JOURNAL OF COSMETIC SCIENCE 220 I/II versus V/VI, and interestingly, the difference in the incidence of basal cell carcinoma (BCC), the most common type of skin cancer, is about 60 (23). The re is an inverse relationship between the increasing FST and the MED on habitually sun-protected skin. However, there is considerable overlap of the MED between FST I and IV. The difference in the MED between FST II and IV on sun-protected skin is about two, which is very modest in SPF terms (24). There have been fewer studies with FST V and VI, but comparisons between I/II and VI show a protection factor of melanin against erythema of about 6–8 (25,26). This is about the same as the protection factor against DNA photodamage for the whole epidermis (23). Stu dies to determine the protective properties of facultative tanning in FST I–IV have shown this to be modest against DNA damage and erythema. Field and laboratory stud- ies have shown protective factors to be in the region of 2–3 (27–29). Other factors such as stratum corneum thickening as a result of solar exposure may be important, but they are poorly understood (30). It is relatively simple to determine the SPF under very controlled laboratory conditions, one of which is the application of the sunscreen at 2 mg/cm2 on skin (14). In practice, people apply very much less with a commensurate typically linear reduction in the SPF (12). For example, one study on Danes on holiday in Egypt showed an average application thickness of 0.79 mg/cm2 (31). This can result in overestimation of protection and, there- fore, overexposure and sunburn (32,33). In general, people do not apply sunscreen very well (34). Correct application of a sunscreen can prevent erythema during a week’s sun holiday with maximal UVI of 9 (18). Sun screens are effective at preventing epidermal CPDs (35), which would be expected because erythema and epidermal CPDs have similar action spectra in the solar UVR range (36). CPDs can be caused by suberythemal exposure, and such DNA damage (as well as erythema) accu- mulates with such exposure daily because epidermal CPDs have a half-life of 33 h (i.e., slow repair) (37). Daily application for 11 d of a low SPF sunscreen, before suberythemal SSR expo- sure, was shown to be effective at reducing cumulative CPDs and erythema (38). A more re- cent study, which detected CPDs by immunostaining and quantitative high pressure liquid chromatography with tandem mass spectrometry, has shown that an SPF 50+ sunscreen was very effective at preventing DNA damage by very high doses of acute and repeated (fi ve con- secutive days) SSR exposure (39). This study also showed that DNA protection, in acute and repeat SSR exposure cases, was dependent on sunscreen application thickness. Signifi cant pro- tection was observed even with application at 0.75 mg/cm2 (equivalent to an SPF of 21), typical of consumer use. Studies have not been designed to determine a DNA protection factor (DNA-PF). A new approach has been recently reported in which an SPF 50+ sunscreen had a DNA-PF of 98.2 [95% confi dence interval (CI) of 51.6–187.2] (40). The re is considerable evidence that the CPD plays a major role in skin cancer, especially keratinocyte cancers (21). Thus, they would be expected to have a benefi cial effect against human skin cancer as has been reported in studies on mouse models (41). Studies show that sunscreen use can inhibit skin cancer (42). The best prospective evidence is from long-term intervention studies in Nambour, Queensland, Australia (discussed in the ar- ticle by A. Green in the same issue). The most robust conclusion from these studies is that a 4.5-y randomized controlled intervention with an SPF 16 product signifi cantly reduced the incidence of melanoma (43) and squamous cell carcinoma (SCC), but not BCC (44). This reduction is very important clinically, but less so numerically for example, the rate

SKIN PHOTOPROTECTION BY PIGMENTATION AND SUNSCREENS 221 ratio (intervention sunscreen vs. discretionary use sunscreen) for SCC in trial and total follow-up (1993–2004) of 0.65 (95% CI 0.45–094), representing a cancer protection fac- tor of ~1.5. In this context, it should be remembered that the real SPF of the product was likely to have been much less than 16 because of the way people typically apply sun- screens. Furthermore, given that the mean age of entry into the study was about 50 y, this rural population will have had a high level of baseline UVR-induced DNA mutations in keratinocytes. It is probable that one of the reasons for the benefi ts of long-term sunscreen use in this study was the immunoprotective effects of sunscreens (45). Photoimmunosuppres- sion is thought to play an important role in skin cancer (46). There is considerable mouse evidence that the CPD initiates the antigen-specifi c cutaneous immunosuppression when a given antigen is presented to the skin after UVR exposure (47). Interestingly, organ transplant patients on drug immunotherapy, which is not antigen-specifi c, are very prone to skin cancer, especially SCC (48). It would be reasonable to suppose that earlier sunscreen intervention, especially in areas of high insolation, would have greater long-term benefi ts. VIT AMIN D SYNTHESIS Vit amin D is important for bone development and maintenance, but there is increasing, though controversial, evidence that vitamin D has many other health benefi ts (49). It is well documented that people with pigmented skins have less good vitamin status than those, at comparable latitudes, with lighter skins. This is usually attributed to an inhibi- tory effect of melanin. Laboratory research on the effect of melanin on vitamin D synthe- sis has given confl icting results (50). A recent laboratory study on FST II–VI has shown that melanin has a very modest effect on vitamin D synthesis. Comparisons of FST II versus VI showed that the melanin inhibition factor was 1.5 (51), but this may be suf- fi cient to explain the epidemiological data. It is also likely that cultural and behavioral factors may also explain differences in vitamin D status in different ethnic groups. The most likely reasons for the differences with the DNA protection data is the spatial rela- tionship between melanin and the UVR target. In the basal layer, nuclei are largely under the melanin, whereas the precursor for vitamin D is throughout the epidermis with high concentrations in the upper less melanized epidermal layers. Two recent reviews concluded that sunscreen use has little or no impact on vitamin D synthesis (52,53). A weeklong fi eld study in Tenerife, under a cloudless sky with a maxi- mum UVI of 9, showed a high level of vitamin D synthesis even when optimal applica- tion of an SPF 15 sunscreen ( 2.0 mg/cm2) prevented erythema (54). This intervention group was in contrast to a discretionary sunscreen use group that had a higher level of vitamin D synthesis but presented with sunburn on multiple body sites. Thus, sunscreens may inhibit vitamin D synthesis, but they still allow considerable synthesis with suber- ythemal exposure. It should be noted that we lack data on the effect of a high SPF sun- screen, especially in temperate climates (55). Sunscreens can enable vitamin D synthesis because the threshold UVR dose for this process is much lower than that for erythema. CON CLUSIONS Con stitutive melanin in deeply pigmented skin is very effective at preventing basal layer DNA photodamage, erythema, and skin cancer. Protection against DNA damage and