HbA1c
(glycated hemoglobin)

HbA1c, or glycated hemoglobin, is the principal biochemical marker of chronic glucose exposure and occupies a central position in both the diagnosis and monitoring of diabetes mellitus. Unlike capillary blood glucose, which represents a point concentration strongly dependent on the time of day, the most recent meal, physical activity, acute stress or treatment at that precise moment, HbA1c reflects the integrated nonenzymatic glycation of hemoglobin within erythrocytes and therefore provides a retrospective measure of average blood glucose over the preceding weeks. This characteristic has made glycated hemoglobin an exceptionally useful clinical tool because it estimates how consistently the patient has been exposed to glucose levels associated with microvascular risk and, through a more complex relationship, macrovascular risk as well.

Its value, however, lies not only in summarizing glucose trends, but in linking that information to concrete therapeutic choices. An elevated HbA1c is not simply a “high” number; it indicates that tissues have been exposed to glucotoxicity for a prolonged period, with effects on the retina, kidneys, peripheral nerves, endothelium, immune system and cardiovascular compartment. Conversely, an apparently satisfactory HbA1c does not always mean optimal control because it may conceal wide fluctuations, recurrent hypoglycemia or biological conditions that alter its meaning. A true understanding of glycated hemoglobin therefore requires starting with its biochemical basis, proceeding through laboratory standardization, clarifying its use in diagnosis and follow-up, and finally recognizing its interpretive limitations with precision.

HbA1c results from the nonenzymatic glycation of hemoglobin A, the predominant hemoglobin in adults. Circulating glucose enters erythrocytes freely and reacts spontaneously with the terminal amino group of the valine residue in the beta chain of hemoglobin. An unstable Schiff base is formed initially and, through an Amadori rearrangement, develops into a more stable product. This final product is the glycated hemoglobin fraction measured as HbA1c. The process is not controlled by a regulated enzyme, but depends on glucose concentration and duration of exposure. The longer erythrocytes remain in contact with elevated glucose concentrations, the greater the proportion of glycated hemoglobin.

This explains why HbA1c is a marker of chronic, rather than acute, glycemia. Erythrocytes have an average lifespan of approximately 120 days, but their contribution to HbA1c is not distributed perfectly evenly across this entire interval. “Younger” red blood cells have accumulated less glycation, whereas “older” cells have accumulated more, and the most recent weeks contribute substantially to the final result. HbA1c is therefore often described as a summary of glucose trends over the previous two to three months, with relatively greater sensitivity to the most recent four to six weeks.

The major clinical strength of this marker is that hemoglobin glycation does not merely measure the presence of glucose in the blood, but translates cumulative exposure to glucose over time. This is pathophysiologically relevant because many chronic complications of diabetes do not depend on a single peak, but on prolonged glycemic burden, advanced glycation end-product formation, oxidative stress, protein kinase C activation, endothelial dysfunction and extracellular matrix alterations. Although glycated hemoglobin is not itself the mechanism of tissue injury, it acts as an excellent tracer of the chronic glycemic pressure that promotes organ damage.

It must nevertheless be remembered that HbA1c is not a direct measure of mean glucose obtained by adding all glucose values over time, but a biological estimate derived from an intracellular process. Individual differences therefore exist because of erythrocyte survival, glycation rate, hemoglobin variants and the laboratory method used. This is fundamental because it makes clear from the outset that glycated hemoglobin is a powerful but not infallible biomarker, and that correct clinical use always requires assessment of the biological context in which it is measured.

The global clinical adoption of HbA1c became possible only through extensive standardization. For many years, laboratories used heterogeneous methods whose results were not always comparable. The problem was not merely technical, but clinical: if a value measured in one laboratory was not equivalent to the values used in studies demonstrating the relationship between HbA1c and complications, the evidence could not be transferred reliably into practice. International harmonization programs were created to address this issue, particularly the National Glycohemoglobin Standardization Program, or NGSP, which linked results to the values used in landmark clinical trials such as the Diabetes Control and Complications Trial and the United Kingdom Prospective Diabetes Study.

In parallel, the International Federation of Clinical Chemistry and Laboratory Medicine, or IFCC, established a primary metrological reference system that is more analytically rigorous. This led to the current dual reporting format: in many settings, HbA1c is expressed both as a percentage according to the NGSP/DCCT tradition and in mmol/mol according to the IFCC standard. These are not two different tests, but two systems for expressing the same result. Dual reporting matters because the percentage remains deeply embedded in clinical practice in countries that historically used the DCCT trial framework, whereas the SI unit improves international laboratory consistency.

