Archives of Dermatological Research

, Volume 299, Issue 1, pp 25–32

Pigmentation after single and multiple UV-exposures depending on UV-spectrum

Authors

    • Department of Dermatology, D92, Bispebjerg HospitalUniversity of Copenhagen
  • H. C. Wulf
    • Department of Dermatology, D92, Bispebjerg HospitalUniversity of Copenhagen
Original Paper

DOI: 10.1007/s00403-006-0728-3

Cite this article as:
Ravnbak, M.H. & Wulf, H.C. Arch Dermatol Res (2007) 299: 25. doi:10.1007/s00403-006-0728-3
  • 119 Views

Abstract

Minimal pigmentation dose (MMD) after a single UV-exposure is well investigated. Whereas only few studies have established MMD after multiple UV-exposures and mainly in fair-skinned persons. The purpose of this study was to establish MMD 1 week after, respectively, one and five UV-exposures in volunteers with a large variation in constitutive pigmentation. A total of 52 volunteers (skin Types II–V) had skin pigmentation quantified by reflectance spectroscopy. They were UV-exposed on the back for 1 and 5 days using a Solar Simulator, narrowband UVB, broadband UVA and UVA1. For all sources a higher dose was needed the more pigmented the skin, except for UVA1. After one UV-exposure, we found a significant positive linear correlation between UV-dose to one MMD, skin type and pre-exposure skin pigmentation. After five UV-exposures the positive linear correlation between UV-dose and MMD and skin type was only significant for narrow band UVB, pre-exposure skin pigmentation was significant also for Solar Simulator. For UVA and particularly UVA1 the MMD was independent of pre-exposure pigmentation. The number of SED to MMD is therefore almost the same for very fair-skinned and dark-skinned persons. Pre-exposure pigmentation was clearly more predictive of MMD than skin type. 50% of MMD equals a pigmentation increase of 1%. The shorter the wavelengths the higher the SED to produce MMD. Solar was the least melanogenic and UVA1 the most melanogenic. For the UVB-sources a higher dose was needed the more pigmented the skin. For UVA the MMD was independent of pre-exposure pigmentation.

Keywords

Single and multiple UV-exposuresPigmentation

Abbreviations

bUVA

Broadband UVA

MED

Minimal erythema dose, the UV dose to elicit just perceptible erythema 24 h after a single UV-exposure

MMD

Minimal melanogenesis dose, the UV dose to elicit just perceptible tanning 7 days after the last UV-exposure. MMD will be expressed in SEDs [7]

nUVB

Narrowband UVB

SED

Standard erythema dose, the UV dose that elicits just perceptible erythema in the most sensitive people in a group of very sun-sensitive but otherwise healthy individuals. One SED is defined as 100 J/m2 at 298 nm using the CIE erythema action spectrum

Solar

Solar simulator

Introduction

Sun and solarium exposure with the purpose of tanning is almost always performed as repetitive behavior. Nevertheless only few studies have investigated the minimal pigmentation dose (MMD) after multiple UV-exposures and then mainly in fair-skinned persons [1, 5, 10].

After the same sub-erythemogenic doses from UV-sources with different emission of UVB and UVA in skin types II and III, it was found that UVA irradiation enhanced pigmentation, whereas UVB irradiation did not [1]. After five repetitive UV-exposures to 0.25, 0.5, 0.75 and one MED Kaidbey and Kligman found melanogenesis 7 days after the fifth UVA-exposure to only 0.25 MED, whereas no pigmentation was evident with UVB [5]. Parrish found as well, in skin types II and III, that UVA was more melanogenic than erythemogenic; evidenced by MMD/MED < 1, whereas the opposite was true for UVB (MMD > MED). This was true for both single and multiple UV-exposures [10].

