Edited by James G. Titus U.S. Environmental Protection Agency
N. Kollias and A. Baqer Department of Dermatology, Al-Sabak Hospital Physical Department, Kuwait University Kuwait
This paper assesses the average amount of pigment in a sample of the population of Kuwait and compares it with the sensitivity of the population to artificially produced UV-B, as well as monochromatic bands of UV-B, to arrive at an estimate of the protection afforded by the native pigment. The pigment level of the population sampled was 2.18 on a scale from 0 to 9, while the level in an equivalent sample of northern Europeans would be 0.6-0.8. In these measurements we estimate the pigment level (melanin concentration) assuming it resides in the epidermis. This implies that the concentration of melanin in the local population is on the order of 3.5 times larger, while the amount of light that enters the dermis after the above absorption is approximately 20 times smaller. The correlation found between the pigment level and the minimum dose for a UVB-induced erythema was very weak, with an average value less than one half the value of a northern European sample. The responses to monochromatic bands of UV-B indicate that the action spectrum of our population is different from that reported by WHO (1982) for a standard population sample, especially at 295 nm.
We conclude that the pigment plays two roles in photoprotection. First, it absorbs light, and the suppression of the erythema effectiveness of the radiation is proportional to the absorbance. Second, it modulates the amount of UV-B that is delivered to the dermis and it appears that carcinogenesis is proportional to this intensity.
Numerous researchers have concluded that people who are more pigmented are less susceptible to UV-B (Urbach 1982; Pathak and Fitzpatrick 1974; Hawk and Parrish 1982). This implies that pigmented people are at a lower risk of developing an erythema reaction and are also at a lower risk for tumor production. Human skin has been classified in six types according to the way it responds to sunlight (Pathak, Fitzpatrick, and Parrish 1982), i.e., always burn easily-tan little or none, usually burn easily-tan minimally, burn moderately and average tan. Such classification is based on clinical history and no mention is made of native pigment. We are looking for an objective criterion that would allow assessment of the risk factor. Native pigment is an obvious candidate.
Amblard et al. (1982) show that native pigment and eye color exhibit a good correlation with the erythema threshold level (minimal erythema dose, MED). A recent study (Shono et al. 1985) suggests that there is a strong correlation between native pigment and erythema reaction with a small statistical sample.
In Kuwait we find the native pigment to be a weak indicator of a person's anticipated response to UV-B. That is to say, given two persons of equal pigment level we can only predict their MED within a factor of two of the measured value--not a precise estimate. Native pigment is a parameter that is perceived by the eye, i.e., using visible radiation. In general, it is not a good practice to deduce the absorbance of any material in the ultraviolet simply by recording its absorbance in the visible, unless, of course, one has identified the compound by its visible spectrum. The major absorber in the skin is melanin, which resides mainly in the epidermis in melanocytes and keratinocytes. It absorbs strongly in the visible and even more strongly in the ultraviolet (UV). The absorption by intact epidermis (Kaidbey et al. 1979) shows a linear relation with wavelength down to about 320 nm; for wavelengths shorter than 320 nm the absorbance increases rapidly. Over the UV-B range the absorbance by the epidermis shows an increase by a factor of two (approximately).
A method has been developed for assessing the melanin concentration in human skin (Kollias and Baqer 1985; 1986). We have used this method to assess the pigment level in a small sample of the population of Kuwait. We then tested the sensitivity of local skin to polychromatic and monochromatic UV-B radiation. Using the statistics available through the Kuwait Cancer Registry, we assessed the UV-B risk and the role that native pigment plays in the erythemogenesis and carcinogenesis.
Measurement of the Pigment Level of a Small Sample of the Population of Kuwait
Over the last two years, we have carried out measurements of skin pigment on 314 volunteers. The measurement was non-invasive and only took 1.5 minutes to complete. The probe that comes in contact with the skin was temperature regulated to cause a minimum of discomfort to the individual volunteer; over 250 of these were patients. The volunteers came from all walks of life and were randomly selected. The measurements on the patients were always taken on uninvolved areas of the skin. We did make measurements on involved areas but those were excluded from this study. We made certain that none of the volunteers were under any medication which could possibly have an effect on their pigmentation. The healthy volunteers were the doctors, nurses, technicians, and personnel of the Skin Department of Al-Sabah Hospital as well as some students from Kuwait University. All volunteers were informed of the nature of the experiment and their consent was obtained.
