AumniVIDA

Hemoglobin A1c (glycated hemoglobin) (HbA1c)

Updated October 31, 2026 · 40 sources · Written by the AumniVIDA team

Units

What is HbA1c?

HbA1c (A1c) is the percentage of hemoglobin in red blood cells that has glucose attached. It reflects average blood glucose over roughly 2 to 3 months, needs no fasting, and diagnoses diabetes at 6.5% (48 mmol/mol)48 mmol/mol (6.5%) or higher. ADA calls 5.7 to 6.4% (39 to 47 mmol/mol)39 to 47 mmol/mol (5.7 to 6.4%) prediabetes .

Cutaway illustration of red blood cells in a vessel, one opened to show hemoglobin molecules with small glucose molecules attached to them.
HbA1c forms when glucose in the blood attaches to hemoglobin inside red cells and stays there for the life of the cell.

Questions people ask about HbA1c

  1. 01HbA1c chart: normal, prediabetes and diabetes rangesA normal HbA1c is below 5.7% (39 mmol/mol)39 mmol/mol (5.7%). Prediabetes is 5.7 to 6.4% (39 to 47 mmol/mol)39 to 47 mmol/mol (5.7 to 6.4%), and 6.5% (48 mmol/mol)48 mmol/mol (6.5%) or higher is in the diabetes range, confirmed by a second test. These ADA thresholds apply at every age; NICE starts its high-risk range at 6.0% (42 mmol/mol)42 mmol/mol (6.0%) .Read the answer
  2. 02How to lower HbA1cWeight loss, diet pattern, structured exercise and medication lower HbA1c most. In type 2 diabetes trials, each kilogram lost lowered it about 0.1 points, a Mediterranean diet 0.47, structured aerobic exercise 0.73 and tirzepatide 2.10 versus placebo. Because red cells turn over slowly, changes take about 3 months to show .Read the answer
  3. 03What does a high HbA1c mean?A high HbA1c means more hemoglobin has glucose attached, usually because average glucose has been raised for months.Read the answer
  4. 04HbA1c to average glucose: conversion chartEstimated average glucose (eAG) in mg/dL equals 28.7 x A1c minus 46.7, from the ADAG study of 507 people.Read the answer
  5. 05HbA1c test: fasting, home kits and how oftenThe HbA1c test needs no fasting and is unaffected by a recent meal, stress or exercise.Read the answer

Key takeaways

  1. 1

    HbA1c summarizes about 3 months of glucose exposure, weighted to the last few weeks, because red cells live about 120 days [1] [2].

  2. 2

    ADA: below 5.7% (39 mmol/mol)39 mmol/mol (5.7%) normal, 5.7 to 6.4% prediabetes, 6.5% (48 mmol/mol)48 mmol/mol (6.5%) or higher diabetes range, confirmed by a second test [3].

  3. 3

    Each 1 point lower HbA1c in type 2 diabetes was linked to 37% fewer microvascular complications, and risk of diabetes and heart disease rises from about 5.5% in people without diabetes [4] [5].

  4. 4

    Iron deficiency can raise HbA1c; blood loss, hemolysis, pregnancy, sickle cell trait and G6PD deficiency can lower it [6] [7] [3].

  5. 5

    Average glucose (mg/dL) is about 28.7 x A1c minus 46.7, with a wide individual spread [8].

What is HbA1c?

HbA1c, or A1c, is the share of hemoglobin in your red blood cells that has glucose attached. It is reported as a percentage (NGSP/DCCT units) or in mmol/mol (IFCC units) and reflects average blood glucose over roughly the past 2 to 3 months [3].

Hemoglobin is the oxygen-carrying protein that fills red blood cells. Glucose in the blood binds to it slowly and stays bound, so the more glucose you carry around day and night, the higher the share of hemoglobin that ends up glycated [1].

Labs report the result in two units. The US and many other countries use percent, traceable to the assay used in the DCCT trial through the National Glycohemoglobin Standardization Program [9]. The UK, Europe and Australia use mmol/mol, the International Federation of Clinical Chemistry (IFCC) unit [10].

The test goes by several names: HbA1c, A1c, hemoglobin A1c, glycated hemoglobin and glycohemoglobin all mean the same measurement.

Doctors use HbA1c for three jobs: to screen for and diagnose diabetes, to flag prediabetes, and to track how well glucose is controlled in people who already have diabetes [3] [11]. It needs no fasting and changes little with a single meal, a stressful day or a short illness, which is why it is so widely used [3].

Minimal illustration of a row of small round cells moving along a gentle curved path, newer cells pale and older cells darker.
HbA1c is a moving summary: each new red cell starts clean, so the number keeps updating as your glucose changes.

How does glucose turn into HbA1c?

Glucose attaches to the end of the hemoglobin beta chain without any enzyme, a process called glycation. The bond forms slowly and lasts for the red cell's life of about 120 days, so HbA1c builds up in proportion to the glucose each cell is exposed to [1].

