HbA_1c Levels Above 8% Pose All-Cause Mortality Risk

MONTREAL — No difference in mortality was found at 4 years between baseline hemoglobin A_1c levels of less than 6.5% and levels of 6.5%–7.0% in a prospective observational study of nearly 3,000 unselected patients with type 2 diabetes.

However, the Diabetes in Germany (DIG) study also found a dramatically increased risk of mortality for those with baseline HbA_1c levels greater than 8%, compared with those who began the study with lower HbA_1c values. Other baseline predictors of mortality included age, smoking, cardiovascular disease, and systolic blood pressure, Dr. Markolf Hanefeld reported at the World Diabetes Congress.

“In a diabetes population rather well controlled for hemoglobin A_1c, smoking status and good blood pressure control are of utmost importance for survival. However, at a level greater than 8%, [the degree of] glucose control becomes a serious risk factor for all-cause mortality,” said Dr. Hanefeld of the Center for Clinical Studies, Technical University, Dresden, Germany.

Of an initial 4,020 unselected patients aged 35–80 years with type 2 diabetes from 238 sites in Germany, 2,784 completed the study at a median of 3.7 years and 175 died during that time. Most (86%) had no history of a major adverse cardiovascular event (MACE) at baseline, while 251 (8.5%) reported a first MACE during follow-up.

The average baseline HbA_1c for the entire group was 7.0%. Thirty-seven percent met the International Diabetes Federation's and American Association of Clinical Endocrinologists' target HbA_1c of less than 6.5%, whereas 57% met the American Diabetes Association's target of less than 7.0%. However, 29% had HbA_1c values above 7.5%. The average HbA_1c level for the entire group did not change over the 4-year period, Dr. Hanefeld said.

Among those who died during the study period, 6% had baseline HbA_1c values of less than 6.5%; 5.3% had values of 6.5%–6.9%; 5.1% had values of 7.0%–7.9%; and 7.6% had values of 8% or higher. The same trend was seen in MACE.

In a multivariate analysis, the most significant factor predicting mortality was MACE at baseline, conferring a twofold greater risk. Also significant were smoking, age, and systolic blood pressure. Female gender cut the risk by half. HbA_1c did not contribute significantly to mortality, he said.

A comparison of these DIG findings with the standard care arms of the recent randomized, controlled glucose-lowering trials ADVANCE (Action in Diabetes and Vascular Disease), ACCORD (Action to Control Cardiovascular Risk in Diabetes), and VADT (Veterans Affairs Diabetes Trial) shows no link between HbA_1c and mortality. In fact, the standard care arm of the ADVANCE study had the highest annual death rate (1.92%) but the second-lowest average HbA_1c (7.5%). The 7.0% average HbA_1c in DIG was the lowest of the four trials, but its annual mortality rate was 1.59% (for the entire group, since all were in “standard” care), higher than the 1.14 annual death rate in the standard care arm of ACCORD. That death rate in ACCORD's standard care arm was the lowest of the four studies, while the mean HbA_1cwas the second highest (8.3%, vs. 9.4% in VADT).

Dr. Hanefeld said he had no conflicts of interest.