|| Checking for direct PDF access through Ovid
Sunscreens have been advocated to prevent burning in the hope that this will decrease the chance of developing melanoma. In a single-centre case-control study in Styria, Austria, we examined the risk of cutaneous malignant melanoma in relation to phenotypic markers, sunlightrelated factors and sunscreen use. In total, 193 melanoma patients and 319 control subjects answered a comprehensive questionnaire regarding phenotypic markers, a variety of sunlight-related factors and sunscreen use. Risk factors for melanoma were examined through the use of unconditional logistic regression analysis, controlling for age and sex. Screening for confounding factors was done by forward and backward elimination of non-significant variables (P<0.05). The resulting set of factors were investigated further for effect modification by introducing interactions into the model. The factor most significantly associated with increased melanoma risk was the use of sunscreens. Subjects who often used sunscreens had an increased odds ratio (OR) of 3.47 (95% confidence interval [CI]1.81-6.64) compared with subjects who never used sunscreens (P=0.001), after adjustment for sex, age and other significant sunlight-related factors. Skin colour and higher numbers of sunbaths were significant protective factors. Subjects with medium skin colour had an adjusted OR of 0.63 (95% Cl 0.41-0.99) compared with subjects with light skin colour (P=0.0022). Subjects who took more than 30 sunbaths per year and subjects who took 20 30 sunbaths per year had, in the absence of sunburn(s), a decreased OR of 0.09 (95% Cl 0.02-0.39) and 0.28 (95% Cl 0.13-0.64), respectively, compared with subjects who took less than 20 sunbaths per year (P=0.0002). However, sunbaths had no protective value when they were associated with sunburns. Although we cannot exclude the presence of an unknown confounding factor, our results suggest that the use of sunscreens does not help prevent melanoma.