Glycation is a chemical reaction, not a biological signal. Glucose reacts with the N-terminal amino group of the beta chain of hemoglobin A, first forming a reversible link and then a stable one. In 1976, Bunn and colleagues used labeled iron to follow newly made red cells and found HbA1c rising slowly over about 60 days, consistent with a non-enzymatic process across the 120-day life of the cell [1].

Two consequences follow:

  • Recent weeks count most. Older red cells carry more glucose but are removed first, and new cells start clean. HbA1c is a weighted average in which the last few weeks weigh more than the months before. When glucose was normalized in newly diagnosed patients, HbA1c fell with a half-time of about 35 days, and its weighting extended back about 100 days [2].
  • Red cell lifespan matters. Anything that shortens red cell life lowers HbA1c, and anything that lengthens it raises HbA1c, even when glucose is unchanged. People with hemolysis had markedly lower HbA1c in the original work [1], and even in healthy people the mean age of red cells varied from 38 to 60 days, enough to change HbA1c in a clinically meaningful way [12].

This is why HbA1c is a strong summary of glucose exposure for most people and a misleading one for some.

Why HbA1c reflects about 3 months, weighted to recent weeks

  1. Counts most
  2. TodayNewest red cellsThey start clean and carry the least glucose, but recent weeks weigh most in the result.
  3. About 35 daysHalf-time of changeAfter glucose is normalized, HbA1c falls by half in about 35 days.
  4. Counts less
  5. About 60 daysGlycation keeps buildingIn labeled new cells, HbA1c rose slowly over about 60 days.
  6. About 100 daysWeighting fades outGlucose from this far back adds little to today's HbA1c.
  7. About 120 daysOldest cells are removedThey carry the most glucose but leave circulation first.

Red cell lifespan differs between healthy people (mean cell age 38 to 60 days in one study), which shifts HbA1c at the same glucose (Cohen 2008).

Red cells live about 120 days and keep collecting glucose, so recent weeks count more than older ones: after glucose drops, HbA1c falls with a half-time of about 35 days.

Why does HbA1c matter?

HbA1c predicts the complications of diabetes. In the DCCT and UKPDS trials, lowering it reduced eye, kidney and nerve damage, and each 1 point lower HbA1c in type 2 diabetes was linked to 37% fewer microvascular complications [13] [4].

The DCCT randomized 1,441 people with type 1 diabetes to intensive or conventional treatment. Over a mean 6.5 years, intensive control reduced new retinopathy by 76%, albuminuria by 54% and clinical neuropathy by 60%, with two to three times more severe hypoglycemia [13].

The UKPDS did the same in newly diagnosed type 2 diabetes. A median HbA1c of 7.0% versus 7.9% over 10 years lowered microvascular endpoints by 25% [14]. In the observational analysis of the same cohort, every 1 point lower updated mean HbA1c was associated with 21% fewer diabetes-related endpoints, 14% fewer heart attacks and 37% fewer microvascular complications, with no threshold below which risk stopped falling [4].

Risk starts below the diabetes range. In the ARIC study of 11,092 adults without diabetes, HbA1c of 5.5 to 5.9% carried 1.86 times the risk of later diagnosed diabetes compared with 5.0 to 5.4%, and 6.0 to 6.4% carried 4.48 times the risk. Coronary heart disease risk rose too, and HbA1c predicted cardiovascular outcomes better than fasting glucose [5].

HbA1c is one part of a metabolic picture. Fasting insulin can show insulin resistance years before HbA1c moves, and triglycerides and ApoB carry the cardiovascular side of the same story.

The studies that gave HbA1c its meaning
StudyWhoWhat it showed
DCCT, 19931,441 people with type 1 diabetesIntensive control cut new retinopathy by 76% and neuropathy by 60%
UKPDS 33, 19983,867 people with new type 2 diabetesMedian 7.0% vs 7.9% cut microvascular endpoints by 25%
UKPDS 35, 20004,585 people with type 2 diabetesEach 1 point lower: 21% fewer diabetes endpoints, 37% fewer microvascular
ARIC, 201011,092 adults without diabetesDiabetes and heart disease risk rose from 5.5%, more than with fasting glucose
ADAG, 2008507 people with and without diabetesAverage glucose = 28.7 x A1c minus 46.7 mg/dL

Future diabetes risk by baseline HbA1c in adults without diabetes

Show the data
Future diabetes risk by baseline HbA1c in adults without diabetes
ItemValue (hazard ratio for diagnosed diabetes)LowHighNote
Below 5.0%0.520.40.69
5.0 to 5.4% (reference)1
5.5 to 5.9%1.861.672.08
6.0 to 6.4%4.483.925.13
6.5% or higher16.4714.2219.08
In the ARIC study, risk of diagnosed diabetes rose steadily with HbA1c: 1.9 times at 5.5 to 5.9% and 4.5 times at 6.0 to 6.4%, compared with 5.0 to 5.4%.

How do you read HbA1c results?

Under 5.7% (39 mmol/mol)39 mmol/mol (5.7%) is normal by ADA criteria, 5.7 to 6.4% (39 to 47 mmol/mol)39 to 47 mmol/mol (5.7 to 6.4%) is prediabetes, and 6.5% (48 mmol/mol)48 mmol/mol (6.5%) or higher is in the diabetes range. A diagnosis needs a confirming test [3].

