Glycated Hemoglobin

Glycated hemoglobin is hemoglobin that has undergone glycation, the nonenzymatic attachment of sugars to a protein. Its principal clinical measure is hemoglobin A1c (HbA1c), a specific form in which glucose-derived groups are attached to the terminal valine residues of hemoglobin’s beta chains. Although “glycated hemoglobin” and “HbA1c” are often used interchangeably in clinical language, glycated hemoglobin is the broader biochemical category. (diabetesjournals.org)

HbA1c is an important biomarker of glucose exposure because it reflects glycemia over approximately the preceding two to three months, rather than the concentration at a single moment. Its measurement supports the diagnosis of diabetes mellitus and the assessment of established diabetes. Standardization connects laboratory results with evidence from major clinical trials, allowing measurements to be compared across laboratories and related to the risk of diabetic complications. (niddk.nih.gov)

The interpretation of HbA1c depends on both measurement accuracy and the biology of red blood cells. Hemoglobin variants, altered cell survival, and several clinical conditions can change the relationship between HbA1c and average glucose. Its biochemical formation, historical development, analytical methods, reporting systems, clinical uses, and relationship to other measures of glycemia therefore form interconnected parts of its significance in laboratory medicine. (ngsp.org)

Biochemical Identity and Terminology

Glycated hemoglobin is a family of hemoglobin molecules bearing sugar-derived modifications covalently attached to amino groups in their globin chains. In the context of blood glucose measurement, these are principally products of the nonenzymatic reaction of glucose with hemoglobin. The broader category, commonly abbreviated GHb and also called glycohemoglobin, includes modifications at several molecular sites. HbA1c, by contrast, identifies a particular glucose-modified species: hemoglobin A with a stable adduct at the N-terminal valine of one or both beta chains. Thus, glycated hemoglobin and HbA1c are related terms, but they are not chemically interchangeable. The 2023 laboratory guidelines by David B. Sacks and colleagues explicitly distinguish GHb as the general category from HbA1c as the specific clinical analyte. (diabetesjournals.org)

Adult hemoglobin A, abbreviated HbA, consists of two alpha and two beta globin chains, expressed as $\alpha_2\beta_2$, and accounts for approximately 97% of hemoglobin in normal adult life. The remaining fraction consists chiefly of hemoglobin A2, whose non-alpha chains are delta chains, and fetal hemoglobin, whose non-alpha chains are gamma chains. HbA1c retains the alpha–beta composition of HbA; its distinguishing feature is a chemical modification of the existing beta chain rather than replacement by another globin-chain type. The designation “N-terminal” locates this modification at the amino end of the chain, specifically on the free amino group of its first amino acid, valine. (ncbi.nlm.nih.gov)

The stable glucose-derived substituent is termed N-(1-deoxyfructos-1-yl), giving the more chemically explicit description beta-N-(1-deoxyfructos-1-yl)hemoglobin. The word “fructosyl” describes the rearranged structure of the attached group; it does not imply that circulating fructose, rather than glucose, was the starting sugar. Formation involves an initial reversible sugar–amino-group reaction followed by rearrangement into a stable ketoamine. Other glucose-derived hemoglobin adducts occur at the N-terminal valines of alpha chains and at internal lysine amino groups. These belong to total glycated hemoglobin but do not, by themselves, define HbA1c. (publications.iupac.org)

The historical expression “glycosylated hemoglobin” remains recognizable but obscures a biochemical distinction. In this setting, glycation denotes nonenzymatic sugar attachment to a protein, whereas glycosylation generally denotes saccharide attachment catalyzed by an enzyme, such as a glycosyltransferase. HbA1c formation does not require such an enzyme. The 1988 IUPAC–IUB nomenclature recommendations specifically identified “glycosylated hemoglobin” and “glucosylated hemoglobin” as inappropriate descriptions of these nonenzymatic reaction products and recommended terminology based on glycation or the deoxyfructosyl modification. This distinction concerns the chemical route and resulting bond structure, not merely a preference between synonyms. (publications.iupac.org)

Related abbreviations also require separation. HbA1 denotes a group of minor hemoglobin fractions comprising HbA1a, HbA1b, and HbA1c; it is not another name for HbA1c alone. A1C, without the hemoglobin prefix, is the simplified clinical name used for the HbA1c test, rather than a separate molecular species. Likewise, an assay described as measuring “total glycated hemoglobin” may recognize a broader collection of glucose–hemoglobin adducts while being calibrated to report an HbA1c-equivalent result. The molecular target, the collection of compounds detected by an assay, and the name attached to its reported result are therefore distinct aspects of the terminology. (diabetesjournals.org)

Formation and Persistence in Red Blood Cells

Circulating red blood cells are continuously exposed to glucose, which enters through the insulin-independent glucose transporter GLUT1; intracellular glucose concentrations therefore remain close to those in plasma. Glucose reacts spontaneously with accessible amino groups on hemoglobin, without requiring an enzyme. For HbA1c, the relevant site is the free amino group of the N-terminal valine on a hemoglobin beta chain. The reaction begins when the carbonyl group of glucose condenses with this amino group, producing a glucose–hemoglobin adduct called a Schiff base, or aldimine. This initial attachment is reversible: the adduct can dissociate back into glucose and unmodified hemoglobin rather than proceeding to stable glycation. (pmc.ncbi.nlm.nih.gov)

