Hemoglobin A1c Measurements:
Methods, Interpretation, and Effect of Hemoglobinopathies

Hemoglobin A1c (HgbA1c), also often colloquially known as glycated hemoglobin (GHb), glycosylated hemoglobin, and glycohemoglobin, refers to a measure of the percentage of hemoglobin A1 that has been glycosylated. However, total glycated hemoglobin and hemoglobin A1c are not the same; the former refers to many glycosylated species of hemoglobin while the latter refers to a well-defined single such species and is the measurement upon which most diabetic control recommendations are based. As discussed below, patients with hemoglobinopathies (e.g., sickle cell disease, hemoglobin CC, hemoglobin EE) do not produce hemoglobin A and hence do not have any hemoglobin A1c. Although some methodologies will nonetheless report out a “hemoglobin A1c”, the measurement must be very carefully interpreted in these patients.

Glucose nonenzymatically attaches itself to amine groups of many blood proteins. Over the course of days, continuously elevated blood glucose levels increase the probability that these reversibly glycosylated products will proceed slowly to adopt a more stable chemical configuration. Since glucose passes freely into red blood cells, hemoglobin becomes glycosylated inside RBCs, forming a compound that persists throughout the red cell’s normal 120 day life span. Therefore, hemoglobin A1c represents a measure of average glucose levels over the preceding 120 days. Because of the kinetics of RBC turnover, however, it is skewed towards more closely representing glucose levels over the last 60-90 days. The assay is routinely available through the Immunology Laboratory and as a point-of-care assay in the adult and pediatric outpatient departments (see below).

Serum proteins with shorter half lives, such as albumin, also become heavily glycosylated in the presence of persistently elevated glucose levels. Total glycosylated serum proteins and glycosylated serum albumin (as measured by the fructosamine assay) can provide an estimation of glycemic control over the preceding 7-14 or 14-20 days. The clinically useful role of this measurement, however, is considered much more limited for the purposes of routine diabetic care. The fructosamine test is therefore not available in-house but can be sent to a reference laboratory via the Chemistry Laboratory.

The American Diabetes Association (ADA) has recognized for some time the importance of measuring Hemoglobin A1c levels in the management of diabetes. Their recommendations from a 1997 consensus report include obtaining a HgbA1c level at initial consultation and regular checks of HgbA1c levels in follow up appointments depending on the specific clinical case - optimally about every 3 months, but no less than once or twice a year for well-controlled diabetic patients. Other authorities have also stressed the importance of HgbA1c checks on a quarterly basis, even in well- controlled patients. Normally, HgbA1c levels are between 4-6%. Recommended treatment goals usually quote a HgbA1c level < 7%. The recommendation is based on the results of the Diabetes Control and Complications Trial (DCCT) that determined risks for complications of nephropathy, neuropathy, and retinopathy in type 1 diabetes, showing they decline proportionally to the HgbA1 level. Subsequent studies have confirmed the sudden exponential increase in risks for complication from hyperglycemia at HgbA1c levels somewhere between 6 and 7% and support the same quoted target levels for type 2 diabetes patients as well. The ADA also recommends that a HgbA1c level > 8% should prompt investigations into treatment regimen modifications.

Several methods are available for determination of HgbA1c, and in the past there has been concern with regard to the correlation of individual assays to the levels in the DCCT study. Since 1996, an effort has been underway to achieve standardization of HgbA1c testing. At this time most major laboratories have tests that correlate well with the DCCT reference method. However, physicians must still be wary when interpreting results from several years ago, or from laboratories where the testing method is unknown.

Our Immunology Laboratory has recently begun to utilize a method of HgbA1c measurement that is based on differing mobilities of the glycated versus non-glycated forms of hemoglobin as detected by HPLC technology on a cationic column. A photometric scanner measures relative peak areas and calculates a percentage of hemoglobin A1c. The method is generally considered to be the “gold standard” technology that correlates as precisely as possible to the DCCT standards. The major alternative technology uses an immunological method for determining glycated hemoglobin - some immunologic methods directly measure hemoglobin A1c while others actually measure total glycated hemoglobin and then a calculation is performed to determine the A1c level. Although such a “calculated” A1c is intrinsically potentially less accurate, most of these latter type of immunoassay systems perform up to levels defined by the ADA.