Guidelines agree on the diabetes threshold and differ on the risk range below it. ADA uses 5.7 to 6.4% for prediabetes [3]. WHO endorses 6.5% for diagnosis but found too little evidence to define any category below it [15]. NICE and the International Expert Committee start the high-risk range at 6.0% (42 mmol/mol)42 mmol/mol (6.0%) [16] [17].

Converting units. The IFCC and NGSP scales are linked by NGSP % = 0.915 x IFCC % + 2.15 [10]. Rearranged, that gives:

  • mmol/mol = 10.929 x (A1c % minus 2.15)
  • A1c % = (mmol/mol / 10.929) + 2.15

So 5.7% is 39 mmol/mol, 6.5% is 48 mmol/mol and 7.0% is 53 mmol/mol. The full chart from 4% to 14%, with estimated average glucose, is on the HbA1c chart page, and the glucose conversion is explained on the average glucose page.

One result is not a diagnosis. Without obvious symptoms, ADA requires two abnormal results, either two different tests from the same sample or the same test repeated, and a clinician should check for anything that could distort either test [3]. For people who already have diabetes, ADA suggests a goal below 7% (53 mmol/mol)53 mmol/mol (7%) for many nonpregnant adults, set individually [11].

HbA1c thresholds by guideline, in % (NGSP/DCCT) and mmol/mol (IFCC)
CategoryADA Standards of Care 2026WHO 2011NICE PH38International Expert Committee 2009
NormalBelow 5.7% (below 39 mmol/mol)No formal threshold below 6.5%Below 6.0% (below 42 mmol/mol)No lower threshold set
Raised riskPrediabetes: 5.7 to 6.4% (39 to 47 mmol/mol)39 to 47 mmol/mol (5.7 to 6.4%)Insufficient evidence for a formal categoryHigh risk: 6.0 to 6.4% (42 to 47 mmol/mol)42 to 47 mmol/mol (6.0 to 6.4%)6.0 to 6.4% (42 to 47 mmol/mol)42 to 47 mmol/mol (6.0 to 6.4%): offer prevention
Diabetes6.5% (48 mmol/mol)48 mmol/mol (6.5%) or higher, confirmed6.5% (48 mmol/mol)48 mmol/mol (6.5%) or higher6.5% (48 mmol/mol)48 mmol/mol (6.5%) or higher, second test6.5% (48 mmol/mol)48 mmol/mol (6.5%) or higher, repeat test
Goal with diabetesBelow 7% (53 mmol/mol)53 mmol/mol (7%) for many adultsNot coveredNot coveredNot covered

HbA1c ranges in percent: normal, prediabetes and diabetes

NormalPrediabetesDiabetes rangeNormal: 4 to 5.7 %Normal: 4 to 5.7 %Normal: 4 to 5.7 %Normal: 4 to 5.7 %Normal: 4 to 5.7 %Normal: 4 to 5.7 %Normal: 4 to 5.7 %Normal: 4 to 5.7 %Normal: 4 to 5.7 %Normal: 4 to 5.7 %Normal: 4 to 5.7 %Normal: 4 to 5.7 %Normal: 4 to 5.7 %Normal: 4 to 5.7 %Normal: 4 to 5.7 %Normal: 4 to 5.7 %Prediabetes: 5.7 to 6.5 %Prediabetes: 5.7 to 6.5 %Prediabetes: 5.7 to 6.5 %Prediabetes: 5.7 to 6.5 %Prediabetes: 5.7 to 6.5 %Prediabetes: 5.7 to 6.5 %Prediabetes: 5.7 to 6.5 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Normal: 4 to 5.7 %Normal: 4 to 5.7 %Normal: 4 to 5.7 %Normal: 4 to 5.7 %Normal: 4 to 5.7 %Normal: 4 to 5.7 %Normal: 4 to 5.7 %Normal: 4 to 5.7 %Normal: 4 to 5.7 %Normal: 4 to 5.7 %Normal: 4 to 5.7 %Normal: 4 to 5.7 %Normal: 4 to 5.7 %Normal: 4 to 5.7 %Normal: 4 to 5.7 %Normal: 4 to 5.7 %Prediabetes: 5.7 to 6.5 %Prediabetes: 5.7 to 6.5 %Prediabetes: 5.7 to 6.5 %Prediabetes: 5.7 to 6.5 %Prediabetes: 5.7 to 6.5 %Prediabetes: 5.7 to 6.5 %Prediabetes: 5.7 to 6.5 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %Diabetes range: 6.5 to 9 %45.7ADA prediabetes: 5.7 %6NICE and IEC high risk: 6 %6.5ADA and WHO diagnosis: 6.5 %7ADA goal for many with diabetes: 7 %9 %
Normal
4 to 5.7
Prediabetes
5.7 to 6.5
Diabetes range
6.5 to 9
  • 5.7 %: ADA prediabetes
  • 6 %: NICE and IEC high risk
  • 6.5 %: ADA and WHO diagnosis
  • 7 %: ADA goal for many with diabetes
ADA calls 5.7 to 6.4% prediabetes; ADA and WHO both diagnose diabetes from 6.5%, and NICE and the IEC flag high risk from 6.0%.

