Association may be iatrogenic

incidence of both melanoma and non-melanoma skin cancer in people with pigmented skin is several times lower than that in people with less pigmented skin.2How do the authors explain an apparent lack of association of lymphoma and leukaemia with skin colour?The authors' reasoning is based on two assump- tions: that the incidence of skin cancer can be used as a marker of exposure to ultraviolet radiation and that immunosuppression has a dominant role in the pathogenesis of cancer induced by ultraviolet radiation.Both assumptions may be false.Al- though ultraviolet radiation has an unequivocal role in the development of skin cancer, the patho- genesis of skin neoplasms is multifactorial.This is shown by the fact that exposure to the sun increases the relative risk of melanoma by only a factor of two to five, which is not impressive in comparison with the 85-1269 times higher risk in people with two family members with melanoma.'Moreover, recent evidence suggests that, in the pathogenesis of skin cancer, mechanisms other than immunosuppression are relevant, such as mutagenesis and suppression of apoptosis induced by ultraviolet radiation.Factors other than expo- sure to ultraviolet radiation-for example, genetic predisposition-probably better explain the association between skin cancer and the occurrence oflymphoma and leukaemia.