Objective: To evaluate the prediction that the observed increasing trend in male bladder cancer mortality in the Netherlands between 1955 and 1988 would change to a decreasing trend.
Materials and methods: Using demographic and mortality data from the Netherlands Central Bureau of Statistics, male and female bladder cancer mortality per 10(5) person-years from 1955 to 1994 were calculated. Changes in the age structure of the Dutch population were adjusted using direct standardization to the European "standard' population. The effects of age, calendar period and birth cohort on the temporal trend in mortality were evaluated using log-linear modelling.
Results: Male mortality from bladder cancer increased from 6.2 per 10(5) in 1955 to 12.6 per 10(5) in 1990, but decreased thereafter. Female mortality from bladder cancer has remained stable throughout the study period, at 2.8 and 2.7 per 10(5) in 1955 and 1994, respectively. The changing trend among men is consistent with an increasing risk for successive birth cohorts until the 1910 cohort and a decreasing risk for cohorts born after 1930. This cohort effect is probably caused by changes in the smoking behaviour of the male Dutch population. There were no differences in risk for successive birth cohorts of females.
Conclusion: Male mortality from bladder cancer in the Netherlands will probably decrease for at least another decade. A future increase in female mortality from bladder cancer, as recorded for lung cancer, is unlikely.