Measurement of the Minimum Erythema Dose to Artificially Produced UV-B Radiation of Human Skin
These measurements were conducted on 26 psoriatic patients over the course of one year. These patients were about to begin treatment with UV-B radiation, which constitutes a very effective line of treatment for psoriasis. Each patient was phototested to determine the minimum UV-B dose necessary to elicit an erythema reaction. Before any radiation was applied, we made sure that all the parameters were recorded and pigment index was measured. The pigment level was evaluated on an area of normal skin. The information obtained in these measurements was of interest to the physician in prescribing the indicated dose of UV-B for the treatment.
The patients were given doses of 50, 100, 150, and 200 mJ/cm on four areas of the back that were 2.5 cm in diameter. Patients with very light complexion were initially given doses of 20, 40, 60, and 80 mJ/cm. The irradiated sites were observed 24 hours after irradiation and the minimum dose required to elicit an erythema reaction was recorded. If no reaction occurred, the test was repeated with four higher doses. If, on the other hand, all areas showed erythema, then the test was repeated with four lower doses. A Waldman Model 80001K upright UVB-UVA unit was used for the irradiations. The patient received radiation in a standing position at a power of 0.4 mW/cm so that the desired doses could be delivered within a few minutes at minimum discomfort to the patient. During the phototest the patients' face and body were covered by a protective robe which allowed only the four areas on the back of the patient to be irradiated. When the dose level was reached for the first areas, the cubicle door was opened and that area was covered. This procedure was repeated until all spots were covered. The UV-B producing lamps in the apparatus were Sylvania 75/85W/UV21 fluorescent lamps.
All the tests and measurements carried out in this section were supervised by Dr. V. Heigy of the Department of Dermatology of Al-Sabah Hospital. All patients were informed of the nature of the tests and the reasons for them; their consent was obtained before the measurements were made.
Measurement of the Erythema Effectiveness of Three Selected Wavelengths of UV-B
In this series of measurements we used 16 healthy volunteers who had skin types common to the area of Kuwait. The wavelengths used were 295, 305, and 315 +/-5 nm. The irradiation sites were on the upper back of each individual and we made sure that we stayed away from the middle of the back since the thickness of tissue under the skin as well as the blood supply differs from the rest of the back. The volunteers were in a sitting position during the test. At each wavelength six spots were irradiated starting with a dose that was considered to be suberythemogenic, progressing to higher doses at 20% increments. Thus, the three wavelength tests generated three rows of spots 6 mm in diameter on the back of each volunteer. We further made sure that none of the subjects had exposed their backs to the sun in the two to three months immediately before the test.
The instrument used to carry out these irradiations was an Applied Photophysics "Clinical Photoirradiator" model UV-90, which is capable of providing a monochromatic output in the UV-B wavelengths used in this study. The output power of this instrument was 2.1 mW/cm at 295 nm, 2.4 mW/cm at 305 nm, and 2.8 mW/cm at 315 nm. The range of doses used was from 12 to 150 mJ/cm at 295 nm, 33 to 270 mJ/cm at 305 nm, and 330 to 2700 mJ/cm at 315 nm. Because of the high doses necessary to elicit an erythema reaction at the longest wavelength the total time for the test to be completed was approximately 1.5 hours. This narrowed the number of people who were willing to volunteer.
Each volunteer had to return to the Phototherapy Unit of Al-Sabah Hospital eight hours and 24 hours after irradiation to assess their skin reaction. The bandpass of the irradiation was +/-5 nm for all the wavelengths. All the tests carried out in this section were supervised by Dr. Yousef Malallah of the Department of Dermatology of Al-Sabah Hospital, who worked very closely with the investigators during these measurements.
