HbA1c and Metabolic Syndrome

G lycated hemoglobin (HbA1c) has assumed a central place in diabetes management. Once an adjunct approach to monitoring glycemic control, it has become the lingua franca for diabetes care over the last 30 years or so. It can be measured either in a formal clinical laboratory or in the office with point-of-care testing. Patients who are otherwise illiterate about their diabetes management ask for, and know, their HbA1c level. Guidelines are issued to define HbA1c goal values, and physicians are rated (and paid) on their skill (or luck) in achieving those goals, regardless of the heated debate on whether outcomes are affected, either positively or negatively, by striving for those goals.

More recently, HbA1c has been accepted as a criterion used to diagnose diabetes or other disorders of glucose metabolism, ¹ i.e., HbA1c 6.5% is considered the threshold for diabetes. The American Diabetes Association has also defined HbA1c of 5.7–6.4 as “prediabetic,” reminding clinicians that while the risk of retinopathy increases sharply at the 6.5% point, the risk of diabetes increases along a glycemic continuum. A formula deriving average blood glucose from HbA1c has recently been validated. ² HbA1c may be a better predictor of cardiovascular risk than other glycemic markers. ³

In this issue of Metabolic Syndrome and Related Disorders, Janghorbani and Amini push the envelope even further. ⁴ They provide data from an academic center in Iran exploring the association of glycohemoglobin and metabolic syndrome in people who have normal glucose tolerance but are at risk because they are first-degree relatives of individuals with type 2 diabetes. Whereas other studies have explored the role of glycohemoglobin and metabolic syndrome, most of these studies have been in developed countries, and very few have studied people with normal glucose tolerance. These authors report that glycohemoglobin, at levels below diabetes, might be able to predict those at risk of developing metabolic syndrome in normoglycemic first-degree relatives of patients with type 2 diabetes. How did we arrive at this point in the evolution of this lab test?

HbA1c, a minor component of hemoglobin A, was identified in 1958. ⁵ Subsequent developments documented that an elevation in glycated hemoglobin was associated with diabetes. ⁶ In 1972, Bunn and colleagues closed the loop by reporting that the cause of the increased glycated hemoglobin in diabetes (predominantly the A1c component) was increased nonenzymatic glycation. ⁷ [By way of review, glycation, also known as nonenzymatic glycosylation, is the result of typically covalent bonding of a protein (e.g., hemoglobin) or lipid molecule with a sugar molecule, such as fructose or glucose. The so-called “browning reactions (usually Maillard-type reactions)” seen in foods, either intentionally for presentation or unintentionally are evidence of these glycations.] By the late 1970s, HbA1c levels were described as correlating with fasting glucose ⁸ and overall glycemic control. ⁹ The resulting explosion in various assay methods was brought under control under the aegis of the Diabetes Control and Complications Trial (DCCT), which required a centralized, consistent measure of HbA1c, ¹⁰ and then more completely through the National Glycohemoglobin Standardization Program (NGSP), whose reference method is ion-exchange high-performance liquid chromatography (HPLC). This is now the standard against which other tests are certified. ¹¹ This standardization process has led to a dramatic improvement in precision—and credibility internationally—because much of the world's A1c testing is referent to DCCT standards. ¹² Despite these advances, there is no truly international standardization program, and challenges remain in terms of this barrier that are beyond the scope of this essay.