We chose to include also skin types IV and V to have a broader skin type (pigmentation) spectrum. The aim of this study was to examine the ability to tan, assessed as MMD, after single and multiple UV-exposures in relation to skin type and pre-exposure skin pigmentation. Moreover, we investigated how the melanogenic potential depends on the UV-spectrum (light-source) and finally by skin reflectance we investigated the objectively measured pigmentation increase.

Materials and methods

Subjects

The local ethics committee approved the study (KF01-020/99), which took place outside the summer period to avoid seasonal variation of skin pigmentation. Sixty-two healthy volunteers, 34 females and 28 males, aged between 19 and 44 years (mean age 25 years) were recruited by advertisements in local papers. The volunteers were enrolled in the study after giving written consent. Skin diseases, sunbathing or exposure to artificial tanning 3 months prior entering into the study excluded participation as did medication with glucocorticoids, non-steroidal anti-inflammatory drugs, anti-histamines or retinoids. Volunteers with different skin types [3] were selected to achieve a broad range of skin pigmentation. A total of 10 volunteers only completed the single UV-exposure and then dropped out. Seven volunteers had self-evaluated skin type II, 16 skin type III, 16 skin type IV and 13 skin type V. The ethnic backgrounds varied from Scandinavians (n = 16), Hispanics (n = 12), and Asians (n = 12) to Indians and Pakistanis (n = 12). The pre-exposure pigmentation on the back varied from 13 to 55% [12].

Study design

At entry, the lower back of each volunteer was sub-divided into four areas. One area was allocated to UV irradiation with narrow-band UVB (nUVB) (TL01) (Philips, Rosendaal, Holland), the second area to UV irradiation with Solar Simulator (Solar)(Solar Light Co., Philadelphia, PA, USA), the third area to UV irradiation with broad-band UVA (bUVA) (Cleo-performance) (Philips) and the fourth area to UV irradiation with UVA1 (TL10) (Philips). To avoid anatomical differences in skin UV-sensitivity, the allocation of the four UVR sources was carried out by randomization. The pre-exposure skin pigmentation in the test areas was measured by a reflectance meter. To calculate the objectively measured pigmentation increase, the skin pigmentation was measured again by the reflectance meter 7 days after, respectively, the single UV-exposure and the five UV-exposures.

To determine the MMD after a single UV-exposure six UV-doses with 25% increments were used. MMD was determined clinically 7 days after the UV-exposure. When the individual MMD was determined, the volunteers were exposed to the four UV-sources in four new areas for five consecutive days in 24 h intervals.

About 7 days after the fifth UV-exposure MMD was evaluated clinically. The daily dose to give the same measurable pigmentation increase after one and five UV-exposures was calculated.

Throughout the 5 days of exposure we used six doses with 25% increments for the Solar Simulator. For the other UV-sources we used 100% increments since we did not know what to expect. Maximum dose for bUVA and UVA1 was two MMD as found after the single exposure. For the Solar Simulator and the nUVB maximum dose was one MMD to avoid burns.

Pigmentation determination

The skin reflectance was measured using an instrument (UV-Optimize 555, Matic, Nærum, Denmark) with peak wavelength at 555 and 660 nm. The reflectance of these wavelengths is giving the melanin content or pigmentation of the skin, when the skin redness is zero. In the most pale persons, when the arm was drained from blood, the reflection of 660 nm light was 70%, which was defined to be 0% pigmentation, and 100% pigmentation corresponds to 0% reflection of the 660 nm light [9].

Dosimetry and UVR sources

The emission spectra and intensity of the radiation sources were measured before the start and regularly during the study using a hand held electronic spectroradiometer (Sola Tell, Sola-Hazard, 4D Controls Ltd., Redruth, UK). The manufacturer calibrated the spectroradiometer. The erythema activity of a UV source was measured in standard erythema dose (SED) as proposed on the 12th International Congress on Photobiology in 1996 [13, 14] and accepted by the CIE [2]. The physical dimension of one SED is 10 mJ/cm2 at 298 nm using the CIE action spectrum [7]. UV doses were therefore expressed in SED.

