Drs Lesser and Libow raise some important points that we are pleased to have the opportunity to address. We agree that the absence of any association between simvastatin and AD in our data set could indicate the presence of confounding variables producing either positive or negative spurious associations. Such variables could lead to a spurious positive association between pravastatin or lovastatin and decreased prevalence of AD, or to a spurious negative association between simvastatin and decreased prevalence of AD. However, a recent report by Jick and colleagues1 shows that patients taking a group of statins, including simvastatin, also have a reduced risk of AD. This supports our overall conclusions and suggests that the negative association between simvastatin and decreased prevalence of AD seen in our study is not significant. Jick and colleagues studied 284 cases with dementia and 1080 controls, using a case-control design, and observed a 70% lower risk of AD in patients taking any of several statins, but most often simvastatin. These data suggest that simvastatin is also protective against AD.