Background
At the 2009 IDF Congress, an international committee of experts proposed a revision of the diagnostic criteria for diabetes, suggesting that HbA1c may be a better diagnostic indicator of diabetes and more appropriate as a diagnostic criterion for diabetes compared to blood glucose (the proposed cut point is defined as HbA1c >= 6.5%).
This view is based on evidence of correlation between microvascular complications in evidence-based medicine and HbA1c, which is at least similar to fasting or post-load glucose, while HbA1c has less daily variability than fasting or post-load glucose and can be measured at any time of the day without fasting or glucose loading. To date, this proposal has been pushed by authoritative organizations such as the European Association for the Study of Diabetes (EASD), the American Diabetes Association (ADA), the International Diabetes Federation (IDF) and the World Health Organization (WHO), as well as leading experts from several countries around the world.
The decision was not an easy one; redefining diagnostic criteria usually means adopting one set of criteria instead of another. Having different diagnostic procedures in different regions at the same time may create confusion and opacity. Therefore, a careful evaluation of the available evidence and the public health, economic, and practical implications of redefining the diagnostic criteria for diabetes is imperative.
Defining the diagnostic cut point
The clinical manifestations of diabetes were first described in the Chinese and Western medical literature more than 2000 years ago, but the first modern standardized, accepted definition of diabetes was adopted by a WHO expert committee 30 years ago. This set of diagnostic criteria has since been revised several times by WHO expert committees in 1985, 1997 and 2006, including in 2009 when WHO reconvened an expert committee to review the criteria for the diagnosis and classification of diabetes.
The accumulation of evidence-based medical evidence, the increasing understanding by experts of the pathogenesis of diabetes, and the increasing understanding of the association of glucose with microvascular/ macrovascular complications are important drivers for the continuous updating and revision of diagnostic criteria.
Why is it so difficult to define the diagnostic cut point for diabetes? This is because being below the cut point would mean that the population diagnosed with diabetes actually includes a significant number of normal people, resulting in a waste of social resources, while being above the cut point would be a missed diagnosis for many people with true diabetes and a loss of optimal treatment, when such a cut point may not exist. The reason for working so hard to control blood glucose in diabetic patients is based on the knowledge of the relationship between blood glucose and vascular complications.
The correlation between microvascular lesions and glucose is now well established by DCCT and has been recognized by diabetologists, so that microvascular complications appear to be more specific for confirming diabetes. In a recent study of 48,418 subjects based on the DETECT-2 database, results showed that there is a range of glucose and HbA1c below which the incidence of diabetes-specific moderate to severe retinopathy is low, and above which the incidence increases significantly with increasing glucose levels.
However, the statistical and mathematical models currently used do not allow for the identification of a clear cut-off point. Because of the complexity of the pathogenesis of type 2 diabetes, UKPDS, however, cannot equally strongly demonstrate the benefit of strict glycemic control on macrovascular events. Indeed as with other risk factors, there is considerable evidence of an increasing risk of macrovascular disease with increasing glucose levels and HbA1c levels 2 hours after OGTT sugar administration.
Only stroke (fatal and non-fatal) is non-linearly correlated with HbA1c, with an increased risk at HbA1c >7.0%, but the number of events per HbA1c tier is low and the range of each tier is wide, and this threshold may not be easy to determine. Thus although macrovascular disease is the most common cause of death in patients with type 2 diabetes, the fact that it is not specific for diabetes may limit its usefulness in defining diagnostic cut points.
Feasibility of using HbA1c to diagnose diabetes mellitus
The idea of using HbA1c to diagnose diabetes was first proposed by an ADA expert committee in 1997. The correlation between HbA1c increased to diabetic levels and retinopathy was similar to the correlation between oral/2-hour glucose and diabetic complications, and the decile distribution of the correlation between the three different glucose measures and retinopathy was similar.
The main issue at that time was the standardization of HbA1c measurement methods. This issue continued until the International Federation of Clinical Chemistry (IFCC), together with EASD and ADA, provided a standardized method, thus offering the possibility of international standardization of HbA1c measurement methods.
