Estimated Average Glucose (eAG): What Your Number Means, and Why It Is Really Your A1c

Sample type: Blood

Other names: Estimated Average Glucose, Estimated Average Glucose (eAG), eAG, eAG (mg/dL), eAG (mmol/L), Estim. Avg Glu (eAG), Estimated Avg Glucose, Estimated Mean Glucose

Optimal result: 68 – 114 mg/dL

Written October 15, 2026

Updated October 15, 2026

By Healthmatters Editorial Team

Peer-reviewed sources

A result above the stored range means your calculated eAG came in over 114 mg/dL (6.33 mmol/L) — and the most useful thing to understand first is what that boundary is.

Put it back through the equation and 114 mg/dL is an A1c of 5.6%. The stored range corresponds to A1c 4.0 – 5.6%, the usual non-diabetic span, so a high flag here begins at roughly the point where ADA's prediabetes band starts (A1c 5.7%, an eAG of about 117). At an A1c of 6.5% — ADA's diabetes diagnostic threshold — the eAG is about 140.

eAG	A1c	ADA category
68 – 114 mg/dL	4.0 – 5.6%	usual non-diabetic range
117 – 137 mg/dL	5.7 – 6.4%	prediabetes
140 mg/dL and above	6.5% and above	at or above the diabetes threshold

A flagged line is not a diagnosis. Diagnosis rests on clinical assessment and confirmatory testing, and this value cannot be the confirmation of its own A1c — they are one result in two units. The figure belongs with the clinician who ordered it.

If you have symptoms of high blood sugar — heavy thirst, passing a lot of urine, blurred vision, unexplained weight loss — do not wait for a repeat test. Be seen promptly.

And the estimate carries real spread. The ADAG study reports a prediction error with a study-wide standard deviation of 15.7 mg/dL, with about 90% of samples within 15% of the calculated value. That spread is about how far your measured average glucose may sit from this estimate — it is not a statement about the precision of the A1c, and it does not make two different A1c results equivalent. A trend across several results is worth more than any single one.

Check the A1c it was calculated from, and check whether that A1c is trustworthy. This number carries no information the A1c did not, and it inherits every one of the A1c's failure modes. ADA notes that factors affecting hemoglobin or red blood cells may affect A1c, listing anemias, G6PD deficiency, recent transfusion, drugs that stimulate red cell production, end-stage kidney disease and pregnancy. Conditions that shorten red cell survival tend to push A1c down, so they are a more common explanation for a falsely reassuring result than a falsely high one — but iron-deficiency anemia can push it up. Where A1c is unreliable, ADA says alternative approaches to monitoring should be used.

What a high value does not tell you. Whether your glucose is high right now — the window is two to three months. Whether your readings are steady or swinging — an average cannot show its own spread. Whether you have had low readings; ADA states that A1C "does not provide a measure of glycemic variability or hypoglycemia". And what caused it: this number is a summary, not an explanation.

If you wear a continuous glucose monitor, expect some difference between its average and this one. ADA says it is unclear how generalizable the A1c-to-eAG conversion is to modern sensors, and describes the glucose management indicator that sensor apps report as a calculated value approximating A1c, "not always equivalent". A large or persistent difference deserves review with your clinician: it may reflect the periods being compared, glucose that has been changing, or limits affecting either measurement.

The useful step is to take this result, the A1c beside it, and any meter or sensor data to whoever ordered the test. A single high eAG is a reason for a conversation and a repeat, not for a conclusion.

Being flagged low here means your calculated eAG came in under the 68 mg/dL (3.77 mmol/L) stored on this record — which corresponds to an A1c of roughly 4.0%, the bottom of the usual non-diabetic range. So a low flag is not the same thing as good control. It means the calculated average sits below the usual non-diabetic range, and what that signifies depends almost entirely on one question.

Do you take any medication that lowers blood glucose?

If you do — insulin, a sulfonylurea such as glipizide or glyburide, or any similar agent — a low average deserves a conversation with whoever prescribes it, and soon. An average is a midpoint, and the same midpoint can come from steady readings or from highs and lows that cancel out. ADA puts it plainly: "A1C does not provide a measure of glycemic variability or hypoglycemia." Your eAG cannot tell you whether you have been having low-glucose episodes, and a low average is one of the patterns that can sit on top of them. ADA treats this as serious: "Severe or frequent hypoglycemia is an absolute indication for the modification of treatment plans." None of that means your result is dangerous. It means this number cannot rule the question out, so it should not be read as reassurance on its own.

