2.1.

Volunteers

The sunscreen was applied on the flexor forearms of six healthy volunteers (in total one-third males and two-thirds females), aged between 23 and $45\text{years}$ (skin phototypes I to III).⁸ The volunteers stayed in an inner room without sun exposure (room temperature about $21°\mathrm{C}$ ) from half an hour before the examination started until the end of the examination. The ethical approval for these experiments was obtained from the Ethics Committee of the Charité Universitaetsmedizin Berlin, Department of Dermatology, Berlin, Germany. All volunteers gave their written informed consent.

2.2.

Sunscreen Application

The investigated sunscreens were bought directly before the measurements were carried out (commercial samples) or were freshly prepared [COLIPA (European Cosmetics Toiletry and Perfumery Association) emulsions and model formulations], thus guaranteeing that all products were used within their date labeling. Samples with the following SPF values were investigated: 4, 6, 8, 12, 13, 16, 20, 25, 26, 30, 50, 55, and 60.

A precleaning of the skin was realized by rinsing the flexor forearm with cold water and drying it with a paper towel afterward. We applied $2\mathrm{mg}∕{\mathrm{cm}}^2$ of the commercial sunscreens or model formulations to an area of $8\times 10{\mathrm{cm}}^2$ corresponding to the COLIPA standard.³

2.3.

Tape-Stripping

The tape-stripping procedure; as described previously,^{9, 10} started $1\mathrm{h}$ after sunscreen application, transferring the stratum corneum layer by layer to the tape strips together with the UV filters. After pressing the adhesive tapes (tesa film, 5529, Beiersdorf, Hamburg, Germany, width: $19\mathrm{mm}$ ) onto the human skin with a stamp (pressure: $15\mathrm{kP}∕{\mathrm{cm}}^2$ ) the strips were quickly removed. Ten tapes were taken from the treated and untreated skin areas.

2.4.

Spectroscopic Measurements

The absorption spectra were recorded immediately after removal — within $15\mathrm{s}$ — to avoid disturbances by diffusion processes inside the adhesive layer, which result in a homogeneous distribution of the UV filters.^{11, 12}

The spectra of the tape strips together with an empty tape as a reference were recorded in the range $240\text{to}500\mathrm{nm}$ using the UV/VIS spectrometer Lambda 5 (PerkinElmer, Frankfurt/Main, Germany) with an integrating sphere and a rectangular beam diameter of $8\times 10{\mathrm{mm}}^2$ .

The software UV Winlab Version 2.70.01 (PerkinElmer, Frankfurt/Main, Germany) was used to correct the corneocyte-correlated influences and to calculate the sum of the transmission spectra.

2.5.

Sum Transmission Spectra and Average Sum Transmission Values

The sum transmission spectra in the complete UV range (see Fig. 1 in the following section) were calculated by adding the spectra of all tape strips with a detectable amount of UV filters on the basis of the absorbance values. The obtained sum spectrum was subsequently changed to the transmission values.

Fig. 1

Determination of the sum transmission spectrum and the corresponding areas beneath the sum spectrum used to determine the average UVA und UVB sum transmissions (sunscreen sample 7 with SPF 16).

The areas beneath the last curve (sum transmission spectrum) are the basis on which we calculated the average sum transmission values, representing the remaining intensity after sunscreen application in the corresponding spectral ranges:

3.1.

Average Sum Transmission Spectrum

The protocol determining the sum transmission spectra was described previously² and in the previous section. Figure 1 illustrates the development of a sum transmission curve taking into account the individual spectra of the tape strips removed one by one.

The area beneath the sum transmission curve and the calculated average UV sum transmission values reflect the influence of the applied UV filters on the incident UV radiation. In the given example, the average UV sum transmission resulted in 12% transmission, the average UVB sum transmission resulted in 2% transmission, and the average UVA sum transmission resulted in 18% transmission.

The corresponding data determined for all investigated sunscreens are taken to discuss the relationship between the characteristic values of the UVB filter — described by the average UVB sum transmission — and the SPF. To understand the influence of the variable UVB/UVA intensity ratio, the average UVA sum transmission was also taken into account.

3.2.

Relation of the Absorption Behavior in the UVB Range to SPF

In a first step, the originally determined average UVB sum transmission values obtained after tape stripping and spectroscopic measurements are compared with the corresponding SPF values declared by the sunscreen producers (Table 1 ).

Table 1

Summary of the average UVB sum transmission values in relation to the SPF values.