When can HbA1c be misleading?

HbA1c misleads when red cells live shorter or longer than usual or when the assay is disturbed. Iron deficiency can raise it; blood loss, hemolysis, pregnancy, some hemoglobin variants and G6PD deficiency can lower it. In those cases ADA advises plasma glucose for diagnosis [3].

Because HbA1c depends on how long red cells circulate, anything that changes red cell turnover changes the number. ADA lists anemia, iron status, splenectomy, blood loss, transfusion, hemolysis, G6PD deficiency, erythropoietin, HIV, cirrhosis, kidney failure, dialysis and pregnancy [3].

  • Iron deficiency raises HbA1c without any change in glucose, while other anemias may lower it [6]. If you have low ferritin, read a borderline HbA1c with care.
  • Hemoglobin variants such as sickle cell trait can interfere with some assays. At the same fasting glucose, African Americans with sickle cell trait had HbA1c about 0.29 points lower [7]. ADA recommends an assay free of variant interference [3].
  • G6PD deficiency. The G202A variant, carried by 11% of Black people in the US, lowers HbA1c by about 0.8 points in homozygous men and 0.7 in homozygous women [3].
  • Pregnancy speeds red cell turnover, so HbA1c runs slightly lower [18].
  • Kidney disease. HbA1c becomes less accurate as kidney function falls, especially with dialysis [19].
  • Ancestry and age. In adults without diabetes, HbA1c ran 0.26 points higher in Black, 0.24 higher in Asian and 0.08 higher in Latino people than in White people [20], and it rises about 0.01 points per year of age [21]. Whether these gaps reflect glucose or red cell biology is still debated, and guidelines use the same thresholds for everyone [3].

What to use instead. When HbA1c is unreliable, ADA says plasma glucose should be used for diagnosis, and fructosamine, glycated albumin or continuous glucose monitoring (CGM) for monitoring [3] [11]. Fructosamine reflects roughly the past 30 days, compared with about 100 days for HbA1c [2]. CGM reports average glucose and time in range directly [22].

What shifts HbA1c away from true average glucose, compiled from guidelines and studies
ConditionDirectionTypical sizeSource
Iron deficiency, with or without anemiaHigherRaised HbA1c with no rise in glucoseEnglish 2015 systematic review
Other anemias, hemolysis, blood loss, transfusionLowerVaries with red cell survivalBunn 1976; ADA 2026
Normal variation in red cell lifespanEitherMean cell age 38 to 60 days in healthy peopleCohen 2008
Sickle cell trait (some assays)LowerAbout 0.3 points at the same glucoseLacy 2017
G6PD deficiency (G202A variant)LowerAbout 0.7 to 0.8 points in homozygous peopleADA 2026
Black, Asian or Latino ancestry vs WhiteHigher0.26, 0.24 and 0.08 pointsCavagnolli 2017
AgeHigherAbout 0.01 points per yearPani 2008
PregnancyLowerSlightly lower from faster red cell turnoverADA 2026 pregnancy
Advanced kidney disease, dialysis, erythropoietinEitherAccuracy falls as kidney function declinesADA 2026; Galindo 2023
HIV, cirrhosis, splenectomyEitherListed as altering the HbA1c to glucose linkADA 2026

When HbA1c and glucose disagree

HbA1c and glucose disagree

  • HbA1c reads higher than glucose suggestsIron deficiency, Black, Asian or Latino ancestry (0.08 to 0.26 points), older age
  • HbA1c reads lower than glucose suggestsBlood loss, hemolysis, transfusion, sickle cell trait (about 0.3 points), G6PD deficiency (0.7 to 0.8 points), pregnancy
  • Either directionKidney disease, dialysis, erythropoietin, HIV, cirrhosis, splenectomy
  • A hemoglobin variant is possibleAsk whether the lab assay is free of variant interference
  • Diagnosis is the questionUse plasma glucose criteria
  • Monitoring is the questionUse fructosamine, glycated albumin or CGM

Sizes are group averages from the cited studies (ADA 2026, Lacy 2017, Cavagnolli 2017). A clinician decides which test fits your situation.

A result that does not fit your glucose readings is a reason to look for interference or a red cell problem, not to ignore either test.

What changes HbA1c?

Weight loss, diet pattern, structured exercise and medication all lower HbA1c, mostly by lowering average glucose. In trials of people with type 2 diabetes, each kilogram lost lowered it by about 0.1 points and structured aerobic exercise by 0.73 points [23] [24].

The levers that move HbA1c are the ones that lower glucose across the whole day:

  • Weight loss has the most consistent dose response: about 0.1 points per kilogram in trials of type 2 diabetes [23].
  • Diet pattern. A Mediterranean diet lowered HbA1c by 0.47 points versus lower-fat diets [25]; low-carbohydrate diets lowered it by 0.47 at 3 months but not by 12 months [26]; more fibre lowered it by 2.0 mmol/mol [27].
  • Exercise. Structured training of more than 150 minutes a week lowered it by 0.89 points [24].
  • Medication. Most oral diabetes drugs lower HbA1c by 0.5 to 1.25 points [28], and GLP-1 receptor agonists by more, up to 2.10 points for tirzepatide versus placebo [29]. Medication decisions belong with a clinician.

