Caution should be exercised in interpreting the conclusions regarding the maximally effective dose of glipizide in the recent important paper by Stenman and colleagues [1]. First, the group of patients studied was markedly hyperglycemic. Furthermore, although they were somewhat improved, at the conclusion of the study they remained hyperglycemic. Can one make judgments regarding the efficacy of this drug in patients who have responded so poorly?

In addition, these patients represent a homogeneous population, whereas patients with type 2 diabetes are heterogeneous. Several studies indicate the possible utility of higher doses of glipizide in selected patients [2, 3]. Anecdotally, members of our ethnically mixed population show different sensitivities to glipizide. Several of our patients who are most sensitive to the effects of glipizide are of Scandinavian descent. Although hardly definitive, this suggests the possibility that other populations might show different dose-response curves to glipizide. At the least, studies such as Stenman and colleagues' should be done in other populations before treatments that may be generalized to all patients with type 2 diabetes are recommended. On the basis of this study, it is premature for physicians to decrease the dose of glipizide to 10 mg in patients receiving larger doses. We agree, however, that insulin therapy should be initiated in patients who remain hyperglycemic on maximally effective doses of sulfonylurea.