OBJECTIVETo investigate the association between cigarette smoking and cancer mortality in urban men in Shanghai and its impact when smoking habit changed during the follow-up period.

METHODSA total of 18 244 male residents aged 45 to 64 years in urban Shanghai were enrolled in the study during January 1, 1986 through September 30, 1989, and was actively followed up on annual visits. Cox proportional hazards model was used to estimate relative risks (RR).

RESULTSBy the end of follow-up program in 2002, 235 762 person-years, averaged 12.9 years per subject in the cohort was reached. 3365 deaths including 1381 cancer deaths were registered during the follow-up period. The mortality rates for cancers of lung, stomach, liver, pancreas, esophagus, head and neck etc. increased significantly among smokers. Compared with data of nonsmokers at the baseline survey, the adjusted RR was 1.49 for all-causes mortality among current smokers at the baseline survey. After excluding subjects who changed their smoking habit during the follow-up period, the RR became 1.78 compared with lifelong-nonsmokers. The corresponding RRs rose from 2.05 to 2.58 for all cancer deaths and from 6.40 to 8.77 for lung cancer deaths. The age-adjusted all-causes and cancer death rates among current smokers at the baseline survey were 1695.6 and 782.0 per 100 000 person-years, respectively. After exclusion of those with smoking habit changed during the follow-up period, the rates among persistent smokers were 2353.7 and 1144.6 per 100 000 person-years, respectively.

CONCLUSIONCigarette smoking is an important predictor for risk of all-causes of death as well as for cancer deaths. The change of smoking habit during the follow-up period could result in underestimating the deleterious effect of cigarette smoking on health.