Low blood glucose levels do not explain the excess deaths seen in the intensive control group of the Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial, according to a report presented at the American Diabetes Association's 69th Scientific Sessions.

"After the original ACCORD report showed an increased mortality risk in the intensive strategy group aiming to lower A1C level under 6%, concern arose that this approach might be dangerous," said Matthew C. Riddle, M.D., Professor of Medicine, Oregon Health Science University and a member of the Glycemia Management Group of ACCORD. He was also a site principal investigator in the study. While he confirmed the estimate reported last year of a 20% increased risk of death associated with ACCORD's intensive treatment strategy, new analyses did not confirm the proposed theory that low A1C levels might be the cause.

"An A1C below 7% alone does not appear to explain the excess deaths in the ACCORD trial and is not necessarily a predictor of mortality risk," said Riddle. "Further, the rate of one-year change in A1C showed that a greater decline in A1C was associated with a lower risk of death." Analyses to identify the cause of the increased mortality, such as hypoglycemia, weight gain, or a particular drug or combination of drugs - are continuing.

"However, we found a 20% higher risk of death for every 1% higher A1C level above 6%, suggesting that lower blood glucose levels may be a worthy target in some patients," said Riddle. That higher risk of death remained even after statistical adjustment for age, duration of diabetes, or prior cardiovascular events. This analysis is consistent with analyses from other epidemiologic studies showing a relationship between higher A1Cs and higher risk of death.

The goal was for the intensive group to reach an A1C level of less than 6% and for the standard group goal to reach between 7 and 7.9%.

A closer look at the role of hypoglycemia was presented by Denise Bonds, M.D., MPH, who was an investigator at ACCORD's Coordinating Center and who is now a medical officer for the trial at the National Heart, Lung, and Blood Institute (NHLBI) at the National Institutes of Health (NIH).

The study defined severe hypoglycemia as a blood glucose level of less than 50 mg/dl or symptoms consistent with hypoglycemia and recovery with glucagon or an oral carbohydrate. It was further divided into those who required any assistance versus those who required emergency medical assistance at home or in a hospital.

"Some predictors of hypoglycemia cannot be changed. For example, we found that more frequently those who required medical assistance for severe hypoglycemia were female or African-American, had a longer duration of diabetes, or had entered the study already receiving insulin," said Bonds. "However, more than half of severe hypoglycemic events were preceded by a missed meal, delayed meal, or a meal with fewer carbohydrates than normal - information that individuals with diabetes can use to modify their behavior and help in preventing and reducing hypoglycemic events."

About one-third of severe hypoglycemic events resulted in confusion or abnormal behavior for the individual, one-third resulted in loss of consciousness, and one-third required hospitalization.

A total of 451 deaths occurred in the ACCORD standard and intensive groups combined, and 7% of those individuals had at least one severe hypoglycemic event requiring medical assistance.

"We found that severe hypoglycemia in both groups was associated with a higher risk of death but, among those who had severe hypoglycemia in the intensive arm, it was associated with a lower risk of death compared to those who had severe hypoglycemia in the standard group," said Bonds.

The explanation for this is unclear, but it must be remembered that patients in the intensive group were seen more frequently, had more phone calls with health care providers, and more frequent but milder hypoglycemic events. It is theorized that these patients learned how to recognize and treat themselves at the first signs of hypoglycemia or perhaps that the hypoglycemia that caused mortality was of a different type.

The ACCORD group also adjudicated each death for the potential role of hypoglycemia. Of the 451 deaths, they were only able to identify one case where it was concluded that severe hypoglycemia definitely played a role. However, there cannot be complete certainty because glucose measures are not available near or at the times of deaths. Continuous glucose monitoring was rarely done during the trial.

ACCORD enrolled 10,251 adults at 77 clinics in the U.S. and Canada. The average age at study entry was 62 years. The 35% who had prior cardiovascular events (myocardial infarction and strokes) were the secondary prevention group, while the balance who were at high risk for such events had problems such as left ventricular hypertrophy, microalbuminuria, or at least two of the following risk factors: high LDL, low HDL, high blood pressure, an elevated BMI, or were smokers.

ACCORD was testing an intensive strategy of glycemic control versus a standard strategy, so intensive group patients had more frequent clinic visits and were required to do more frequent self-monitoring of blood glucose because they modified certain medications themselves in response to self-monitoring results. The same menu of medications was available to physicians treating both groups, but patients in the intensive group tended to be prescribed more of them, based on the individual physician's clinical judgment, to get to the target. A broad comparison is that 52% of intensive patients vs. 16% of those in the standard group were likely to be on insulin plus three oral agents.

ACCORD's comparison of two forms of glycemic therapy ended after 3.4 years of follow up due to the excess deaths in the intensive group, and that experience forms the basis for all reports to date. However, the trial has continued and, together with other studies for blood pressure and lipid control, will end June 30th, with complete analyses of all three studies expected in the next year.

ACCORD is primarily sponsored by NHLBI, with additional funding and scientific expertise from the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK). Other components of the NIH - the National Institute of Aging and National Eye Institute - as well as the Centers for Disease Control and Prevention, provide support substudies.