Glycated haemoglobin (HbA1c) for the diagnosis of diabetes (2024)

Glycated haemoglobin (HbA1c) was initially identified as an “unusual” haemoglobin in patients with diabetes over 40 years ago (12). After that discovery, numerous small studies were conducted correlating it to glucose measurements resulting in the idea that HbA1c could be used as an objective measure of glycaemic control. The A1C-Derived Average Glucose (ADAG) study included 643 participants representing a range of A1C levels. It established a validated relationship between A1C and average glucose across a range of diabetes types and patient populations (13). HbA1c was introduced into clinical use in the 1980s and subsequently has become a cornerstone of clinical practice (14).

HbA1c reflects average plasma glucose over the previous eight to 12 weeks (15). It can be performed at any time of the day and does not require any special preparation such as fasting. These properties have made it the preferred test for assessing glycaemic control in people with diabetes. More recently, there has been substantial interest in using it as a diagnostic test for diabetes and as a screening test for persons at high risk of diabetes (16).

Owing in large part to the inconvenience of measuring fasting plasma glucose levels or performing an OGTT, and day-to-day variability in glucose, an alternative to glucose measurements for the diagnosis of diabetes has long been sought. HbA1c has now been recommended by an International Committee and by the ADA as a means to diagnose diabetes (16). Although it gives equal or almost equal sensitivity and specificity to a fasting or post-load glucose measurement as a predictor of prevalent retinopathy (17), it is not available in many parts of the world. Also, many people identified as having diabetes based on HbA1c will not have diabetes by direct glucose measurement and vice versa.

The relationship between HbA1c and prevalent retinopathy is similar to that of plasma glucose, whether glucose and HbA1c are plotted in deciles (18), in vigintiles (Figure 1) or as continuous variables (Figure 2). This relationship was originally reported in the Pima Indians (19) and has also been observed in several other populations including Egyptians (20), the NHANES study in the USA (21),, in Japanese (22) and more recently in the DETECT-2 analysis (Figures 1 and 2). Overall, the performance of HbA1c has been similar to that of fasting or 2-h plasma glucose. For all three measures of glycaemia, the value above which the prevalence of retinopathy begins to rise rapidly has differed to some extent between studies. Although HbA1c gives equal or almost equal sensitivity and specificity to glucose measurement as a predictor of prevalent retinopathy, it is not available in many parts of the world and in general, it is not known which is the better for predicting microvascular complications.

It is unclear whether HbA1c or blood glucose is better for predicting the development of retinopathy, but a recent report from Australia has shown that a model including HbA1c for predicting incident retinopathy is as good as or possibly better than one including fasting plasma glucose (23).

The use of HbA1c can avoid the problem of day-to-day variability of glucose values, and importantly it avoids the need for the person to fast and to have preceding dietary preparations. These advantages have implications for early identification and treatment which have been strongly advocated in recent years.

However, HbA1c may be affected by a variety of genetic, haematologic and illness-related factors (Annex 1) (24). The most common important factors worldwide affecting HbA1c levels are haemoglobinopathies (depending on the assay employed), certain anaemias, and disorders associated with accelerated red cell turnover such as malaria (16;25).

The utility and convenience of HbA1c compared with measures of plasma glucose for the diagnosis of diabetes needs to be balanced against the fact that it is unavailable in many countries, despite being a recognized valuable tool in diabetes management. In addition the HbA1c assay is not currently well enough standardized in many countries for its use to be recommended universally at this time. However, there will be countries where optimal circ*mstances already exist for its use. Factors influencing HbA1c assays are presented in Annex 2 and 3.

There are aspects of the measurement of HbA1c that are problematic. Although in some laboratories the precision of HbA1c measurement is similar to that of plasma glucose, global consistency with both assays remains a problem (16). Whether it is the glucose or HbA1c assay that is used, consistent and comparable data that meet international standards are required. This is starting to happen in many countries but obviously is still not standard across the world. Within any country, it is axiomatic that results for glucose and HbA1c should be consistent between laboratories.

