVAPF
0
UVAPF kc
Mean SD % Var Mean SD % Var Mean SD % Var
UVA1 31.8 3.3 10.5 24.8 3.2 12.8 378 1.3 0.3
UVA2 23.2 1.8 7.9 19.9 2.1 10.6 376 1.3 0.3
UVA3 22.3 2.1 9.4 18.1 1.1 6.1 377 1.3 0.3
UVA4 15.8 1.2 7.7 13.6 1.3 9.2 378 1.3 0.3
UVA5 3.8 0.2 5.1 3.5 0.2 4.6 376 1.2 0.3
UVA6 4.4 0.3 6.6 4.3 0.3 7.7 373 1.5 0.4
UVA7 18.4 1.6 8.9 13.5 1.4 10.7 379 1 0.3
S2 14.2 1 7.3 11.5 0.9 7.9 378 2 0.5
2012 Society of Cosmetic Scientists and the Societe Francaise de Cosmetologie
International Journal of Cosmetic Science, 35, 3540 37
The Revised COLIPA in vitro UVA method D. Moyal et al.
The mean UVAPF values obtained for the reference formulation
S2 using the 2-lm or the 5-lm plates were both in the acceptance
range, based on in vivo UVAPF values.
Both before and after irradiation, mean UVAPF
0
values across
all laboratories for the 2-lm plates were lower than those for the
5-lm plates, whatever the product (P < 0.01).
The percentage of variability in data between laboratories, before
and after irradiation, across the eight products was approximately
8% using the 2-lm plates and approximately 12% using the 5-lm
plates.
A summary of in vitro UVAPF values obtained using the 2-
lm and the 5-lm plates and corresponding in vivo UVAPF val-
ues is presented in Fig. 1, and a plot in vitro values against cor-
responding in vivo values, along with tted linear regression
curves, are presented in Fig. 2. The correlation between in vitro
UVAPF values and in vivo UVAPF values was highly signicant
(P < 0.01) for both plate plates (Pearson r
2
correlation values of
0.98 for the 2-lm plates and 0.98 for the 5-lm plates). Interest-
ingly, the value of the slope of the regression curve was 0.83
for the 2-lm plates and 1.05 for the 5-lm plates (indicating a
closer agreement with absolute in vivo UVAPF values for the 5-
lm plates).
After irradiation, critical wavelength values on the 2-lm plates
were lower compared with values obtained using the 5-lm plates
across all laboratories, whatever the product (P < 0.01). Variability
of results between laboratories was very low for either plate type
(<1.0%).
Discussion
Different types of sunscreen formulation, including creams, lotions,
uids and oils with in vivo UVAPF values from 3 to 29, were tested
using two different types of plates and application rate. Using 5-lm
roughness plates and a 1.3 mg cm
2
dose rate, UVAPF
0
values
were higher than those obtained using the 2-lm plates and a
0.75 mg cm
2
dose rate which, we hypothesize, demonstrates the
signicant inuence of these different test conditions on the shape
of the absorbance curves. The UVR dose applied in the irradiation
step (using D = 1.2 J cm
2
9 UVAPF
0
) to the 5-lm plates was on
average 15% higher than that applied to the 2-lm plates. Because
of the relative photostability of the formulations tested, however,
measured UVAPF values for the 5-lm plates were higher than
those measured for the 2-lm plates.
The variability of the results between laboratories was higher for
the 5-lm plates compared with the 2-lm plates (12% vs. 8%
respectively), possibly due to less familiarity with the new plate
type by technicians (a signicant factor to consider in in vitro test-
ing). Notwithstanding, the slightly higher variance seen with the
5-lm plates is still lower than that achieved with the in vivo
method and, we believe, acceptable. These results agree with previ-
ous observations made by Ferrero et al. [13].
The slope and correlation (0.83; P < 0.01) obtained with the
low-roughness plates were in good agreement with previous results
obtained for this plate type by COLIPA [4].
Conclusion
Based on the results obtained, the COLIPA in vitro Task Force has
recommended using only one type of plate to limit possible variabil-
ity. The group has selected the higher roughness (5 lm) because of
further improved accuracy against the in vivo method. Based on
these data, therefore, COLIPA published revised guidelines for the
method for in vitro determination of UVA protection in April 2011
[7], specifying new specications for plates roughness (5 lm) and
application rate (1.3 mg cm
2
).
1
6
11
16
21
26
31
36
UVA1 UVA2 UVA3 UVA4 UVA5 UVA6 UVA7 S2
U
V
A
P
F
Figure 1 Mean in vitro UVA protection factor (UVAPF) values obtained on
the low ( )- and high ( )-roughness plates compared with corresponding
mean in vivo UVAPF values ( ).
y = 0.8358x
y = 1.0519x
1.0
6.0
11.0
16.0
21.0
26.0
31.0
36.0
1 6 11 16 21 26 31
I
n

