Diabetes doctors debate the best diagnostic tests

Diabetes rates have climbed steeply in this country in the last two decades. Fortunately, scientists' knowledge of how best to manage the disease is advancing as well. Researchers gathered at the American Diabetes Assn. meeting last week in New Orleans to consider the latest developments, including a new test for diagnosing the disease and a better understanding of the risks of aggressive control of blood sugar.

The current techniques for diagnosing Type 2 diabetes involve either a fasting plasma glucose test or, less commonly, an oral glucose tolerance test. In the former, the patient must fast for 12 to 14 hours before the physician measures the level of glucose in his or her blood. In the latter, the patient drinks a solution of concentrated sugar, and blood sugar levels are measured two hours later.

The first test can be affected by eating during the fast, and both can be affected by illness.

Many doctors now believe that a blood analysis for glycated hemoglobin, also known as hemoglobin A1c, is a better test. A1c is a measure of the patient's blood sugar levels for the two to three months preceding the test. The test requires only a sample of blood.

But diabetes groups have been reluctant to recommend wider use of it for diagnosis because only recently have most labs begun using a standardized test for A1c.

A 21-member committee including scientists from the American Diabetes Assn., the International Diabetes Federation and the European Assn. for the Study of Diabetes has been studying literature on the test for several months and, at the meeting, recommended that it be adopted as the primary diagnostic test.

"This is the first major departure from the way diabetes has been diagnosed in the past 30 years," Dr. David Nathan of Harvard Medical School, chairman of the committee, told a news conference.

The committee also concluded that the A1c test is a more accurate predictor of a diabetic's likelihood of developing complications of the disease. Healthy people have an A1c level of 6% or lower; diabetics typically have a level well above 7%. The committee recommended that a cutoff level of 6.5% be used for a diagnosis.

Assessing risks

A1c levels have also been a crucial end point in two large trials studying aggressive control of diabetes.