The relation of smoking, alcohol use and obesity to risk of sigmoid colon and rectal adenomas.

We conducted a case-control study, using 429 cases with histologically confirmed sigmoid adenoma, 75 cases with rectal adenoma, and 3101 controls showing normal colonoscopy at least up to 60 cm from the anus. The subjects were male Self-Defense Forces personnel aged 48-56 who received a retirement health examination including a routine sigmoid- or colonoscopy. Lifestyle characteristics were ascertained by a self-administered questionnaire. Smoking in the recent past (< or = 10 years preceding the colonoscopy) and smoking in the remote past (> 10 years before the colonoscopy) were both significantly associated with risk of sigmoid adenoma but not with rectal adenoma as a whole. After reciprocal adjustment for smoking in the two periods, only smoking in the recent past was associated with both sigmoid colon and rectal adenomas. Odds ratios (OR) of sigmoid adenoma (and 95% confidence interval) for the categories of 0, 1-150, 151-250 and > or = 251 cigarette-years were 1.0 (reference), 1.9 (1.3-2.8), 2.1 (1.4-3.0) and 3.0 (1.9-4.7), respectively (P for trend < 0.01), and those for rectal adenoma were 1.0 (reference), 1.2 (0.4 3.2), 3.5 (1.4-8.5) and 2.0 (0.6 6.7), respectively (P for trend = 0.03). Alcohol use was significantly positively associated with sigmoid adenoma, and insignificantly associated with rectal adenoma. Body mass index was significantly positively associated with sigmoid adenoma, especially large ones. No such association was found for rectal adenoma. These findings suggest that smoking, especially in the recent past, and alcohol use are common risk factors for sigmoid colon and rectal adenomas while obesity may be exclusively related to the growth of sigmoid adenoma.