Because of the slow turnover of red cells, a real change takes about 3 months to show fully [2]. The how to lower HbA1c page ranks every lever by effect size. AumniVIDA's glucose stability, Mediterranean, keto and low-carb and time-restricted eating protocols track the habits behind these numbers.

How much each lever lowers HbA1c

Show the data
How much each lever lowers HbA1c
ItemValue (percentage points lower)LowHighNote
Tirzepatide vs placebo2.11.742.47Network meta-analysis of 76 trials (Yao 2024)
Lose 10 kg1Modeled at 0.1 points per kg across 58 trials (Gummesson 2017)
Structured exercise over 150 min a week0.89Umpierre 2011
Structured aerobic exercise0.730.41.06Umpierre 2011
Mediterranean diet vs lower-fat diet0.470.380.56Esposito 2015
Low-carbohydrate diet, at 3 months0.470.230.71No difference by 12 months (Sainsbury 2018)
Normal vs short sleep0.230.10.36Observational association (Lee 2017)
More dietary fibre0.180.060.32.0 mmol/mol converted to % points (Reynolds 2020)
Average HbA1c reductions from meta-analyses, mostly in people with type 2 diabetes; effects in people without diabetes are smaller because there is less to lower.

Where did the HbA1c test come from?

Samuel Rahbar reported a raised hemoglobin fraction in diabetes in 1968. By 1976 it was shown to track glucose control, the DCCT and UKPDS tied it to complications, and standardization let ADA and WHO adopt 6.5% for diagnosis in 2010 and 2011 [30] [31] [15].

The path from discovery to diagnosis took four decades:

  • 1968: Rahbar found an abnormal fast-moving hemoglobin in the red cells of people with diabetes [30].
  • 1976: Bunn showed that HbA1c forms slowly over the red cell's life [1], and Koenig showed that it fell from 9.8% to 5.8% as patients' glucose control improved [31].
  • 1993 and 1998: The DCCT and UKPDS made HbA1c the standard measure of treatment success [13] [14].
  • 1996 to 2004: The NGSP aligned US labs to the DCCT assay [9], and the IFCC built a reference method linked to it [10].
  • 2008 to 2011: The ADAG study translated HbA1c into average glucose [8], the International Expert Committee proposed 6.5% for diagnosis [17], ADA adopted it with a 5.7% risk threshold [32], and WHO followed [15].

From a lab curiosity to a diagnostic test

  1. 1968

    Rahbar finds a raised hemoglobin fraction in diabetes

  2. 1976

    HbA1c shown to form slowly over the red cell's life and to track glucose control

  3. 1993

    DCCT links lower HbA1c to fewer complications in type 1 diabetes

  4. 1996

    NGSP makes US results traceable to the DCCT assay

  5. 1998

    UKPDS shows the same in type 2 diabetes

  6. 2004

    IFCC reference method links % to mmol/mol

  7. 2008

    ADAG study converts HbA1c to average glucose

  8. 2009

    International Expert Committee backs 6.5% for diagnosis

  9. 2010

    ADA adopts HbA1c for diagnosis and the 5.7% risk threshold

  10. 2011

    WHO endorses 6.5% as the diagnostic cut point

  11. 2018

    CGM estimate renamed glucose management indicator

HbA1c took four decades to move from a 1968 lab finding to a standardized test that diagnoses diabetes worldwide.

What does the evidence not show about HbA1c?

The evidence does not define a healthy HbA1c below the prediabetes range, does not prove that lowering a normal HbA1c prevents disease, and does not set age-specific or ancestry-specific thresholds. WHO found too little evidence to interpret values below 6.5% [15].

  • No proven optimal number in people without diabetes. Risk in ARIC rose from 5.5% [5], but no trial has shown that lowering an HbA1c of, say, 5.4% to 5.0% prevents disease. Claims of an optimal longevity range go beyond the data.
  • Very low is not clearly better. In ARIC, HbA1c below 5.0% was linked to 32% higher all-cause mortality than 5.0 to 5.7%, likely because low values flag other conditions such as anemia [33].
  • Lifestyle effect sizes come mostly from type 2 diabetes. People starting near normal have less to lower, and higher baseline HbA1c predicts a bigger drop [23] [28].
  • Short-term diet effects can fade. Low-carbohydrate diets lowered HbA1c at 3 and 6 months, but the difference was gone at 12 and 24 months [26].
  • Age and ancestry. HbA1c rises with age and differs by ancestry in people without diabetes [21] [20], but there is no agreement on separate thresholds [3].
  • CGM estimates do not replace the lab test. In people without diabetes, the CGM-based GMI ran 0.59 points above lab HbA1c on average [34].

The research

Foundational studies

The work that established what this marker means.

