OBJECTIVE: Male smoking prevalence is still high in Japan, and quantitative information for tobacco control is scarce. The aim of the present study was to project cancer mortality among Japanese males under different future scenarios of smoking prevalence. METHODS: The target population comprised Japanese males aged 40-79 years in 2007, whose smoking prevalence was 35%. On the basis of the pooled data from three large-scale cohort studies in Japan, the effects of age, years of smoking and years after smoking cessation on the time to all-cancer or lung cancer death were estimated by an accelerated failure time model. The parameter estimates were used to project the annual number of deaths from all cancers and lung cancer by running simulations for different future scenarios of smoking prevalence. Each scenario was evaluated by the cumulative number of avoided deaths when compared with the status quo and by the percent change (from the baseline year) in age-standardized rate of mortality. RESULTS: Reducing the smoking prevalence from 35% in 2007 to 0% in 2017 was estimated to avoid 333 900 all-cancer deaths and 171 100 lung cancer deaths in 20 years. Even when we shortened the projection period to 10 years, these numbers of avoided deaths would be 81 100 and 38 800, respectively. The age-standardized rate of all-cancer mortality was estimated to decrease by 9.6% in 10 years and 18.1% in 20 years. CONCLUSIONS: Reducing the prevalence of smoking among males would be effective in reducing the cancer burden even within 10 years in countries with a high male smoking prevalence.
OBJECTIVE: Male smoking prevalence is still high in Japan, and quantitative information for tobacco control is scarce. The aim of the present study was to project cancer mortality among Japanese males under different future scenarios of smoking prevalence. METHODS: The target population comprised Japanese males aged 40-79 years in 2007, whose smoking prevalence was 35%. On the basis of the pooled data from three large-scale cohort studies in Japan, the effects of age, years of smoking and years after smoking cessation on the time to all-cancer or lung cancer death were estimated by an accelerated failure time model. The parameter estimates were used to project the annual number of deaths from all cancers and lung cancer by running simulations for different future scenarios of smoking prevalence. Each scenario was evaluated by the cumulative number of avoided deaths when compared with the status quo and by the percent change (from the baseline year) in age-standardized rate of mortality. RESULTS: Reducing the smoking prevalence from 35% in 2007 to 0% in 2017 was estimated to avoid 333 900 all-cancer deaths and 171 100 lung cancer deaths in 20 years. Even when we shortened the projection period to 10 years, these numbers of avoided deaths would be 81 100 and 38 800, respectively. The age-standardized rate of all-cancer mortality was estimated to decrease by 9.6% in 10 years and 18.1% in 20 years. CONCLUSIONS: Reducing the prevalence of smoking among males would be effective in reducing the cancer burden even within 10 years in countries with a high male smoking prevalence.