Abstract 12985: Incident Oral Antidiabetic Drugs and Risk of Cardiovascular Events
Objective: Effect of oral antidiabetic drugs (OAD) on cardiovascular disease (CVD) are uncertain particularly among those with CVD. Our aim was to compare time to CVD (acute myocardial infarction [AMI], stroke or death) among newly diagnosed diabetic patients initiating sulfonylurea or metformin stratified by CVD history.
Methods: We identified a national retrospective cohort of veterans from Veterans Health Administration (VHA) databases linked with Medicare files. Patient entered the cohort with their first prescription for either metformin or sulfonylurea (10/1/2001thru 9/30/2008) after 1 year of no hypoglycemics filled. All patients were stratified by CVD history at the index OAD prescription and followed until: hospitalization for AMI, stroke or death; end of study or censoring event. Analyses were adjusted for demographics, baseline medications, cholesterol, glycosylated hemoglobin, creatinine, blood pressure, body mass index, healthcare utilization and co-morbidities. Adjusted Hazard ratios (aHR) and 95% confidence intervals (CI) are reported.
Results: There were 285359 patients with incident prescriptions; 27.4% had a history of CVD. In those without CVD (N=207227) there were 50225 metformin initiators (58327 person-years of follow-up and 417 events) and 24855 sulfonylurea initiators (26079 person-years of follow-up and 280 events) included in the adjusted analysis. In this group, metformin users had 7.1 events/1000 person-years vs. 10.7 events/1000 person-years among sulfonylurea users (aHR 1.05 [0.90, 1.24]). Among patients with a past history of CVD (N= 78,132) there were 19380 metformin initiators (23175 person-years of follow-up and 351 events) and 12349 sulfonylurea initiators (13634 person-years of follow-up and 350 events) in adjusted analysis. In those with CVD, metformin users had 15.1 events/1000 person years vs. 25.7 events/1000 person years among sulfonylurea users (aHR 1.33 [1.13, 1.55]). When evaluating CV events (AMI and Stroke) and censoring deaths, results were similar among the no CVD (aHR 1.04 [0.87, 1.25]) and CVD cohort (aHR 1.30 [1.09, 1.55])
Conclusions: Among patients with CVD we found an increased risk of cardiovascular events and death among new use of sulfonylurea compared with new use of metformin
- © 2011 by American Heart Association, Inc.