Background: Site-specific trend analysis is probably the most effective method available for assessing how the long-term trend in melanoma rates relates to changes in sun exposure and behaviour. New Zealand has very high incidence of and mortality from melanoma and the fraction of melanoma cases and deaths with a site specified has been comparatively high.
Methods: Trends in incidence and mortality from melanoma in New Zealand were analysed between 1969 and 1993, by sex and body site. A graphical representation of the trend by birth-cohort and age-period-cohort modelling were used.
Results: For all sites combined, the annual increase in incidence was 6.7% (95% CI : 6.3-7.1%) in men and 3.1% (95% CI : 2.3-3.7%) in women. The increase was significantly greater at each site for males. The largest increases occurred for the upper limbs in males (7.3% a year) and the trunk in females (3.8% a year). Incidence rates slowed appreciably in the later years (currently about 26/100 000 for each sex) and no further increase in lifetime risk of melanoma was observed among post World War II generations. Mortality trends paralleled those for incidence with a 25-year gap, with a more modest rate of increase (2-3% per annum for each sex), essentially due to the increased risk among generations born up to 1919 or 1924. Age-standardized death rates have now stabilized in New Zealand at about 5.5/100 000 (men) and 3.2/100 000 (women). Trends between cohorts were the most marked for sites with a likely intermittent pattern of exposure, and were consistent overall for the trunk and the limbs.
Conclusions: Results support the hypothesis that changes in lifestyle factors resulted in a pattern of carcinogenic exposures that explains both the upsurge in melanoma in the last few decades and the current levelling off in incidence.