A central issue raised by Autier and Doré is that epidemiological studies that purport to show that sunscreen use is a risk factor for melanoma have not adequately controlled for confounding variables. Fair-skinned individuals prone to sunburn will be heavily overrepresented in any random group of sunscreen users.1,2 Therefore, studies that do not adequately account for sun protection factor value, frequency of application, duration of sunscreen use, past and present sun exposure, and susceptibility to sun damage are not measuring sunscreen effect alone. Sunscreen use then becomes a surrogate for confounding factors associated with increased melanoma risk. Although the EORTC study3 attempted to control for susceptibility to sun damage, and to some extent, sun exposure, it did not take into account the sun protection factor value or the frequency of sunscreen use. Note that "regular" sunscreen in the "ever regular sunscreen only" group referred to sunscreens that do not contain psoralens, not use of sunscreens on a regular basis.
Naylor M, Farmer K. The Case for Sunscreens Revisited—Reply. Arch Dermatol. 1998;134(4):510-511. doi: