Is sun exposure a major cause of melanoma? Yes

Site and sensitivity

Firstly, anatomical site of low and high sun exposure predicts patterns of melanoma. In general, the relative density of melanoma is highest on body sites receiving more sun exposure in both sexes and lowest on sites receiving little (scalp in women and buttocks in both sexes).1 2 Furthermore, the difference in the patterns of sun exposure between the sexes is consistent with differences in the most common position of melanoma (trunk in men, lower extremities in woman).3

The incidence of melanoma is also much higher in people of races who tend to burn rather than tan.1 The age standardised incidence of melanoma in non-Hispanic white people (populations of mainly European origin) in New Mexico is an order of magnitude greater than that in Hispanic whites, with similar results in non-Hispanic whites versus people of Hispanic, black, and Asian ethnic groups in Los Angeles.1

Within the lightly pigmented populations, studies using skin phototype (colour of non-exposed skin and ability to tan) found a relative risk of melanoma of 3.1 for the lightest quartile and 3.5 for no tanning ability, which is similar to the risk of people with a history of non-melanoma skin cancer. This is consistent with sun exposure being a strong independent predictor of risk of melanoma.1

Association with exposure

Studies looking at melanoma incidence as a function of ambient geographical ultraviolet levels avoid inaccuracies of recall of exposure. When considering race as a variable, ambient ultraviolet index and decreasing latitude were associated with increased incidence of melanoma, but only in non-Hispanic white people in the US and not in other dark skinned races.4 This is consistent with other studies showing incidence of melanoma increasing with decreasing latitude where racial differences (other than populations of mainly European origin) are less observed.5 Such racial differences explain the general increasing incidence of melanoma with decreasing latitudes within countries but not necessarily within continents.6

The increased incidence of melanoma in mid-European countries in people residing in sunny areas, particularly before 10 years of age,7 is consistent with previous migration and geographical residence studies showing that either early exposure or longer exposure in an environment with high ambient solar radiation leads to an increased risk of melanoma.8 Finally, increasing evidence suggests that the incidence of many cancers is inversely related to ambient solar ultraviolet B radiation exposure. However, the pattern for melanoma is reversed, with a positive association between solar ultraviolet B exposure and incidence.9

Case-control studies confirm intermittent sun exposure and sunburn as risk factors for melanoma. Two meta-analyses of case-control studies found that chronic exposure was either negatively associated (odds ratio=0.86)10or not associated.11 In contrast, there were positive associations with intermittent exposure, lifetime sunburn, and childhood sunburn.10 11 Although studies in children are unclear whether total or intermittent exposure conveys risk of melanoma,8 all studies published since the last meta-analysis support the positive association of sun exposure and melanoma risk.12 13 14 15 These studies lack objective measures of exposure, which may reduce the association between personal exposure and melanoma.16 In contrast, studies attempting to measure total sun exposure by cutaneous microtopography show a significant association with melanoma for high grade solar damage, freckling as a child, and history of solar keratoses.1

Epidemiological and mutational analyses strongly support at least two divergent pathways to induce melanoma: those induced by chronic exposure (preferential head and neck site, associated with a history of non-melanoma skin cancer, no mutation in the BRAF gene (which controls the proliferation of melanocytes) and fewer naevus counts) and those induced by intermittent exposure (related to naevus density, BRAF mutation, and a preferential trunk site).17 18 Epidemiological studies may produce confusing results if these divergent pathways are not taken into account.

Genetic evidence

In young adults with xeroderma pigmentosum, who have a defect in the repair mechanism of ultraviolet radiation induced thymidine dimers, the incidence of melanoma is 1000 times higher than in controls, although the anatomical distribution is the same.19

Analysis of melanoma mutations found in the suppressor oncogene CDKN2A (P16/INK4a) and ras oncogene family is consistent with induction by ultraviolet radiation.20 Evidence suggests that although ultraviolet radiation is required to induce BRAF mutations found in melanoma and acquired naevi, other factors are also necessary.21 22

Protection

Overall, case-control studies have not shown a reduction in the incidence of melanoma with sunscreen use.23 24 However, sunscreens are used to prolong intentional sun exposure,25 they tend to be used by people at higher risk of melanoma, and the studies did not assess the sun protection factor or correct application of sunscreens. Nevertheless, the incidence of melanoma among young adults in Australia fell from 1983 to 1996, coinciding with strong public health messages to use sun protection.1

In white populations of mainly European origin there is evidence of clear associations between sun exposure and melanoma incidence. A comparison of US white people and black people, in which the melanoma incidence of black people was taken as the non-exposed incidence in white people , gave an estimate that 96% of melanoma in men and 92% in women was caused by sun exposure.26 Sun exposure is clearly a major cause of this disease.