If you do not take glucose-lowering medication, an unexpectedly low eAG still needs interpreting alongside the A1c it came from, any symptoms, and anything affecting your red cells. Absence of medication does not by itself establish that a low result is benign.

That second question is the one worth pressing, because eAG is not measured. It is your A1c put through a fixed equation, so anything that makes an A1c too low makes your eAG exactly as low, with nothing left over to signal the problem. ADA notes that "Factors that affect hemoglobin or red blood cells may affect A1C" — anemias, G6PD deficiency, recent blood transfusion, medications that stimulate red cell production, end-stage kidney disease and pregnancy can all "interfere with the accuracy of A1C." Blood loss and anything that shortens red cell survival pull it the same way. Where A1c is unreliable, ADA says "alternative approaches to monitoring glycemic status should be used" — a direct glucose measurement, or continuous monitoring.

And the estimate has its own width. In the ADAG study the prediction error had a study-wide standard deviation of 15.7 mg/dL (0.87 mmol/L), and about 90% of samples fell within 15% of the calculated figure. That describes how far your measured average glucose may sit from this estimate — it is not a statement about the precision of the A1c itself. Read the figure alongside the A1c it came from rather than to more decimal places than it carries.

What to do with it. If you take glucose-lowering medication, raise it with your prescriber rather than acting on it. If you do not, bring it to your next appointment along with the A1c it came from — the two are one result, and the A1c is the one that was actually measured. There is nothing to treat in a low eAG by itself.

Estimated average glucose is your HbA1c converted into the units a glucose meter uses. No glucose is measured — the value is computed from the A1c by a fixed equation published by the ADAG study, eAG (mg/dL) = 28.7 × A1c − 46.7. LOINC records the derivation in the test's own method field.

The range stored here, 68 - 114 mg/dL (3.77 – 6.33 mmol/L), corresponds to an A1c of 4.0 to 5.6%, so a high flag begins near ADA's 5.7% prediabetes threshold. A diagnosis needs confirmatory testing, and this value cannot confirm the A1c it came from.

The estimate has a width: the ADAG study reported a study-wide prediction-error standard deviation of 15.7 mg/dL. That describes uncertainty about your measured average glucose, not the precision of your A1c or which category it falls in.

It reflects two to three months, and it cannot show glucose swings or low readings — ADA states that A1C "does not provide a measure of glycemic variability or hypoglycemia".

Because it is A1c rescaled, every condition that can make A1c unreliable can make this unreliable too, invisibly: anemias, recent transfusion, end-stage kidney disease, pregnancy and hemoglobin variants among them. Whether one does depends on the condition and the assay. A continuous monitor's average often differs; a large or persistent gap is worth raising.

At a Glance

This is not a blood sugar measurement. It is your HbA1c put through a fixed equation and reported in the units a glucose meter uses. No separate sample is drawn, no glucose is measured, and nothing is tested that the A1c did not already test.

The conversion comes from the ADAG study, which compared A1c against continuous and fingerstick glucose monitoring in 507 people and published the formula laboratories have used ever since:

eAG (mg/dL) = 28.7 × A1c − 46.7 · eAG (mmol/L) = 1.59 × A1c − 2.59

So your eAG and your A1c are one result wearing two outfits. If you have both on a report, they cannot disagree, and neither one is a second opinion on the other.

How to read your result

The range stored on this record is 68 - 114 mg/dL (3.77 – 6.33 mmol/L). Put those two numbers back through the ADAG equation and they correspond to an A1c of 4.0% and 5.6% — the usual non-diabetic span.

That correspondence is worth knowing, because it tells you what being flagged actually means here:

Your eAG	The A1c behind it	Where that sits
below 68 mg/dL	below ~4.0%	below the usual range — see the low-result section
68 – 114 mg/dL	4.0 – 5.6%	within the usual non-diabetic range
117 – 137 mg/dL	5.7 – 6.4%	the band ADA defines as prediabetes
140 mg/dL and above	6.5% and above	at or above ADA's diabetes diagnostic threshold

A high flag on this record therefore begins at roughly the prediabetes threshold, not at some arbitrary line. That is useful. What it is not is a diagnosis. Diagnosis rests on clinical assessment and on confirmatory testing, and an A1c and the eAG calculated from it are one result, not two — they cannot confirm each other. A single flagged line on a dashboard is not a diagnosis.

One exception runs the other way. If you have symptoms of high blood sugar — heavy thirst, passing a lot of urine, blurred vision, unexplained weight loss — that is a reason to be seen promptly rather than to wait for a repeat test.

Read the figure against your own report's reference range, which may differ from the one stored here, and against the A1c printed beside it.

