The incidence rate of cutaneous melanoma has been increasing faster than that of any other cancer in white-skinned populations over the past decades. The main risk factors for melanoma (i.e. exposure to sunlight, naevus count, phototype, and family history of melanoma) may not wholly explain the epidemiological trends observed for this cancer. The light-at-night theory postulates that increasing use of artificial light-at-night may contribute to the increasing breast cancer incidence through suppressed secretion of melatonin (a hormone produced in the dark and inhibited by light, which regulates circadian rhythms). Here, we postulate that this theory may also apply to melanoma and that it may explain a part of this cancer burden. Consistent with our hypothesis is evidence from experimental studies suggesting a lightening effect of melatonin on frog skin and mammal hair during seasonal changes, its antioxidant and anti-carcinogenic effects in skin melanocytes, as well as the expression of melatonin receptors in melanocytes. Also, epidemiological data suggest lower melatonin concentrations in melanoma patients compared with controls; a potential therapeutic effect of melatonin in patients with metastatic disease; a higher prevalence of melanoma in pilots and aircrews, with increased risks with higher time zones travelled; and increased melanoma risks in office workers exposed to fluorescent lighting. Moreover, melanoma incidence and seasonal patterns are consistent with a reduction of melatonin secretion with intensity of exposure to light, although it remains difficult to distinguish the effect of melatonin disruption from that of sun exposure on the basis of ecological studies. Finally, the reported associations between hormonal factors and melanoma are consistent with melatonin inhibition increasing the risk of melanoma by increasing circulating oestrogen levels. Despite the existing suggestive evidence, the light-at-night hypothesis has never been directly tested for melanoma. Very few studies examined the potential associations between melanoma risk and shift work or melatonin concentrations, and we found no studies reporting on the relationship between melanoma and number of sleeping hours, use of melatonin supplements, blindness, night-time city light levels, bedroom light levels, or clock genes polymorphisms. Therefore, since several observations support our hypothesis and very little research has been undertaken on this subject, we strongly encourage analytic epidemiological studies to test the light-at-night theory for melanoma causation.
(c) 2010 Elsevier Ltd. All rights reserved.