The nUVB, the bUVA and the UVA1 consisted of a bank of six fluorescent tubes. The nUVB tubes emit 81% in the UVB range and 19% in the UVA range with peak emission at 311 nm (one SED corresponds to 181 mJ/cm2). bUVA emit 99% in the UVA range and 1% in the UVB range (one SED corresponds to 8,400 mJ/cm2). UVA1 tubes emit 99.9% in the UVA range and 0.1% in the UVB range (one SED corresponds to 29,500 mJ/cm2). The emission spectra of the four UV-sources are shown in Fig. 1. Six 2 × 2 cm2 squares, each square representing one UV dose, were arranged as 2 × 3 openings in an UV impermeable mask. During irradiation the distance from the skin to the tubes was 40 cm for nUVB and 10 cm for bUVA and UVA1.
https://static-content.springer.com/image/art%3A10.1007%2Fs00403-006-0728-3/MediaObjects/403_2006_728_Fig1_HTML.gif
Fig. 1

Emission spectrum of the four UV-sources

The Solar Simulator is equipped with a filtered xenon lamp that emits 8.7% in the UVB range and 91.3% in the UVA range (one SED corresponds to 1,490 mJ/cm2). The Solar Simulator conducts UV irradiation through six individually adjustable liquid light-guides with 10 mm Ø. During the irradiation the distance from the skin to the end of the light-guides was about 1 mm. Due to some instability and fluctuations in the intensity of the xenon lamp the intensity was measured just before and just after the exposure, and the average UV dose was used for the statistical analyses.

UV-exposure was interrupted in a specific area at erythema grade +++ or complaints of burning sensation irrespective of the erythema grade.

Statistics

Regression analyses were performed to find the relation between UV-dose to one MMD and skin type and pre-exposure skin pigmentation. Not all the data were normally distributed, therefore we performed paired non-parametric tests (Wilcoxon signed rank test) to calculate if the differences in SED to one MMD between the UV-sources were significant.

Skin type and pre-exposure pigmentation cannot be compared by r-values as skin type is categorical data and pigmentation is continuous data. Not all the data were normally distributed, therefore, the relation between MMD and skin type and pre-exposure pigmentation was analyzed by the non-parametric Spearman’s rank correlation, and comparison can be performed of r and P-values in Table 1 (parts a and b).
Table 1

Correlation coefficient r and P-values from Spearman’s rank correlation test of the relation between MMD and skin type (a) and between MMD and pre-exposure pigmentation for different light sources (b)

Number of exposures

Solar

nUVB

bUVA

UVA1

r

P

r

P

r

P

r

P

(a) MMD and skin type

 1

0.30

0.02

0.41

0.001

0.38

0.002

−0.19

NS

 5

0.17

NS

0.43

0.002

0.01

NS

−0.05

NS

(b) MMD and pre-exposure pigmentation

 1

0.65

<0.001

0.63

<0.0001

0.41

0.0009

−0.20

NS

 5

0.32

0.03

0.50

0.0003

0.22

NS

0.01

NS

A paired t-test was used to test if the differences in objectively measured pigmentation increase between the UV-sources were significant, when data were normally distributed, otherwise Wilcoxon signed rank test was performed. Moreover, for each of the UV-sources we used the Wilcoxon signed rank test to test if the objectively measured pigmentation increase was significantly higher after five compared to one UV-exposure.

Covariance analysis was performed to test for gender differences.

We used the Wilcoxon signed rank test to test if the difference in SED dose to one MMD between Solar Simulator (the least melanogenic) and UVA1 (TL10) (the most melanogenic) was significantly different after a single UV-exposure compared to five UV-exposures. P < 0.05 were considered to be significant.

Results

Four of the Scandinavian persons (skin Type II) did not tan after a single UV-exposure to narrowband UVB or Solar Simulator and could therefore not be exposed for these UV-sources for 5 days.

