CIGARETTE-SMOKING AND PANCREAS CANCER - A CASE-CONTROL STUDY BASED ONDIRECT INTERVIEWS

Background: Cigarette smoking is the most consistently reported risk f actor for pancreas cancer, yet the dose-response relationship in many pancreas cancer studies is weak. Because of the poor prognosis for pan creas cancer, many case-control studies have been based largely on int erviews with proxy respondents, who are known to report less reliable information on detailed smoking habits than original subjects. Purpose : Our purpose was to evaluate cigarette smoking as a risk factor for p ancreas cancer based on data obtained only from direct interviews and to estimate the effects of quitting smoking and of switching from nonf iltered to filtered cigarettes on risk. Our objective also was to esti mate the contribution of cigarette smoking toward explaining the highe r pancreas cancer incidence experienced by black Americans compared wi th white Americans. Methods: A population-based, case-control study of pancreas cancer was conducted during 1986-1989 in Atlanta, Ga., Detro it, Mich., and 10 counties in New Jersey. Direct interviews were succe ssfully completed with 526 case patients and 2153 control subjects age d 30-79 years, making this the largest population-based, case-control study of pancreas cancer to date based only on direct interviews. Resu lts: Cigarette smokers had a significant, 70% increased risk of pancre as cancer compared with the risk in nonsmokers. A significant, positiv e trend in risk with increasing duration smoked was apparent (P<.0001) , with long-term (greater than or equal to 40 years) smokers experienc ing a modest 2.1-fold risk. We also observed a negative trend in risk with increasing years quit smoking. Smokers who quit for more than 10 years experienced about a 30% reduction in risk relative to current sm okers; quitters of 10 years or less experienced no risk reduction. Swi tching from nonfiltered to filtered cigarettes did not appear to decre ase risk. Compared with nonsmokers, subjects who smoked only filtered cigarettes had a 50% elevated risk and those who smoked only nonfilter ed cigarettes had a 40% elevated risk. The proportion of pancreas canc er attributable to cigarette smoking was 29% in blacks and 26% in whit es. Conclusions: The relationship between cigarette smoking and pancre as cancer risk is likely to be causal, despite the weakness of the dos e-response data. Long-term smoking cessation clearly reduces risk, whe reas switching from nonfiltered to filtered cigarettes may not be bene ficial. Cigarette smoking appears to explain little of the excess panc reas cancer risk experienced by blacks. Implications: Elimination of c igarette smoking would eventually prevent approximately 27% of pancrea s cancer, saving 6750 lives in the United States annually.