The National Glycohemoglobin Standardization Program (NGSP) (26) was established following the completion of the Diabetes Complications and Control Trial (DCCT). For many years it was the sole basis for improved harmonization of HbA1c assays. More recently the International Federation of Clinical Chemists (IFCC) established a working group on HbA1c in an attempt to introduce an international standardization program (27). An important part of this effort was establishment of reference method procedures for HbA1c. Currently, both the NGSP and the IFCC base their evaluations on reference method procedures that have further enhanced the harmonization of HbA1c assays across manufacturers. Finally in the USA, the College of American Pathologists (CAP) has mandated more stringent criteria for individual assays to match assigned values for materials provided in the CAP proficiency programme (28).

A further major factor concerns costs and availability of HbA1c assays in many countries. Also, the situation in several of these countries will be exacerbated by high prevalences of conditions such as haemoglobinopathies, which affect HbA1c measurement, as discussed earlier.

A report published in 2009 by an International Expert Committee on the role of HbA1c in the diagnosis of diabetes recommended that HbA1c can be used to diagnose diabetes and that the diagnosis can be made if the HbA1c level is ≥6.5%(16). Diagnosis should be confirmed with a repeat HbA1c test, unless clinical symptoms and plasma glucose levels >11.1mmol/l (200 mg/dl) are present in which case further testing is not required. Levels of HbA1c just below 6.5% may indicate the presence of intermediate hyperglycaemia. The precise lower cut-off point for this has yet to be defined, although the ADA has suggested 5.7 – 6.4% as the high risk range (29). While recognizing the continuum of risk that may be captured by the HbA1c assay, the International Expert Committee recommended that persons with a HbA1c level between 6.0 and 6.5% were at particularly high risk and might be considered for diabetes prevention interventions.

The WHO consultation reviewed the evidence on the relationship between HbA1c and prevalent and incident microvascular complications presented in the systematic review. Tables 1 and 2 show HbA1c and glucose cut-off points associated with prevalent and incident microvascular complications in available studies. GRADE tables of evidence are presented in Tables 3 and 4. In view of the above and of the advances in technology over recent years, members of the consultation agreed that HbA1c may be used to diagnose diabetes providing that appropriate conditions apply, i.e. standardized assay, low coefficient of variability, and calibration against IFCC standards. Furthermore, each country should decide whether it is appropriate for its own circ*mstances. The choice of diagnostic method will depend on local considerations such as cost, availability of equipment, population characteristics, presence of a national quality assurance system etc. Policy-makers are advised to ensure that accurate blood glucose measurement be generally available at the primary health care level, before introducing HbA1c measurement as a diagnostic test. The consultation concluded that there is insufficient evidence to make any formal recommendation on the interpretation of HbA1c levels below 6.5%.

Long term prospective studies are required in all major ethnic groups to establish more precisely the glucose and HbA1c levels predictive of microvascular and macrovascular complications. A working group should be established to examine all aspects of HbA1c and glucose measurement methodology.

The diagnosis of diabetes in an asymptomatic person should not be made on the basis of a single abnormal plasma glucose or HbA1c value. At least one additional HbA1c or plasma glucose test result with a value in the diabetic range is required, either fasting, from a random (casual) sample, or from the oral glucose tolerance test (OGTT). The diagnosis should be made by the best technology available, avoiding blood glucose monitoring meters and single-use HbA1c test kits (except where this is the only option available or where there is a stringent quality assurance programme in place).

It is advisable to use one test or the other but if both glucose and HbA1c are measured and both are “diagnostic” then the diagnosis is made. If one only is abnormal then a further abnormal test result, using the same method, is required to confirm the diagnosis.

More and more asymptomatic subjects are being detected as a result of screening programmes so that diagnostic certainty is paramount. If such tests fail to confirm the diagnosis of diabetes, it will usually be advisable to maintain surveillance with periodic re–testing until the glycaemic status becomes clear.