Skin Cancer Statistics
The skin cancer statistics were provided by the Kuwait Cancer Registry and included data collected from hospitals where biopsies were obtained. Cases of patients who sought treatment overseas do not appear in these statistics.
Measurement of the Pigment Level of a Small Sample of the Population of Kuwait
The results of these measurements are displayed in Figure 1. The mean is 2.18 +/- 0.08 with a standard deviation of 1.40. It could be argued that this is neither a random sample of the population nor a sufficiently large one. These data are not presented as a true average but rather as an unbiased indication of the pigment level of the population.
Measurement of the Minimum Erythema Dose to Artificially Produced UV-B Radiation of Human Skin
In the tests conducted we found erythemogenic doses to be from 50 to 350 mJ/cm, while the pigmentation index varied from 0.5 to 5.2. The correlation between the pigmentation index and the log of the minimum erythema dose for all the volunteers is displayed in Figure 2. The correlation is not a strong one; however, it is significant. The bandpass of the irradiation was +/-5 nm for all the wavelengths.
Although we feel that the above results are reliable, we by no means wish to indicate that they are truly representative of what the MED of the Kuwait population would be. Such a conclusion could only be arrived at through a much more thorough study in which we would need to include a well-represented and large sample of the population of Kuwait.
Skin Cancer Statistics
These statistics are presented in Table 1. They were supplied to us by the Kuwait Cancer Registry and they include only cases that were biopsied in the hospitals of Kuwait. The first numerical column indicates the total number of citizens of the Arabian Gulf countries and the second column shows the grand total. Of the 125 cases in the first reporting period, 94 were on exposed areas of the skin; this constitutes 75% of the total. Of the 78 cases in the second period, 58 were on exposed areas, also constituting 75% of the total.
Skin cancer will occur with repeated suberythemogenic doses of UV-B (Van der Leun 1984); this is a situation that prevails with our population. The data of the Kuwait Cancer Registry indicate that the most frequent forms of neoplasm are the basal cell carcinoma, with 43% of all cases reported between 1974 and 1980, and squamous cell carcinoma with 42%. For Gulf Cooperation Council nationals, 48% were squamous cell carcinomas and 37% basal cell carcinomas. The remainder of the cases were of various types of skin malignancies (see Table 1). What we consider very interesting and alarming is that skin cancer exists in this area even though people do not "burn," and that 75% of the neoplasms are on exposed areas of the skin. As the population of Kuwait exposes only the hands and the face to the sun, a reasonable estimate of the exposed area of the skin of the population would be approximately 10%. Thus 75% of the neoplasms occur on 10% of the skin. Therefore, it is reasonable to conclude that solar UV-B does contribute to the occurrence of skin cancer in the Kuwaiti population, as in other populations.
Scotto, Fears, and Fraumenti (1982) report that the annual age-adjusted rates of skin cancer in New Mexico are ten times higher for "Anglos" than for "Hispanics." Therefore, we would expect a reduced rate for darker pigmented persons. Furthermore, Scotto and Fraumenti (1982) report that the age-adjusted rate for skin cancer for American whites was 232.6 per 100,000 while for blacks it was 3.4 per 100,000. For all other cancers the rate for whites was 318.9 and for blacks 347.3 per 100,000; in other words, the latter was not as significantly different as the former. However, it has been found that experimental animals "...kept in a heated environment rapidly developed more UV-induced tumors than mice living in a temperate environment" and that "...wind and increased humidity caused increased acute UV-induced damage and acceleration of tumor formation." It is therefore likely that these environmental factors could outweigh the natural protection that darker people have for UV-induced neoplasms. Kagetsu et al. (1985) report that UV-A-induced erythema is definitely enhanced by increased skin surface temperature in humans. It is clear that more work needs to be done to evaluate the additional risk by these climatic conditions (that is, heat and humidity) as well as the genetic protection, if any, associated with skin type.
