Many professional societies, including the American Diabetes Association (ADA), recommend using Hemoglobin A1c in screening and diagnosis of diabetes because it reflects long-term glycemic status and does not require fasting prior to sample collection. However, since ADA’s recommendation in 2010, considerable debate has surrounded the universal cutoff of ≥6.5%. A large body of evidence has demonstrated higher HbA1c values in certain ethnic populations, but the controversy centers on whether these increases indicate higher levels of glycemia or if they reflect differences in the rate of glycation between ethnic groups.
If HbA1c measurements are affected by non-glycemic factors, is the use of a single cutoff of ≥6.5% appropriate for all patients? Or should we take a more nuanced approach by defining different cutoffs for different patient populations?
In an afternoon symposium today, “Influence of Non-Glycemic Factors on Hemoglobin A1c: Fact and Fiction”, David Sacks, MD, William Herman, MD, and Elizabeth Selvin, PhD will discuss HbA1c measurement from a laboratory perspective and debate the topic of ethnic factors that impact HbA1c measurements independently of blood glucose.
Sacks will begin by highlighting factors other than glycemia that alter HbA1c values, including red blood cell lifespan, hemolysis, and hemoglobin variants. He will describe analytical challenges associated with measuring HbA1c and illustrate how certain patient characteristics can lead to HbA1c measurements that may not reflect their glycemic status. Herman will follow with an analysis of how race alters HbA1c independently of glycemia.
Several studies, including NHANES III and ARIC, showed higher average levels of HbA1c in Hispanics and blacks relative to whites. It is hypothesized that differences in erythrocyte permeability, different rates of intraerythrocytic glucose metabolism or glucose attachment and release may explain the disparities in mean HbA1c reported between different ethnic groups. Herman will point out that a standard cutoff of 6.5% may result in overdiagnosis of diabetes in ethnic groups shown to have a higher mean HbA1c value.
As an alternative, he will advocate a stepwise approach where HbA1c values <5.5% or >7.0% may be accepted as diagnostic in all ethnic groups, but fasting glucose measurements should be used to diagnose individuals whose HbA1c values fall in the gray zone, just above or below the cutoff of 6.5%.
Selvin will conclude the symposium by asserting that while there may be ethnic differences that affect HbA1c independently of glycemia, their relative impact on HbA1c values is not clinically significant. She will note that higher mean HbA1c values in non-whites are mainly accounted for by lifestyle and socioeconomic status, with only a minor contribution from genetic factors associated with particular ethnic groups.
To support her argument, she will point to studies that document elevations in alternate markers of glycemia in non-whites, suggesting that higher HbA1c values can be attributed to higher levels of blood glucose— potentially indicating delays in diagnosis—rather than non-glycemic factors. She will advocate against ethnic-specific HbA1c cutoffs on the grounds that race does not modify the relationship between HbA1c and adverse cardiovascular outcomes.
It is unlikely that the controversy surrounding the use of HbA1c in screening and diagnosis of diabetes will be settled anytime soon. In the meantime, it is essential that clinical laboratorians understand the strengths and limitations of current HbA1c testing methodologies and comprehend non-glycemic factors that can influence HbA1c values in order to assist clinicians in the diagnosis and management of patients with diabetes.