The Schiff-base intermediate is often termed labile HbA1c, labile A1C, or pre-A1c. A proportion undergoes an Amadori rearrangement, which reorganizes the glucose-derived portion into a substantially more stable ketoamine: stable HbA1c. The distinction concerns chemical structure and persistence, not merely two names for the same substance. In a 1981 kinetic study, investigators separately measured formation and dissociation of the labile intermediate and its conversion into stable HbA1c. They used radiolabeled glucose and cyanoborohydride to trap the Schiff base during incubations lasting up to eight hours, whereas determination of the rearrangement rate used longer incubations of six to twenty-one days. These experimental durations describe the study’s measurement procedures, not fixed waiting periods for HbA1c formation in circulating cells. (sciencedirect.com)

Labile and stable fractions consequently respond differently to changing glucose exposure. The labile fraction can increase rapidly when glucose rises and decrease when glucose falls, while stable HbA1c accumulates over a longer interval. James M. Hempe and colleagues demonstrated this distinction experimentally in 2012 by incubating erythrocytes at 37 °C with glucose concentrations ranging from 0 to 55.6 mmol/L, then transferring glucose-loaded cells into glucose-free buffer. Unstable hemoglobin complexes increased in the presence of glucose and decreased in its absence; separated unstable subfractions showed different equilibration times, approximately four and twenty hours. The clinical HbA1c analyte is the stable product, not the sum of stable HbA1c and these readily dissociable precursors. (sciencedirect.com)

Stable HbA1c is effectively retained over the remaining life of the erythrocyte, rather than disappearing promptly when plasma glucose decreases. Red-cell survival is conventionally described as approximately 120 days, although individual cells and individuals differ. Under comparable glucose exposure, older cells have had more time to accumulate glycation than newly released cells. A blood sample therefore contains a mixture of cells with different ages and different HbA1c contents. Removal of older erythrocytes and replacement by younger cells continually renew this mixture, making cell survival as well as glucose exposure a determinant of the measured fraction. (pmc.ncbi.nlm.nih.gov)

This dependence on cell age is measurable even without an overt blood disorder. A 2008 study of twelve participants with and without diabetes found mean circulating red-cell ages ranging from 38.4 to 59.5 days; these are population mean ages, not the complete lifespans of individual erythrocytes. The observed survival differences were sufficient to alter HbA1c at a given mean glucose concentration. Thus, HbA1c’s role as a biomarker follows from accumulated chemical exposure and erythrocyte turnover, not from HbA1c acting as a feedback regulator that controls glucose concentration or determines red-cell survival. (ashpublications.org)

Relationship to Average Blood Glucose

HbA1c reflects integrated exposure to blood glucose over approximately two to three months, rather than the glucose concentration at the moment of sampling. This interval arises from the accumulation of glucose-modified hemoglobin in circulating red blood cells, whose lifespan averages about 120 days. The measurement is not an equally weighted arithmetic average of the preceding 90 or 120 days: glucose exposure during the most recent 30 days contributes more strongly than exposure several months earlier. Recent glucose affects cells across the circulating age distribution, whereas earlier exposure is retained only in cells that have survived since that time. Consequently, a sustained change in glucose can alter HbA1c before an entire red-cell population has been replaced. (niddk.nih.gov)

The empirical basis for expressing HbA1c as estimated average glucose (eAG) was established by the A1c-Derived Average Glucose (ADAG) study, published in 2008 by David M. Nathan and colleagues. Its final analysis included 507 adults from 10 international centers: 268 with type 1 diabetes, 159 with type 2 diabetes, and 80 without diabetes. During three months, participants underwent at least two days of continuous glucose monitoring on four occasions and performed seven-point fingerstick glucose profiles on at least three days each week. Approximately 2,700 glucose observations per participant were used to estimate average glucose, which was compared with HbA1c measured in a central laboratory at the end of the observation period. (pmc.ncbi.nlm.nih.gov)

Linear regression yielded the relationship

with $R^2=0.84$. Here, HbA1c is entered as its numerical percentage—for example, 7.0 rather than 0.07. The equation gives approximately 126 mg/dL (7.0 mmol/L) for HbA1c of 6.0%, 154 mg/dL (8.6 mmol/L) for 7.0%, and 183 mg/dL (10.2 mmol/L) for 8.0%. An eAG reported alongside HbA1c is therefore a calculated translation into familiar glucose units, not a second measurement of glucose in the blood specimen. Nor is it necessarily identical to an average calculated from a person's glucose-meter readings, which may sample particular times of day disproportionately. (pmc.ncbi.nlm.nih.gov)