Skin type

After one UV-exposure, we found a highly significant positive linear correlation between UV-dose to one MMD and skin type [see Table 1 (part a)]. This relation was found for all UV-sources except for the UVA1-source (most melanogenic) (Fig. 2a–d). After five UV-exposures the relation between UV-dose to one MMD and skin type was only significant for the nUVB-source (most erythemogenic) [Table 1 (part a)]. See Fig. 3a–d.
https://static-content.springer.com/image/art%3A10.1007%2Fs00403-006-0728-3/MediaObjects/403_2006_728_Fig2_HTML.gif
Fig. 2

SED to give one MMD on the back of individuals with different skin types after a single UV-exposure to: a nUVB (n = 58), b solar simulator (n = 58), c bUVA (n = 62) and d UVA1 (n = 62)

https://static-content.springer.com/image/art%3A10.1007%2Fs00403-006-0728-3/MediaObjects/403_2006_728_Fig3_HTML.gif
Fig. 3

SED to give one MMD on the back of individuals with different skin types after five UV-exposures to: a nUVB (n = 49), b solar simulator (n = 49), c bUVA (n = 52) and d UVA1 (n = 52)

Pre-exposure pigmentation

After one UV-exposure, we found a highly significant positive linear correlation between UV-dose to one MMD and pre-exposure pigmentation [Table 1 (part b)]. This relation was found for all UV-sources except for the UVA1 (Fig. 4a–d). For UVA1 (Fig. 4d) the number of SED to one MMD was independent of the pre-exposure pigmentation, being ∼1 SED for all skin types. After five UV-exposures this relation was only significant for (the most erythemogenic UV-sources) nUVB and the Solar Simulator (Fig. 5a–d). For UVA1 (Fig. 5d) the number of SED to one MMD was independent of the pre-exposure pigmentation, each dose being 0.4 SED for all skin types [Table 2 (part a)].
https://static-content.springer.com/image/art%3A10.1007%2Fs00403-006-0728-3/MediaObjects/403_2006_728_Fig4_HTML.gif
Fig. 4

SED to give one MMD on the back of individuals with different pre-exposure skin pigmentation after a single UV-exposure to: a nUVB (n = 58), b solar simulator (n = 58), c bUVA (n = 62) and d UVA1 (n = 61)

https://static-content.springer.com/image/art%3A10.1007%2Fs00403-006-0728-3/MediaObjects/403_2006_728_Fig5_HTML.gif
Fig. 5

SED to give one MMD on the back of individuals with different pre-exposure skin pigmentation after five UV-exposures to: a nUVB (n = 49), b solar simulator (n = 49), c bUVA (n = 52) and d UVA1 (n = 52)

Table 2

Average UV-dose (SED) to give a minimal pigmentation (MMD) after one and five UV-exposures (a) and average UV-dose (mJ/cm2) to give a minimal pigmentation (MMD) after a single UV-exposure (b) in skin types II–V

 

Solar

nUVB

bUVA

UVA1

(a) Average UV-dose (SED)

1 UV-exposure

9.3 (3.9–14.8)

6.5 (3–13)

2.2 (1.2–5)

0.9 (0.7–1.6)

5 UV-exposures

5.6 (2.0–11.1)

4.0 (1.6–7.6)

1.1 (0.4–2.6)

0.4 (0.1–0.9)

(b) Average UV-dose (mJ/cm2)

 

13857 (5811–22052)

1177 (543–2353)

18480 (10080–42000)

26550 (20650–47200)

Mean and range

Wavelength

The UV-dose (SED) needed to produce a minimal pigmentation (one MMD) after, respectively, one and five UV-exposures are listed in Table 2 (part a) for each of the four UV-sources. The physical UV-dose (mJ/cm2) to MMD after a single exposure equivalent to SED is calculated in Table 2 (part b). Table 2 (part a) shows that a much lower UV-dose (SED) is needed to produce a minimal pigmentation as the UVB-content of the light source declines and the UVA-content increases. Except for Solar Simulator compared to nUVB.

