Purpose: It is currently unknown whether intensive insulin treatment of diabetes decreases the risk for cardiovascular complications by lowering glucose levels or increases the risk by postulated direct atherogenic effects. This article reviews published data from two long-term, randomized clinical trials that compared cardiovascular outcomes associated with different exogenous insulin treatment regimens.
Study Selection and Data Sources: The University Group Diabetes Program (UGDP) and the Diabetes Control and Complications Trial (DCCT) were selected as the only available randomized intervention trials with long-term follow-up results. Data reviewed were from the major publications of these two trials.
Results: The UGDP compared the effects of a fixed-dose standard insulin regimen, a variable-dose insulin regimen, and a diet plus oral placebo regimen in over 600 patients with non–insulin-dependent diabetes mellitus (NIDDM) who were followed for up to 13 years (1962 to 1975). Plasma glucose levels were 1.7 to 2.2 mmol/L lower in the variable-dose insulin regimen group than in the other two treatment groups. No significant differences were found in the final prevalence or the cumulative incidence of total deaths, cardiovascular disease deaths, or myocardial infarctions among the three treatment groups, even when outcomes were adjusted for pertinent baseline cardiovascular risk factors. There was a slight suggestion only from post hoc analysis that patients in both insulin treatment groups, who were defined as having good glucose control, had fewer cardiovascular events than those with fair or poor control.
The DCCT compared intensive insulin treatment with conventional insulin treatment (the mean hemoglobin A1c [HbA1c] level was 7.2% and 9.0%, respectively) in over 1400 patients with insulin-dependent diabetes (IDDM) followed for up to 10 years (1983 to 1993). Three major cardiovascular events occurred in the intensive treatment group as compared with 14 in the conventional treatment group (P > 0.05), and low-density lipoprotein cholesterol and triglyceride levels were substantially lower in the intensive treatment group. However, body mass index increased substantially more with intensive than with conventional therapy.
Conclusions: The UGDP trial was flawed by inadequate power, uncertainties in compliance with treatment regimens, insufficient separation of glycemic levels, and ignorance of smoking history as a possible confounder. In the DCCT, the number of cardiovascular events was few because the patients were young and had a relatively short duration of diabetes at baseline. In addition, total daily insulin doses were similar in the two DCCT treatment groups. For these reasons, neither trial provides a definitive answer to the question about the effects of intensive insulin therapy. A better designed clinical trial is needed to determine whether insulin treatment has beneficial or adverse effects, or even offsetting beneficial and adverse effects, on the risk for cardiovascular disease in NIDDM and IDDM.