PURPOSE: The aim of this study was to examine the relationships between smo
king and alcohol use and risk of colorectal adenoma. METHODS: Information a
bout smoking, alcohol use, and other lifestyle variables mere obtained pros
pectively from 14,427 male and 17,125 female residents in a city of Gifu Pr
efecture, Japan, by a self-administered questionnaire in September, 1992. C
olorectal adenomas mere newly diagnosed in 181 men and 78 women in this coh
ort between January, 1993 and December, 1995 by colonoscopic examination at
two major hospitals of the city. Gender-specific and site-specific relativ
e risks and 95 percent confidence intervals adjusted for age and for age pl
us other potential confounding factors were calculated by using logistic re
gression models. RESULTS: Thirty or more years of smoking was significantly
associated with risk: of adenoma in general compared with never having smo
ked in both men and women (relative risk, 1.60; 95 percent confidence inter
val, 1.02-2.62 and relative risk, 4.54; 95 per cent confidence interval, 2.
04-9.08, respectively). Effect of smoking was stronger in the proximal colo
n. After adjusting for age and carbohydrate intake, total alcohol intake wa
s not associated with risk of adenoma in any site in the colon in men. Sake
drinkers were at significantly increased risk of adenoma in general, but t
he dose-response relationship was not statistically significant. Risk of ad
enoma in the rectum was not significantly increased for those who consumed
>30.3 g/day of ethanol (relative risk, 5.7). CONCLUSION: These data suggest
that smoking is a risk factor of adenoma in Japanese men and women. The ro
le of alcohol, however, is less clear.