Associations of cigarette smoking, alcohol drinking and body mass index with survival after colorectal cancer diagnosis by anatomic subsite: a prospective patient cohort study in Japan.

Background: Cigarette smoking, alcohol drinking and obesity are known to be risk factors for colorectal cancer. These factors may affect survival after diagnosis, but evidence has been inconsistent. We investigated subsite-specific associations between prediagnosis smoking, alcohol drinking and body mass index and survival in colorectal cancer.

Methods: Subjects were 1300 patients (colon 778; rectum 502; concurrent 20) with histologically confirmed colorectal cancer diagnosed during 1997-2013 at a single institution in Japan. Histories of smoking and alcohol drinking, height and prediagnosis weight were assessed using a self-administered questionnaire. Using Cox proportional hazards model, hazard ratios and 95% confidence intervals of mortality were estimated.

Results: During a median follow-up period of 6.7 years, 479 deaths were documented. Ever-smoking was associated with an increased risk of all-cause death among patients with colon cancer (hazard ratio: 1.47; 95% confidence interval: 1.07-2.02 compared with never-smoking). According to colon subsite, this increased risk was clear in patients with proximal colon cancer (hazard ratio: 2.09; 95% confidence interval: 1.28-3.40). There was no association between smoking and rectal cancer survival. Alcohol drinking was not associated with survival for either colon or rectal cancer. Among patients with rectal cancer, higher body mass index was associated with a lower risk of all-cause (Ptrend = 0.0006) and disease-specific death (Ptrend = 0.02). For colon cancer, lower body mass index tended to be associated with a higher risk of all-cause death (Ptrend = 0.05).

Conclusions: The results indicate that lifestyles identified as risk factors for colorectal cancer may impact differently on patient survival according to anatomic subsite.