A brief glucose excursion generally has little influence on HbA1c because it represents only a small portion of cumulative exposure over weeks to months. A single post-meal rise therefore cannot be reconstructed from HbA1c, although repeated or prolonged elevations can contribute substantially. Conversely, a sustained rise or fall over several weeks can affect the result appreciably because recent exposure receives greater weight. HbA1c does not preserve the sequence or timing of glucose values: similar results can accompany relatively steady glucose or alternating high and low concentrations. This is a limitation of an integrated biomarker, rather than evidence that short-term fluctuations did not occur. (niddk.nih.gov)

The strong population-level correlation does not establish an exact conversion for every individual. The ADAG regression retained residual variation, and red-cell survival can change HbA1c independently of average glucose. In a 2008 study, Robert M. Cohen and colleagues used biotin labeling to track red cells in six people with diabetes and six controls without diabetes. Mean circulating red-cell age ranged from 39 to 56 days in the diabetes group and from 38 to 60 days in controls, despite otherwise normal hematological findings. The observed survival differences were sufficient to produce materially different HbA1c values at a given mean glucose. Thus, the ADAG equation describes an expected relationship across a studied population; it does not remove individual variation in the duration of hemoglobin's exposure to glucose. (pmc.ncbi.nlm.nih.gov)

Discovery and Development as a Clinical Test

The separation of minor hemoglobin fractions preceded recognition of their clinical significance. In 1958, David W. Allen, W. A. Schroeder, and Joan Balog used cation-exchange chromatography to investigate the heterogeneity of normal adult hemoglobin, resolving minor components from the predominant fraction. These investigations established that apparently normal blood contained several distinguishable hemoglobin fractions rather than a single homogeneous substance. Subsequent structural work focused on the component designated HbA1c: in 1968, Robert M. Bookchin and Paul M. Gallop identified a hexose-associated modification at the N-terminal region of its beta chains. Thus, HbA1c was already an identifiable biochemical fraction before its elevation in diabetes became established. (doi.org)

Samuel Rahbar reported an unusual hemoglobin in the red cells of people with diabetes mellitus in 1968. On August 22, 1969, Rahbar, Olga Blumenfeld, and Helen M. Ranney published Studies of an unusual hemoglobin in patients with diabetes mellitus in Biochemical and Biophysical Research Communications. They found that the unusual fraction resembled HbA1c isolated from normal subjects and reported an approximately twofold increase in diabetic patients. Their structural interpretation still included the possibility of an amino sugar attached to hemoglobin; the definitive glucose-derived chemistry emerged through subsequent work. The important clinical observation was therefore a quantitative increase in an existing minor fraction, not the discovery of a hemoglobin confined to people with diabetes. (diabetesjournals.org)

In 1975, H. Franklin Bunn, David N. Haney, Kenneth H. Gabbay, and Paul M. Gallop provided evidence that glucose initially attached to the beta-chain N-terminal amino group through a Schiff-base linkage, followed by rearrangement into a more stable ketoamine. Bunn and colleagues then investigated its formation in living red cells in a study published in June 1976. After administering iron-59-labelled transferrin to two individuals, they observed that labelled major hemoglobin appeared promptly, whereas labelled HbA1c accumulated slowly, reaching comparable specific activity at about day 60. This demonstrated that HbA1c arose through gradual modification after hemoglobin synthesis, providing a mechanistic basis for interpreting it as a record of exposure during erythrocyte circulation rather than as a contemporaneous glucose measurement. (researchgate.net)

The clinical application became explicit in Ronald J. Koenig, Anthony Cerami, and colleagues’ August 1976 study of five hospitalized diabetic patients. Before improved glycemic control, their mean fasting blood glucose was 343 mg/dL and mean HbA1c was 9.8%; during improved control, these values were 84 mg/dL and 5.8%, respectively. The investigators concluded that HbA1c reflected mean glucose exposure over preceding weeks to months and proposed periodic measurement to document sustained metabolic control. Although small, this longitudinal study supplied evidence that HbA1c changed with treatment, moving its role beyond an association with diabetes toward use as a monitoring biomarker. (pubmed.ncbi.nlm.nih.gov)

Commercial HbA1c assays became available in 1978, and testing expanded during the 1980s. Routine adoption exposed substantial comparability problems: laboratories used different methods, measured different glycated fractions, and applied different calibration systems. Following publication of the Diabetes Control and Complications Trial results in 1993, the American Association for Clinical Chemistry established a committee to develop a standardization protocol. The National Glycohemoglobin Standardization Program (NGSP), launched in 1996, implemented a reference-laboratory network that aligned routine results with the assay used in that trial. Its practical purpose was to make an HbA1c value measured in an ordinary laboratory comparable to the values underlying clinical-trial evidence, without requiring every laboratory to adopt the same analytical technique. (diabetesjournals.org)

Laboratory Measurement and Analytical Methods

HbA1c is measured in whole blood, obtained by venipuncture or finger-stick capillary sampling, because the analyte is contained within red blood cells. Fasting is unnecessary. Collection tubes require an anticoagulant compatible with the assay; EDTA is commonly acceptable. Storage limits are method-dependent: the 2023 laboratory guidelines by David B. Sacks and colleagues describe whole-blood specimens as generally stable for up to one week at $4^\circ\mathrm{C}$, but manufacturer-specific instructions determine acceptable handling. Capillary collection systems using filter paper or stabilizing solutions require validation against conventional specimens for the particular assay. Sample preparation includes cell lysis to release hemoglobin, either before analysis or within the instrument. (diabetesjournals.org)

