The Fallacy of Average: How Using HbA1c Alone to Assess Glycemic Control Can Be Misleading

Abstract

HbA_1c is a valuable metric for comparing treatment groups in a randomized trial, for assessing glycemic trends in a population over time, or for cross-sectional comparisons of glycemic control in different populations. However, what is not widely appreciated is that HbA_1c may not be a good indicator of an individual patient's glycemic control because of the wide range of mean glucose concentrations and glucose profiles that can be associated with a given HbA_1c level. To illustrate this point, we plotted mean glucose measured with continuous glucose monitoring (CGM) versus central laboratory-measured HbA_1c in 387 participants in three randomized trials, showing that not infrequently HbA_1c may underestimate or overestimate mean glucose, sometimes substantially. Thus, if HbA_1c is to be used to assess glycemic control, it is imperative to know the patient's actual mean glucose to understand how well HbA_1c is an indicator of the patient's glycemic control. With knowledge of the mean glucose, an estimated HbA_1c (eA1C) can be calculated with the formula provided in this article to compare with the measured HbA_1c. Estimating glycemic control from HbA_1c alone is in essence applying a population average to an individual, which can be misleading. Thus, a patient's CGM glucose profile has considerable value for optimizing his or her diabetes management. In this era of personalized, precision medicine, there are few better examples with respect to the fallacy of applying a population average to a specific patient rather than using specific information about the patient to determine the optimal approach to treatment.

Figure 1

Plot of CGM-measured mean glucose concentration vs. laboratory-measured HbA_1c. The shaded area represents the 95% prediction interval (analogous to an individual CI) for a patient’s mean glucose concentration for a measured HbA_1c level, demonstrating the wide range of mean glucose concentration values that are possible for any HbA_1c value.

Figure 2

Plot of laboratory-measured HbA_1c vs. CGM-measured mean glucose concentration used to derive eA1C. The shaded area represents the 95% CI for the population mean HbA_1c estimated from a mean glucose concentration. Equation to estimate HbA_1c for a given mean glucose concentration: eA1C = 3.38 + 0.02345 × [mean glucose] (23,24).

Figure 3

AGPs from four patients with laboratory-measured HbA_1c of 8.0%. AGPs are shown for four adults with type 1 diabetes using multiple daily injections of insulin, all with an HbA_1c of 8.0% from the same central reference laboratory. Displayed are 2 weeks of CGM data (up to 288 CGM values/day) for each patient measured just prior to the HbA_1c laboratory test. The CGM is displayed as a modal or standard day. Shown are the median lines of all the glucose values over 2 weeks, the 25% and 75% lines (enclosing the shaded interquartile range [IQR]), and the 10% and 90% lines (dashed). The hatched area is the target glucose range of 70–180 mg/dL. Clinical note 1: Although each of the four patients has the same HbA_1c, the AGP patterns are very different and would suggest different insulin and or lifestyle interventions. Clinical note 2: The mean glucose varies from 156 to 195 mg/dL among the four patients. For patients 1 and 2, eA1C (8.4%) based on mean CGM glucose (195 mg/dL) is slightly higher than the measured HbA_1c (8.0%), indicating that the measured HbA_1c is slightly underestimating the mean glucose and that despite the same mean glucose, the daily pattern varies considerably. For patients 3 and 4, eA1C (7.0% and 7.3%, respectively) based on the mean CGM glucose (156 and 163 mg/dL, respectively) is substantially lower than the measured HbA_1c (8.0%), indicating that the measured HbA_1c is overestimating the mean glucose. Again, despite similar mean glucose concentrations, the daily pattern has considerable variation between the two patients. Avg, average.

Figure 3

AGPs from four patients with laboratory-measured HbA_1c of 8.0%. AGPs are shown for four adults with type 1 diabetes using multiple daily injections of insulin, all with an HbA_1c of 8.0% from the same central reference laboratory. Displayed are 2 weeks of CGM data (up to 288 CGM values/day) for each patient measured just prior to the HbA_1c laboratory test. The CGM is displayed as a modal or standard day. Shown are the median lines of all the glucose values over 2 weeks, the 25% and 75% lines (enclosing the shaded interquartile range [IQR]), and the 10% and 90% lines (dashed). The hatched area is the target glucose range of 70–180 mg/dL. Clinical note 1: Although each of the four patients has the same HbA_1c, the AGP patterns are very different and would suggest different insulin and or lifestyle interventions. Clinical note 2: The mean glucose varies from 156 to 195 mg/dL among the four patients. For patients 1 and 2, eA1C (8.4%) based on mean CGM glucose (195 mg/dL) is slightly higher than the measured HbA_1c (8.0%), indicating that the measured HbA_1c is slightly underestimating the mean glucose and that despite the same mean glucose, the daily pattern varies considerably. For patients 3 and 4, eA1C (7.0% and 7.3%, respectively) based on the mean CGM glucose (156 and 163 mg/dL, respectively) is substantially lower than the measured HbA_1c (8.0%), indicating that the measured HbA_1c is overestimating the mean glucose. Again, despite similar mean glucose concentrations, the daily pattern has considerable variation between the two patients. Avg, average.