What the number is good for

A1c, and therefore eAG, reflects glucose exposure over roughly the preceding two to three months. Its value is that it is hard to game: a single fasting glucose tells you about one morning, while this tells you about a season.

For some people the glucose units are simply easier to think in. An A1c of 7.0% is abstract; 154 mg/dL is a number you can compare with what your meter shows. That translation is the entire reason eAG exists — the ADAG authors compared it to reporting creatinine as an estimated GFR.

The estimate has a width, and it is about the wrong quantity to read precisely

The equation describes a population, not you. In the ADAG study it explained most but not all of the variation — R² = 0.84 — and the paper is specific about the spread: "The correlation has an SD of prediction error of 15.7 mg/dl (0.87 mmol/l)", with about 90% of samples falling within 15% of the calculated figure.

What that spread is about matters. It describes how far your measured average glucose can sit from the figure the equation predicts. It is not a statement about how precisely your A1c was measured, and it does not blur the line between one A1c result and another: an A1c of 5.6% and an A1c of 5.8% are genuinely different measurements.

So: your real average glucose can differ from this estimate, sometimes substantially. A borderline result is interpreted through the A1c it came from and the follow-up your clinician judges appropriate, not by reading the estimate to more precision than it has.

When A1c — and therefore eAG — may be misleading

Because eAG is A1c rescaled, anything that makes an A1c wrong makes the eAG exactly as wrong, and there is no extra information in the eAG to reveal it. Whether any of the following actually affects a given result depends on the condition, on the assay the laboratory used, and on the clinical picture — these are reasons to interpret carefully, not automatic invalidations.

ADA states that "Factors that affect hemoglobin or red blood cells may affect A1C", and lists conditions that "can interfere with the accuracy of A1C": anemias, G6PD deficiency, recent blood transfusion, drugs that stimulate red cell production, end-stage kidney disease and pregnancy. On hemoglobin variants it is blunter — "A1C cannot be measured in individuals with sickle cell disease (HbSS) or other homozygous hemoglobin variants." Where A1c is unreliable, ADA says "alternative approaches to monitoring glycemic status should be used."

The ADAG study itself never included these people: subjects with erythrocyte disorders and with chronic renal or liver disease "were excluded from the study", and on children and pregnant women the authors wrote that "additional data in these groups are needed to confirm the established relationship." Their own summary is that "the current results are only directly applicable to this population."

If you wear a continuous glucose monitor

Your sensor reports an average too, and it often will not match your eAG. Some difference is expected.

They are different estimates built from different data. ADA says of the A1c-to-eAG conversion table that "It is unclear how generalizable these estimates are to mean glucose measurements obtained using modern CGM systems", and describes the glucose management indicator your sensor app reports as a "Calculated value approximating A1C (not always equivalent)."

They are also two different quantities: a sensor's mean glucose is an average of measured readings, while the glucose management indicator an app reports is a calculated stand-in for A1c. Comparing the wrong pair accounts for some apparent mismatches on its own.

A small difference is unremarkable. A large or persistent one is worth raising, because it can reflect the periods being compared, glucose that has been changing, limits of either measurement, or a stable non-glycemic factor affecting the A1c. That last possibility matters: an analysis of A1c against continuous monitoring warns that failing to recognize such a factor "may lead to chronic under- or overtreatment" where A1c guides management.

If you remember only one thing

eAG and A1c are the same result in different units, so this number adds no new measurement — but the units make it readable, and the stored range here corresponds to an A1c of 4.0 to 5.6%.

Two people, same result

Two people are both reported at 154 mg/dL, from an A1c of 7.0%.

The first has type 2 diabetes, takes metformin, and checks her glucose occasionally. For her the figure is a reasonable summary of a quarter, and the conversation with her clinician is about whether 7.0% is the right target for her — ADA is explicit that targets are individualized, and that less stringent goals suit some people.

The second has the same number on a panel run after a hospital admission for gastrointestinal bleeding, and he was transfused. Blood loss and transfusion both disturb the red cells A1c is measured on, so his 7.0% — and the 154 computed from it — may not describe his last three months at all. ADA lists recent transfusion among the factors that interfere with A1c accuracy.

Same figure, and for one of them it does not mean what it says. Nothing in the number marks the difference.

What this result cannot tell you

Whether you have had low blood sugar. ADA states it directly: "A1C does not provide a measure of glycemic variability or hypoglycemia." An average cannot show the swings underneath it.

What your glucose is doing today. The window is months. A change made last week is barely in it.