Ion-exchange high-performance liquid chromatography (HPLC) separates hemoglobin fractions by differences in charge and their interaction with a charged stationary phase. The resulting chromatogram permits quantification of the HbA1c fraction and inspection of other peaks. Capillary electrophoresis instead separates fractions through their differing migration in an electric field. Both approaches can provide information about abnormal hemoglobin fractions, although their resolution and identification capabilities depend on the analytical system. Boronate-affinity methods exploit a different property: boronate groups bind the cis-diol structures of glucose-derived modifications. They capture a broader glycated-hemoglobin fraction rather than exclusively the beta-chain modification defining HbA1c; calibration expresses the measurement as an HbA1c-equivalent result. (diabetesjournals.org)

Immunoassays use antibodies recognizing the glycated N-terminal region of the hemoglobin beta chain. Depending on the design, antibody binding produces or inhibits particle agglutination, which is measured optically; a separate total-hemoglobin measurement commonly supplies the denominator. Enzymatic assays replace antibody recognition with selective biochemical reactions. For example, an FDA-reviewed assay submitted in 2013 uses protease digestion to release fructosyl-valyl-histidine from the beta-chain N-terminus. Fructosyl peptide oxidase then generates hydrogen peroxide, producing a measurable signal related to the glycated peptide concentration. Total hemoglobin is measured separately, and the instrument calculates the reported result. Such assays can operate on automated chemistry analyzers without generating a hemoglobin-fraction chromatogram. (accessdata.fda.gov)

Central-laboratory testing generally uses automated systems operated within laboratory quality-management arrangements. Point-of-care testing brings the assay into a clinic or other patient-care setting and can return a result during the visit; it describes the operational setting, not a unique measurement principle. A 2024 method-validation study in Uganda, for example, compared laboratory electrophoresis and immunoassay with a boronate-affinity point-of-care platform. Decentralized testing makes operator competency, reagent storage, lot changes, and local quality-control practices particularly important. In the United States, oversight depends on the test’s regulatory complexity category: CLIA-waived testing does not carry the same mandatory proficiency-testing requirements as nonwaived laboratory testing. Manufacturer certification therefore does not, by itself, establish the performance achieved at every testing site. (pmc.ncbi.nlm.nih.gov)

Analytical precision describes the dispersion of repeated measurements; bias describes systematic departure from an assigned reference value. The 2023 guidelines specify goals of within-laboratory coefficient of variation below 1.5%, between-laboratory variation below 2.5%, and ideally no measurable bias. These relative coefficients are distinct from percentage-point changes in a reported HbA1c result. Chromatographic peaks, electrophoretic patterns, and instrument flags can reveal an unexpected fraction or problematic separation, but are not definitive identification of a hemoglobin variant. Conversely, a numerical result without a warning does not establish freedom from interference. Systems differ in which variants they resolve or detect, so comparability depends on the particular method and its documented performance rather than on universal interchangeability within an assay class. (diabetesjournals.org)

Standardization, Traceability, and Reporting Units

The NGSP, established in 1996 as the National Glycohemoglobin Standardization Program, aligns routine HbA1c measurements with the assay used in the Diabetes Control and Complications Trial (DCCT). Its primary reference laboratories maintain the Bio-Rex 70 cation-exchange chromatographic method used in that trial; secondary reference laboratories compare results with this reference and support manufacturers’ standardization and certification. This preserves the numerical scale underlying the DCCT’s clinical findings. The International Federation of Clinical Chemistry and Laboratory Medicine (IFCC) provides a complementary, higher-order reference measurement system based on a chemically defined analyte. NGSP alignment therefore preserves continuity with clinical evidence, whereas IFCC standardization supplies the reference framework for metrological traceability. The two networks are linked through specimen comparisons and a monitored conversion relationship. (ngsp.org)

The IFCC reference procedures, approved in July 2001 and published in 2002 by Jan-Olof Jeppsson and colleagues, measure the glycated and nonglycated N-terminal peptides of the hemoglobin beta chain. After hemolysis, the enzyme endoproteinase Glu-C releases six-amino-acid peptides from these chains. In one procedure, high-performance liquid chromatography separates the peptides before electrospray-ionization mass spectrometry quantifies them. In the other, chromatography isolates the relevant peptide fraction, followed by capillary electrophoresis and quantitative detection. Mixtures of purified HbA1c and nonglycated HbA0 serve as primary calibrators. The peptide measurements establish the amount-of-substance fraction $\mathrm{HbA1c}/(\mathrm{HbA1c}+\mathrm{HbA0})$, reported in millimoles per mole. These procedures assign reference values rather than serving as routine patient-testing methods. (ngsp.org)

NGSP results are conventionally reported as percentages, whereas IFCC results are reported as mmol/mol; the IFCC recommended the latter units in 2007 to distinguish the scales. Their relationship is expressed by the master equation:

