How Reliable Is Hemoglobin A1c?

April 6, 2017

Grafik_blutkreislaufBy Alexa Yuen

Peer Reviewed 

According to the CDC, there are 22 million people with an established diagnosis of diabetes mellitus in the United States, but more frightening is that the rate of diabetes continues to rise both nationally and globally, with predictions that 7.7% of the world’s adult population will be afflicted with the disease in 2030.[1],[2] Preventing, monitoring, and managing this disease is of utmost importance in responding to the growing epidemic. The hemoglobin A1c (HbA1c) blood test has become a powerful tool in this effort, but it is imperative that healthcare providers be aware of both its strengths and limitations.

Diabetes is currently diagnosed in one of four ways:

–Fasting plasma glucose ≥126 mg/dL

–Two-hour oral glucose tolerance test plasma glucose ≥200 mg/dL

–HbA1c of ≥6.5%,

–Random plasma glucose of ≥200 mg/dL with symptoms of hyperglycema.[3]

For each test, a positive result needs to be repeated for confirmation unless the diagnosis of diabetes is obvious. HbA1c has several advantages over the other tests.  It is a measure of glycemia over a 2-3 month period and sheds more light on overall glucose levels than a single plasma glucose measurement. Consequently, it does not require fasting and is unaffected by acute perturbations like stress or recent exercise. In addition to convenience, the pre-analytic stability of HbA1c is superior to that of the plasma glucose test.[4]  After blood is drawn and centrifuged, glucose consumption occurs and glucose concentration decreases at 5% to 7% per hour, with a greater rate at higher temperatures.[5], [6] Often, blood samples are processed hours after withdrawal and, as a result, glucose readings are falsely lowered.  The pre-analytical variability of fasting plasma glucose is 5% to 10%, while the pre-analytic variability for HbA1c is negligible.4 Most importantly, HbA1c is useful in preventing, diagnosing, and monitoring diabetes. HbA1c can identify individuals with prediabetic results between 5.7% and 6.4% as being at higher risk for diabetes, prompting preventive measures. For people with diabetes, HbA1c is useful in reflecting the effect of lifestyle changes and medications on glycemic control over the past few months, and whether further changes are needed for management.

HbA1c measures hemoglobin glycation from exposure to glucose over the 120-day lifespan of a red blood cell (RBC). As a result, conditions that affect the duration of RBC circulation or the production of RBCs can produce HbA1c readings that do not correspond to actual plasma glucose concentrations.

Conditions that reduce the average age of red blood cells in circulation can falsely lower HbA1c values.  These conditions include hemolytic anemia; hemodialysis and chronic kidney disease (which are associated with a 50% reduction in RBC lifespan); erythropoietin therapy, which increases red blood cell production; blood loss; and transfusions.

Anemia from deficiencies in iron, folate or vitamin B12 reduce the rate of red blood cell production, increase average red cell age, and thereby falsely elevate HbA1c readings. Functional asplenia increases average red cell age due to failure to clear older red cells from circulation.[7]  Additionally, chemical changes to hemoglobin, including acetylation of hemoglobin following high-dose aspirin use (500 mg/day) and carbamylation of hemoglobin from elevated urea in renal failure can make it difficult to distinguish affected hemoglobin from glycated hemoglobin in the HbA1c assay, thus falsely elevating values.[8] Furthermore, hemoglobinopathies like sickle cell anemia and thalassemia can interfere with the HbA1c assay and produce misleading results that are either too high or too low.3

Even without any pathology present, a 2008 study found that HbA1c levels vary between hematologically normal individuals based on the heterogeneous range of RBC lifespan.[9] It was also found that, despite controlling for factors including age, sex, body-mass index, and fasting glucose levels, HbA1c levels vary between ethnic groups.[10] The phenomenon is poorly understood and its implications for interpretation have yet to be defined.

When HbA1c results are suspected to be unreliable, alternative tests to monitor glycemic control over a period of time include fructosamine, glycated albumin, and 1,5-anhydroglucitol (1,5-AG). Fructosamine and glycated albumin are measures of overall protein glycation or albumin glycation, respectively. They reflect glycemic control over the preceding 2-3 weeks, versus HbA1c, which reflects 2-3 months.[11] However, both tests’ results are compromised by hypoproteinemia and hypoalbuminemia in conditions like nephrotic syndrome or severe liver disease. They are also affected by altered protein metabolism in hyper- or hypothyroidism.[12]  1,5-AG is a dietary polyol used as a marker for overall glucose levels over the preceding 2-14 days. The kidneys normally reabsorb it, but at glucose levels above 180 mg/dL, glucose is preferentially reabsorbed and competitively inhibits 1,5-AG.[13] Consequently, 1,5-AG levels and glucose levels are negatively correlated. This test is dependent on kidney function and is limited in renal failure, pregnancy, and chronic liver disease.11

Despite all the benefits that HbA1c provides in the management of diabetes, fasting plasma glucose, the oral glucose tolerance test, and random glucose plasma levels maintain important roles in combating the diabetes epidemic. These tests provide the glucose concentration at a single point in time, compared to the average glucose level reflected by the HbA1c assay. They are staples of diabetes diagnosis and management.

