Since many smokers are unable or unwilling to totally quit smoking, the concept of harm reduction is raised. This study, with up to 31 years of follow up, studied the smoking habits of 11 151 men and 8563 women aged between 20 and 93 years. 864 cases of primary lung cancer occurred during follow up. Study participants were classified as heavy smokers (⩾15 g/day), light smokers (1–14 g/day), ex-smokers, and never smokers. Consumption was calculated by equating a cigarette to 1 g of tobacco, a cheroot to 3 g, and a cigar to 5 g. Smoking reduction was defined as a reduction of 50% or more in the amount smoked between the two follow up visits.

For follow up, six groups were defined: continued heavy smokers, reducers (heavy smokers who had a reduction of at least 50% without quitting), continued light smokers, quitters, stable ex-smokers, and never smokers. Cox regression analysis was performed to calculate adjusted hazard ratios (HR). Compared with persistent heavy smokers, the HR for lung cancer in reducers was 0.73 (95% confidence interval (CI) 0.54 to 0.98). For light smokers the HR was 0.44 (CI 0.35 to 0.56), for quitters 0.50 (CI 0.36 to 0.69), for stable ex-smokers 0.17 (CI 0.13 to 0.23), and for never smokers 0.09 (CI 0.06 to 0.13).

The study shows that, in heavy smokers, smoking reduction by 50% or more significantly reduces the risk of lung cancer. However, the authors conclude that smoking cessation should still be advocated as reducing smoking has not been shown to decrease the risk of other smoking related disease.