The Solar Simulator was the least melanogenic of the four UV-sources and UVA1 was the most melanogenic. This means that the melanogenic effect is highly dependent on wavelength. The differences between the UV-sources were highly significant (P < 0.0001).

Single versus multiple UV-exposures

When giving five UV-exposures the daily dose needed to produce a minimal pigmentation (MMD evaluated visually) was ∼50% lowered compared to a single UV-exposure [Table 2 (part a)]. Therefore the cumulative dose is still higher after five UV-exposures.

The daily MMD dose to give the same measurable pigmentation percentage increase after a single and five UV-exposures was also calculated. It was found to be only ∼39% of the daily dose needed after a single UV-exposure.

Objectively measured pigmentation increase

Table 3 shows the objectively measured pigmentation increase corresponding to 100 and 50% of visible MMD. For both one and five UV-exposures and for 100% as well as 50% of MMD, there were no significant differences in objectively measured pigmentation increase between the UV-sources. A total of 50% of MMD corresponds to a pigmentation increase of ∼1% (Table 3). In other words, if the pigmentation should increase by pigmentation 1%, a dose equal to 50% of MMD is needed. The pigmentation is thus measurable before it becomes evident clinically.
Table 3

The dose to 100% of MMD and 50% of MMD corresponds to the indicated objectively measured pigmentation percentage increase (reflectance)

MMD

UV-exposures

nUVB

Solar

bUVA

UVA1

Mean

P

100%

1

3.1

3.0

3.1

2.2

2.9

NS

5

5.3

3.7

4.1

3.7

4.2

NS

P

0.007

NS

NS

0.03

0.003

 

50%

1

0.5

1.1

1.5

1.7

1.2

NS

5

1.4

−0.2

1.4

0.9

0.9

NS

P

NS

NS

NS

NS

NS

 

Fifty percent of MMD corresponds to a pigmentation increase of approximately pigmentation 1%. The objectively measured pigmentation percent is more pronounced after five UV-exposures than after one exposure for 100% MMD. However, this is statistically significant only for nUVB (P = 0.007) and UVA1 (P = 0.03). Average of all volunteers (skin types II–V)

After five UV-exposures the measurable pigmentation increase is on average 45% higher than after a single UV-exposure (for 100% MMD).

Ideally they should have been equal if the objective (measured) and clinical (visual) evaluation were identical. The daily dose to give the same measurable pigmentation increase after a single and five UV-exposures was calculated. It was found to be only ∼39% of the dose needed after a single UV-exposure.

To determine which of the parameters; skin type and pre-exposure pigmentation, that was more predictive of dose to one MMD, we compared the r and P-values obtained from Spearman’s rank correlation [Table 1 (parts a and b)]. The table shows a remarkable difference, pre-exposure pigmentation clearly preferable to skin type.

No gender differences were found in UV-dose to one MMD, except for bUVA. The slope of the bUVA curve for MMD in relation to pre-exposure pigmentation after one UV-exposure was steeper for females compared to males (P = 0.001), but no significant difference in pigmentation was found.

Discussion

The Solar Simulator emits shorter waved UVB compared to nUVB (TL01). The short-wave UVB counts heavily to the SED and therefore the UVA dose only contributes to the SED by 14%. This is the reason why the Solar was less melanogenic than nUVB despite its higher UVA-content. Moreover does the Solar Simulator emit short-wave UVA that is less melanogenic than UVA1. nUVB was the most erythemogenic and a lower SED was needed to produce MMD than for the Solar Simulator [4].

Parrish found, in skin types II and III, that UVA was more melanogenic than erythemogenic; evidenced by MMD/MED < 1, whereas the opposite was true for UVB. This was true for both single and multiple UV-exposures on the buttocks [10].