Advantages of using HbA1c for diabetes diagnosis
Underdiagnosis of diabetes is a major problem in most countries, with epidemiological findings showing that 30% to 50% of patients with diabetes are underdiagnosed, which is why screening for diabetes is recommended by some countries and organizations. However, the diagnosis of existing diabetes screening is not simple. Choosing fasting glucose may miss a significant number of patients with postprandial hyperglycemia, and another problem in diagnosing diabetes with plasma glucose is the accuracy of the test and standardization of the process.
To properly measure plasma glucose, blood samples need to be stored immediately in ice water and centrifuged within 30-60 minutes to separate the plasma, but this is rarely done in daily clinical practice, and if blood samples are not processed correctly and in a timely manner, cells in the blood will continue to depend on glucose for survival, causing glucose concentrations to fall by an average of at least 0.5 mmol/L. In addition, inter-individual variation in fasting plasma glucose is approximately 12% to 15%, with even greater variability in plasma glucose 2h postprandial. Therefore, an OGTT is required, but this test is time-consuming, expensive and inconvenient, with poor reproducibility, and is rarely used in scientific studies and investigations.
HbA1c became a convenient alternative?HbA1c testing can be performed at any time of the day, and blood can be drawn and measured locally or sent elsewhere without altering the results, with only 2% inter-individual variation in HbA1c. Therefore, changing to the use of HbA1c for the diagnosis of diabetes would facilitate the early detection of diabetes, which has important public health implications.
Limitations of using HbA1c to diagnose diabetes mellitus
From an economic point of view, diagnosing diabetes with HbA1c is more expensive than glucose measurement, making it unaffordable in many parts of the world. From a societal perspective this issue is less important because two morning fasting measurements are required to confirm diabetes in asymptomatic individuals, and a large proportion of individuals require an OGTT, and adding these costs to the loss of lost time from work (which for most people results in two 1 to 3 hours of lost work time before diagnosis) the economic advantage of using glucose to diagnose diabetes is less clear .
If HbA1c is recognized as a diagnostic test, market competition can bring down the cost of HbA1c measurement. The cost and efficacy can simply be entered into models to examine the economic impact in different contexts.
Correctness of measurement and standardization of methods remains an issue in HbA1c measurement. Although the IFCC in conjunction with the EASD and ADA has proposed standardized methods, thus offering the possibility of international standardization of HbA1c measurement methods, it is still not fully adopted at this time.
More importantly all patients with abnormal hemoglobin cannot be diagnosed with diabetes using HbA1c assays, such as HbS, HbC, HbF and HbE, because they interfere with some HbA1c assays. and conditions with altered red blood cell cycles (such as hemolytic anemia, chronic malaria, massive blood loss or transfusion) will affect the HbA1c results. These individuals with questionable HbA1c levels must still be diagnosed using traditional glucose-based diagnostic methods. There is also the new onset of diabetes within 3 months where the sensitivity of glucose is clearly greater than that of HbA1c.
Prospects
Regardless of the approach used to diagnose diabetes, the ultimate goal is to achieve a universal approach to diabetes diagnosis. This method has sufficient evidence-based medical evidence and is simple, practical and inexpensive. However, given the clinical circumstances that interfere with the accurate measurement of HbA1c, it is unlikely that HbA1c will become this target method in the near future without important technical improvements. Therefore, it is premature to recommend the use of HbA1c as the sole diagnostic criterion for diabetes worldwide. Glucose measurement will continue to play an important role in the diagnosis of diabetes for quite some time to come.
This means that globally we could see different diagnostic methods used in different countries. Even in a given country, there should be a nationally accepted protocol for the diagnosis of diabetes to avoid confusion in the perception of patients and medical personnel. When deciding whether to choose HbA1c as a diagnostic method, one also needs to consider feasibility, cost, the accuracy of local HbA1c measurements, and the prevalence of clinical conditions that interfere with this method.
While the “diabetes diagnosis” debate continues, the inclusion of HbA1c in the diagnostic criteria is a step forward in achieving a universal diagnosis of diabetes and in reducing missed diagnoses worldwide.