Guidelines and reviews

Where expert bodies and systematic reviews have landed.

Recent research

Notable work from the last three years.

What well-known voices say about HbA1c

Their positions, summarized in our words. Follow each link for the full context.

  • Peter AttiaPhysician, host of The Drive podcast and author of Outlive

    Attia treats HbA1c as an indirect predictor of average glucose that can mislead in either direction because red cell lifespan varies; he says his own result runs higher than his CGM average because of a thalassemia trait. He prefers CGM and an oral glucose tolerance test for judging metabolic health and argues that a lower value within the normal range is still better.

    The Drive, AMA #24: Deep dive into blood glucose (June 2021)

  • Elizabeth SelvinEpidemiologist at the Johns Hopkins Bloomberg School of Public Health and lead author of the ARIC HbA1c studies

    Selvin calls A1C one of the most important clinical biomarkers because it varies less than glucose and reflects overall exposure. She rejects the idea that CGM should replace it, since both tools are imperfect and measure different things, and she found Black and White differences in A1C too small to change how results are read.

    ADA Meeting News: Kelly West Award lecture (June 2020)

  • Roy TaylorProfessor of medicine at Newcastle University and co-lead of the DiRECT trial

    Taylor argues that type 2 diabetes is not necessarily lifelong: substantial weight loss removes excess fat from the liver and pancreas, lets them return to normal function and can bring HbA1c below the diabetes threshold without medication.

    Newcastle University press release: Type 2 diabetes is not for life (December 2017)

Watch

Talks and explainers worth your time. Nothing loads from YouTube until you press play.

  • A1C Test for Diabetes, Animation

    Alila Medical MediaOpen on YouTube

    An animated walk through how glucose binds hemoglobin and why the test reflects months rather than a moment.

  • Why Is Hemoglobin A1c Important?

    Cleveland ClinicOpen on YouTube

    A short clinician explanation of what the number means for diabetes care, from a major academic medical center.

  • A1C Levels

    Khan AcademyOpen on YouTube

    A step-by-step whiteboard lesson linking A1c percentages to blood glucose.

  • Your Diabetes Health Checks: HbA1c | Learning Zone | Diabetes UK

    Diabetes UKOpen on YouTube

    The UK view of the test, useful for readers whose results come in mmol/mol.

  • Hemoglobin A1C

    Association for Diagnostics & Laboratory MedicineOpen on YouTube

    The laboratory side of the test, from the professional society for clinical lab scientists.

Protocols that act on HbA1c

Common questions

What does A1c stand for?

A1c is short for hemoglobin A1c, the fraction of adult hemoglobin (hemoglobin A) that carries glucose. It was named for the order in which it separated in early lab methods, and it is also called HbA1c, glycated hemoglobin or glycohemoglobin.

Is A1c the same as blood sugar?

No. A blood sugar reading measures glucose at one moment, while A1c measures how much hemoglobin has been glycated over about 2 to 3 months [3]. They are related through the ADAG formula, eAG (mg/dL) = 28.7 x A1c minus 46.7 [8].

Does A1c increase with age?

Yes, slightly. In adults without diabetes, A1c rose about 0.01 to 0.014 points per year of age [21]. Diagnostic thresholds are still the same at every age [3].

Can an A1c test be wrong?

It can be misleading when red cell lifespan is unusual or an assay is disturbed, for example with iron deficiency, blood loss, pregnancy, kidney failure or some hemoglobin variants [3] [6]. A result that does not fit your glucose readings is worth discussing with a clinician.

Can A1c be too low?

In people without diabetes, an A1c below 5.0% was linked to higher mortality in the ARIC study, probably because it often reflects conditions such as anemia [33]. In people taking glucose-lowering drugs, pushing A1c very low brings more hypoglycemia, which is why ADA sets goals individually [13] [11].

Can A1c change in a month?

Partly. Because recent weeks weigh most, A1c can move noticeably within a month, and after glucose drops it falls by half of the gap in about 35 days [2]. The full effect of a change takes about 3 months.

What is the difference between A1c and GMI?

GMI, the glucose management indicator, is an estimate of A1c calculated from CGM average glucose: GMI (%) = 3.31 + 0.02392 x mean glucose in mg/dL [35]. It often differs from a lab A1c; in people without diabetes it ran 0.59 points higher on average [34].

Why is A1c not used to diagnose diabetes in pregnancy?

Red cells turn over faster in pregnancy, so A1c runs slightly lower and can miss raised glucose [18]. ADA advises plasma glucose criteria in pregnancy and other conditions that alter red cell turnover [3].

Does A1c test for insulin resistance?

No. A1c reflects average glucose, and insulin resistance can exist for years while glucose stays normal because the pancreas makes more insulin. Fasting insulin and triglycerides are closer markers of insulin resistance.

Can stress raise your A1c?

A single stressful day does not move A1c much, because it is unaffected by short-term stress, illness or activity [3]. Sustained stress that keeps glucose higher for weeks can raise it, as can poor or short sleep [36].

