DRUGS AND COLON-CANCER

In a case-control study of colon cancer conducted in three geographic regions of the United States, 1993 case subjects and 2410 control subj ects were interviewed. In addition to queries regarding other known or suspected risk factors, subjects were asked about their use of eight drugs or drug groups. Two of these, aspirin and other nonsteroidal ant i-inflammatory drugs (NSAIDs), have been inversely associated with ris k in other studies. Three others-asthma medications, digitalis prepara tions, and phenmetrazine were positively associated and the last three -diazepam, penicillin, an phenformin-were negatively associated with r isk of colon cancer in an earlier study that screened pharmaceuticals for possible carcinogenic effects. Reported use of aspirin and NSAIDs were both inversely related to risk with essentially the same odds rat ios (0.7, 95% confidence interval 0.6-0.8) for both drugs in both univ ariate and multivariate analyses controlling for use of each other and for other colon cancer risk factors. Subdivision by age at starting t he drug, duration of use, latency interval, sex, race, family history of colon cancer, or proximal versus distal cancer revealed no substant ial differences among subgroups for either aspirin or NSAIDs, but redu ced risk was associated primarily with recent aspirin use. Phenformin showed a strong positive association but the data concerning this drug appeared to be inaccurate. The other drugs and drug groups showed ess entially no association with colon cancer risk. (C) 1998 John Wiley & Sons, Ltd.