Glycated hemoglobin A_1c (HbA_1c) and diabetes: a new era?

Abstract

In January 2011 the American Diabetes Association (ADA) published the latest guidelines for the diagnosis and treatment of diabetes mellitus (DM)^1,2. Despite some controversies, glycated hemoglobin A_1c (HbA_1c), an established marker of long-term glycemia traditionally used to assess the quality of DM management, remained an independent criterion for the diagnosis of DM, and indeed now appears to be well established in the USA. This has far-reaching implications for clinical practice worldwide.

Keywords::

• American Diabetes Association
• Glycated hemoglobin A_1c
• Diabetes mellitus
• Diagnosis
• Glucose

HbA_1c and diagnosis of DM

HbA_1c was first introduced as a diagnostic marker of DM in 2010 by the ADA3, with an HbA_1c ≥ 6.5% as an alternative criterion for the diagnosis of DM, in addition to the long established glucose related diagnostic criteria1. In equivocal cases, a second abnormal value is required to confirm the diagnosis of DM. Furthermore, HbA_1c testing should be performed using a method certified by the National Glycohemoglobin Standardization Program (NGSP) and standardized to the Diabetes Control and Complications Trial (DCCT)4,5, as emphasized in the new guidelines1,2.

Apart from the ADA, the European Association for the Study of Diabetes (EASD) and the World Health Organization (WHO) also accept an HbA_1c ≥ 6.5% for the diagnosis of DM, provided that measurements are performed by a standardized method6. In this context, in 2010 the ADA, EASD, International Diabetes Federation (IDF), International Federation of Clinical Chemists (IFCC) and International Society for Pediatric and Adolescent Diabetes produced a consensus statement on the worldwide standardization of HbA_1c measurement. They highlighted the importance of quality assessments in all laboratories to ensure the accuracy of measurements6.

The use of HbA_1c as a diagnostic criterion for DM has several advantages over glucose measurements, given that the latter are affected by several factors (biological, preanalytical and analytical), as recently reported7,8. Furthermore, fasting samples are not needed for HbA_1c measurement, HbA_1c values have less biological variability (<2% vs 12–15% for fasting plasma glucose FPG), the sample remains stable and unaffected by acute factors (e.g. exercise), and the assay is standardized7,8. However, there are also certain limitations regarding the implementation of this criterion, mainly the need for a standardized evaluation method, possible racial/ethnic differences, effects of hemoglobinopathies and inability to diagnose type 1 DM with a rapid development9.

Furthermore, the IFCC Working Group on Standardization of HbA_1c (WG-HbA_1c) recommended since 2007 that HbA_1c results should be reported in both mmol/mol (Standard International [SI] units) and derived NGSP units (%), using the IFCC-NGSP master equation10,11; all new instruments sold after January 1, 2011 should report these units10. The use of both units will facilitate the setting of targets and comparisons between the results of clinical studies worldwide. However, confusion is likely to occur and, therefore, patients and health care providers should be adequately informed12.

It should be noted that when HbA_1c was used as a diagnostic marker of pre-diabetes, almost 50 million Americans were re-classified, i.e. from having impaired fasting glucose (IFG) to not being pre-diabetic and from not having IFG to being pre-diabetic13. Moreover, a recent study14 has found that the prevalence of undiagnosed DM, overt DM and high risk for DM were substantially lower when estimated with the ADA recommended HbA_1c criterion than with FPG or 2 h glucose cut-off points. Similarly, this HbA_1c criterion was less sensitive for identifying pre-diabetes than IFG and impaired glucose tolerance in a substudy of the Insulin Resistance Atherosclerosis Study (IRAS)15. As indicated above, such discrepancies are not surprising. Part of the problem is that glucose and HbA_1c are continuous variables in the population and in setting diagnostic criteria they are converted into dichotomous ones, based on the value of a particular threshold to predict future risk. In the case of DM the risk prediction relates to epidemiological studies of retinopathy1,2. While arguments have been proposed against the use of retinopathy, it remains the complication that is most specifically related to glycemia and is also diagnosed on the basis of sound criteria. HbA_1c is related more intimately to the risk of retinopathy compared with single or episodic measurements of glucose levels9 and therefore it may be considered as a better glycemic marker in terms of predicting the risk of this diabetic complication1,2,9.

