抄録
Objectives: We aimed to determine the effects of age, period, and birth cohort on cervical cancer mortality rate trends in Japanese women, by age-period-cohort (APC) analysis. Additionally, we analyzed projected mortality rates. Methods: We obtained data on the number of cervical cancer deaths in Japanese women from 1975–2011 from the national vital statistics and census population data. A cohort table of mortality rate data was analyzed on the basis of a Bayesian APC model. We also projected the mortality rates for the 2012–2031 period. Results: The period effect was relatively limited, compared with the age and cohort effects. The age effect increased suddenly from 25–29 to 45–49 years of age and gently increased thereafter. An analysis of the cohort effect on mortality rate trends revealed a steep decreasing slope for birth cohorts born from 1908–1940 and a subsequent sudden increase after 1945. The mortality rate projections indicated increasing trends from 40 to 74 years of age until the year 2031. Conclusions: The age effect increased from 25–29 years of age. This could be attributable to the high human papilloma virus (HPV) infection risk and the low cervical cancer screening rate. The cohort effect changed from decreasing to increasing after the early 1940s. This might be attributable to the spread of cervical cancer screening and treatment before 1940 and the high HPV infection risk and reduced cervical cancer screening rate after 1945. The projected mortality rate indicated an increasing trend until the year 2031.