Thus, an NGSP result of $7.0\%$ corresponds to approximately $53\ \mathrm{mmol/mol}$. This is not a simple conversion of percentage into millimoles per mole: the nonzero intercept reflects the relationship between two differently established measurement scales. Multiplying an NGSP percentage by ten would therefore give an incorrect IFCC value. Reporting either scale—or both together—requires an explicit unit designation. (ngsp.org)

Calibration traceability is the documented chain connecting routine assay results, through calibrators and assigned values, to the reference system. In the IFCC hierarchy, primary calibrators calibrate the reference procedure, which assigns values to secondary materials used to standardize manufacturers’ assays. Manufacturer certification is a separate demonstration that a specified analytical system meets comparison criteria. Under the NGSP protocol, manufacturers compare 40 fresh or fresh-frozen blood samples with a secondary reference laboratory; at least 36 results must be within 5% of the reference laboratory’s means. Certificates apply to the tested instrumentation, reagent lots, and calibrator lots, remain valid for one year, and require annual renewal. Certification does not establish the performance of every installation using that system. (ifcchba1c.org)

Laboratory quality control addresses continuing performance at the testing site rather than the origin of calibrator values. The 2023 laboratory guidelines endorsed by the American Association for Clinical Chemistry and American Diabetes Association describe controls at two HbA1c levels, representing lower and higher concentrations, for detecting analytical instability. External proficiency testing independently evaluates laboratory results against assigned targets and other participants’ performance. Accuracy-based testing with suitable whole-blood specimens can reveal bias that manufacturer certification alone does not exclude. Processed materials, including lyophilized specimens, may show matrix effects, so agreement on such materials cannot automatically establish equivalent agreement on patient blood. Calibration traceability, certification, internal control, and external assessment consequently address distinct parts of the measurement process. (diabetesjournals.org)

Screening and Diagnosis of Diabetes

The American Diabetes Association (ADA), in its Standards of Care in Diabetes—2026, defines an HbA1c concentration of ≥6.5% (≥48 mmol/mol) as a diagnostic criterion for diabetes mellitus in nonpregnant individuals, subject to confirmation when hyperglycemia is not unequivocal. Alternative criteria are fasting plasma glucose ≥126 mg/dL (≥7.0 mmol/L), or a two-hour plasma glucose ≥200 mg/dL (≥11.1 mmol/L) during a 75-g oral glucose tolerance test (OGTT). HbA1c requires no fasting, making it convenient for screening. These tests identify overlapping but nonidentical populations; a result below the HbA1c threshold does not exclude diabetes established by glucose criteria. The tests establish glycemic status but do not, by themselves, distinguish type 1 from type 2 diabetes or other etiological categories. (diabetesjournals.org)

The ADA increased-risk category, commonly termed prediabetes, includes HbA1c 5.7–6.4% (39–47 mmol/mol). This is a guideline-defined interval, not a biological boundary at which risk suddenly appears: progression to type 2 diabetes remains possible below 5.7%, and risk increases toward the upper end of the interval. HbA1c-defined prediabetes, impaired fasting glucose, and impaired glucose tolerance do not necessarily identify the same individuals. The NIDDK therefore describes HbA1c, fasting glucose, and the two-hour OGTT as complementary screening options rather than interchangeable measurements. A value in the prediabetes range denotes increased risk, not established diabetes or inevitable progression. (diabetesjournals.org)

In the absence of unequivocal hyperglycemia, diagnosis requires two abnormal results: repeated measurements of the same test, or two different tests meeting their respective diabetes thresholds, obtained together or at separate time points. When HbA1c and fasting glucose disagree, the test above its diagnostic threshold is repeated. For example, two HbA1c results ≥6.5% can confirm diabetes despite fasting glucose below 126 mg/dL, provided the HbA1c measurements are valid. Persistent, substantial discordance warrants investigation of analytical interference or an altered HbA1c–glucose relationship. Confirmation is unnecessary when classic hyperglycemic symptoms or a hyperglycemic crisis accompany random plasma glucose ≥200 mg/dL (≥11.1 mmol/L). Results close to a diagnostic boundary are generally reassessed within three to six months under ADA guidance. (niddk.nih.gov)

Diagnostic HbA1c testing requires a method certified by the National Glycohemoglobin Standardization Program (NGSP) and traceable to the Diabetes Control and Complications Trial reference assay. The 2026 ADA standards restrict point-of-care screening and diagnosis to devices approved for diagnosis by the U.S. Food and Drug Administration, operated by trained personnel in Clinical Laboratory Improvement Amendments–certified laboratories performing moderate-complexity or higher testing. Authorization for monitoring established diabetes does not automatically establish suitability for diagnostic use. (diabetesjournals.org)

Plasma glucose criteria are more appropriate when HbA1c no longer reliably represents glycemic exposure. Conditions affecting erythrocyte survival are particularly important: shortened survival after hemolysis or acute blood loss can lower HbA1c independently of glucose, regardless of assay method. Significant iron-deficiency anemia can instead increase HbA1c, while hemoglobin variants may produce method-specific interference. These mechanisms require different responses—selecting a noninterfering assay can address some variant-related errors, but cannot correct abnormal red-cell turnover. (ngsp.org) Diagnostic thresholds also serve a different purpose from subsequent monitoring targets: after diagnosis, serial HbA1c measurements assess sustained glycemia and treatment response, with targets determined by clinical circumstances rather than simply by the 6.5% diagnostic boundary. (niddk.nih.gov)