In the Finnish Diabetes Prevention study by Pajunen and colleagues, only 39% of diabetic individuals newly diagnosed via two consecutive OGTTs had an HbA1c of ≥6.5%. They also found that participants with an HbA1c ≥6.5% were more obese, had a higher 2-hour OGTT glucose level, and had higher fasting glucose.[14]

In comparison, a study by Carson and colleagues found that an HbA1c of ≥6.5% identified about 30%-40% of previously undiagnosed diabetics, while fasting plasma glucose levels ≥126 mg/dL diagnosed about 50% and 2-hour plasma glucose levels ≥200 mg/dL detected 90%.[15] The poor sensitivity of HbA1c is likely due its tendency to become elevated later in the course of the disease than elevated fasting or post-prandial glucose. This means that a large proportion of truly diabetic patients may have an HbA1c below 6.5% and that without plasma glucose testing, they remain undiagnosed and have their treatment delayed.

It is important to understand the main methods of monitoring glucose control in diabetes, recognize the limitations of each, and be aware of alternatives. There is no doubt that HbA1c plays a critical role in the management of diabetes and will likely become more significant as standardization continues to improve. Factors influencing HbA1c, including variation in red blood cell lifespans, ethnicity, and hemoglobin subtype are becoming better understood and will be increasingly incorporated into interpretation of results. However, a variety of measurements are useful for a more complete understanding of diabetes risk and progression unique to each patient.

Alexa Yuen, is a 3rd year medical student at NYU Langone Medical Center

Peer reviewed by Michael Tanner, MD, Associate Editor, Clinical Correlations

Image courtesy of Wikimedia Commons


[1]. Centers for Disease Control and Prevention website. Number (in millions) of civilian, non-institutionalized persons with diagnosed diabetes, United States, 1980-2014. Updated December 1, 2015. Accessed October 19, 2016.

[2]. Shaw JE, Sicree RA, Zimmet PZ. Global estimates of the prevalence of diabetes for 2010 and 2030. Diabetes Res Clin Pract. 2010:87(1):4-14.

[3]. Classification and diagnosis of diabetes. Diabetes Care. 2015:38(Suppl. 1):S8-S16.

[4]. Bonora E, Tuomilehto J. The pros and cons of diagnosing diabetes with A1C. Diabetes Care. 2011:34(Suppl. 2):S184-S190.

[5]. Mikesh LM, Bruns DE. Stabilization of glucose in blood specimens: mechanism of delay in fluoride inhibition of glycolysis. Clin Chem. 2008:54(5):930–932.

[6]. Bruns DE, Knowler WC. Stabilization of glucose in blood samples: why it matters. Clin Chem. 2009:55(5):850–852.

[7]. Shepard J, Airee A, Dake A, McFarland M, Vora A. Limitations of A1c interpretation. South Med J. 2015:108(12):724-729.

[8]. Nathan DM, Francis TB, Palmer JL. Effect of aspirin on determinations of glycosylated hemoglobin. Clin Chem. 1983:29(3):466-469.

[9]. Cohen RM, Franco RS, Khera PK, et al. Red cell life span heterogeneity in hematologically normal people is sufficient to alter HbA1c. Blood. 2008:122(10):4284-4291.

[10]. Herman W, Cohen R. Racial and ethnic difference in the relationship between HbA1c and blood glucose: implications for the diagnosis of diabetes. J Clin Endocrinol Metab. 2012:97(4):1067-1072.

[11]. Radin MS. Pitfalls in hemoglobin A1c measurement: when results may be misleading. J Gen Intern Med. 2014:29(2):388-394.

[12]. Armbruster DA. Fructosamine: structure, analysis, and clinical usefulness. Clin Chem. 1987:33(12):2153-2163.

[13]. Kohnert KD, Heinke P, Vogt L, Salzseider E. Utility of different glycemic control metrics for optimizing management of diabetes. World J Diabetes. 2015:6(1):17-29.

[14]. Pajunen P, Peltonen M, Eriksson JG, et al. HbA1c in diagnosing and predicting type 2 diabetes in impaired glucose tolerance: the Finnish Diabetes Prevention Study. Diabet Med. 2011:28(1):36–42.

[15]. Carson AP, Reynolds K, Fonseca VA, Muntner P. Comparison of A1C and fasting glucose criteria to diagnose diabetes among U.S. adults. Diabetes Care. 2010:33(1):95-97.