Background: Animal experiments and epidemiologic data have suggested that the use of nonsteroidal antiinflammatory drugs (NSAIDs) may decrease the incidence of large bowel carcinoma. Our purpose was to assess the relation of the use of aspirin and nonaspirin NSAIDs with the risk of large bowel carcinoma.
Methods: A population-based case-control study of colon and rectal carcinoma was conducted in Massachusetts from 1992 to 1994. Data on NSAID use and risk factors for large bowel carcinoma were collected by interview from 1201 incident cases of large bowel carcinoma and 1201 controls matched by age, gender, and area of residence.
Results: Regular NSAID use that continued into the year before diagnosis was associated with a significantly decreased relative risk estimate overall (0.7; 95% confidence interval [CI], 0.5-0.8) and among Stage II-IV tumors (0.6; 95% CI, 0.4-0.7). There was no reduction in risk for discontinued use. The inverse association with regular continuing use was present across age and gender and for both colon and rectal carcinoma. Similar inverse associations were present for regular continuing use of aspirin and nonaspirin NSAIDs. There was no significant evidence of a trend for the relative risk to decrease as the duration of use increased, nor was there a trend across the dose of aspirin, which ranged from less than one-half of a 325 mg tablet per day to > or = 2 tablets per day. Discontinuation of use in response to symptoms of carcinoma did not appear to explain the inverse association, nor did bias related to diagnosis of the carcinoma.
Conclusions: These data add to the growing body of evidence that suggests a protective effect of NSAIDs against large bowel carcinoma.