In addition, the term ‘pre-diabetes’ implies the likely development of diabetes in the relatively near term. Obviously, this ‘prediction’ yields both false positives and false negatives. However, it is important to recognize the ability of HbA_1c to predict the development of health outcomes even when below the cut off used to diagnose diabetes. Although retinopathy was the main criterion for the selection of the HbA_1c threshold for DM diagnosis, HbA_1c was also shown to be predictive of cardiovascular disease (CVD) incidence and mortality as well as of DM development, as further discussed below.

HbA_1c in specific patient populations

The interpretation of HbA_1c values (in SI or as a %) in certain populations has limitations. For example, patients with anemia or renal disease may present false HbA_1c values. Furthermore, intravenous iron and erythropoietin-stimulating agents have been shown to reduce HbA_1c levels without altering glycemic control16. Haptoglobin phenotype17 and other hemoglobinopathies18–20 may also affect HbA_1c levels. It should be noted that at least 10% of the African Americans aged over 17 years old are estimated to have hemoglobin (Hb) S or HbC trait21. Furthermore, anemia is more frequent in diabetic patients with nephropathy compared with non-diabetic individuals with kidney disease22,23.

In such cases, measurement of other markers for monitoring long-term glucose control (e.g. serum fructosamine concentration24, glycated albumin25 or even mean glucose level of 4–5 daily home measurements26) have been suggested. However, the fructosamine assay is limited due to the short half-life of fructosamine and the difficult standardization of the assay21, whereas glycated albumin may be influenced by various factors i.e. inflammation, obesity27, smoking28 and hyperuricemia29. Such limitations should be taken into account when determining these markers.

The performance of certain HbA_1c assays (i.e. Roche Diagnostics 1st and 2nd generation HbA_1c immunoassays, with the latter method showing better performance)21 represents another option. As hemoglobin variants and chemical derivatives adversely affect the accuracy of HbA_1c measurements30, the use of the latter assays may improve the precision of the HbA_1c calculations. Furthermore, immunoassays express HbA_1c as a percentage of total Hb, whereas HPLC assays as a percentage of total HbA, thus possibly reporting lower HbA_1c values21.

The glycation gap (defined as the difference between measured HbA_1c and HbA_1c predicted from fructosamine31) and hemoglobin glycation index32 have also been used to quantitate the relationship between plasma glucose and HbA_1c33.

In general, the presence of a Hb variant should be evaluated in any case where the HbA_1c values do not correspond to clinical evaluation and measurements of fasting blood glucose, glycemic control could be better monitored by the use of an alternative method30. The true prevalence of such discrepancies is currently unknown, and clinical suspicion is needed to identify appropriate individuals based on clinical and demographic characteristics of the patient.

HbA_1c and predicting diabetic complications

The rationale of using HbA_1c as a diagnostic tool for DM was based on the fact that this marker better reflects glucose levels over time (approximately 2–3 months) and is more closely correlated with micro- and macrovascular complications than random measurements of plasma glucose9. Indeed, several studies have shown that HbA_1c levels were significantly associated with development of retinopathy34–37, nephropathy35,36,38, neuropathy35,39,40 and left ventricular dysfunction41 as well as CVD42–44. There is also a continuous relationship between HbA_1c and coronary heart disease risk; this extends to below the diagnostic threshold for DM45. The latter association remains unclear in terms of HbA_1c target levels, i.e. intensive vs conventional glycemic control43,46–48. A lower HbA_1c has also been shown to predict decreased healing time for diabetic foot ulcers49. A link with the risk of stroke has also been reported50.

HbA_1c and monitoring glycemic control of DM

The definition of specific HbA_1c levels as a treatment target was based on the results of the DCCT51 and the United Kingdom Prospective Diabetes Study (UKPDS)52,53. These studies reported improved microvascular outcomes in relation to intensive glycemic control in patients with type 1 or type 2 DM, respectively. Consequently, certain HbA_1c cut-off points were determined as targets for DM treatment, namely HbA_1c ≤ 7% (by the ADA2) and ≤6.5% (by the IDF54).

Achieving strict (≤6.5%) HbA_1c levels has been associated with lower mortality rates55. However, aggressive glucose control was not always shown to improve the cardiovascular prognosis of diabetic patients56. Briefly, in the Veterans Affairs Diabetes Trial (VADT)57, the Action in Diabetes and Vascular Disease: Preterax and Diamicron Modified Release Controlled Evaluation (ADVANCE) trial48 and the Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial58, reaching normoglycemic HbA_1c levels was related to either similar, better or even worse cardiovascular outcomes59. Systematic reviews60 and meta-analyses of these studies43 have shown that intensive glycemic control reduced the risk for CVD morbidity but not mortality. Furthermore, tight glucose control may be harmful in specific populations (e.g. the elderly, those with DM over a prolonged period and those with CVD) as it can lead to more frequent and more severe hypoglycemic episodes2,61.

