ORLANDO, Fla.—Despite a recent recommendation that hemoglobin A1c (HbA1c) be used as a standard diagnostic test for diabetes, the proposed diagnostic criteria fail to pick up a significant number of cases of diabetes and prediabetes, investigators reported at the 70th Annual Scientific Sessions of the American Diabetes Association.
“The bottom line is that if you use an HbA1c measurement as a screening test, you’ll have a lot of false-negatives and some false-positives,” said Mary Rhee, MD, Assistant Professor of Endocrinology at Emory University in Atlanta. “And in this day and age when diabetes is a worsening public health epidemic not only in the U.S. but also around the world, we can’t afford to end up with more cases of undiagnosed diabetes.”
She emphasized, however, that the HbA1c remains a useful tool for monitoring diabetes.
Dr. Rhee and colleagues compared the identification of prediabetes, dysglycemia (diabetes or prediabetes), and diabetes by the recommended HbA1c-based screening with the conventional oral glucose tolerance test (OGTT).
“Early identification of diabetes and prediabetes can allow the initiation of treatment aimed at preventing complications or delaying disease progression,” Dr. Rhee pointed out. ”However, most U.S. healthcare systems do not have systematic screening programs, and many Americans with diabetes or prediabetes remain undiagnosed.”
The ADA recently proposed new diagnostic criteria based on HbA1c levels, she added. However, the new criteria have not been rigorously examined.
According to the proposed criteria, an HbA1c of 6.5% or higher indicates the presence of diabetes, and an HbA1c of 5.7% to 6.4% means that the patient is pre-diabetic.
In a cross-sectional analysis, the researchers compared HbA1c and gold standard OGTT measurements in non-Hispanic white and black adults without known diabetes drawn from three populations.
Overall, the analysis examined data from 1581 patients in the Screening for Impaired Glucose Tolerance (SIGT) study, 2014 patients who participated in the Third National Health and Nutrition Examination Survey (NHANES III), and 1111 patients who participated in NHANES 2005-2006.
Overall, HbA1C testing found lower prevalences of diabetes and prediabetes than identified by the OGTT. OGTTs identified prediabetes in 35.8% and diabetes in 5.2% of the combined patients in the three populations, whereas the HbA1c criteria identified only 19.5% of the subjects as prediabetic and 2.2% as having diabetes.
In addition, if HbA1c were used to screen, about 70% of cases of prediabetes or diabetes identified by OGTT would be missed and roughly 12% of patients with normal glucose tolerance would be falsely identified as having prediabetes or diabetes.
“If we assume that NHANES data are representative of the U.S. population, our findings mean that about 5.9 million non-Hispanic adult Americans with unrecognized diabetes and 43 to 52 million with prediabetes would be missed by screening with the proposed HbA1c criteria,” Dr. Rhee said.
The study also revealed that the screening ability of HbA1c in non-Hispanic Americans differed by race. For example, researchers observed more false- negative results and fewer false-positive results in whites than blacks.
Given that the proposed HbA1c diagnostic for screening are both insensitive and racially discrepant, what is the optimal way to screen for diabetes? Dr. Rhee recommended that doctors perform the standard 75 gram OGTT, or, at the very least, a fasting glucose level followed by the OGTT in individuals whose levels are abnormal (100 mg/dL and above). Findings from a recent study, she said, suggest that the 50 gram glucose challenge test, a test used to screen for gestational diabetes in pregnancy, may represent a promising alternative to HbA1c as a screening test for diabetes or prediabetes (Phillips et al. Diabetologia. 2009;52:1798-1807) because it has been shown to be accurate and can be readily performed during office visits at any time of the day and without a prior fast.