PAPER www.rsc.org/pps | Photochemical & Photobiological Sciences The FDA proposed solar simulator versus sunlight Robert M. Sayre a,b and John C. Dowdy b Received 29th October 2009, Accepted 8th February 2010 First published as an Advance Article on the web 1st March 2010 DOI: 10.1039/b9pp00148d The US Food and Drug Administration is in the process of formulating final rules for sunscreen labeling and testing. They have adopted a version of the solar simulator standard proposed by COLIPA, a European cosmetic products trade association. From our files we have selected spectral data on several solar simulators that comply with the proposed rules and have compared these sources both one to another and to several standard solar spectra of Air Mass 1.0, 1.5, and 2.0. In doing so we have used additional spectral analysis procedures including examining the goodness of fit between each solar simulator spectrum and an Air Mass 1.0 (0 ◦ zenith angle) solar spectrum. The index of goodness of fit ranges from ~78% to just over 90% compared to solar spectra representing other Air Masses of 1.5 and 2.0, the goodness of fit is lower. Unfortunately, one may not assume that complying with a standard assures that other solar simulators also complying will produce identical results. In fact, by our analysis, none of the solar simulators we examined would be expected to produce the same SPF as sunlight. Introduction The US Food and Drug Administration (FDA) began developing rules for labeling and testing the sun protection factor (SPF) of sunscreen products with the issuance of a report and proposed standard in 1978. 1 As of today, while several tentatively-final and a final monograph have been issued and withdrawn, a final rule has yet to be established. In the first sunscreen monograph, the solar simulator was described simply as having an ultraviolet (UV) spectrum “similar” to sunlight with a 10 ◦ zenith angle. The FDA solar simulator evaluated in this study is spectrally described in the 2007 proposed rule. 2,3 It is similar to the solar simulator described by the initial COLIPA method 4 issued in 1994. Both spectral specifications have adopted the use of percent erythemal contribution or percentage relative cumulative erythemal effectiveness (%RCEE) to evaluate the suitability of the of the solar simulator for testing. Both the FDA and COLIPA evaluate the %RCEE at 310, 320, 330, 340 nm and at 400 nm where by definition it is 1.0 or 100%. The FDA includes a single check point in the UVA-1, an integration from 290 nm to 350, whereas the 2006 COLIPA-International Harmonization solar simulator specification 5 includes a short wavelength 290 to 300 nm additional UVB check point. In essence, both undertake to ensure that the short wavelength UV is relatively similar to a standard solar spectrum but longer UVA wavelengths are not considered. The solar simulators used to test sunscreens today still use essentially 1970 technology when the FDA OTC review first began. In fact the spectral emissions are little changed since the first published description. 6 While differences in solar simulator optical filtration have been shown to potentially cause up to 50% differences in SPF test results, 7 neither the FDA nor the sunscreen industry seem to have considered any need to modernize the solar simulator. With the a Division of Dermatology, Department of Medicine, University of Tennessee Health Sciences Center, Memphis, TN, 38103, USA. E-mail: rptl@aol.com b Rapid Precision Testing Laboratories, Cordova, TN, 38016, USA published need and sales promotions of sunscreens containing UVA protection and regulatory urging of the need for highly protective sunscreens, we are examining the consequences of the failure to require similarity of solar simulator spectra to actual solar spectra. In 2004 Schauberger et al. published a useful statistical approach to compare UV sources, including solar simulators, to a reference standard solar spectrum. 8 In their investigation they echo the use of “cumulated relative spectral effective irradiance” extending the method used to specify the UV spectrum of solar simulators in the 1994 COLIPA method for sun protection factor (SPF) testing. 4 Accompanying this they also have proposed statistical controls for determining compliance, such as the goodness of fit index. 8 In this work we have applied this approach to evaluate the spectra of solar simulators used in clinical SPF testing relative to the UV occurring in terrestrial sunlight. Methods Solar simulators Three solar simulators that passed the FDA and COLIPA- International Harmonization standards are used in this study, designated as A, B and C. Each solar simulator spectrum was measured from 250 to 800 nm at 1 nm increments with either OL- 754 or OL-756 spectroradiometers using ~1 nm band pass slits calibrated against a NIST traceable quartz halogen standard. The fourth solar simulator spectrum, A¢, is A shifted 3 nm to the red. This solar simulator spectrum would meet the US FDA standard, but does not meet the more rigorous UVB COLIPA-International Harmonization Standard. Solar spectra The three solar spectra Air Mass 1.0, 1.5, and 2.0 are described fully in CIE Technical Report #85 9 and were extrapolated through shorter wavelengths an additional ~1 decade of irradiance to This journal is © The Royal Society of Chemistry and Owner Societies 2010 Photochem. Photobiol. Sci., 2010, 9, 535–539 | 535