It is obvious that the UVB sum absorption obtained for sunscreens with identical SPF values varies to a high extent.

In a previous publication,¹³ the UVB sum transmission as a measure of the active UVB radiation intensity was correlated to the SPF demonstrating the principal relation of the spectroscopic data and the SPF. The connection of the data was described by an exponential trend line.

Enlarging the number of sunscreens by taking into account formulations with a broader variation of the UVB/UVA ratio, including the “historical” formulations with an extremely low UVA protection, give new insights.

In the graph presented in Fig. 2 , the reciprocal UVB transmissions are used to obtain linear trend lines of these data to the SPF values. In the graph, three trend lines can be recognized with a higher variation of the individual points around the middle rank. The existence of the different trend lines becomes understandable when we take into account the interaction of the UVB and the UVA radiation producing the erythema.

Fig. 2

Relation of the reciprocal values of the average UVB sum transmissions to SPF.

The available spectroscopic data enabled the calculation of the sample specific UVB/UVA relation factors $F_{\mathrm{UVB}∕\mathrm{UVA}}$ dividing the average UVB sum transmission by the average UVA sum transmission.

Two samples belonging to the upper and the lower trend lines are both shown in Fig. 3 to illustrate the differences in the UVB/UVA relation.

Fig. 3

Sum transmission spectra of sunscreens belonging to the upper and the lower trend lines in Fig. 2: line A, sample 7, $\mathrm{SPF}=13$ ; line B, sample 11, $\mathrm{SPF}=20$ .

The calculated factors given in Table 2 quantify the absorption behavior of the investigated sunscreens in the UVB and the UVA range, thus describing the possible influence of photoaugmentation and/or photoaddition. The given mean values of the UVB/UVA relation (row 2 of the table) clearly describe the affiliation of the UVB transmissions to one of the three trend lines with a stronger variation for the samples collected in the middle trend line. This underlines a clear influence of the accompanying radiation intensity in the UVA range on the UVB/SPF correlation. Without an absorption in the UVA range, the UVB filter concentration must be much higher to obtain the same SPF value.

Table 2

Correlation of the position of the reciprocal values of the average UVB sum transmission and the SPF values (Fig. 2) to the UVB/UVA factors.

As well as these clear dependencies, two exceptions exist in the group of the investigated sunscreens. First, a small amount of overlap between the ranges found for the $F_{\mathrm{UVB}∕\mathrm{UVA}}$ factors collected in the upper and the middle trend lines (lines 2 and 3 in the third row of Table 2). Second, the two samples marked by arrows in Fig. 2, both with an SPF of 20, vary considerably in the UVB/UVA factor (upper point $F_{\mathrm{UVB}∕\mathrm{UVA}}=0.13$ , lower point $F_{\mathrm{UVB}∕\mathrm{UVA}}=0.36$ ). These exceptions hint to the influences of parameters affecting the formation of the erythema in addition to the changes in the UVB/UVA relation, e.g., ingredients with antioxidizing or cell-protecting qualities, etc.^{14, 15, 16, 17, 18}.

Generally, the different trend lines reflect, illustratively, the historical development of the filter types used in sunscreens. Originally, sunscreens were restricted to contain only a UVB filter, resulting in the four points on the upper curve measured for historical formulations not now on the market, These historical sunscreens contained high amounts of UVB filter to obtain reasonable SPF values, because of the missing UVA absorption.

Later in sunscreen development, UVA filters were added, resulting in a higher UVB/UVA relation. This situation is reflected by the sunscreens found in the middle trend line. In this group of formulations, a relatively broad variation around the trend line occurs.

The lowest trend lines summarize three highly effective modern sunscreens with a high absorption in the UVA range, lowering the absorption necessary in the UVB range to realize the measured SPF. This underlines the fact that spectroscopic data are suited to estimate, in an orientated manner, the size of the SPF values taking into account the absorption behavior in the UVB and the UVA ranges. The application of this method is of special interest for the investigation of filter substances not yet approved, not using human but porcine skin.^{19, 20}

The discussion concerning the SPF is limited to one biological response to the formation of erythema with an individual dependence on the intensity of the incident radiation. It is to be expected that corresponding investigations considering other described sun-induced injuries will provide additional and possibly quite different insights. Therefore, it is important to characterize the efficacy of sunscreens not by different effects of biological responses in the human organism, but by noninvasive spectroscopic measurements, as described in this paper.