Standardization does not eliminate the need to know the analytical method. HbA1c can be measured using high-performance liquid chromatography, immunoassays, enzymatic methods, capillary electrophoresis and other techniques. These methods are not identical in their susceptibility to interference. Hemoglobin variants, elevated fetal hemoglobin, carbamylation in kidney failure and other factors may affect different systems differently. When an HbA1c result is inconsistent with the clinical picture, it is therefore not enough simply to repeat the test; the laboratory method must also be considered.

Clinically, the key point is that glycated hemoglobin should be interpreted only when it is produced by laboratories using methods that are certified and traceable to recognized standards. Otherwise, the number may appear formally precise but may not actually be comparable with international treatment targets or diagnostic thresholds. This is especially important in diagnosis, where an apparently small difference around 6.5% or the corresponding mmol/mol value can change the patient’s classification.

Another often-overlooked issue is that HbA1c analytical quality does not concern only the central laboratory. Point-of-care systems are also available and can be very useful for immediate patient management in the clinic. Not all rapid systems, however, have the characteristics required for diagnostic use. Organizational convenience must therefore never take precedence over metrological validity: the usefulness of the result depends on reliability before speed.

Glycated hemoglobin is not only a follow-up marker, but also a diagnostic test for diabetes mellitus, provided it is measured using standardized methods and no conditions are present that compromise reliability. The adoption of HbA1c for diagnosis was important because it offers practical advantages over plasma glucose or an oral glucose tolerance test. It does not require fasting, has less intraindividual variability than an isolated glucose measurement and reflects prolonged glucose exposure, reducing the risk that a single incidental measurement influenced by transient factors will be mistaken for a stable metabolic state.

From a classification standpoint, a value of 6.5%, corresponding to 48 mmol/mol, is the diagnostic threshold for diabetes when the test is valid and the clinical picture does not suggest interference. Prediabetes is identified by intermediate values, generally 5.7% to 6.4%, or 39 to 47 mmol/mol, a range that indicates an increased risk of future progression to diabetes and a greater likelihood of metabolic comorbidity. This intermediate range does not represent fully established disease, however, but a state of pathophysiological vulnerability in which varying degrees of insulin resistance, beta-cell dysfunction and abnormal regulation of postprandial and fasting glucose coexist.

Diagnostic use of HbA1c requires caution. In forms of rapid-onset diabetes, such as some cases of autoimmune diabetes, glucose may rise abruptly before HbA1c has had time to increase proportionally. In these scenarios, a normal or only modestly abnormal value does not exclude diabetes when the patient has classic symptoms, marked hyperglycemia or ketosis. Conversely, altered erythrocyte turnover can produce falsely low or falsely high results and compromise classification.

Diagnosis based on HbA1c must therefore always be integrated with the clinical context. In the presence of typical symptoms, overt hyperglycemia, a catabolic state or suspected type 1 diabetes, the priority is not the average over recent weeks but immediate documentation of the metabolic abnormality. In less acute situations, HbA1c can be very useful for identifying clinically significant chronic dysglycemia. Even when used for screening, interpretation should take account of age, ethnicity, comorbidities, current treatment, hematologic status and the biological plausibility of the result.

• Normal value: generally below 5.7%, or below 39 mmol/mol.
• Prediabetes: 5.7% to 6.4%, or 39 to 47 mmol/mol.
• Diabetes mellitus: at least 6.5%, or 48 mmol/mol or higher, when the method is valid and there is no relevant interference.

The most established role of HbA1c remains the long-term monitoring of diabetes. In this setting, the test helps determine whether pharmacological treatment, nutritional education, physical activity, self-monitoring and overall adherence are genuinely reducing average glucose exposure. Its main advantage is that HbA1c does not capture a single moment, but the integrated effect of thousands of glucose fluctuations over time, many of which would never be detected by a few isolated capillary measurements.

In practical terms, a reduction in glycated hemoglobin indicates that chronic glycemic pressure on tissues has decreased. This is why HbA1c became the principal efficacy endpoint of diabetes treatment throughout the history of diabetology. Its relationship with the risk of retinopathy, nephropathy and neuropathy is particularly strong, especially in the early and intermediate stages of disease. In major clinical trials, improvement in HbA1c was associated with fewer microvascular complications, making it a direct bridge between laboratory data and clinical prognosis.

