Smoking has been implicated in many malignant diseases, but its association with colorectal cancer (CRC) is controversial. We quantitatively evaluated the relation between smoking and incidence of CRC in a meta-analysis of cohort studies.
Full publications of prospective cohort studies were identified in MEDLINE and EMBASE from 1950 to 2008. Subjects were classified as current smokers, former smokers, or never smokers. The quantity of smoking was assessed by number of cigarettes per day, years of smoking, and pack-years. The reported relative risks of CRC were pooled by random-effects model. Sensitivity analysis was conducted, and publication bias was evaluated.
A total of 1,463,796 subjects were recruited in 28 prospective cohorts from America, Europe, and Asia, with median follow-up of 13 years (range, 4-30 years). Current smokers showed a modestly higher risk of CRC (relative risk [RR], 1.20; 95% confidence interval [CI], 1.10-1.30) than never smokers. The risk of CRC among male smokers (RR, 1.38; 95% CI, 1.22-1.56) was more significant than among female smokers (RR, 1.06; 95% CI, 0.95-1.19). Rectal cancer was more closely related to smoking (RR, 1.36; 95% CI, 1.15-1.61) than colonic cancer. Former smokers still carried a higher CRC risk than never smokers. The increased risk of CRC was related to cigarettes per day, longer years of smoking, or larger pack-years.
Smoking was associated with a significantly increased risk of CRC. The associated risk was higher for men and for rectal cancers. The association of tobacco consumption and CRC risk appeared to be dose-related.