Monitoring Diabetes and Evidence from Clinical Trials

Serial HbA1c measurements are used in type 1 diabetes and type 2 diabetes to assess sustained glycemia and response to treatment. Because each measurement reflects approximately two to three months of glucose exposure, successive results show whether a change in treatment is associated with a maintained change in glycemia rather than a single favorable blood glucose reading. HbA1c therefore complements, rather than replaces, measurements used for day-to-day treatment decisions. (doi.org)

The Diabetes Control and Complications Trial (DCCT), conducted from 1983 to 1993, randomly assigned 1,441 participants with type 1 diabetes to intensive or conventional treatment and followed them for an average of 6.5 years. Intensive treatment used an insulin pump or at least three daily injections, guided by frequent glucose measurements. Average HbA1c was approximately 7% with intensive treatment and 9% with conventional treatment. Intensive treatment reduced the relative risk of developing retinopathy by 76% among participants without retinopathy at entry and reduced its progression by 54% among those with existing disease. Neuropathy risk was reduced by 60%, but severe hypoglycemia occurred approximately three times as frequently. These were effects of randomized treatment strategies, not a comparison in which HbA1c itself was independently manipulated. (niddk.nih.gov)

The UK Prospective Diabetes Study (UKPDS) provided corresponding evidence in newly diagnosed type 2 diabetes. Its principal sulfonylurea–insulin comparison, published in 1998 as UKPDS 33, achieved median HbA1c values of 7.0% with intensive treatment and 7.9% with conventional treatment over a median ten years. Intensive treatment reduced the relative risk of microvascular endpoints by 25%, including endpoints involving retinal photocoagulation, and reduced the aggregate diabetes-related endpoint by 12%. The initial trial did not establish a statistically significant reduction in myocardial infarction. Post-trial follow-up subsequently demonstrated persistent microvascular benefit and emerging reductions in myocardial infarction and mortality despite convergence of HbA1c between groups—the phenomenon termed a “legacy effect.” Such findings concern average effects in study populations; they do not assign an individual a fixed probability of complications from one HbA1c result. (doi.org)

The American Diabetes Association’s 2026 Standards of Care distinguish monitoring frequency from the choice of glycemic target. Glycemic status is assessed at least twice yearly when stable and more frequently, commonly every three months, after treatment changes or when goals are unmet. The guideline identifies HbA1c below 7% (53 mmol/mol) as appropriate for many nonpregnant adults, while allowing lower goals, such as below 6.5% (48 mmol/mol), when health, function, treatment risk, and burden permit. Less stringent goals may be appropriate with frailty, major comorbidity, cognitive or functional limitations, or disproportionate treatment harms. These are guideline categories, not personal prescriptions or biological thresholds separating safety from danger. (doi.org)

HbA1c does not describe the timing or amplitude of glucose fluctuations. Similar results can accompany relatively steady glucose or alternating high and low concentrations; the value alone cannot establish the frequency, duration, or severity of hypoglycemia. Direct glucose measurements and continuous monitoring supply this otherwise missing temporal information. In clinical trials, HbA1c is a surrogate endpoint: a measurable treatment-response marker used in place of waiting for some clinical complications to develop. Its validation for microvascular risk reduction does not make it a direct measurement of retinal injury, kidney function, cardiovascular events, survival, or treatment safety. Consequently, lowering HbA1c cannot by itself demonstrate improvement in every clinically important outcome. (diabetesjournals.org)

Biological Influences and Assay Interference

Analytical interference and biological distortion of the HbA1c–glucose relationship are distinct problems. Analytical interference occurs when an assay misidentifies or inaccurately quantifies glycated hemoglobin; its presence and direction depend on the measurement method. Biological distortion occurs when changes in erythrocyte survival or age distribution alter the time available for glycation at a given glucose concentration. Such changes can affect results obtained by every assay, even when the measurement itself is analytically accurate. Consequently, changing methods can resolve some interferences but cannot restore the usual relationship between HbA1c and glycemia when red-cell turnover is abnormal. The 2023 laboratory guidelines led by David B. Sacks explicitly distinguish these limitations. (doi.org)

The structural variants HbS, HbC, HbE, and HbD can produce falsely high or low results with susceptible assays. Charge-based chromatographic methods may separate variant fractions imperfectly, whereas interference in immunoassays depends partly on whether an altered amino-acid sequence affects antibody recognition. HbE and HbD substitutions lie away from the beta-chain N-terminus, explaining why many immunoassays are unaffected by these variants. Elevated fetal hemoglobin, HbF, presents a separate problem: it can interfere with some immunoassay, boronate-affinity, and enzymatic methods, with tolerance depending on the particular assay and HbF concentration. Chemically modified hemoglobin, including carbamylated hemoglobin associated with renal failure, is another potential interferent. The NGSP interference tables therefore classify individual methods rather than assigning a universal effect to each variant. (ngsp.org)

