Takayuki Seino1, Hiroto Nakadaira, Kazuo Endoh, Masaharu Yamamoto. 1. Department of Community Preventive Medicine, Niigata University Graduate School of Medical and Dental Sciences, 757 Asahimachi-dori, Ichi-bancho, Niigata, Japan. setaka@msj.biglobe.ne.jp
Abstract
OBJECTIVES: The 5-year survival rate for pancreatic cancer is known to be lower than that for cancer at any other site in the body, and the proportion of pancreatic cancer deaths among all cancer deaths has been increasing in Japan. The aim of this study was to investigate pancreatic cancer mortality in the light of temporal and geographical trends in the 47 prefectures of the country between 1968 and 2002. METHODS: To survey the geographical aspects of pancreatic cancer mortality, we mapped the direct age-adjusted mortality rates of persons aged 40 years and older by sex in seven 5-year periods (1968-1972 to 1998-2002). We also evaluated the changes in period and birth cohort trends using estimable functions based on the age-period-cohort models in each prefecture. RESULTS: During the observation period the Hokkaido and Tohoku regions had high mortality rates for both sexes. No significant increase in period trends was observed from 1973 to 2002, but significant increases in cohort trends were observed from 1913 to 1962-in two prefectures, for males, and in four prefectures, for females. CONCLUSIONS: The results of this study reveal a combination of time trends in pancreatic cancer mortality and changes in period or birth cohort trends. The changes in cohort trends in each prefecture were more variable than the period trends. This finding probably indicates the need for further investigation of the cohort-related factors involved in the prevalence of pancreatic cancer. Further research on mortality in the 47 prefectures needs to be conducted while taking the two time effects into account.
OBJECTIVES: The 5-year survival rate for pancreatic cancer is known to be lower than that for cancer at any other site in the body, and the proportion of pancreatic cancer deaths among all cancer deaths has been increasing in Japan. The aim of this study was to investigate pancreatic cancer mortality in the light of temporal and geographical trends in the 47 prefectures of the country between 1968 and 2002. METHODS: To survey the geographical aspects of pancreatic cancer mortality, we mapped the direct age-adjusted mortality rates of persons aged 40 years and older by sex in seven 5-year periods (1968-1972 to 1998-2002). We also evaluated the changes in period and birth cohort trends using estimable functions based on the age-period-cohort models in each prefecture. RESULTS: During the observation period the Hokkaido and Tohoku regions had high mortality rates for both sexes. No significant increase in period trends was observed from 1973 to 2002, but significant increases in cohort trends were observed from 1913 to 1962-in two prefectures, for males, and in four prefectures, for females. CONCLUSIONS: The results of this study reveal a combination of time trends in pancreatic cancer mortality and changes in period or birth cohort trends. The changes in cohort trends in each prefecture were more variable than the period trends. This finding probably indicates the need for further investigation of the cohort-related factors involved in the prevalence of pancreatic cancer. Further research on mortality in the 47 prefectures needs to be conducted while taking the two time effects into account.