The measurements on the "pigmentation index" showed a value of 2.18 +/- 0.08 for a population of 314 volunteers. This is not a large enough statistical sample but it is large enough to provide us with an estimate of where we are. Similar measurements do not exist for any other population yet since the technique for carrying out these measurements was just published in January 1986 (Kollias and Baqer 1986). Based on current experience, we would guess that the pigmentation index for English people would be approximately 0.8 +/- 0.2.
If we assume that the remitted intensity from white Caucasian skin is approximately 20% of the incident intensity (Kollias and Baqer 1985), then because the Beer Lambert law can be considered valid for intact epidermis (Bruls et al. 1984) we have the following equation:
I / I0 = exp( -0.8 x A) = 0.20 for white skin and
I / I0 = exp( -2.2 x A) for Kuwaiti skin
where A is an arbitrary parameter representing the product of the thickness of the absorber and the factor that relates the pigmentation index with the absorption coefficient times the concentration. Solving the first equation for A, we obtain A = 2.0. Substituting this value of A into the second equation, we obtain a remitted intensity approximately 1% of the incident intensity. This simple calculation implies that the intensity that arrives at the dermis of a white-skinned Caucasian is approximately 20 times larger than that of a typical inhabitant of Kuwait.
The validity of this calculation hinges on the validity of the assumption that the remitted light is attenuated in the epidermis and that UV-B absorbance by the epidermis is linearly related to the visible absorbance.
If the incident intensity is attenuated so strongly by the resident pigment in the epidermis, then we should be able to predict the dose at which erythema will be induced by artificially produced UV-B in accordance with the concentration of pigment. These measurements on 26 psoriatic patients do not substantiate this assumption. In Figure 2 we see a gentle and general correlation between the pigment level and the minimum erythema dose. [The reason that it is plotted against the logarithm of the MED is because it has been determined that the MED dose for a population does not form a normal distribution while the logarithm of the MED does form a normal curve (Mackenzie 1983; Amblard et al. 1982)].
It is known that patients tend to have a higher sensitivity to UV-B than healthy volunteers. From these data we can conclude that the native pigment does not offer a great deal of photoprotection against UV-induced erythema, although it obviously absorbs strongly the incident UV-B. The mechanism for erythema production seems to be mediated through a photoproduct that is produced in the upper epidermis. We are currently experimenting in order to establish a correlation between the native pigment and some measurable attenuating factor.
Because artificial sources tend to have a spectral output that is usually different from that of the sun, we tested for the sensitivity of human skin (local) to selected wavelengths in the UV-B range. The wavelengths selected were 295, 305, 315 +/- 5 nm. In these measurements no effort was made to correlate the MED and the pigment level; rather phototests were carried out on 16 volunteers with pigment level similar to the average determined previously. The results obtained were compared with those of the photobiology unit of Dundee, Scotland (Mackenzie 1983).
The log (MED) ratio for 295 nm was 1.6 times larger for the Kuwait subjects. The ratio for 305 nm was 1.1. The ratio for 315 nm was 0.99. It follows that the difference in the MED is maximum for the 295 band and is almost the same for the other two bands. The waveband for which the MED's ratio is the largest is the one that is the weakest in the solar insolation. We can thus conclude that the erythemal effectiveness of monochromatic UV-B wavelengths is similar at the long wavelength end of the populations of Kuwait and Scotland with the difference becoming maximum for the 295 nm band.
Since the erythema effectiveness maximum factor of five is different at the shortest wavelength [a factor of 1.6 for Log (MED)], it cannot possibly account for the factor of one hundred difference for the carcinogenic effect of UV-B on deeply pigmented versus white individuals. These observations provide further evidence that the erythema effectiveness of UV-B wavelengths is not necessarily the same as the effectiveness for carcinogenesis or for photokeratosis.
