Hemoglobin A1c Not Reliable in Diagnosing Type 2 Diabetes

With Maria Mercedes Chang Villacreses, MD, and Elena Christofides, MD, FACE

The test that doctors most often rely on to detect a person’s risk for prediabetes and type 2 diabetes—the hemoglobin A1c (A1c) blood test—too often delivers a poor reading, thereby missing the diagnosis in nearly three out of four at-risk individuals,1 according to research presented at the Endocrine Society meeting in New Orleans, Louisiana.

The investigators compared the accuracy of the A1c test with the other screening method used to assess patients’ risk of diabetes—the oral glucose tolerance test (OGTT)—to arrive at this startling conclusion.1

A1c test misses 3 in 4 people at risk for diabetes.The common use of the hemoglobin A1c test to screen for prediabetes and type 2 diabetes should be skipped in place of more reliable tests. Photo: 123rf

"The A1c missed almost 73% of the people with diabetes in comparison to the oral glucose tolerance test," says Maria Mercedes Chang Villacreses, MD, a clinical endocrinology fellow at the City of Hope Diabetes and Metabolism Research Institute in Duarte, California, who introduced the findings at the annual meeting of the Endocrine Society but are considered preliminary until published in a peer-reviewed medical journal.

The A1c lab value represents a look back at average blood glucose levels over the previous three months. It has been commonly used to detect diabetes for the last decade. Unlike other tests, no fasting is needed before you have blood drawn to assess your A1c level, making it more convenient for people.2

Since the fasting blood test requires a person to fast, the OGTT measures the body's response to sugar; it requires a person to fast overnight. First, blood is taken, then the person drinks a sugary drink. Blood is taken again two hours later.

Large Trial Compared Usual Tools for Assessing for Diabetes in Adults

The researchers looked at data from 9,000 adults, ages 20 years and older, from the 2005-2014 National Health and Nutrition Examination Survey (NHANES). The information collected by the research team included body weight (based on body mass index [BMI]) and blood test results.

Based on the fasting blood glucose test and the OGTT, 765 patients were diagnosed as having type 2 diabetes (T2D). However, only about 27% of these individuals were classified as having diabetes based on their A1c levels, which is how Dr. Villacreses and her team determined that nearly three-quarters of those at risk for diabetes were not aware that they had this chronic disease, and therefore were not receiving treatment.1

"Most worrisome, 73 % of patients would have missed out on early intervention and treatment," she tells EndocrineWeb. While the A1c test is convenient, ''we recommend that we do not rely solely on this number," Dr. Villacreses says.

The guidelines for diagnosis and treatment of type 2 diabetes from the American Diabetes Association (ADA) already advise against relying solely on A1c,3 she says. “While the American Diabetes Association guidelines specify that diabetes can be diagnosed based on fasting plasma glucose (FPG), the OGTT, or the A1c, our findings confirm that reliance on A1c remains the least reliable method for assessing diabetes risk.”

The choice of which tests to use is of course left to the discretion of the physician, Dr. Villacreses says. "We wanted to bring this up to emphasize the gap [in the accuracy of detection]," she says, between the A1c and the other tests.1

Closer attention to ethnic and racial differences is critical when checking patients for blood glucose levels as part of an evaluation of risks for diabetes. The individuals included in this study represented a diverse patient population, which is very important, according to the researchers; they found that both race and ethnicity introduce differences that affected the accuracy of the A1c testing.1

When they looked at the accuracy of the A1c test at measuring normal blood glucose levels across the entire patient population, they discovered that the results were lowest in African Americans at 71%, followed by 85% in Latinos overall and 86% among Mexican Americans. The accuracy was highest for patients who were Caucasian, at 89%.

Yet, Dr. Villacreses said that they couldn’t explain the reasoning behind the racial and ethnic differences in A1c results but that it’s important for both patients and doctors to be aware of this.

Best to Avoid Reliance on A1c in Making an Initial Diagnosis of Diabetes

The shortcomings of the A1c test are known, says Elena Christofides, MD, FACE, an endocrinologist and CEO of Endocrinology Associates, in Columbus, Ohio, upon reviewing the study for EndocrineWeb.

While she takes issue with some of the researchers statistical methods, such as how they grouped those with an abnormal glucose tolerance test together with those with a full-blown diabetes (eg, having a diagnosis of T2D) for the analysis, she says they are not the first to conclude that the A1c test is not a great tool to capture diabetes diagnoses accurately for most individuals.

When the ADA guidelines were in development,3 Dr. Christofides tells EndocrineWeb, ''there were significant arguments back and forth about even permitting the continued use of the HbA1c test…due to the fact that it would indeed miss a number of individuals. That is why the use of the A1c is still being used but the OGTT is still thought to be the more definitive test."

"The guidelines also state that if the A1c is normal and there is still a suspicion of prediabetes or diabetes, doctors are advised to do the OGTT anyway," she says. So an A1c can be done so long as you are then advised to have another test to check your blood sugar levels if there is any concern regarding your risks.

To fully understand the reason that problems arise with the reliability of the A1c test, Dr. Christofides explains that when circulating glucose builds up in the blood, it binds to the hemoglobin molecule in the red blood cell, and the A1c measures how much glucose is bound.

As the differences found among different patient populations, she says that genetic mutations are more likely to occur in certain ethnicities, which alter the way the glucose binds to the hemoglobin molecule—thereby giving false readings more often. African Americans, for instance, carry a variant form of a gene for the sickle cell anemia that is involved in abnormal binding of blood glucose, she says, so the A1c test is artificially lower than expected in these patients.

What Does This Mean for People Wondering About Their Diabetes Risk

For individuals who haven’t received a diagnosis of diabetes, Dr. Villacreses says, one important lesson from these fidings is that ''you should not feel 100% reassured if your A1c number is less than 6.5% that you are do not have diabetes.” A test result that is greater than 6.4% defines the beginning of diabetes, so you may have prediabetes or have already progressed to type 2 diabetes.

The sooner you receive a diagnosis regarding your diabetes status, the earlier treatment can begin. This is important because by understanding your risks, and making the necessary lifestyle changes, you can prevent prediabetes from progressing or even reverse the type 2 diabetes.

In addition, she says, patients should feel comfortable asking their doctor about the choice of tests and indicate your desire to skip the A1c, and have one of the other more reliable screening methods, while more time consuming, to determine if you are at risk for pre-diabetes or diabetes.

Dr. Villacreses has received research funding and speaking fees from a variety of pharmaceutical companies but none that pose a conflict in this research. Dr. Christofides has no relevant financial disclosures.

Last updated on
Continue Reading
Diabetic Hyperlipidemia: Managing High Cholesterol When You Have Diabetes