Accessible Search Form           Advanced Search

Skip left side navigation and go to content

Health Professionals

Elizabeth G. Nabel, M.D.

Director, National Heart, Lung, and Blood Institute

Wednesday, February 6, 2008

Thank you and good morning. Today we are sharing with you a major finding and decision regarding the ACCORD Trial. ACCORD is a large, important clinical trial designed to determine the best ways to decrease the high rate of cardiovascular disease among adults with type 2 diabetes who are at especially high risk for heart attack and stroke. Type 2 diabetes is a complex metabolic disease that results in elevated blood sugar levels. The ACCORD trial, which began in 2001, is testing three treatment approaches: intensive lowering of blood sugar levels compared to a standard blood sugar treatment; intensive lowering of blood pressure compared to a standard blood pressure treatment; and treatment of blood lipids by a fibrate plus a statin compared to a statin alone.

This morning, we are announcing the outcome of one of the treatment strategies and a change in how the study will be conducted. After thoroughly reviewing the data collected to date, ACCORD investigators found that among these adults with type 2 diabetes who are at especially high risk of cardiovascular disease, a medical treatment strategy to intensively lower their blood sugar levels below the current guidelines increased the risk of death compared to standard blood sugar lowering treatment.

Because of this finding, and at the recommendation of the study's Data and Safety Monitoring Board, the NHLBI is stopping the intensive blood sugar lowering treatment part of the study 18 months early. Although we have stopped this treatment, we will continue to care for all study participants. Participants who were receiving the intensive blood sugar-lowering treatment will now receive the standard blood sugar lowering treatment, which aims for blood sugar control similar to that achieved in the general population of similar adults treated for type 2 diabetes, which is an A1C level of about 7.5 percent. As you know, A1C is a measure of blood sugar levels.

The treatment approaches for testing blood pressure and lipid control will continue until the study ends as planned in June 2009.

As always, our primary concern is to protect the safety of our study volunteers. We made the decision to stop the intensive treatment approach in the ACCORD study after a thorough review of the health risks to study participants. We will continue to monitor the health of all of the study participants, seek to understand the underlying causes for this finding, and carry on the other important research within ACCORD.

The findings from ACCORD will inform treatment decisions for the millions of individuals with type 2 diabetes who are at especially high risk of cardiovascular disease. It is important to note, however, that the results apply only to those individuals who are similar to the study participants. All 10,251 participants had type 2 diabetes on average for 10 years, when they enrolled in the study. In addition, ACCORD participants had blood sugar levels that were higher than most type 2 diabetes patients in the U.S. today, that is, on entry, their A1C levels were about 8.2 percent. To be eligible for the study, participants also had known heart disease or at least two risk factors, in addition to diabetes, including high blood pressure, high cholesterol levels, obesity, and smoking. In other words, they had diabetes plus other risk factors which place them at even higher risk for heart disease than if they just had diabetes alone. In this population of individuals with type 2 diabetes at especially high risk for heart disease, it has been observed that the risk of death is approximately 50 deaths per 1000 individuals per year ( about 5 percent per year).

All ACCORD participants were randomly assigned to either an intensive medical treatment strategy, with a goal to lower A1C levels to less than 6 percent, or to a standard medical treatment strategy to lower A1C levels to 7 to 7.9 percent. A variety of FDA-approved medications were used to try to reach the assigned blood sugar goals.

Here is what the ACCORD researchers found:

  • The median A1C level achieved in the intensive treatment group was 6.4 percent, while the median A1C level in the standard treatment group was 7.5 percent.

  • In the standard treatment group, we observed 11 deaths per 1000 individuals per year, on average, over 4 years of follow-up. In the intensive treatment group, we observed 14 deaths per 1000 individuals per year. First and foremost, it is important to recognize that this death rate is lower than what has been previously observed in individuals with type 2 diabetes at especially high risk for heart disease.

  • Nonetheless, in the intensive treatment group, there were 257 deaths and in the standard group, 203 deaths. This is a difference of 54 deaths, or 3 per 1,000 participants per year, over an average of 4 years of treatment. Because of this difference, the increased risk between the two groups outweighed potential benefits of the intensive treatment strategy on nonfatal events. Accordingly, the NHLBI has made the decision to stop this intensive treatment approach of the trial.

  • This is an important finding which shows that if you have type 2 diabetes and are at especially high risk for heart disease, very intensive glucose lowering treatments aimed at normalizing blood glucose to an A1C of less than 6 percent may be detrimental.

  • These preliminary findings of the ACCORD trial are consistent with recommendations from the American Diabetes Association in 3 aspects:
    1. First and foremost, individuals with diabetes should not change their diabetes treatment without consulting with their healthcare provider.

    2. Second, we concur with the general recommendation of the ADA that advises people with diabetes to aim for an A1C level of less than 7%.

    3. However, for this special group of individuals with diabetes, as exemplified in the ACCORD population, which were average age of 62, had diabetes for an average of 10 years, and had know heart disease or were at high risk, less stringent A1C goals are likely appropriate, with an aim for around 7%, understanding that treatment should be individualized.

ACCORD researchers have extensively analyzed the data available to date and have not identified any specific cause for the higher death rate among the intensive blood sugar treatment group. We know that the higher death rate is not due to episodes of low blood sugar, known as hypoglycemia, or due to any single drug, including rosiglitazone, or combination of drugs.

As I stated earlier, ACCORD researchers will continue to monitor participants and conduct additional analyses to try and explain the findings while continuing other important research studies which are a part of ACCORD.