v
i
t
r
o

U
V
A
P
F
In vivo UVAPF
Figure 2 Correlation between mean in vitro UVA protection factor (UVAPF)
values for the high ( )- and low ( )-roughness plates and corresponding
mean in vivo UVAPF values.
Table III UVAPF
0
, UVAPF and critical wavelength results on the 5-lm
plates, mean values obtained across six laboratories, standard deviation (SD)
and % variance between laboratories
Product
UVAPF
0
UVAPF kc
Mean SD % Var Mean SD % Var Mean SD % Var
UVA1 34 5.1 15.1 32.1 4.5 14.1 379 1.6 0.4
UVA2 26.1 3.6 13.7 23.5 1.8 7.8 377 0.9 0.2
UVA3 28.4 5.7 20.2 22.2 4 18 379 1.6 0.4
UVA4 17.5 3.2 18.3 15.8 2.7 16.9 379 1.8 0.5
UVA5 4.3 0.2 4.8 3.9 0.3 7.5 378 1 0.3
UVA6 6.8 0.7 10.7 6.6 0.8 12.5 378 1 0.3
UVA7 21.8 2.7 12.3 19.1 2.6 13.8 380 2 0.4
S2 15.2 0.7 4.3 14.1 1.2 8.1 381 1 0.3
2012 Society of Cosmetic Scientists and the Societe Francaise de Cosmetologie
International Journal of Cosmetic Science, 35, 3540 38
The Revised COLIPA in vitro UVA method D. Moyal et al.
Acknowledgement
The authors acknowledge their companies for funding their partici-
pation in the studies reported herein.
References
1. Method for the in vitro determination of
UVA protection provided by sunscreen
products, COLIPA Guideline (2007).
2. Measurement standards for protection efcacy,
Japan Cosmetic Industry Association Guideline
(1996).
3. Moyal, D., Chardon, A. and Kollias, N. UVA
protection efcacy of sunscreens can be deter-
mined by the persistent pigment darkening
(PPD) method (Part 2). Photodermatol. Photo-
immunol. Photomed., 16, 250255 (2000).
4. Matts, P.J., Alard, V., Brown, M.W. et al. The
Colipa in vitro UVA method: a standard and
reproducible measure of sunscreen UVA pro-
tection. Int. J. Cosmet. Sci. 32, 3546 (2010).
5. European Commission, Enterprise and
Industry Directorate-General. Commission
recommendation of 22 September 2006 on
the efcacy of sunscreen products and the
claims made relating thereto, 2006/647/EC.
J. Eur. Union L265, 3943 (2007).
6. In vitro Method for the determination of the
UVA protection factor and critical wave-
length values of sunscreen products, COLI-
PA Guideline (2009).
7. In vitro Method for the determination of the
UVA protection factor and critical wave-
length values of sunscreen products, COLI-
PA Guideline (2011).
8. International Sun Protection Factor (SPF)
Test Method, COLIPA Guideline (2006).
9. DIN 67501: Experimental evaluation of the
protection from erythema of external sun-
screen products for the human skin, The
Lighting Technology Standards Committee
(FNL), DIN Deutsches Institut fu r Normung
e. V. (1999).
10. Guidelines for monitoring UV radiation
sources, COLIPA Guideline (2007).
11. Stout, K.J., Sullivan, P.J., Dong, W.P.,
Mainsah, E., Lou, N., Mathia, T. and Zahou-
ani, H. The Development of Methods for The
Characterisation of Roughness in Three
Dimensions, Report EUR 15178 EN. EC
Brussels (1993).
12. Pissavini, M., Marguerie, S., Dehais, A.,
Ferrero, L. and Zastrow, L. Characterizing
Roughness: a new substrate to measure
SPF. Cosmet. Toilet. 124, 5664 (2009).
13. Ferrero, L., Pissavini, M., Dehais, A.,
Marguerie, S. and Zastrow, L. Importance of
substrate roughness for In vitro Sun protec-
tion measurement. IFSCC Mag. 9, 97108
(2006).
Appendix
Equations used in the method in this article:
Calculation of the in vitro SPF
in vitro
SPF
in vitro