Measurement frequency should be individualized. Patients with stable diabetes, good control and unchanged treatment can generally be assessed at wider intervals, whereas those who are not at target, change therapy or have metabolic instability require more frequent testing. This does not mean using HbA1c as the sole arbiter of follow-up, but as a central axis around which capillary glucose values, continuous monitoring, symptoms, weight, comorbidities and treatment safety are integrated.

A crucial point is that HbA1c describes average exposure, not the distribution of that exposure. Two patients with the same glycated hemoglobin may have very different profiles: one relatively stable and the other characterized by wide swings between hyperglycemia and hypoglycemia. From the laboratory perspective they may appear “the same,” but clinically they are not. This explains why HbA1c remains fundamental but insufficient, especially in the era of continuous glucose monitoring.

The greatest value of the test emerges when it is read sequentially rather than in isolation. The trajectory of HbA1c over time often conveys more than a single result. A moderately elevated HbA1c that is steadily improving after a therapeutic intervention has a different meaning from an identical value that is progressively worsening. Mature clinical practice therefore does not merely classify patients according to the current number, but evaluates the direction of change, speed of response and consistency with daily glucose data.

Interpretation of HbA1c cannot be separated from individualized targets. In the past, a single ideal value was often applied to everyone, but this approach is now outdated. The glycemic target depends on the balance between expected benefit and treatment risk: the probability of reducing chronic complications without increasing hypoglycemia, frailty, treatment burden or adverse effects. For many nonpregnant adults, a target below 7%, or below 53 mmol/mol, is a reasonable reference, but it is not universal.

A tighter target may be appropriate for younger people with a long life expectancy, low vulnerability to hypoglycemia, strong motivation and access to effective monitoring technology. In these cases, early reduction of chronic glucose exposure has a compelling prognostic rationale because cumulative time spent in hyperglycemia is a major determinant of future complication burden. Intensification makes sense, however, only when it can be achieved safely and sustainably.

Conversely, a less aggressive target may be more appropriate in frail older adults with multiple comorbidities, limited life expectancy, cognitive impairment, a history of severe hypoglycemia or a high treatment burden. In these patients, the immediate risk of an overly intensive strategy may exceed the theoretical benefit of a further reduction in HbA1c. Individualization is therefore not a concession to poor adherence, but a principle of therapeutic proportionality.

It should also be remembered that an HbA1c target does not automatically correspond to the target for every individual glucose reading. HbA1c is an average marker, so two patients with the same goal may require different strategies depending on whether fasting glucose, postprandial glucose, variability or hypoglycemia is the dominant problem. In practice, HbA1c defines the overall destination, while daily data help choose the route to reach it.

The correct goal is therefore not the lowest possible number, but the best possible clinical balance among glycemic control, safety, quality of life and sustainability. In this sense, HbA1c should be used as a shared decision-making tool rather than as an abstract report card. Its value is greatest when the target is explicit, justified and periodically reassessed.

To facilitate clinical understanding and communication, HbA1c can be translated into an estimated average glucose, or eAG. This concept is based on the observation that a reasonably robust relationship exists between HbA1c and mean glucose measured through repeated assessments. The conversion can be educationally useful because it links a laboratory marker expressed as a percentage or mmol/mol with the units patients encounter in daily life, namely mg/dL or mmol/L.

This translation must, however, be used intelligently. eAG is not a true average calculated from the individual patient’s glucose readings, but a derived estimate based on population-level statistical relationships. Some individuals show important discrepancies between HbA1c and mean glucose observed with continuous or capillary monitoring because of biological differences in glycation or erythrocyte survival. eAG is therefore valuable for explaining the concept, but it must not replace assessment of actual glucose values.

Comparison between HbA1c and daily monitoring is extremely informative. If HbA1c is higher than expected from capillary profiles, it is necessary to ask whether the patient is checking only at “favorable” times, whether substantial undetected hyperglycemia is occurring—especially postprandially or overnight—or whether a biological condition is distorting the test. If HbA1c appears unexpectedly low relative to actual data, shortened erythrocyte lifespan, hemoglobinopathies, bleeding, hemolysis or other causes of an artificially low result should be considered.

Methodologically, glycated hemoglobin and daily monitoring answer different questions. HbA1c asks “how high was glucose on average,” whereas daily monitoring asks “when, in which direction and with what variability does glucose move?” This distinction is decisive. A patient may have an acceptable HbA1c while spending considerable time in both hypoglycemia and hyperglycemia. Another may have a similar HbA1c with a far more stable profile. Advanced practice therefore integrates the methods rather than setting them against one another.

