BACKGROUND: The rate of decline in lung cancer risk after smoking cessation among male population and the importance of the magnitude of the early decline were not sufficiently defined in the earlier studies. We evaluated the detailed duration-response relationship between years since smoking cessation and lung cancer risk across major histological types in a population-based case-referent study. METHODS: We recruited 1208 consecutive incident cases of primary lung cancer among Chinese males from the largest oncology center in Hong Kong during 2004-2006, and 1069 male community referents frequency-matched in 5-year age groups. We performed unconditional multiple logistic regression and generalized additive model incorporating smoothing spline to model the potential nonlinear effect of years since cessation on lung cancer. RESULTS: All histological types of lung cancer were strongly associated with current smoking. We observed a rapidly decreasing odds ratio of lung cancer (>50%) across all major histological types of lung cancer (except for the large cell type) within the first 5 years of quitting; the odds ratio continued to decrease but at a slower rate in the subsequent years. CONCLUSION: The substantial benefits obtainable within a short period of 5 years' abstinence should convey an encouraging message to chronic smokers, clinicians, and public health workers.
BACKGROUND: The rate of decline in lung cancer risk after smoking cessation among male population and the importance of the magnitude of the early decline were not sufficiently defined in the earlier studies. We evaluated the detailed duration-response relationship between years since smoking cessation and lung cancer risk across major histological types in a population-based case-referent study. METHODS: We recruited 1208 consecutive incident cases of primary lung cancer among Chinese males from the largest oncology center in Hong Kong during 2004-2006, and 1069 male community referents frequency-matched in 5-year age groups. We performed unconditional multiple logistic regression and generalized additive model incorporating smoothing spline to model the potential nonlinear effect of years since cessation on lung cancer. RESULTS: All histological types of lung cancer were strongly associated with current smoking. We observed a rapidly decreasing odds ratio of lung cancer (>50%) across all major histological types of lung cancer (except for the large cell type) within the first 5 years of quitting; the odds ratio continued to decrease but at a slower rate in the subsequent years. CONCLUSION: The substantial benefits obtainable within a short period of 5 years' abstinence should convey an encouraging message to chronic smokers, clinicians, and public health workers.