Carrier states must be distinguished from hemoglobin disorders that also change cell survival. In a heterozygous carrier with normal erythrocyte turnover, an assay unaffected by the relevant variant can often measure HbA1c accurately. By contrast, HbSS and HbSC forms of sickle cell disease, and HbCC disease, may involve anemia, hemolysis, and transfusion requirements. Their HbA1c results can be biologically misleading regardless of analytical specificity; individuals without hemoglobin A also lack the substrate for conventional HbA1c measurement. Thus, detecting a variant does not by itself establish either the magnitude or direction of error: genotype, red-cell survival, and assay characteristics are separate determinants. (niddk.nih.gov)

Hemolysis and recovery from acute blood loss reduce mean erythrocyte age, generally lowering HbA1c relative to the prevailing glucose concentration because younger cells have accumulated less glycation. Erythropoietin treatment similarly increases the proportion of newly produced cells. Transfusion introduces donor erythrocytes with a different glycation history, so the resulting value no longer represents exposure of the recipient’s original cell population alone. A 2011 report by David H. Spencer, Brenda J. Grossman, and Mitchell G. Scott documented decreases in HbA1c after red-cell transfusion in patients with diabetes; this observation does not establish a universal correction for all transfusion settings. (ngsp.org)

Iron-deficiency anemia is commonly associated with higher HbA1c, and iron replacement can lower it without a corresponding improvement in glycemia. This differs from the lowering associated with hemolytic anemia, making “anemia” alone an insufficient description of the expected effect. During pregnancy, increased erythrocyte turnover generally lowers HbA1c; the American Diabetes Association’s 2026 standards identify this physiological effect in people both with and without diabetes. However, iron deficiency in late pregnancy can act in the opposite direction, so pregnancy does not imply a uniform bias under all conditions. (ngsp.org)

Chronic kidney disease combines potentially assay-dependent interference from hemoglobin derivatives with biological influences from renal anemia, shortened cell survival, iron treatment, and erythropoiesis-stimulating therapy. In advanced disease, particularly during dialysis, HbA1c may underestimate glycemic exposure, but this is not a universal effect across kidney-disease stages. A 2010 study in patients with diabetes and stage IIIb–IV kidney disease found mean HbA1c reductions of approximately 0.4 percentage points after intravenous iron and 0.7 percentage points after erythropoiesis-stimulating treatment, without changes in measured mean glucose. These findings illustrate how treatment-related shifts in red-cell populations can resemble improved glycemic control. (diabetesjournals.org)

Complementary and Alternative Measures of Glycemia

A laboratory plasma glucose measurement or capillary blood glucose reading describes glycemia at the time of sampling; repeated measurements reveal changes associated with meals, exercise, or treatment. HbA1c instead integrates exposure over approximately two to three months and cannot identify the timing of individual excursions. Continuous glucose monitoring (CGM) supplies a different record: sensors estimate glucose in interstitial fluid every few minutes, producing day-and-night profiles rather than isolated blood samples. These profiles can reveal nocturnal or asymptomatic hypoglycemia and postprandial rises that an HbA1c value does not disclose. Interstitial measurements are not identical to simultaneous blood measurements, and blood glucose checks remain relevant when sensor readings conflict with symptoms or appear inaccurate. (pubmed.ncbi.nlm.nih.gov)

CGM summaries describe both average exposure and its distribution. For nonpregnant individuals with type 1 or type 2 diabetes, time in range commonly denotes the percentage of recorded time at 70–180 mg/dL (3.9–10.0 mmol/L). Time below range reports exposure below 70 mg/dL and, separately, below 54 mg/dL (3.0 mmol/L); time above range quantifies hyperglycemia. Glucose variability can be expressed as the coefficient of variation, $100 \times \text{standard deviation}/\text{mean glucose}$, distinguishing relatively stable glucose from wide fluctuations around a similar mean. The American Diabetes Association’s 2026 standards describe approximately 14 days of monitoring with at least 70% active data as a basis for pattern assessment. These metrics therefore characterize a specified observation window, not automatically the entire interval represented by HbA1c. (diabetesjournals.org)

The glucose management indicator (GMI) translates mean CGM glucose into an HbA1c-like percentage through an empirical relationship; it does not measure glycated hemoglobin. Richard Bergenstal and colleagues introduced the term in 2018 to replace “estimated A1C,” emphasizing that the calculated result need not match laboratory HbA1c. Discordance can reflect differences between the monitoring periods, individual erythrocyte survival and glycation, or measurement limitations. Thus, GMI summarizes sensor-derived mean glucose, whereas laboratory HbA1c measures a chemical modification within circulating red cells; agreement between them is not a prerequisite for either result to contain useful information. (pmc.ncbi.nlm.nih.gov)