The principal interpretive limitation of HbA1c is that the test assumes normal erythrocyte survival and a hemoglobin structure compatible with the method used. When these assumptions are not met, the value may not faithfully represent actual glucose exposure. All conditions that shorten red-cell lifespan, including hemolysis, recent bleeding, some hemolytic anemias, erythropoietin treatment or recovery after hemorrhage, tend to lower HbA1c falsely because circulating blood contains a greater proportion of young erythrocytes that have been exposed to glucose for less time.

Conversely, conditions that prolong erythrocyte survival or alter production dynamics can falsely increase the value. A classic example is iron deficiency, in which slower red-cell turnover may be associated with a higher-than-expected HbA1c. Some forms of anemia, asplenia and other states in which erythrocytes remain in circulation longer can also lead to overestimation of metabolic control. In practice, an elevated HbA1c in a patient with anemia should never be interpreted automatically without evaluating hematologic status.

Hemoglobinopathies are another major category. Variants such as HbS, HbC, HbD and HbE, or elevated fetal hemoglobin, may interfere with some methods but not others. The problem is twofold: the variant may genuinely alter glycation or erythrocyte survival, or the laboratory method may measure the glycated fraction incorrectly. A known or suspected hemoglobinopathy therefore requires knowledge of the method used by the laboratory and, when necessary, use of alternative markers.

Chronic kidney disease can also affect test reliability through complex mechanisms. Uremia, hemoglobin carbamylation, erythropoietin use, transfusions, anemia and altered red-cell turnover can make HbA1c less representative of actual glycemic control. Pregnancy is another critical setting because erythrocyte turnover changes and the relationship between HbA1c and glucose may differ from that in nonpregnant individuals. Recent transfusions, major bleeding, transplantation or hematologic disease may make the result substantially uninterpretable.

The most useful signal for the clinician is discordance. When HbA1c is inconsistent with capillary glucose, continuous monitoring, the clinical picture or the patient’s symptoms, reality should not be forced to fit the number. The biological or analytical explanation for the discordance should instead be sought. In these situations, the most serious error is not having an imperfect test, but ignoring its limitations and making treatment decisions on the basis of an unreliable value.

The use of HbA1c varies considerably among populations. In type 1 diabetes, the test remains fundamental for estimating chronic control, but its interpretation almost always needs to be integrated with capillary or continuous monitoring because glycemic variability and the risk of hypoglycemia carry greater weight than in many forms of type 2 diabetes. In other words, two patients with identical HbA1c values in type 1 diabetes may have very different clinical profiles, and the average alone is insufficient to guide optimal management.

In type 2 diabetes, HbA1c retains a central role in both diagnosis and follow-up. Its meaning nevertheless changes with the stage of disease. Early on, it can be particularly informative for documenting the chronic hyperglycemic burden and guiding treatment intensification. In advanced stages, especially in the presence of kidney failure, cardiovascular comorbidities, anemia or polypharmacy, the value must be interpreted more cautiously within an overall strategy of individualized targets.

During pregnancy, HbA1c is useful mainly for preconception assessment or initial classification, but it has clear limitations as the primary monitoring tool. Control requirements are stringent, pregnancy-related hematologic changes affect the test and the potential harm of even modest hyperglycemia makes daily monitoring more important. Glycemic management during pregnancy therefore cannot rely on glycated hemoglobin alone.

In a frail older adult, HbA1c must be interpreted in relation to frailty, risk of hypoglycemia, cognitive capacity, functional independence and realistic goals of care. A value that is not perfectly “at target” may be acceptable or even preferable if it prevents dangerous acute events. Individualization is especially important in this age group because the prognostic meaning of tight control differs substantially from that in a younger person with a long life expectancy.

In people with hemoglobinopathies, chronic anemia, liver disease, advanced nephropathy, recent transfusions or treatments that alter erythropoiesis, HbA1c loses some reliability and complementary tools such as fructosamine, glycated albumin or continuous glucose monitoring may be needed. The choice of an alternative marker should not be automatic, but guided by the type of interference and the clinical question.

One of the most important skills in interpreting HbA1c is recognizing when the value does not credibly “tell the story” of what is happening to the patient. Clinical-laboratory discordance can take many forms. The most common is an elevated HbA1c in a patient whose home glucose readings appear good. In this situation, the first suspicion should not immediately be laboratory error, but incomplete glucose data: tests performed only while fasting, no postprandial assessment, omission of the worst days or failure to detect nocturnal hyperglycemia.

