Snoep et al are concerned that a “healthy user bias” may account for our finding that regular use of aspirin reduces the risk of mortality among women. They reason that women who regularly used aspirin for cardiovascular prophylaxis are more likely to have favorable lifestyle characteristics. However, as we originally reported, less than 10% of aspirin users in this cohort indicated that their reason for use was to prevent cardiovascular disease. In fact, most of the aspirin users in our study were using the drug for headaches or musculoskeletal pain.1 As Snoep et al suggest, the association of aspirin use with mortality is not likely to be confounded by other lifestyle factors among such a population. In addition, data suggesting a potential cardiovascular benefit to routine aspirin use did not emerge until the 1990s: we did not observe any material difference in the effect of aspirin when we confined our analyses to the 1980s or the 1990s. Moreover, controlling for a wide range of known or suspected risk factors for mortality did not appreciably influence our results.
Chan A, Fuchs CS. A Possible Overestimation of the Effect of Aspirin—Reply. Arch Intern Med. 2007;167(21):2373. doi:10.1001/archinte.167.21.2373