Background: In contrast to most observational studies, the randomized Physicians' Health Study found no association between aspirin use and colorectal cancer after 5 years.
Objective: To determine the effect of randomly assigned aspirin treatment and self-selected aspirin use on the incidence of colorectal cancer after 12 years and to identify factors influencing the self-selection of regular aspirin use.
Design: Randomized clinical trial and prospective cohort study.
Setting: Male physicians throughout the United States.
Patients: 22 071 healthy male physicians who were 40 to 84 years of age in 1982.
Intervention: 325 mg of aspirin every other day. In 1988, the aspirin arm of the randomized trial was stopped early. Participants then chose to receive either aspirin or placebo for the rest of the study.
Measurements: Annual questionnaires asking about aspirin use and other variables, including occurrence of cancer.
Results: Colorectal cancer was diagnosed in 341 patients during the study period. Over 12 years of follow-up, random assignment to aspirin was associated with a relative risk for colorectal cancer of 1.03 (95% CI, 0.83 to 1.28). Various gastrointestinal symptoms and diagnoses were strong predictors of less frequent aspirin use in 1988. The relative risk for colorectal cancer in persons who used aspirin frequently after 1988 was 1.07 (CI, 0.75 to 1.53).
Conclusions: In the Physicians' Health Study, both randomized and observational analyses indicate that there is no association between the use of aspirin and the incidence of colorectal cancer. The low dose of aspirin used and the short treatment period may account for the null findings. However, other characteristics associated with the use of aspirin in observational studies remain a plausible alternative explanation.