Two–Cut-Point Method Improves Accuracy of Hemoglobin A1c in Diagnosing Diabetes

This study assesses a strategy that I think is quite reasonable, and was suggested in the American Association of Clinical Endocrinologists’ position statement a number of months ago.

One must recognize that a "negative" hemoglobin A1c level (below 6.5%) misses from one-third to one-half of those with diabetes by glucose tolerance test criteria, whereas a "positive" value (6.5% or greater) may not be the result of diabetes in persons who have greater degrees of hemoglobin glycation. Because high glycation is present in blacks, older populations, and people with iron deficiency, and also is a common variant in the overall population, I would even suggest that blood glucose confirmation – although not necessarily with glucose tolerance testing – should be done in all persons with high A1c, regardless of the level.

Similarly, there are people whose degree of hemoglobin glycation is lower than average. Thus, if there is clinical reason to look for diabetes, it is reasonable to perform glucose tolerance testing even with rather low A1c levels.

Given this inherent variability in glycation, just as the 6.5% diagnostic cutoff is incorrect for many persons whose diabetes status is being ascertained, the use of a specific A1c goal of, say, 6.5% or 7.0%, may not be appropriate for all patients with known diabetes. Again, assessment of actual blood glucose levels is crucial in the management of diabetes.

Zachary T. Bloomgarden, M.D., of the Mount Sinai School of Medicine in New York, is on the speakers bureau for Merck, Novo Nordisk, and GlaxoSmithKline; serves on an advisory panel for Merck, Bristol-Myers Squibb, AstraZeneca, Boehringer Ingelheim, and Biodel; is a consultant for Merck, Novartis, Dainippon Sumitomo Pharma America, and Forest Laboratories; and is a stock shareholder of Covidien, C.R. Bard, Novartis, Roche, and Stryker Corp.