Examination of solar simulators used for the determination of sunscreen UVA efficacy.

The U.S. FDA recently proposed both in vivo and in vitro UVA efficacy tests for sunscreen products with the lower result used to establish the sunscreen's labeled UVA protection claim. The FDA stated their rationale for dual tests was concern that the in vivo test method overemphasizes UVA-2 (320-340 nm) photoprotection. We attribute FDA's observation to the relative lack, compared to sunlight, of UVA-1 (340-400 nm) radiation in the current JCIA UVA solar simulator specification, allowing the method to generate higher UVA protection factors than sunscreens will provide in sunlight. Our work is based upon comparisons of Air Mass 1.0 sunlight to variously filtered UVA solar simulators. Sources near the JCIA UVA-2/UVA limits (8-20%) had a goodness of fit to solar UVA of only 67-79%. We propose that instead of using ratios of UVA-2 to UVA the standard should be a goodness of fit to the UVA region of an Air Mass 1 solar reference spectrum. As the spectral distribution of solar UVA varies much less than UVB, sunlight of reasonable zenith angles of < or = 60 degrees will have similar spectral shapes and approximate risk spectrum. Goodness of fit to this spectrum will produce UVA protection values predictive to those actually achieved in sunlight of different zenith angles.