R
k 400nm
k 290nm
Ek Ik dk
R
k 400nm
k 290nm
Ek Ik 10
A0k
dk
1
where: E(k) = Erythema action spectrum (CIE 1987 [17]); I
(k) = Spectral irradiance received from the UV source; A
0
(k) = Mean
monochromatic absorbance of the test product layer before UV expo-
sure; Dk = Wavelength step (1 nm).
Calculation of the adjusted SPF
in vitro,adj
and
determination of the coefcient of adjustment C
C is the coefcient of adjustment, iteratively determined to adjust
the calculated in vitro SPF value to the labelled (in vivo) SPF value.
It is recommended that C ranges between 0.8 and 1.2.
SPF
in vitro;adj
SPF label
R
k 400nm
k 290nm
Ek Ik dk
R
k 400nm
k 290nm
Ek Ik 10
A0kC
dk
2
where E(k), I(k), A
0
(k) and dk are dened in eqn (1).
Calculation of UVAPF
0
UVAPF
0
is calculated for each plate individually.
UVAPF0
R
k 400nm
k 320nm
Pk Ik dk
R
k 400nm
k 320nm
Pk Ik 10
A0kC
dk
3
Where P(k) = PPD action spectrum; I(k) = Spectral irradiance
received from the UV source (UVA 320400 nm for PPD testing);
A
0
(k) = Mean monochromatic absorbance of the test product layer
before UV exposure; C = Coefcient of adjustment previously deter-
mined in eqn (2); dk = Wavelength step (1 nm).
Calculation of UVAPF of plates after UV irradiation of
the sample
UVAPFi
R
k 400nm
k 320nm
Pk Ik dk
R
k 400nm
k 320nm
Pk Ik 10
AkC
dk
4
where P(k), I(k), C and dk are dened in eqn (3). A(k) is the mean
monochromatic absorbance of the test product layer after UV expo-
sure.
Calculation of critical wavelength (kc) of plates after
UV irradiation of the sample
The critical wavelength k
c
value for the test product is dened as
that wavelength where the area under the absorbance spectrum
for the irradiated product (obtained using the method described
above) from 290 nm to k
c
is 90% of the integral of the absorbance
spectrum from 290 to 400 nm, and is calculated in the following
way:
2012 Society of Cosmetic Scientists and the Societe Francaise de Cosmetologie
International Journal of Cosmetic Science, 35, 3540 39
The Revised COLIPA in vitro UVA method D. Moyal et al.
A series of absorbance values (dependent on the wavelength
increment) are calculated for each of the three separate plates to
which the test product has been applied. Absorbance at each wave-
length increment (A
k
) is calculated thus:
A
k
logC
k
=P
k
5
where
C
k

p
nc
k
1 c
k
2 . . . c
k
n
P
k

p
np
k
1 p
k
2 . . . p
k
n
6
c
k
[n] = Arithmetical mean of transmission measurements taken
at measurement point n and at wavelength k for the reference sam-
ple (glycerine-treated roughened PMMA plate)
p
k
[n] = Arithmetical mean of transmission measurements taken
at measurement point n and at wavelength k for irradiated, sun-
screen product-treated sample (roughened PMMA plate)
Critical wavelength k
c
is calculated for each irradiated plate as
follows:
Z
kc
290
A
k
d
k
0:9
Z
400
290
A
k
d
k
7
The nal critical wavelength value for each tested sunscreen prod-
uct is the mean of values recorded for each measured, irradiated,
product-treated PMMA plate.
2012 Society of Cosmetic Scientists and the Societe Francaise de Cosmetologie
International Journal of Cosmetic Science, 35, 3540 40
The Revised COLIPA in vitro UVA method D. Moyal et al.