Sources

  1. 1.Bunn HF, Haney DN, Kamin S, Gabbay KH, Gallop PM. The biosynthesis of human hemoglobin A1c. Slow glycosylation of hemoglobin in vivo. Journal of Clinical Investigation, 1976. doi:10.1172/JCI108436
  2. 2.Tahara Y, Shima K. Kinetics of HbA1c, glycated albumin, and fructosamine and analysis of their weight functions against preceding plasma glucose level. Diabetes Care, 1995. doi:10.2337/diacare.18.4.440
  3. 3.American Diabetes Association Professional Practice Committee for Diabetes. 2. Diagnosis and Classification of Diabetes: Standards of Care in Diabetes-2026. Diabetes Care, 2026. doi:10.2337/dc26-S002
  4. 4.Stratton IM, Adler AI, Neil HA, et al. Association of glycaemia with macrovascular and microvascular complications of type 2 diabetes (UKPDS 35): prospective observational study. BMJ, 2000. doi:10.1136/bmj.321.7258.405
  5. 5.Selvin E, Steffes MW, Zhu H, et al. Glycated hemoglobin, diabetes, and cardiovascular risk in nondiabetic adults. New England Journal of Medicine, 2010. doi:10.1056/NEJMoa0908359
  6. 6.English E, Idris I, Smith G, Dhatariya K, Kilpatrick ES, John WG. The effect of anaemia and abnormalities of erythrocyte indices on HbA1c analysis: a systematic review. Diabetologia, 2015. doi:10.1007/s00125-015-3599-3
  7. 7.Lacy ME, Wellenius GA, Sumner AE, et al. Association of Sickle Cell Trait With Hemoglobin A1c in African Americans. JAMA, 2017. doi:10.1001/jama.2016.21035
  8. 8.Nathan DM, Kuenen J, Borg R, Zheng H, Schoenfeld D, Heine RJ; A1c-Derived Average Glucose Study Group. Translating the A1C assay into estimated average glucose values. Diabetes Care, 2008. doi:10.2337/dc08-0545
  9. 9.Little RR, Rohlfing CL, Wiedmeyer HM, Myers GL, Sacks DB, Goldstein DE. The national glycohemoglobin standardization program: a five-year progress report. Clinical Chemistry, 2001. PubMed 11673367
  10. 10.Hoelzel W, Weykamp C, Jeppsson JO, et al. IFCC reference system for measurement of hemoglobin A1c in human blood and the national standardization schemes in the United States, Japan, and Sweden: a method-comparison study. Clinical Chemistry, 2004. doi:10.1373/clinchem.2003.024802
  11. 11.American Diabetes Association Professional Practice Committee for Diabetes. 6. Glycemic Goals, Hypoglycemia, and Hyperglycemic Crises: Standards of Care in Diabetes-2026. Diabetes Care, 2026. doi:10.2337/dc26-S006
  12. 12.Cohen RM, Franco RS, Khera PK, et al. Red cell life span heterogeneity in hematologically normal people is sufficient to alter HbA1c. Blood, 2008. doi:10.1182/blood-2008-04-154112
  13. 13.Diabetes Control and Complications Trial Research Group; Nathan DM, Genuth S, et al. The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. New England Journal of Medicine, 1993. doi:10.1056/NEJM199309303291401
  14. 14.UK Prospective Diabetes Study (UKPDS) Group. Intensive blood-glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). Lancet, 1998. PubMed 9742976
  15. 15.World Health Organization. Use of glycated haemoglobin (HbA1c) in diagnosis of diabetes mellitus: abbreviated report of a WHO consultation. World Health Organization, 2011. Link
  16. 16.National Institute for Health and Care Excellence. Type 2 diabetes: prevention in people at high risk (PH38). NICE public health guideline, 2012. Link
  17. 17.International Expert Committee. International Expert Committee report on the role of the A1C assay in the diagnosis of diabetes. Diabetes Care, 2009. doi:10.2337/dc09-9033
  18. 18.American Diabetes Association Professional Practice Committee for Diabetes. 15. Management of Diabetes in Pregnancy: Standards of Care in Diabetes-2026. Diabetes Care, 2026. doi:10.2337/dc26-S015
  19. 19.Galindo RJ, de Boer IH, Neumiller JJ, Tuttle KR. Continuous Glucose Monitoring to Optimize Management of Diabetes in Patients with Advanced CKD. Clinical Journal of the American Society of Nephrology, 2023. doi:10.2215/CJN.04510422
  20. 20.Cavagnolli G, Pimentel AL, Freitas PA, Gross JL, Camargo JL. Effect of ethnicity on HbA1c levels in individuals without diabetes: Systematic review and meta-analysis. PLoS One, 2017. doi:10.1371/journal.pone.0171315
  21. 21.Pani LN, Korenda L, Meigs JB, et al. Effect of aging on A1C levels in individuals without diabetes: evidence from the Framingham Offspring Study and the National Health and Nutrition Examination Survey 2001-2004. Diabetes Care, 2008. doi:10.2337/dc08-0577