The A1c-Derived Average Glucose (ADAG) study, which was international in scope, studied people with normal glucose tolerance as well as patients with type 1 and type 2 diabetes and was based on large numbers of blood samples per individual, supported the hypothesis that there is a linear relationship between mean blood glucose and HbA1c. ² The relationship was not perfect; the correlation coefficient was 0.84. Part of the variation may have been methodological, but there are other issues: Children, Asians, and Pacific Islanders were not studied. Nonetheless, in ADAG, although there were no statistically significant differences between ethnic groups, the study was not designed to evaluate these possible differences. Other, smaller studies have suggested an effect of ethnicity on A1c at different levels of glycemia (e.g., blacks may have slightly higher glycated hemoglobin levels than whites), but these studies must be considered preliminary at this point. ^13,14

There are other issues limiting the utility of glycated hemoglobin. Artificially low values have been reported to occur with certain hemoglobinopathies (sickle cell disease, thalassemia), or states of increased erythrocyte turnover such as hemolytic anemia, or chronic kidney disease patients receiving erythropoietin. ¹⁵ Conversely, conditions that prolong erythrocyte survival, such as iron deficiency, splenectomy, aplastic anemia, and some hemoglobinopathies, can elevate HbA1c. ¹⁶ But the factors affecting HbA1c can be complex and poorly understood. In some parts of the world, such as India, iron deficiency is the most common form of anemia. In normoglycemic iron-deficient patients, HbA1c was lower than in nonanemic controls; treatment of anemia raised HbA1c values. ¹⁷ Another study from India reported a spuriously high prevalence of prediabetes based on HbA1c not supported by glucose tolerance testing, explained in part by hematological factors, including iron deficiency anemia. ¹⁸ These concerns, coupled with the ethnic issues raised above, must be considered when glycated hemoglobin becomes a diagnostic criterion, particularly in prediabetic states where the cutoff points are somewhat arbitrary.

Recently, it has been suggested that because HbA1c reflects the prior 2–3 months of glucose control other glycosylation markers might have better short-term utility. Glycated albumin, ^19,20 fructosamine, ²⁰ and 1,5-anhydroglucitol ²⁰ have all been proposed as short-term indices. These compounds have neither been internationally validated nor achieved wide use outside of isolated clinical circumstances and will not be discussed further.

In the study by Janghorbani and Amini, ⁴ over 3000 (25.7% men) first-degree relatives (FDR, siblings or adult children) of patients with type 2 diabetes were screened. The authors focused on those with normal glucose tolerance, as defined in reference 1, determined by an oral glucose tolerance test (OGTT) administered to those individuals with normal fasting glucose. A total of 1386 individuals were ultimately studied (mean age 42 years, 74% women). Glycated hemoglobin (GHb) was measured by an NGSP-certified method (ion-exchange HPLC). Subjects were stratified into quintiles of GHb. A complex statistical analysis [considering age, gender, body mass index (BMI), waist circumference, lipid values, and systolic blood pressure] was performed to determine the ability of GHb to predict the presence of metabolic syndrome; this was examined by receiver operating characteristic (ROC) curves. The top quintile of GHb had a value >5.6%. The risk of metabolic syndrome increased across all quintiles; the odds ratios varied slightly for the different groups depending on whether the ratios were adjusted for age, age and gender, or underwent multivariate analysis. Nonetheless, the consistent significant impact of GHb was seen only with the highest quintile. It should be pointed out that this quintile >5.6% overlaps with the American Diabetes Association (ADA) “prediabetic” range of 5.7%–6.4%.

It is already known that HbA1c predicts cardiovascular risk. Janghorbani and Amini now tell us that HbA1c predicts metabolic syndrome. Using glycohemoglobin to predict metabolic syndrome has not been highly studied. While the authors have done a service extending to residents of central Iran a risk analysis that had until now largely been done in developed countries, and demonstrating the universal utility of GHb in analyzing cardiovascular risk (and validating the ADA classification of “prediabetes” based on HbA1c in a population not specifically addressed in the ADA guidelines), they raise perhaps a larger question: Does identifying someone with metabolic syndrome add to that risk assessment? In other words, is a patient with metabolic syndrome more likely to be identified as needing intervention than a patient with “prediabetic” HbA1c? It should be recalled that patients defined as having metabolic syndrome not only have a cluster of readily measured risk factors, but also a cluster of factors hidden from view, such as a prothrombotic and proinflammatory state, that can also raise cardiovascular risk. ²¹