Objective: To review recent literature on the limitations of hemoglobin A(1c) (HbA(1c)) as a marker of glycemic control.
Methods: English-language literature published between 1985 and 2011 was reviewed specific to analyses of major trials relating glycemic control to complications of diabetes mellitus, as expressed through HbA(1c) as a marker of glycemic control.
Results: HbA(1c) has been accepted as the most fundamental biomarker in diabetes, if not all of medicine, as it clearly predicts risk for diabetes-related complications. What is not generally appreciated is that HbA(1c) is a crude marker of glycemia with many limitations. It is now accepted that HbA(1c) does not reflect mean glucose for many people, and even for those it does, any level could represent a wide range of glycemia. While we have learned HbA(1c) is not a perfect biomarker, we also know that in the Diabetes Control and Complications Trial, HbA(1c) could only explain 11% of the variation in retinopathy risk between the conventional and intensive therapy groups. This important finding suggests that other glycemic and nonglycemic factors may be responsible for the pathogenesis of diabetes-related complications. One candidate is glycemic variability, which must be differentiated from postprandial hyperglycemia since hypoglycemia can also result in inflammatory activation. Importantly, although it is clear that in insulin-requiring patients glycemic variability is associated with hypoglycemia, we require a definitive prospective trial to confirm glycemic variability's association with one or more vascular complications.
Conclusions: What is abundantly clear is that the HbA(1c) message, as we know it, is too simplistic. While certain wholesome concepts such as motherhood and apple pie are accepted by all, the HbA(1c) message may be more complex than originally appreciated, and it may be time to reevaluate our most basic premise in diabetes.