[The effects of various factors on cerebrovascular disease mortality rates in the 20th century and future trends in Japan].

PURPOSES

To analyze the outcomes of measures designed to decrease cerebrovascular diseases (CVDs) in Japan and to project CVD mortality trends into the 21st century based on an analysis of rates observed in the 20th century.

METHODS

The numbers of CVD deaths and population sizes from 1920 to 2003 (excluding 1940 to 1946) by sex, year, and 5-year age group (from 20 to 79 years old) were used and effects of various factors on CVD mortality rates were estimated using Nakamura's Bayesian age-period-cohort model. The numbers of CVD deaths up to the year 2050 were projected based on estimates of age, cohort, and future period effects under three scenarios: (i) values remaining constant after year 2003; (ii) linearly extrapolated values; and (iii) quadratically extrapolated values, we obtained using a regression line for period effects from 1995 to 2003.

RESULTS

The age, cohort, and period effects on CVD mortality rates were large and in order of the magnitude of their ranges. There were small differences between males and females. The age effect increased with aging and the period effect started decreasing after 1970. The cohort effect was high for birth cohorts born from the 1840s to the 1890s and low for those born from the 1920s to the 1970s. There were some differences in the cohort effect between males and females for birth cohorts born after 1940s; for females there was a gradual decrease, while for males there was a slight increase, after which it remained almost constant. According to the three scenarios, CVD deaths: (i) had upward trends through the projected period and peaked at around 2025 and 2045; (ii) remained almost constant at the present level for males, and decreased slightly for females; (iii) decreased for both males and females.

CONCLUSIONS

The outcomes of measures designed to decrease CVDs were observed as period effects after 1970. Exposure to these measures is associated with prevention of CVD deaths. Nevertheless, in the first half of the 21st century, the number of CVD deaths is projected to increase due to the aging of the baby boomers and upward trends in the cohort effect for males. It would be necessary to adopt and develop both population strategies to decrease future period effects and high-risk strategies to decrease cohort effects for younger males who are currently in their twenties and thirties.