RESPONSE: more about: sunscreen use, wearing clothes, and number of nevi in 6- to 7-year-Old european children

The recent article by Autier et al. (1) and the accompanying editorial (2) have heightened interest in the relationship between sunscreen use and the incidence and prevalence of melanocytic lesions, including nevi and cutaneous melanoma. In view of the concern about sunscreens and melanoma risk (3,4), we have analyzed data from the Western Canada Melanoma Study. This study is one of the largest and most detailed studies of melanoma, and it demonstrated clear associations with pigmentation characteristics and tendency to sunburn (skin type), as well as with recreational and occupational sun exposure (5,6). It compared incident cases of melanoma diagnosed from 1979 through 1981 with population-based controls, using a home interview. Subjects were asked, “When exposed to sun on your skin, other than face and arms, do you use a suntan or anti-sunburn lotion or cream almost always, sometimes, only for the first few hours, or almost never?” Only three control subjects and no case patients responded “almost never.” Before this analysis, we hypothesized that if sunscreens directly increase melanoma risk, risks will be highest in the “almost always” users, whereas if sunscreens are used to permit more intense sun exposure, risks will be highest in the “only for the first few hours” category. The risks were assessed as odds ratios for melanoma of the trunk or lower limb (369 case–control pairs), after adjustment for host factors and skin type, and for measures of sun exposure, using multiple logistic regression. As shown in Table 1, compared with those using sunscreens “sometimes,” which we interpret as the lowest degree of use, those reporting use “almost always” had an unchanged risk of melanoma, both in simple analysis and after adjustment for host factors (hair, eye, skin color, and skin type) and for measures of sun exposure. In contrast, those who reported use of sunscreens “only for the first few hours” had a statistically significantly increased risk of melanoma that was only slightly lowered and remained statistically significant, after adjustment for host factors and for sun exposure. Analyses of men and women separately showed very similar results, the adjusted odds ratios for sunscreen use only for the first few hours being 1.68 in men and 1.70 in women. The results for the “almost always” users argue strongly against any direct increased risk of sunscreen use. The results for those who use sunscreens only for the first few hours suggest that such use provides inadequate protection against the increased risks due to sun exposure or host characteristics. This study was done before extensive publicity on the risks of sun exposure, which makes response bias less likely, although the types and usage patterns of sunscreen use may differ from current