CARDIOVASCULAR EVENTS AND CORRELATES IN THE VETERANS AFFAIRS DIABETESFEASIBILITY TRIAL - VETERANS AFFAIRS COOPERATIVE STUDY ON GYLCEMIC CONTROL AND COMPLICATIONS IN TYPE-II DIABETES

Background: The risks and benefits of intensive therapy in non-insulin -dependent diabetes mellitus (NIDDM) need to be defined. In preparatio n for a long-term trial, a feasibility study of 153 men in 5 medical c enters compared standard vs intensive insulin therapy. Objective: To a ssess the rate of development of new cardiovascular events and their c orrelates. Methods: Patients with a mean+/-SD age of 60+/-6 years and diagnosis of NIDDM for 7.8+/-4.0 years were randomly assigned to a sta ndard (1 insulin injection every morning) or to an intensive treatment arm (stepped plan from 1 evening injection of insulin, alone or with glipizide, to multiple daily injections) designed to attain near-norma l glycemia levels. A 2.07% separation of glycosylated hemoglobin (HbA( 1c)) was sustained for a mean follow-up of 27 months (P<.001). Predefi ned cardiovascular events were assessed by a committee unaware of trea tment assignment. Results: Mild and moderate hypoglycemic events were more frequent in the intensive than in the standard treatment arm (16. 5 vs 1.5 per patient per year, respectively). Mean insulin dose was 23 % lower in the standard treatment arm (P<.001). There were 61 new card iovascular events in 24 patients (32%) in the intensive treatment arm and in 16 patients (20%) in the standard treatment arm (P=.10). There was no difference in total and cardiovascular mortality (n=5 and n=3 i n the intensive and standard treatment arms, respectively) or in new e vents in patients with cardiovascular history (n=10 in each arm). In C ox regression analysis, the only significant correlate for new cardiov ascular events was previous cardiovascular disease (P=.04). Entering i n the analysis any baseline cardiovascular abnormality, the regression model indicated a lower HbA(1c) level prior to the event as the only correlate for new cardiovascular events (P=.05). Conclusion: A long-te rm prospective trial is needed to assess the risk-benefit ratio of int ensive insulin therapy for NIDDM in patients who require it.