The opposite situation—an HbA1c that is surprisingly low despite frequently elevated daily values—requires early consideration of accelerated erythrocyte turnover, hemolysis, bleeding, transfusions, erythropoietin therapy or hemoglobin variants. It may be useful to compare HbA1c with the mean obtained from continuous monitoring and with the GMI, or glucose management indicator, which translates sensor data into a value analogous to HbA1c. When GMI and HbA1c diverge consistently, the explanation may be biological rather than behavioral.

The correct response to discordance is not to change treatment immediately, but to validate the result. The laboratory method should be reviewed; hematologic disorders or anemia should be assessed; creatinine, ferritin, complete blood count, hemoglobin status and transfusion history should be checked; and glucose logs or sensor traces should be reexamined. Only after this verification can it be determined whether the problem is real, underestimated or overestimated.

In some situations, the discordance itself becomes valuable clinical information. A patient with HbA1c within the reference range but frequent hypoglycemia and wide hyperglycemic excursions has only apparently good control. Another with a moderately elevated HbA1c but a rapidly improving profile may be on the right trajectory and may not require immediate intensification. Mature interpretation of glycated hemoglobin means never treating it as self-sufficient, but placing it within a coherent metabolic narrative.

In contemporary diabetology, HbA1c is no longer considered the sole indicator of control quality, but its role remains central when integrated with other tools. SMBG, or self-monitoring of blood glucose, provides point values that can immediately guide action; CGM, or continuous glucose monitoring, allows analysis of trends, time in range, time below range, time above range and glycemic variability; HbA1c provides a biological summary of chronic glucose exposure. None of these tools is sufficient by itself for every patient.

The most useful integration comes from comparing what each indicator does best. HbA1c is particularly strong in linking metabolic control to the risk of microvascular complications. CGM is superior for documenting hypoglycemia, postprandial excursions, nocturnal trends and variability. Capillary self-monitoring retains a role in confirmation and immediate decision support. In the individual patient, best practice consists of making these data interact rather than replacing one with another.

There are situations in which continuous monitoring adds decisive information to an apparently reassuring HbA1c. A patient whose glycated hemoglobin is at target but who has wide variability may spend many hours outside range, a condition that the biological average tends to smooth over. Conversely, a patient whose HbA1c is not yet optimal but whose time in range has improved substantially and whose variability has decreased may be experiencing a favorable treatment response that glycated hemoglobin has not yet fully registered because of its temporal inertia.

HbA1c has therefore not been made obsolete by technology, but its role has been redefined. It remains the most robust marker of chronic glycemia and continues to serve as a common language in many clinical trials, guidelines and quality-of-care programs. Its best interpretation today, however, is not isolated but multicentric, combining average exposure, variability, safety and clinical context.

The advantages of HbA1c are clear. It is a stable, standardized, nonfasting test with less day-to-day variability than an isolated glucose measurement, a strong relationship with the risk of microvascular complications and usefulness in both diagnosis and follow-up. Its apparent simplicity has supported enormous clinical adoption and made it possible to establish shared targets, audit programs and treatment-intensification strategies based on a common marker.

Its limitations are equally real. HbA1c does not describe glycemic variability, does not detect hypoglycemia well, can be distorted by numerous hematologic or methodological conditions and does not always reflect the individual patient’s mean glucose precisely. Its prognostic value, although robust, should not be oversimplified. Cardiovascular risk in a patient with diabetes also depends on blood pressure, lipids, inflammation, kidney function, smoking, obesity, thrombotic status, duration of diabetes and the overall risk profile, not only on glycated hemoglobin.

Prognostically, HbA1c retains enormous importance particularly as a cumulative measure of glucose exposure. A long history of elevated HbA1c is associated with a greater likelihood of microvascular damage and contributes to the concept of metabolic memory: early and sustained control leaves a favorable imprint on future outcomes. This explains why optimizing HbA1c, when feasible and safe, has value beyond the numerical result at a single point in time.

In conclusion, glycated hemoglobin is one of the most useful and most misunderstood biomarkers in metabolic medicine. Its strength lies in its ability to summarize the chronology of glycemia; its risk lies in being interpreted as a self-sufficient number. Correct interpretation always requires understanding the biological mechanism that generates it, the method by which it was measured, the population in which it is applied and its consistency with other clinical data. When used in this way, HbA1c remains an irreplaceable tool in diabetes care.