Now, I'd like to take a minute to describe how the decision to change the blood sugar treatment part of the study was made. The ACCORD Data and Safety Monitoring Board, or DSMB, is an independent group of 10 experts who were appointed by the NHLBI to regularly examine study outcomes and safety data. The DSMB has expertise in diabetes, cardiovascular disease, statistics, ethics, epidemiology, and clinical trials. This group is responsible for providing recommendations to the NHLBI on starting, continuing, or stopping the study or portions of the study. The DSMB carefully considers the safety and efficacy of the study treatments and monitors the overall conduct of the study. The DSMB's recommendations are based on safeguarding the interests of study participants.

Since the study began in 2001, the DSMB has met regularly, generally every six months, to monitor study conduct and to review ACCORD data. In its regular review of the study data, the ACCORD DSMB noted an unexpected higher total death rate from any cause among participants who had been randomly assigned to the intensive blood sugar treatment group compared to those assigned to the standard blood sugar treatment group---the difference of 54 deaths which I described a moment ago.

Although there appeared to be some benefit of an overall lower death rate in both groups, the DSMB recommended stopping the intensive treatment because of the difference in deaths between the intensive and standard treatment groups; that is, the harm of the intensive treatment outweighed the potential benefit. NHLBI accepted the DSMB's recommendation to stop the intensive treatment group and decided to continue treating all ACCORD participants at the standard treatment approach as well as to continue the blood pressure and lipid treatment parts of the study. We will continue to monitor all study participants for an additional 18 months as planned until the study ends in June 2009.

As I emphasized earlier, on the whole, the death rates in both blood sugar treatment groups were lower than those seen in similar populations. That is, although the death rate was higher in the intensive treatment group than the standard group, it was still lower than death rates reported in other studies of type 2 diabetes in similar adults.

As people with diabetes learn the results of the ACCORD trial, we advise them to consult with their healthcare professional before making any changes to their treatment. This is an important message we will repeat several times today.

The NHLBI felt it was important to inform study participants and the diabetes community soon after reaching this decision. Accordingly, letters were sent to all study participants by their clinical site principal investigators on Monday February 4th with recommendations regarding further follow up care. Investigators have begun to prepare a report of these initial findings, which they will submit for publication in a peer-reviewed medical journal. We anticipate that these findings will be published shortly.

To summarize, ACCORD is the first major clinical trial to study whether lowering a raised blood sugar level to a level similar to that seen in people without diabetes reduces the risk of cardiovascular disease. We now have one part of the answer to this question. The study will continue to examine other ways to lower the risk of cardiovascular disease in high-risk adults with type 2 diabetes using blood pressure and lipid lowering approaches.

Our message to individuals with type 2 diabetes who are at especially high risk for heart disease is to target your A1C level to about 7 percent and not to more intensive levels. No one with diabetes should change their treatment without consulting with their healthcare professional first and we concur with the ADA recommendations that treatment must be individualized.

We have several experts with us today who have contributed to this research, and who will now provide additional detail about the ACCORD study and these important findings:

Dr. William Friedewald, Clinical Professor of Public Health and Medicine at Columbia University, and chairman of the ACCORD Steering Committee.

Dr. Hertzel Gerstein, professor at McMaster University and Hamilton Health Sciences in Ontario, where he holds the Population Health Institute Chair in Diabetes Research. Dr. Gerstein is principal investigator of one of the seven ACCORD clinical center networks and led the group that designed the blood sugar treatment approaches.

Dr. Judith Fradkin is the director of the Division of Diabetes, Endocrinology and Metabolic Diseases at the National Institute for Diabetes and Digestive and Kidney Diseases here at the NIH and the key person from NIDDK involved in ACCORD over its duration.

Also with us to answer your questions will be Dr. John Buse, Professor of Medicine and Chief of Endocrinology at the University of North Carolina School of Medicine and vice chair of the ACCORD Steering Committee; Dr. Buse also serves as President of Medicine and Science of the American Diabetes Association. We also have Dr. Robert Byington, professor in the Department of Epidemiology and Prevention in the Division of Public Health Sciences at Wake Forest University School of Medicine, who leads the ACCORD Coordinating Center, and Dr. Denise Simons-Morton, project officer for ACCORD at the NHLBI and a member of the ACCORD steering committee.

Other Prepared Remarks

Summary Remarks

Thank you, Judy. And, thank you to all of our speakers this morning. In summary, we have discontinued the intensive blood-sugar treatment strategy in the ACCORD trial, and will now treat all the ACCORD participants according to the standard blood-sugar treatment strategy. The ACCORD blood pressure and lipid trials are continuing.

The findings we are reporting to you today are extremely important for the care of individuals with diabetes around the world. They indicate that in older patients with diabetes who also have existing cardiovascular disease or two or more CVD risk factors (such as high blood pressure or elevated blood cholesterol), care should be taken to not intensively lower their blood sugar levels to a near-normal level using combinations of medications available today.

Now we will be pleased to take your questions. Please identify yourself and your organization.

Question and Answer Period

Closing Remarks

I want to thank all of you who have called in today. You have a vital role in communicating this important health information to the public.

A recording of this press conference will be available at the following phone number within about 30 minutes.

(800) 475-6701, Access Code: 909685

The press release and Questions and Answers sheet have been posted to the NHLBI website, at A transcript of this teleconference will be posted in the next two days.

Last updated February 6, 2008

Twitter iconTwitterimage of external icon Facebook iconFacebookimage of external icon YouTube iconYouTubeimage of external icon Google+ iconGoogle+image of external icon