Retrospective cohort study of smoking and lung cancer incidence in rural prefecture, Japan.

OBJECTIVES We conducted an epidemiological study of the relationship between lung cancer incidence and smoking, with special reference to the benefits of smoking cessation for reducing lung cancer incidence, to promote a local smoking control program. METHODS The study was a retrospective cohort study. The population studied was 16,383 male examinees of lung cancer health examinations in 1995 in Tottori Prefecture, Japan. Smoking status from the questionnaire during the health examination was used as the exposure variable. Endpoint (lung cancer incidence) was obtained from the Tottori population-based cancer registry. A multivariable analysis using the Cox proportional hazard model was adapted for statistical analysis. The average follow-up period was 4.3 years. RESULTS The hazard ratio of current smokers for the incidence of lung cancer was 4.9, whereas that of ex-smokers was 2.2. The dose-response relationship between lung cancer incidence and lifetime cigarette consumption (pack year) was determined. The ratio increased among younger subjects (under 65 years old). The hazard ratio of ex-smokers decreased with years just after quitting smoking, and reached the level of never smokers after 10-19 years from smoking cessation. CONCLUSIONS We reconfirmed that the magnitude of risk estimates of smoking for lung cancer incidence was similar to those of previous studies, and smoking cessation was effective for reducing lung cancer risk.