We appreciate the comments by Sprague and Wood and Kumar et al and the opportunity to reply. While in vitro and animal data suggest that statins may protect against breast cancer, it is important to weigh all available evidence before conducting a randomized controlled trial of statin use in the prevention of breast cancer. Thus far the cumulative evidence is not compelling, with epidemiologic evidence to date leaning toward no association, as noted by Sprague and Wood. As we noted in our article,1 we agree that further study is needed to evaluate this association with longer durations of use and by estrogen receptor status of the tumor. Given the relatively recent widespread use of statins, even in a large observational study such as ours, the number of long-term users is still small, such that there are relatively few breast cancer cases among users. Also as we noted, we were unable to address the importance of hydrophilic vs lipophilic statins, or specific statin drugs, within the Nurses' Health Study. However, in a recent large meta-analysis, Dale et al2 found no association between statin use and overall cancer incidence (odds ratio [OR], 1.02) or mortality (OR, 1.01), and the null association was evident for both hydrophilic and lipophilic statins. In addition, no significant association was observed with breast cancer incidence, and the OR was not suggestive of a protective effect (OR, 1.33; 95% confidence interval, 0.79-2.26).