  22. 22.Battelino T, Danne T, Bergenstal RM, et al. Clinical Targets for Continuous Glucose Monitoring Data Interpretation: Recommendations From the International Consensus on Time in Range. Diabetes Care, 2019. doi:10.2337/dci19-0028
  23. 23.Gummesson A, Nyman E, Knutsson M, Karpefors M. Effect of weight reduction on glycated haemoglobin in weight loss trials in patients with type 2 diabetes. Diabetes, Obesity and Metabolism, 2017. doi:10.1111/dom.12971
  24. 24.Umpierre D, Ribeiro PA, Kramer CK, et al. Physical activity advice only or structured exercise training and association with HbA1c levels in type 2 diabetes: a systematic review and meta-analysis. JAMA, 2011. doi:10.1001/jama.2011.576
  25. 25.Esposito K, Maiorino MI, Bellastella G, Chiodini P, Panagiotakos D, Giugliano D. A journey into a Mediterranean diet and type 2 diabetes: a systematic review with meta-analyses. BMJ Open, 2015. doi:10.1136/bmjopen-2015-008222
  26. 26.Sainsbury E, Kizirian NV, Partridge SR, Gill T, Colagiuri S, Gibson AA. Effect of dietary carbohydrate restriction on glycemic control in adults with diabetes: A systematic review and meta-analysis. Diabetes Research and Clinical Practice, 2018. doi:10.1016/j.diabres.2018.02.026
  27. 27.Reynolds AN, Akerman AP, Mann J. Dietary fibre and whole grains in diabetes management: Systematic review and meta-analyses. PLoS Medicine, 2020. doi:10.1371/journal.pmed.1003053
  28. 28.Sherifali D, Nerenberg K, Pullenayegum E, Cheng JE, Gerstein HC. The effect of oral antidiabetic agents on A1C levels: a systematic review and meta-analysis. Diabetes Care, 2010. doi:10.2337/dc09-1727
  29. 29.Yao H, Zhang A, Li D, et al. Comparative effectiveness of GLP-1 receptor agonists on glycaemic control, body weight, and lipid profile for type 2 diabetes: systematic review and network meta-analysis. BMJ, 2024. doi:10.1136/bmj-2023-076410
  30. 30.Rahbar S. An abnormal hemoglobin in red cells of diabetics. Clinica Chimica Acta, 1968. doi:10.1016/0009-8981(68)90372-0
  31. 31.Koenig RJ, Peterson CM, Jones RL, Saudek C, Lehrman M, Cerami A. Correlation of glucose regulation and hemoglobin AIc in diabetes mellitus. New England Journal of Medicine, 1976. doi:10.1056/NEJM197608192950804
  32. 32.American Diabetes Association. Diagnosis and classification of diabetes mellitus. Diabetes Care, 2010. doi:10.2337/dc10-S062
  33. 33.Aggarwal V, Schneider AL, Selvin E. Low hemoglobin A(1c) in nondiabetic adults: an elevated risk state?. Diabetes Care, 2012. doi:10.2337/dc11-2531
  34. 34.Shah VN, Vigers T, Pyle L, Calhoun P, Bergenstal RM. Discordance Between Glucose Management Indicator and Glycated Hemoglobin in People Without Diabetes. Diabetes Technology and Therapeutics, 2023. doi:10.1089/dia.2022.0544
  35. 35.Bergenstal RM, Beck RW, Close KL, et al. Glucose Management Indicator (GMI): A New Term for Estimating A1C From Continuous Glucose Monitoring. Diabetes Care, 2018. doi:10.2337/dc18-1581
  36. 36.Lee SWH, Ng KY, Chin WK. The impact of sleep amount and sleep quality on glycemic control in type 2 diabetes: A systematic review and meta-analysis. Sleep Medicine Reviews, 2017. doi:10.1016/j.smrv.2016.02.001
  37. 37.Knowler WC, Barrett-Connor E, Fowler SE, et al. Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin. New England Journal of Medicine, 2002. doi:10.1056/NEJMoa012512
  38. 38.Lean ME, Leslie WS, Barnes AC, et al. Primary care-led weight management for remission of type 2 diabetes (DiRECT): an open-label, cluster-randomised trial. Lancet, 2018. doi:10.1016/S0140-6736(17)33102-1
  39. 39.Semnani-Azad Z, Khan TA, Chiavaroli L, et al. Intermittent fasting strategies and their effects on body weight and other cardiometabolic risk factors: systematic review and network meta-analysis of randomised clinical trials. BMJ, 2025. doi:10.1136/bmj-2024-082007
  40. 40.Sandforth A, von Schwartzenberg RJ, Arreola EV, et al. Mechanisms of weight loss-induced remission in people with prediabetes: a post-hoc analysis of the randomised, controlled, multicentre Prediabetes Lifestyle Intervention Study (PLIS). Lancet Diabetes and Endocrinology, 2023. doi:10.1016/S2213-8587(23)00235-8

AumniVIDA explains health data for general wellness. Talk to your clinician about your own results, medication and diagnosis.

Track your HbA1c over time.

Bring every lab report into one history, compare results with your own baseline, and ask the coach what changed. Join the iPhone beta.

Join the iPhone beta. We send TestFlight invitations in batches.