COMMENTS AND RESPONSES Concerns About Meta-analysis of Glucose Control and Cardiovascular Disease in Type 2 Diabetes

TO THE EDITOR: The proposition that glucose lowering affects car-diovascular risk in patients with type 2 diabetes is very important. Although we recognize the value of the article by Kelly and col-leagues (1), we would like to draw attention to errors in their meta-analysis that we believe affect the conclusions reached. First, Kelly and colleagues combined data from 3 recent studies (2–4) with 2 substudies of the UKPDS (United Kingdom Prospec-tive Diabetes Study) (5, 6) to suggest that many outcomes differed between the “early ” and “late ” studies. In so doing, they did not notice that the 411 participants and their associated events in the conventional treatment (diet) group of the metformin substudy, UKPDS 34 (6), were also included in the conventional treatment group in the main study, UKPDS 33 (5). By using the same studies and methodology as Kelly and colleagues, but without double-counting, we analyzed UKPDS 33 and 34 as 1 study, including th