Background: Reasons underlying time changes in cutaneous melanoma mortality in light-skinned populations are not well understood. An analysis of long-term time trends in melanoma mortality was carried out after regrouping countries in homogeneous regions.
Methods: Using the World Health Organisation (WHO) mortality database, age-period-cohort models were fitted for seven regions where the majority of population is light-skinned. Cohort effects are denoted as changes in rates occurring at different times in steadily older age groups. Period effects are denoted as changes in rates occurring simultaneously in several age groups.
Results: Cohort effects better explained changes in melanoma mortality over time than period effects. Lifetime risk to die from melanoma increased in successive generations from 1875 until a peak year. Peak years were for subjects born in 1936-1940 in Oceania, 1937-1943 in North America, 1941-1942 in Northern Europe, 1945-1953 in the United Kingdom (UK) and Ireland, 1948 in Western Europe and 1957 in Central Europe. After peak years, lifetime risk of melanoma death gradually decreased in successive generations and risks of subjects born in 1990-1995 were back to risk levels observed for subjects born before 1900-1905. In Southern Europe, birth years with highest lifetime risk of melanoma death have not yet been attained. As time passes, melanoma deaths will steadily rarefy in younger age groups and concentrate in older age groups, for ultimately fade away after 2040-2050.
Conclusion: Independently from screening or treatment, over next decades, death from melanoma is likely to become an increasingly rare event. The temporary epidemic of fatal melanoma was most probably due to excessive UV-exposure of children that prevailed in 1900-1960, and mortality decreases would be due to progressive reductions in UV-exposure of children over the last decades.
Keywords: Age–period–cohort; Melanoma; Mortality; Time trends; Ultraviolet.
Copyright © 2015 Elsevier Ltd. All rights reserved.