A large new observational study finds more evidence of an association between daily aspirin use and modestly lower cancer mortality, but suggests any reduction may be smaller than that observed in a recent analysis. The study, appearing early online in the Journal of the National Cancer Institute (JNCI), provides additional support for a potential benefit of daily aspirin use for cancer mortality, but the authors say important questions remain about the size of the potential benefit.

A recent analysis pooling results from existing randomized trials of daily aspirin for prevention of vascular events found an estimated 37% reduction in cancer mortality among those using aspirin for five years or more. But uncertainty remains about how much daily aspirin use may lower cancer mortality, as the size of this pooled analysis was limited and two very large randomized trials of aspirin taken every other day found no effect on overall cancer mortality.

For the current study, American Cancer Society researchers led by Eric J. Jacobs, Ph.D., analyzed information from 100,139 predominantly elderly participants in the Cancer Prevention Study II Nutrition Cohort who reported aspirin use on questionnaires, did not have cancer at the start of the study, and were followed for up to 11 years. They found daily aspirin use was associated with an estimated 16% lower overall risk of cancer mortality, both among people who reported taking aspirin daily for at least five years and among those who reported shorter term daily use. The lower overall cancer mortality was driven by about 40% lower mortality from cancers of the gastrointestinal tract (such as esophageal, stomach, and colorectal cancer) and about 12% lower mortality from cancers outside the gastrointestinal tract.

The reduction in cancer mortality observed in the current study is considerably smaller than the 37% reduction reported in the recent pooled analysis of randomized trials. The authors note that their study was observational, not randomized, and therefore could have underestimated or overestimated potential effects on cancer mortality if p