Fructosamine and glycated albumin provide shorter-term biochemical measures, generally reflecting the preceding two to four weeks. Fructosamine assays quantify glycated serum proteins, predominantly albumin, whereas glycated albumin assays report the glycated fraction of total albumin. Their shorter interval arises from serum-protein turnover rather than red-cell turnover. Neither measure supplies the timing of hypoglycemia or glucose excursions. Although these markers avoid dependence on hemoglobin and erythrocyte lifespan, protein metabolism introduces different limitations: fructosamine depends partly on circulating protein concentrations, and accelerated albumin turnover can alter both measures independently of glucose. Significant urinary protein loss or albumin loss associated with peritoneal dialysis can therefore complicate interpretation; expressing glycated albumin as a proportion does not eliminate altered-turnover effects. (diabetesjournals.org)

The evidence for serum-protein markers includes longitudinal associations rather than an outcome-trial foundation equivalent to that of HbA1c. In a 2014 analysis led by Elizabeth Selvin, the Atherosclerosis Risk in Communities study measured fructosamine and glycated albumin in 11,348 adults without diabetes and 958 with diagnosed diabetes, using samples collected in 1990–1992. Both markers were associated with retinopathy and subsequent chronic kidney disease, but such associations do not establish that treatment directed to marker-specific targets improves outcomes. The 2026 ADA standards note that few clinical trials support these markers compared with HbA1c. Their clinical role includes monitoring when hemoglobin-based assessment is uninterpretable or unavailable—for example, with some homozygous hemoglobin variants—while direct glucose measurements and CGM provide complementary evidence without relying on either erythrocyte or serum-protein turnover. (pubmed.ncbi.nlm.nih.gov)

References

1. NGSP: IFCC Reference Systemngsp.org
2. HBA1 hemoglobin subunit alpha 1 — NIH Genetic Testing Registryncbi.nlm.nih.gov
3. ATP-dependent Mechanism Protects Spectrin against Glycation in Human Erythrocytespmc.ncbi.nlm.nih.gov
4. The A1C Test & Diabetes - NIDDKniddk.nih.gov
5. Translating the A1C Assay Into Estimated Average Glucose Values - PMCpmc.ncbi.nlm.nih.gov
6. NGSP: HbA1c and eAGngsp.org
7. Translating the A1C Assay - PMCpmc.ncbi.nlm.nih.gov
8. Red cell life span heterogeneity in hematologically normal people is sufficient to alter HbA1c - PMCpmc.ncbi.nlm.nih.gov
9. Red cell life span heterogeneity in hematologically normal people is sufficient to alter HbA1c - PubMedpubmed.ncbi.nlm.nih.gov
10. The Biosynthesis of Human Hemoglobin A1cdm5migu4zj3pb.cloudfront.net
11. The biosynthesis of human hemoglobin A1c. Slow glycosylation of hemoglobin in vivo.jci.org
12. Correlation of glucose regulation and hemoglobin AIc in diabetes mellituspubmed.ncbi.nlm.nih.gov
13. NGSP Protocol Overviewngsp.org
14. Impact of haemoglobin variants on the diagnostic sensitivity of glycated haemoglobin (HbA1c) assay methodologies in sub-Saharan Africa: a laboratory-based method validation studypmc.ncbi.nlm.nih.gov
15. 510(k) Substantial Equivalence Determination Decision Memorandum: A1C EZ Glycohemoglobin Analysis Systemaccessdata.fda.gov
16. Glycated Hemoglobin (HbA1c)accessdata.fda.gov
17. 510(k) Substantial Equivalence Determination Decision Summaryaccessdata.fda.gov
18. Diabetes in America: Classification and Diagnosis of Diabetesniddk.nih.gov
19. NGSP: Factors that Interfere with HbA1c Test Resultsngsp.org
20. NGSP Background Informationngsp.org
21. NGSP: IFCC Standardization Overviewngsp.org
22. C1RMP_P16_MS - IFCC method for HbA1c (HPLC / MS-ESI)jctlmdb.org
23. IFCC Standardization: IFCC and NGSPngsp.org
24. IFCC reference systemifcchba1c.org
25. NGSP Protocol: Certificationngsp.org
26. Diabetes Tests & Diagnosis - NIDDKniddk.nih.gov
27. Diabetes & Prediabetes Tests - NIDDKniddk.nih.gov
28. Recommended Tests for Identifying Prediabetes - NIDDKniddk.nih.gov
29. Blood Glucose Control Studies for Type 1 Diabetes: DCCT & EDICniddk.nih.gov
30. Intensive blood-glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33)doi.org
31. UK Prospective Diabetes Study: Overviewwww2.dtu.ox.ac.uk
32. HbA1c as a Surrogate Endpointfda.gov
33. Sickle Cell Trait & Other Hemoglobinopathies & Diabetesniddk.nih.gov
34. Red cell transfusion decreases hemoglobin A1c in patients with diabetespubmed.ncbi.nlm.nih.gov
35. Continuous Glucose Monitoring - NIDDKniddk.nih.gov
36. Glucose Management Indicator (GMI): A New Term for Estimating A1C From Continuous Glucose Monitoringpmc.ncbi.nlm.nih.gov
37. Fructosamine and glycated albumin for risk stratification and prediction of incident diabetes and microvascular complications: a prospective cohort analysis of the Atherosclerosis Risk in Communities (ARIC) studypubmed.ncbi.nlm.nih.gov