News & Research

Translating the A1C assay into estimated average glucose values, by David M. Nathan and colleagues. Diabetes Care 31:1473?1478, 2008

What is the problem and what is known about it so far?

The A1C test gives health care providers an idea of how high a patient?s blood glucose levels have been during the past 3 months. It is used to judge how well a patient?s diabetes treatment regimen is working. However, patients can become confused because A1C results are given as a percentage (your most recent A1C may be 7.6%, for example), whereas day-to-day self-monitoring of blood glucose results are given in the units of milligrams per deciliter or millimoles per liter (your meter may show that your blood glucose before lunch on Tuesday is 112 mg/dl or 6.2 mmol/l). A current proposal for one set of units for A1C testing around the world is different from the way A1C is usually reported in the United States, and that could lead to even more confusion for patients and health care providers alike.

Why did researchers do this particular study?

It would be helpful to translate A1C goals and results into an average blood glucose level, which would use the same milligrams-per-deciliter units that patients are used to seeing from checking their blood glucose levels. The researchers wanted to find a reliable way to do that.

Who was studied?

The study included more than 500 people of various racial and ethnic groups, including some with type 1 diabetes, some with type 2 diabetes, and some without diabetes, from 10 medical centers around the world.

How was the study done?

The researchers performed 2 days of continuous glucose level checking on the patients four times during a 3-month period and also measured patients? blood glucose with a meter seven times a day at least 3 days per week during the 3-month period. From this data, they calculated the patients? average blood glucose over the 3 months and compared it to their A1C results at the end of the 3 months.

What did the researchers find?

The researchers were able to perform statistics tests to turn blood glucose monitoring results into an estimated average glucose, or eAG, value that correctly reflects A1C levels. These numbers remained correct regardless of patient characteristics, such as race, ethnic background, and age.

What were the limitations of the study?

Some ethnic/racial groups were not fully represented. Also, because blood glucose was measured with both continuous glucose monitoring and blood glucose meters, combining those numbers required some statistical adjustments that may have affected the results. Finally, because all of the diabetic patients in the study had stable blood glucose control and no red blood cell disorders, the results only directly apply to similar patients. More research will be needed to see if the eAG calculations are accurate for other groups, including children and pregnant women, who were not included.

What are the implications of the study?

A1C test results can be accurately reported as eAG levels in milligrams per deciliter instead of as percentages. For patients used to measuring their blood glucose, it may be easier to aim for an average blood glucose of 154 milligrams per deciliter than to aim for an A1C of 7%, even though the two indicate the same level of blood glucose control.

FOR MORE INFORMATION