The difference in SED to one MMD between Solar Simulator (the least melanogenic) and UVA1 (the most melanogenic) is more pronounced after five UV-exposures compared to a single UV-exposure (P = 0.001). Conversely to Parrish’ study [10], we did not use (sub-) erythemogenic doses, but equal melanogenic doses (MMD doses). Therefore our UVA-doses were relatively smaller, whereas our UVB-doses were relatively larger. The skin on the back is more sensitive to UV-irradiation than the buttock skin [11]. We exposed our volunteers on the back and therefore presume that a smaller UV-dose was needed than the UV-dose to give minimal pigmentation on the buttocks.

Sub-erythemogenic UVA-doses induce pigmentation, whereas tanning from UVB-irradiation only occurs when preceded by erythema [1]. This is in accordance with the finding of Kollias et al. that MED > MMD at 365 nm, and that MED < MMD at 295 and 305 nm and for UVB irradiation at 315 nm MED = MMD after a single UV-exposure in skin types I, II and V [6].

The larger dose intervals (100% increments) for the nUVB, bUVA and UVA1 compared to the Solar Simulator (25% increments) might be expected to result in a larger spread in these data than for the Solar Simulator. This was not the case, in fact, the range in SED to MMD for the Solar Simulator data was larger than for the other UV-sources [Table 2 (part a)].

The relation between UV-dose to one MMD, skin type and pre-exposure pigmentation was stronger after a single UV-exposure than after five UV-exposures [Table 1 (parts a and b)]. The reason for this is probably different initiation of protective mechanisms such as pigmentation and thickening of stratum corneum, already during the exposures. The relation to the pre-exposure pigmentation therefore becomes weaker during repeated exposures as the pigmentation may evolve with different speed and degree in individuals. We do not know if the minimum daily dose to tan after five exposures is reached or may be even lower after more exposures.

After, respectively, one and five exposures the objective (measured) and the clinical (visual) evaluation of pigmentation was identical for the different UV-sources (Table 3).

After five UV-exposures the objectively measured increase in pigmentation percentage was higher than after one UV-exposure, but this was only statistically significant for some UV-sources, nUVB (P = 0.007), UVA1 (P = 0.03) and for a mean of all UV-sources (P = 0.003). If the objective and the clinical evaluation of pigmentation had been identical, this difference would not occur. This means that the reflectance instrument (UV-optimize) is more sensitive at distinguishing pigmentation than the eye. The reason why more pigmentation is needed after five compared to one UV-exposure to be perceived by the eye is unknown. Regarding nUVB, we speculate that thickening of stratum corneum induced by the UV-exposures changes the optics of the skin and thereby changes how pigmentation is perceived. As thickening of stratum corneum mainly is caused by UVB [8] another yet unknown phenomena may play a role for UVA1. However, it is worth noting, that this did not apply to bUVA and the Solar Simulator.

In summary, the melanogenic effect of the UV-sources was highly dependent on wavelength: a lower SED dose was needed to produce a minimal pigmentation the longer the wavelengths. Solar was the least melanogenic and UVA1 the most melanogenic.

The UV-dose to one MMD was well correlated to skin type and pre-exposure skin pigmentation for one UV-exposure, except for the UVA1-source. For five UV-exposures MMD was only well correlated to skin type for nUVB whereas the relation to pre-exposure skin pigmentation was significant for both nUVB and Solar. This study showed that pre-exposure pigmentation was clearly more predictive of induction of pigmentation than skin type [Table 1 (parts a and b)].

Generally, for nUVB and Solar: the more pigmented a person was the higher SED dose was needed to tan. For UVA and particularly UVA1 the dose to one MMD was independent of pre-exposure pigmentation. A total of 50% of MMD corresponded to an invisible pigmentation increase of ∼1%, and was thus just measurable.

Acknowledgments

We thank engineers Jacob Heydenreich and Peter Alshede Philipsen, technicians Benedicte Wulf and Trine Ravn and medical students Pernille Fog Svendsen and Malene Rahbek Holm for their help in the project. This work was supported by Unilever, NJ, USA.

Copyright information

© Springer-Verlag 2007