The association between cigarette smoking and risk of colorectal cancer in a large prospective cohort from the United States.

BACKGROUND: Many studies have reported a 20% to 60% increase in risk of colorectal cancer associated with active smoking. However, neither the U.S. Surgeon General nor the IARC have classified the relationship as causal because of concern about residual confounding.

METHODS: In a prospective study of 184,187 people followed from 1992 to 2005, we used Cox proportional hazard models to examine the relationship of cigarette smoking to incident colorectal cancer, controlling for screening and multiple known and putative risk factors. Information on smoking and time-varying covariates was updated in 1997, 1999, 2001, and 2003.

RESULTS: The incidence of colorectal cancer was significantly higher in current [hazard ratios (HR), 1.27; 95% confidence intervals (CI), 1.06-1.52] and former smokers (HR, 1.23; 95% CI, 1.11-1.36) compared with lifelong nonsmokers in analyses that controlled for 13 covariates, including screening. The relative risk was greatest among current smokers with at least 50 years of smoking (HR, 1.38; 95% CI, 1.04-1.84). Among former smokers, risk of colorectal cancer decreased with greater time since cessation (P trend = 0.0003), and also decreased with earlier age at cessation (P trend = 0.0014). No association was seen among former smokers who had quit before age of 40 years or abstained for 31 years or more.

CONCLUSIONS: Long-term cigarette smoking is associated with colorectal cancer, even after controlling for screening and multiple other risk factors.