Increased risk of fatal prostate cancer may explain the rise in mortality in The Netherlands.

Abstract

Several lines of evidence suggest that, as a result of improved diagnostic techniques, the increase in incidence of prostate cancer is due largely to increased detection of subclinical cases. Between 1971 and 1989, a considerable increase in incidence was found in Southeastern Netherlands among men aged under 60 years without an improvement in prognosis. We hypothesized that in addition to the increase due to increased detection, a genuine increase in incidence has occurred in the last two decades and that this should be reflected in national mortality rates. Age-specific and age-adjusted mortality rates were calculated to determine whether mortality due to prostate cancer continued to increase after 1990. Using log-linear Poisson modelling according to Clayton and Schifflers, we estimated the contribution of period and cohort effects to prostate cancer mortality between 1955 and 1994. The age-adjusted mortality increased from 22 in 1955-1959 to 33 per 10(5) in 1990-1994 (European standardized rate). For men under 65, the rates stabilized after 1989. The age-cohort model fitted the data better than the age-period model. Therefore, the increase in mortality can be explained largely by the increasing risk for successive birth cohorts for men born until 1930. However, more frequent reporting of prostate cancer as the underlying cause of death (partly attributable to a decline in competing causes of death) may have occurred as well. Our findings suggest an increased risk of fatal prostate cancer in The Netherlands between 1955 and 1994.