JOURNAL OF COSMETIC SCIENCE 194 terms of both people newly affected (RR = 0.65 95% CI 0.45–0.94) and new SCC tumor development (RR = 0.62 95% CI 0.38–0.90) in those allocated to daily sunscreen versus discretionary users (9). MELANOMA The effect of regular application of sunscreen in the trial on subsequent occurrence of primary melanoma was evaluated 15 years post-trial. Diagnostic pathology slides of all fi rst primary melanomas (in situ and invasive) that occurred (confi rmed in the Queensland Cancer Registry) in participants between 1993 and 2006 were obtained and were re- viewed by two expert dermatopathologists who were unaware of sunscreen allocation status of the 33 persons affected (19 in situ, 14 invasive, and 0 metastatic). A borderline- signifi cant 50% reduction in risk of melanoma across all sites was observed in people who had been randomly assigned to the sunscreen intervention compared with controls [hazard ratio (HR) = 0.50 95% CI 0.24–1.02], with the estimated risk reduction slightly more on the head, neck, and upper limbs although this was not statistically signifi cant because of the small number of cases involved (HR = 0.46, 95% CI 0.17–1.20) (10). There was no difference between treatment groups in the risk of in situ melanomas (HR = 0.73 95% CI 0.29–1.81), but the risk of invasive melanoma in the intervention group was reduced by more than 70% (HR = 0.27 95% CI = 0.08–0.97) (10). SAFETY In the short-term, some in the intervention group complained of skin irritation (n = 25), skin greasiness (n = 10), and stinging eyes on facial perspiration (n = 6), and most of these were resolved by switching to another sunscreen of similar SPF (8). At the end of the trial, there was no difference in serum vitamin D status between those randomized to daily sunscreen versus discretionary use (14). On long-term follow-up in 2014, there was no difference in deaths in the sunscreen intervention group (n = 160, 59 of which were due to cardiovascular disease) and the control group (n = 170, 76 of which were due to cardio- vascular disease) with an overall HR = 0.94 and 0.77, respectively, for deaths from car- diovascular disease (15). LIMITATIONS Evidence about sunscreen’s ability to protect against skin cancer is limited to this single Nambour Trial conducted in a subtropical population, so there is uncertainty about both the repeatability of the fi ndings and their applicability in populations living in temperate climates. Trial participants in the intervention group applied the sunscreen too thinly on average i.e., at far less that the thickness of 2 mg/cm2 recommended for maximum effec- tiveness (16). In addition, the intervention sunscreen had an SPF 16 rating, whereas many sunscreens today have higher SPF ratings, so the size of protective effects may have been underestimated. Finally, the actual trial period was only 5 years, which is relatively short for the assessment of cancer prevention, and moreover it was conducted among adults, so the effectiveness of a sunscreen intervention earlier in life, carried out for longer, is un- known but potentially greater. The longer term effects of the trial were estimated by in- tention-to-treat analyses (9,10), so although post-trial sunscreen use lessened after the

SUNSCREENS AND SKIN CANCER PREVENTION 195 trial ceased, this would not have introduced bias but would have diluted the strength of the fi ndings. DISCUSSION AND CONCLUSIONS Although there was evidence that sunscreen may prevent the occurrence of multiple BCCs over time, the apparent lack of protective effect of sunscreen use on BCC occurrence overall may be partly explained by the dispersed anatomic site distribution of BCC, where around one-third occur on sites other than the specifi ed trial application sites [although 25% of the intervention group were applying sunscreen regularly to other sites, namely, the trunk and lower limbs, at the end of the trial (16)]. It could also mean that regular sunscreen use for a limited period in adulthood is “too little, too late” to prevent BCC development if BCCs are initiated early in life and are promoted not only by solar UV radiation but by other factors as well. On the other hand, the large reductions in risk of SCC and melanoma with regular sun- screen use are consistent with other trials (17–20) that have shown regular sunscreen use can prevent development of actinic keratosis [the common benign tumors caused by cu- mulative sun exposure that are strongly associated with both SCC and melanoma (21)] and melanocytic naevi [common pigmented lesions that are the strongest known predic- tors of melanoma risk (22)], respectively. The protective effect of sunscreen on skin cancer is also supported by experimental evidence that consistently shows that application of sunscreen prevents DNA damage in human skin (23). In summary, based on the evidence from a single randomized controlled trial conducted in an Australian community, regular sunscreen use by adults does not appear to protect against BCC but does appear to reduce the risk of developing cutaneous SCC and mela- noma, consistent with other fi ndings. Sunscreen use is but one of a suite of sun protection measures, however, and it is important for healthcare agencies and practitioners to en- courage the use of clothing cover and shade in addition to promoting sunscreen (3). Users also need clear instructions on the proper application and reapplication of sunscreen to achieve optimal protection (16). REFERENCES (1) D. Rigel, Cutaneous ultraviolet exposure and its relationship to the development of skin cancer, J. Am. Acad. Dermatol., 58, S129–S132 (2008). (2) A. C. Green, C. M. Olsen, and D. Hunter, “Epidemiology of cancer of the skin,” in A Textbook of Can- cer Epidemiology and Control, 3rd Ed., A. Adami, A. Trichopoulos, and D. Hunter. Eds. (Oxford University Press, New York, NY, 2018), pp. 355–381. (3) M. Janda and A. C. Green, “Primary prevention of skin cancer”, in Evidence-Based Dermatology, 3rd Ed., H. Williams, M. Bigby, A. Herxheimer, L. Naldi, B. Rzany, R. Dellaville, Y. Ran, and F. Furue. Eds. (Wiley-Blackwell, Hoboken, NJ, 2014), pp. 223–230. (4) C. S. Rueegg, J. S. Stenehjem, M. Egger, R. Ghiasvand, E. Cho, E. Lund, E. Weiderpass, A. C. Green, and M. B. Veierod, Challenges in assessing the sunscreen melanoma association, Int. J. Canc., 144, 2651–2668 (2019). (5) A. C. Green and G. M. Williams, Sunscreen use is a safe and effective approach to skin cancer preven- tion, Cancer Epidemiol. Biomark. Prev., 16, 1921–1922 (2007). (6) A. Green, D. Battistutta, V. Hart, D. Leslie, G. Marks, G. Williams, P. Gaffney, P. Parsons, L. Hirst, F. Frost, E. Orrell, K. Durham, and C. Lang The Nambour Prevention Study Group, Nambour skin cancer and