Because the mobility of abnormal hemoglobins (such as sickle hemoglobin S, hemoglobin C and hemoglobin E) is different from that of Hgb A, the new Hgb A1c assay is able to incidentally identify the presence of heterozygous and homozygous hemoglobinopathies in patients whose blood work is sent for hemoglobin A1c measurement. It is usually not possible to identify the specific type of hemoglobinopathy (e.g. Hgb AS vs Hgb AC) - such a determination requires carrying out either HPLC under different conditions or carrying out a hemoglobin electrophoresis. In cases where there is a reduced amount of Hgb A secondary to a heterozygous hemoglobinopathy (e.g. Hgb AS or Hgb AC), the assay still accurately determines the percentage of the available hemoglobin A that is glycosylated. In these cases, the laboratory reports the hemoglobin A1c level and also includes a comment on the report alerting the physician that the patient likely has a hemoglobinopathy, although the exact type cannot be accurately determined without additional testing. Most heterozygous hemoglobinopathies (“carrier states”) are associated with a relatively normal red cell survival and hence the hemoglobin A1c level reported is generally interpretable in the light of standard ADA and other guidelines.

However, in cases of homozygous hemoglobinopathies (e.g. Hgb SS) or in cases of double heterozygous hemoglobinopathies (e.g. Hgb SC), the patient has no Hgb A or glycosylated Hgb A. Without any endogenous hemoglobin A1c, any hemoglobin A1c present must be the result of recent transfusions. In these cases the Immunology Laboratory reports that the glycated hemoglobin cannot be determined by the reference HPLC method.

Other methods, however, respond variably to the presence of hemoglobinopathies depending on the hemoglobinopathy type. The method utilized in the HgbA1c point of care devices in use at YNHH relies on the antibody based detection of hemoglobins and their glycosylated counterparts. This system is not totally specific for A1c but will measure the total hemoglobin and total glycated hemoglobins, including HgbS1c and HgbC1c, and calculate a percentage of glycated hemoglobin, even in the presence of a homozygous hemoglobinopathy. In order to provide maximum information to our clinicians, when the HPLC method will not produce a glycated hemoglobin result, we will determine such a level (reflecting S1c, C1c, or the like) by the immune based method. Our laboratory policy is then to have one of our clinical pathologists contact the ordering clinician, note the patient’s hemoglobinopathy, and explain the limitations of the reported glycated hemoglobin values (which will also include an explanatory comment on the report form). It is important to note that the values reported cannot be accurately interpreted within the context of ADA and other guidelines because the RBC lifespan and the abnormal hemoglobin half life is markedly reduced in these cases - thus the glycated hemoglobin percentage represents a measure of average glucose levels over a time period much shorter than 60-90 days. In addition, although one might think that one could at least use these values to monitor diabetic control on an individual patient basis by comparing the same patient’s results over time, individual results are confounded by recent transfusions and by the variability in red cell survival that occurs in individual patients. For example, a recent sickle crisis may result in marked shortening of even a particular patient’s “usual” RBC survival and hence a “falsely low” glycated hemoglobin value. Hence, if these values are to be followed at all, extreme care must be taken in interpretation for this patient group. Since other plasma proteins are not similarly affected by increased RBC turnover, fructosamine measurements are, at least theoretically, a better diabetes control marker in hemoglobinopathy cases where red cell survival is expected to be reduced.

It should also be noted that immunologic method based instruments, such as those used in our point-of-care sites at YNHH, cannot determine that a patient has a homozygous or double heterozygote hemoglobinopathy and hence will produce a “hemoglobin A1c” result without any accompanying “warning” concerning its utility (or the fact that it is not really A1c). Typically, this is not a major problem since the issues arise only in patients with homozygous and double heterozygote disease, which is usually known to the clinician caring for the patient. The possibility of a hemoglobinopathy should be explored in cases where “hemoglobin A1c” measurements are lower than expected based on daily glucose monitoring. Finally, caution should also be exercised in the interpretation of hemoglobin A1c measurements in patients with hemolytic anemias due to causes other than hemoglobinopathy (e.g., hereditary spherocytosis, autoimmune hemolytic anemia) since under these circumstances, although an A1c level will be accurate, the fact that the patient's red cell survival is shortened will affect its interpretation with respect to ADA guidelines.

Please direct any questions to the Laboratory Medicine Resident, to the Immunology Laboratory at 8-2440, or to Dr. Brian Smith.

References:

1. American Diabetes Association: clinical practice recommendations 1997. Diabetes Care. 20 Suppl 1: S1-70, 1997.

2. Edelman SV, Importance of glucose control. Med Clin North Am. Jul; 82(4): 665-87, 1998.

3. Textbook of Clinical Chemistry, 3rd Ed., Tietz NW, Burtis CA (Ed), Ashwood ER (Ed), W.B. Saunders Co., p. 790-796, 1999.

Richard Torres, MD
Brian Smith, MD