176 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS RESULTS The results of this study (Table I) show that in the case of the 4% PABA standard sunscreen the SPF values obtained differed significantly between the two filter configurations (P 0.005 by paired student "t" test). A higher SPF value was obtained for the 4% PABA standard sunscreen when tested with the WG-305 filter, i.e., when the incident UV spectrum contained shorter wavelengths of UV-B radiation. The SPF value obtained with the WG-305 filter configuration was not significantly different from the SPF value obtained with that same filter configuration in 1977. Table I Results of SPF Value Determinations SPF Values Obtained (Mean + s.d.) 1977 Study WG-305, 2 mm WG-320, 1 mm Sunscreen Tested (n = 13) (n = 12) • (n = 12) • 4% PABA Standard 6.51 _+ 1.04 5.72 _+ 0.99 4.56 + 1.12 8% HMS Standard 3.75 -+ 0.62 3.48 _+ 0.56 3.70 + 0.76 •Both filters were tested on the same individual for each of the standard sunscreens in the present study. In the case of the 8% HMS standard sunscreen, the SPF values obtained for the two filter configurations did not differ significantly nor did either result differ significantly from SPF values obtained for the 8% HMS standard suncreen formulation in the 1977 study. DISCUSSION Human erythemal sensitivity to UV radiation increases with decreasing wavelength through the UV-B region (3-6). The short wavelength UV-B content of natural sunlight reaching the earth's surface varies considerably with latitude, altitude, time of day, and with atmospheric conditions (7,8), thus risk to sunburn varies with these conditions (9). Our results suggest that the protectiveness of sunscreen formulations may also vary with such changes in the solar spectrum. The SPF value for the 4% PABA standard formulation tested increased when the amount of incident short wavelength UV-B was increased. Kaidbey and Kligman (10) tested formulations containing PABA or PABA esters and reported differences of the same nature but greater magnitude when efficacy values obtained using a bank of fluorescent suniamps were compared with values obtained using a xenon arc solar simulator with a 2-mm WG-320 cut-off filter. The spectral output of fluorescent .suniamps extends to shorter wavelengths than does the solar spectrum however, and does not conform to the proposed FDA specifications for solar simulators. This effect of increasing protectiveness probably does not occur with all sunscreening agents. Indeed, the HMS formulation's SPF value decreased slightly, but not significantly, when determined using the WG-305 filter configuration which passed the greater amount of short wavelength UV-B light. We have not yet tested other sunscreening agents in this manner. In the interest of increasing the accuracy of product labeling and thus the protection

EFFECTIVENESS OF SUNSCREENS 177 of the consumer, we would recommend that the proposed FDA regulations regarding solar simulators more narrowly specify the wavelengths of UV light to be included. Further, in the interest of insuring the greatest protection from sunburn and skin damage, we would recommend that the spectral conditions simulated in sunscreen product effectiveness testing should approximate the solar spectral conditions yielding the shortest wavelengths of UV light which might commonly be encountered by individuals requiring sunscreen protection. Our results also suggest that similar problems of results differing due to different spectra of natural sunlight at different testing localities, times of day, or times of year may be encountered during outdoor testing of sunscreen efficacy. The rules for outdoor testing of sunscreens may also need reexamination. REFERENCES 1. Sunscreen Drug Products for Over-The-Counter Human Use, Federal Register, 43, 38206-38269 (1978). 2. D. S. Berger, Specification and design of solar ultraviolet simulators, J. Invest. Dermatol., 53,192-199 (1969). 3. D. S. Berger, F. Urbach and R. E. Davies, The action spectrum of erythema induced by ultraviolet radiation, in Proc. 13th Conf Int. Dermatologiae, Munchen, 1967, Springer-Verlag: Berlin, 1968 pp 1112-1117. 4. D.J. Cripps and C. A. Ramsay, Ultraviolet action spectrum with a prism-grating monochromator, Br. J. Derre., 82, 584-592 (1970). 5. M. A. Everett, R. L. Olson and R. M. Sayre, Ultraviolet erythema, Arch. Dermatol., 92, 713-719 (1965). 6. R. G. Freeman, D. W. Owens, J. M. Knox and H. T. Hudson, Relative energy requirements for erythemal response of skin to monochromatic wavelengths of ultraviolet present in the solar spectrum,J. Invest. Dermatol., 47, 586-592 (1966). 7. J. Scotto, T. R. Fears and G. B. Gori, Measurements of Ultraviolet Radiation in the United States and Comparisons with Skin Cancer Data, U.S. Department of Health, Education and Welfare, DHEW No. (NIH) 76-1029 (1975). 8. L. Koller, The physics of the atmosphere, in The Biologic Effects of Ultraviolet Radiation, F. Urbach, Ed., Pergamon Press: N.Y., 1969 pp 329-333. 9. A. E. S. Green, T. Mo. and J. H. Miller, A study of solar erythema radiation doses, Photochem. Photobiol., 20, 473-482 (1974). 10. K. H. Kaidby and A.M. Kligman, Laboratory methods for appraising the efficacy of sunscreens,J. $oc. Cosmet, Chem., 29, 525-536 (1978).