Action to Control Cardiovascular Risk in Diabetes - ACCORD


The goal of the trial was to evaluate intensive glycemic control through currently available means (i.e., glycated hemoglobin <6%), compared with standard glycemic control (i.e., glycated hemoglobin 7.0-7.9%) among patients with type 2 diabetes mellitus with known cardiovascular disease or with additional risk factors for cardiovascular disease.


Intensive glycemic control (i.e., glycated hemoglobin <6%) will be more effective in preventing the first occurrence of a major cardiovascular event in diabetic patients at high risk of death and stroke.

Study Design

Patients Enrolled: 10,251
Mean Follow Up: 5.6 years
Mean Patient Age: 62 (range 40-82)
Female: 38

Patient Populations:

• Type 2 diabetic patients with glycated hemoglobin >7.5%
• Age 40-79 years with pre-existing heart disease, or
• Age 55-79 years with anatomical evidence of atherosclerosis, albuminuria, left ventricular hypertrophy, or two or more additional risk factors for heart disease (dyslipidemia, hypertension, current smoker, or obesity)


• Frequent hypoglycemic episodes
• Unwillingness to perform home glucose monitoring or inject insulin
• Body mass index >45 kg/m2
• Serum creatine >1.5 mg/dl

Primary Endpoints:

• First occurrence of nonfatal mycardial infarction, nonfatal stroke, or cardiovascular death.
• Cardiovascular death was defined as that due to myocardial infarction, heart failure, arrhythmia, invasive cardiac procedure, cardiovascular event after noncardiac surgery, stroke, unexpected death presumably due to ischemia, and death from other vascular disease.

Secondary Endpoints:

Treatment differences in other cardiovascular outcomes, all-cause mortality, microvascular outcomes, health-related quality of life, and cost-effectiveness

Drug/Procedures Used:

A total of 10,251 patients with type 2 diabetes were randomized to intensive glycemic control (n = 5,128) versus standard glycemic control (n = 5,123).

Concomitant Medications:

For glycemic control, both groups could be treated with metformin, sulfonylureas, meglitinides, thiazolidinediones, α-glucosidase inhibitors, insulin, insulin analogues, and exenatide (added to the trial in April 2007). Patients were also given lifestyle intervention.

At baseline, in the intensive therapy group, the use of insulin was 34%, metformin was 60%, any sulfonylurea was 51%, any thiazolidinedione was 20%, any antihypertensive agent was 85%, angiotensin-converting enzyme inhibitor was 53%, aspirin was 55%, beta-blocker was 29%, any thiazide diuretic was 27%, and statin was 62%.

At follow-up, in the intensive therapy group, the use of insulin was 77%, metformin was 95%, any sulfonylurea was 87%, any thiazolidinedione was 92%, α-glucosidase inhibitor was 23%, and exenatide was 12%.

Principal Findings:

The median duration of diabetes was 10 years, mean glycated hemoglobin was 8.3%, prior cardiovascular event was observed in 35% of participants, and body mass index was 32 kg/m2.

Glycated hemoglobin levels dropped in both groups within 4 months of randomization and was stable from 1 year to the extent of follow-up. At 1 year, glycated hemoglobin was 6.4% versus 7.5%, respectively.

Patients in the intensive arm were exposed to more drugs from every class than the standard therapy group. The mean number of medication changes per year was 4.4 versus 2.0, respectively.

Significant hypoglycemia requiring medical assistance occurred in 10.5% versus 3.5% (p < 0.001), respectively. More than 10 kg weight gain occurred in 28% versus 14% (p < 0.001), respectively.

At a mean follow-up of 3.5 years, the primary composite outcome occurred in 6.7% of the intensive therapy group versus 7.2% of the standard therapy group (p = 0.16). All-cause mortality was 5.0% versus 4.0% (p = 0.04), cardiovascular mortality was 2.6% versus 1.8% (p = 0.02), nonfatal myocardial infarction was 3.6% versus 4.6% (p = 0.004), nonfatal stroke was 1.3% versus 1.2% (p = 0.74), and fatal or nonfatal congestive heart failure was 3.0% versus 2.4% (p = 0.17), respectively.


The National Heart, Lung, and Blood Institute stopped the trial 17 months early when excess deaths were detected in the intensive glycemic control group. Among patients with type 2 diabetes at high risk for death and stroke, a strategy of lowering glycated hemoglobin to a mean of 6.4% is associated with excess mortality.

There was no identifiable cause for the excess deaths, from either prespecified subgroups or exploratory analyses. Possible explanations include the strategy used to achieve intensive control (for example, an unrecognized harmful drug or drug combination effect), the rate of decline in glycated hemoglobin, or the absolute level of glycated hemoglobin achieved.

This trial serves as a reminder that while very intensive glycemic control seemed intuitive, a randomized clinical trial was needed to answer this question properly. It remains possible that certain subgroups (for example, no prior cardiovascular event or lower baseline glycated hemoglogin level) could still benefit from intensive glucose control; however, this will need to be studied.


National Heart, Lung, and Blood Institute Press Release. For Safety, NHLBI Changes Intensive Blood Sugar Treatment Strategy in Clinical Trial of Diabetes and Cardiovascular Disease. February 6, 2008. Available at: (click here). Accessed 02/06/2008.

The Action to Control Cardiovascular Risk in Diabetes Study Group. Effects of intensive glucose lowering in type 2 diabetes. N Engl J Med 2008;358:2545-59.

Keywords: Hypertrophy, Left Ventricular, Myocardial Infarction, Stroke, National Heart, Lung, and Blood Institute (U.S.), Atherosclerosis, Diabetes Mellitus, Type 2, Coronary Disease, Blood Pressure, Weight Gain, Hypoglycemia, Glycated Hemoglobin A, Dyslipidemias, Body Mass Index, Blood Glucose, Heart Failure, Obesity, Hypertension

< Back to Listings