There is no doubt that the population of Kuwait is naturally protected from the intense rays of the sun more than Caucasian people. The question is, are they adequately protected from their severe UV-B environment? The absence of severe photodermatoses as well as the rarity with which one observes severe sunburns is adequate testimony to some protection. The skin of our population is more pigmented than that of northern Europe and the sensitivity of the local skin to artificially produced solar-simulated radiation is at least half that of Caucasians (Y. Malallah, private communication). The results presented in the previous section relate to the log (dose) and consequently they appear small. The actual energy dose is 12.5 mJ/cm for Dundee and 54 mJ/cm for Kuwait for 295 nm to produce an erythema reaction. At 305 nm and at 315 nm the differences remain small even on a linear scale.
Skin cancer incidence according to the statistics that are available to us is at the level of 2.6 per 100,000 for Kuwaiti males and 1.3 per 100,000 for Kuwaiti females for the 1979-81 period. It is interesting that the levels in the previous reporting period were much lower, i.e., 1.1 per 100,000 for Kuwaiti males and 1.0 per 100,000 for Kuwaiti females. This trend indicates either a significant increase in skin cancer incidence or a significant increase in the reporting of skin cancer cases to the Registry. Another factor that is completely beyond the control of the experimenters and is difficult to account for is the number of suspected skin cancer patients who seek treatment overseas. These cases are missing from the local statistics and could make the numbers smaller than they actually are. As the majority of skin neoplasms appear in exposed areas it would not be surprising if a good number of people seek plastic or reconstructive surgery elsewhere.
Our conclusion from these results is that melanin in the epidermis plays two roles: one in the case of carcinogenesis, and another in the case of UVB-induced erythema. It appears that in the case of carcinogenesis, melanin (the primary pigment in human skin) acts as an absorber, attenuating the incident intensity. Thus, tumor initiation appears to be proportional to the amount of UV-B that arrives to the basal cell layer. In the case of UV-B-induced erythema the correlation between the concentration of melanin and the MED dose is very weak, which implies that there must be some other mediator for the production of erythema. The role of melanin in this case would possibly be as a scavenger of free radicals generated by the UV-B radiation in the epidermis. The radicals that are not compensated would then migrate into the dermis through the basal cell layer and generate the erythema reaction. Thus, the concentration of melanin will be a determining factor in the erythema. In a paper that we are currently submitting for publication we discuss the different forms that melanin molecules can take up in the skin and the different functions that these molecules might perform.
The estimated average pigmentation index of the population of Kuwait is 2.2, while that of European Caucasians is estimated to be less than 1.0 and assumed to be 0.8. The pigmentation index is a parameter that varies from 0 to 9 (there are darker people but we have not measured their pigmentation yet), and is directly related to the absorbance of the epidermal melanin.
A weak relationship exists between the minimum erythema dose of UV-B and the pigment level. Considering the average values of MED and pigment level we are led to the conclusion that the relative protection that the pigment renders is related to the absorbance or pigment level; i.e., as the pigment level increases by a factor of 2+ the MED decreases by a factor of 2+. It is unfortunate that for these preliminary results we had to use data obtained from psoriasis patients as the data are probably slightly biased to a lower MED value. MED values determined for normal Kuwaiti subjects would render more definite answers.
Variation in the sensitivity of Kuwaiti skin and that of Caucasians is demonstrated with smaller wavelengths of solar UV-B. At 295 nm, sensitivity varies by a factor of four while at longer wavelengths it is not very different. This is similar to results described in Nakayama et al. (1974). This means that to properly evaluate the UV-B risk, we need to establish the spectral relative sensitivity (action spectrum) of Kuwaiti skin to UV-B. This means that the Effective Spectral Irradiance is not the same as that of Caucasian skin just attenuated by an appropriate factor. It is essential that the action spectrum for UV-B wavelengths of Kuwaiti nationals and typical expatriates be determined.
Analysis of the cancer statistics shows that skin cancer in the Kuwaiti population relates strongly with UV-B exposure. It should be noted that 75% of the skin neoplasms reported between 1974 and 1980 occurred on exposed areas of the skin, i.e., on 10% of the skin. This is an unmistakable and alarming signal. Statistical evaluation of the results is not complete because we are in the process of obtaining more information from the Kuwait Cancer Registry.