While the importance of HbA_1c in assessing the quality of DM management is undeniable, postprandial glucose levels can, in addition to HbA_1c and FPG, contribute to the effective management of DM62.

HbA_1c and predicting the onset of type 2 DM

An International Expert Committee appointed by the ADA, EASD and IDF in 2009 recommended the use of HbA_1c for the identification of those at high risk of developing DM; individuals with HbA_1c ≥ 6.0% but <6.5% are likely to develop DM9 (note that those with HbA_1c ≥ 6.5% are already diagnosed as having DM). Importantly, the expert committee did not propose these HbA_1c levels as fixed cut-off points to initiate preventive treatment. However, the ADA in its recent guidelines1 suggests the implementation of preventive interventions in individuals with HbA_1c in the range of 5.5–6%. In a recent longitudinal cohort study63, the probability of progression to DM was similar for those individuals with pre-diabetes assessed by HbA_1c 5.7–6.4% and those with IFG, although the HbA_1c criterion identified fewer patients at high risk. The importance of treating other DM risk factors such as hypertension, hypertriglyceridemia and obesity in order to prevent or delay the development of DM has also been pointed out9.

Interestingly, Schaufler et al.64 have recently reported that screening programmes for type 2 DM in Germany were cost-effective for the general population and cost-saving for those diagnosed with pre-diabetes who started treatment immediately. Furthermore, early diagnosis of type 2 DM was associated with fewer DM-related adverse events and longer life expectancy64. More recently, the Anglo–Danish–Dutch Study of Intensive Treatment In People with Screen Detected Diabetes in Primary Care (ADDITION) has shown that early intensive management of individuals with newly diagnosed (by screening) T2DM succeeded in a small, although insignificant, reduction of cardiovascular morbidity and mortality compared with the routine care group⁶⁵. It should also be noted that screening policies for DM are especially beneficial when early diagnosis is followed by prompt initiation of therapy in terms of lifestyle changes and pharmacological intervention66. In this context, similarly to fewer micro- and macrovascular complications as a result of early detection and treatment of DM, screening for pre-diabetes may prevent/delay the development of both DM and CVD67. These beneficial effects can be accomplished by the implementation of multifactorial interventions (i.e. lipid-lowering, antihypertensive, hypoglycemic and weight-reducing, both via lifestyle measures and drug treatment) in high-risk populations such as patients with pre-diabetes, metabolic syndrome (MetS) or abnormal liver function tests/non-alcoholic fatty liver disease (NAFLD)68–71.

Conclusions

Overall, the prevalence of obesity, DM and CVD is increasing worldwide. Thus, there is an unmet need for early detection and prompt treatment of high-risk patients. In this endeavor, the use of HbA_1c as a diagnostic criterion of DM appears especially promising with several advantages over glucose measurements. While it is true that this new criterion has its limitations, it may contribute to the earlier diagnosis of DM. This, in turn, may help improve management and reduce the burden of both micro- and macrovascular complications. The clinical implications of this novel diagnostic criterion remain to be established in long-term prospective studies.

Transparency

Declaration of funding

This editorial was written independently. The authors did not receive financial or professional help with the preparation of the manuscript.

Declaration of financial/other relationships

D.P.M. has given talks, attended conferences and participated in advisory boards and trials sponsored by Pfizer, MSD, Genzyme and Astra-Zeneca. N.K. has received a grant from the Hellenic Atherosclerosis Society. N.P. has been an advisory board member of TrigoCare International; has participated in sponsored studies by Novo Nordisk and Novartis; received honoraria as a speaker for Novo Nordisk and Pfizer; and attended conferences sponsored by TrigoCare International, Novo Nordisk, Sanofi-Aventis and Pfizer. V.A.F. is supported in part by the Tullis-Tulane Alumni Chair in Diabetes and the Susan Harling Robinson Fund supporting diabetes research at Tulane University Health Sciences Center. V.A.F. and Tulane University have Research Support Grants and honoraria for consulting and lectures from Glaxo Smith Kline, Novo Nordisk, Takeda, Astra-Zeneca, Pan American Laboratories, Sanofi-Aventis, Eli Lilly, Daiichi-Sankyo, Reata Inc., Xoma and Abbott.