Cancer in the very elderly Dutch population.

BACKGROUND Cancer incidence and mortality rates rarely are studied in people age > 85 years. Usually, patients ages 65 years, 75 years, and 85 years of age are combined into 1 group because of small numbers. The number of people age > or = 85 years in the Netherlands increased from 99,000 in 1976 to 203,000 in 1995 (an increase of 105%). The growth of the total population in this period was only 13%. This study addressed cancer incidence and mortality rates among the very elderly in the Netherlands. METHODS Cancer mortality data (1976-1995) and population data were obtained from Statistics Netherlands, whereas cancer incidence data (1989-1995) were provided by the Netherlands Cancer Registry. Cancer incidence and mortality rates were calculated and trends in cancer mortality were studied. RESULTS Total cancer incidence rates were highest in the age group 85-94 years, in men and women (3466/100,000 person-years and 1604/100,000 person-years, respectively). Prostate carcinoma was the most frequent cancer in men ages 85-94 years, followed by colorectal carcinoma. In women ages 85-94 years, colorectal carcinoma was most frequent, closely followed by breast carcinoma. In the 95+ years age group squamous cell skin carcinoma was the most frequent cancer in both men and women, followed by prostate carcinoma in men and breast carcinoma in women. Cancer mortality rates increased with increasing age to nearly 3700/100,000 person-years in men age 95+ years and 2500/100,000 person-years in women age 95+ years. In men, lung carcinoma was the most frequent cancer-related cause of death in patients age < or = 85 years, whereas in older men this applied to prostate carcinoma. In women, breast carcinoma was the most frequent cancer-related cause of death in all age groups > 55 years. Cancer as a cause of death became less prominent with increasing age. Over the period 1991-1995, 42% of deaths in men ages 55-64 years were attributed to cancer versus 52% of deaths in women (total population); these proportions in the 95+ years age group were 11% and 7%, respectively. CONCLUSIONS Peak incidence rates of major cancers were found in the very elderly population in the Netherlands. Different trends in age specific mortality rates of individual cancer sites were found, with stable rates in the middle age groups and increasing rates in the oldest age groups. This may reflect a real increase caused for instance by changes in mortality from other diseases and/or an artifactual increase caused by increased cancer detection rates in the (very) elderly.