This work was supported by the Environmental Protection Council of Kuwait. We wish to thank Dr. M. M. Selim for his support and trust. We also express our thanks to Dr. Y. Malallah and Dr. V. Heigy for their valuable cooperation.
Amblard P., J. Beani, R. Gautron, J. Reymond, and B. Doyon. 1982. Statistical study of individual variations in sunburn sensitivity in 303 volunteers without photodermatosis. Arch. Dermatol. Res. 274: 195-206.
Bruls, W.A.C., and J. C. van der Leun. 1984. Forwarding scattering properties of human epidermal layers. Photochem. Photobiol. 10:231-42.
Hawk, J.L.M., and J.A. Parrish. 1982. Responses of normal skin to ultraviolet radiation. In The science of photomedicine, eds. J.D. Reagan and J.A. Parrish, 219-260. New York: Plenum Press.
Kagetsu N., R.W. Gange, and J.A. Parrish. 1985. UVA-induced erythema, pigmentation, and skin surface temperature changes are irradiance dependent. J. Invest. Dermatol. 85:445-47.
Kaidbey, K.H., P. Poh Agin, R. M. Sayre, and A.M. Kligman. 1979. Photoprotection by melanin: A comparison of black and Caucasian skin. J. Am. Acad. Dermatol. 1:249-60.
Kollias, N., and A. Baqer. 1985. Spectroscopic characteristics of human melanin in vivo. J. Invest. Dermatol. 85:38-42.
Kollias, N., and A. Baqer. 1986. The assessment of melanin in human skin in vivo. Photochem. Photobiol. 43 :49-54.
Mackenzie, L.A. 1983. The analysis of the ultraviolet radiation doses to produce erythemal responses in normal skin. Brit. J. Dermatol. 108:1-9.
Nakayama, Y., F. Morikawa, M. Fukuda, M. Hamano, K. Toda, and M. Pathak. 1974. Monochromatic radiation and its application: Laboratory studies on the mechanism of erythema and pigmentation induced by psoralen. In Sunlight and man, cons. ed. T.B. Fitzpatrick, eds. M.A. Pathak, L.C. Harber, M. Seiji, and A. Kukita, 591-611. Tokyo: University of Tokyo Press.
Pathak, M.A., and T.B. Fitzpatrick. 1974. The role of natural photoprotective agents in human skin. In Sunlight and man, cons. ed. T. B. Fitzpatrick, eds. M.A. Pathak, L.C. Harber, M. Seiji, and A. Kukita, 725-50. Tokyo: University of Tokyo Press.
Pathak, M.A., T.B. Fitzpatrick, and J.A. Parrish. 1982. Topical and systemic approaches to protection of human skin against harmful effects of solar radiation. In The science of photomedicine, eds. J.D. Regan and J.A. Parrish, 441-73. New York: Plenum Press.
Scotto, J., T.R. Fears, and J.F. Fraumenti, Jr. 1982. Solar radiation. In Cancer epidemiology and prevention, eds. D. Schottenfeld and J.F. Fraumenti Jr., 254-276. Philadelphia: W. B. Saunders Co.
Scotto, J., and J.F. Fraumenti, Jr. 1982. Skin cancer (other than melanoma). In Cancer epidemiology and prevention, eds. D. Schottenfeld and J.F. Fraumenti, Jr., 996-1011. Philadelphia: W.B. Saunders Co.
Shono, S., M. Imura, M. Ota, S. Ono, and K. Toda. 1985. The relationship of skin color, UVB-induced erythema, and melanogenesis. J. Invest. Dermatol. 84:265-67.
Urbach, F. 1982. Photocarcinogenesis. In The science of photomedicine, eds. J.D. Reagan and J.A. Parrish, 261-292. New York: Plenum Press.
van der Leun, J.C. 1984. UV-carcinogenesis. Photochem. Photobiol. 39:61-68.
World Health Organization. 1982. Environmental health criteria 23, lasers and optical radiation.