Whether your readings are steady or swinging. Two people with identical eAG can have very different days. This is why the page below talks about variability and why a meter or sensor answers a question this cannot.

Whether you have diabetes, on its own. Diagnosis needs confirmatory testing, and the eAG cannot serve as confirmation of the A1c it was calculated from.

Anything your A1c could not tell you. There is no independent information here. It is the same measurement, rescaled.

Common interpretation mistakes

Treating eAG as a measured glucose. No glucose was measured. LOINC encodes this in the test's own method field: estimated from glycated hemoglobin.

Comparing eAG with a fasting glucose and concluding one is wrong. A fasting result is one moment; eAG summarises months. They answer different questions and routinely differ.

Reading a low eAG as good control. It may be, and it may be the average of highs and lows cancelling out, or an A1c made falsely low by something unrelated to glucose.

Expecting eAG and a CGM average to agree, or dismissing it when they do not. Some difference is expected; a large or persistent one is worth raising.

Reading the estimate's uncertainty as uncertainty about your A1c. The spread around the conversion describes how far your measured average glucose may sit from it. Your A1c is its own measurement, with its own result.

Questions your doctor may ask

What was the A1c this was calculated from? Do you take any medication that lowers blood glucose, and has it changed recently? Any anemia, kidney disease, recent transfusion, significant blood loss, or known hemoglobin variant — each can distort the A1c this number is calculated from? Are you pregnant? Do you use a meter or a sensor, and what does it show? Is there an earlier A1c to compare with? And what decision would this number change?

Read together with

HbA1c is the measurement this value is computed from, and the one to read first. If the two ever appear to disagree, the arithmetic has gone wrong somewhere, because they cannot.

Fasting glucose samples a single moment and is part of the same diagnostic framework, with its own thresholds.

A complete blood count matters more here than people expect, because anemia and abnormal red cell turnover distort A1c and therefore this number.

Continuous glucose monitoring, where available, is what shows variability and low readings — the two things this value is structurally unable to report.

Clinical pearls

eAG is a presentation layer over A1c, not a second analyte. One result, two units.

The stored interval corresponds to A1c 4.0 – 5.6%, so the high flag begins near the prediabetes threshold.

The estimate predicts measured average glucose with real spread — a study-wide SD of 15.7 mg/dL — so it should not be read to more precision than it has.

Every A1c confounder is an eAG confounder, silently.

An average is not a safety signal: it cannot show hypoglycemia, which is the clinically urgent thing it hides.

Clinical Takeaway

Estimated average glucose is HbA1c expressed in glucose units via the ADAG regression, 28.7 × A1c − 46.7 for mg/dL. It is calculated, not measured, and LOINC 27353-2 carries that in its method field. The range stored here, 68 - 114 mg/dL, corresponds to A1c 4.0 – 5.6%, so a high flag starts near ADA's 5.7% prediabetes threshold, while a diagnosis rests on clinical assessment and confirmatory testing — and this value cannot confirm its own A1c. The estimate predicts measured average glucose with a study-wide prediction-error SD of 15.7 mg/dL, and that spread is about the measured average rather than about the A1c. It reflects two to three months of glucose exposure and cannot show variability or hypoglycemia — ADA states that explicitly. Conditions that disturb hemoglobin or red cells — anemias, G6PD deficiency, recent transfusion, erythropoiesis-stimulating drugs, end-stage kidney disease, pregnancy, hemoglobin variants — can interfere with A1c accuracy and therefore with this value, with no outward sign; whether they do depends on the condition and the assay, and the ADAG study excluded such patients. A CGM average often differs from it, and a large or persistent difference deserves review. Read it with the A1c it came from, and with whoever ordered the panel.

In one sentence

Your A1c in blood-sugar units — the same result, more readable, and no more informative.

Bottom line

If this line is flagged, the number to act on is the A1c behind it, and the question to ask is whether that A1c is trustworthy for you. If it is not flagged and you feel well, it is a reasonable summary of the last few months — and still blind to the highs and lows inside them.

Organize and understand all your lab results

Start organizing — $15/mo →

For Health Professionals Healthmatters Pro Multiple clients’ results in one dashboard. See Pro plans →

Estimated Average Glucose (eAG)

Every lab result you’ve ever had, charted and explained.

Healthmatters is a tracker for your lab results. Upload a report from any lab, any country — every marker charted and explained.

Start with my report →

Frequently Asked Questions

Is eAG the same as a blood sugar test?

No, and this is the most useful thing to know about it. No glucose was measured to produce this number. Your eAG is your HbA1c result put through a fixed equation and reported in the units a glucose meter uses: eAG in mg/dL = 28.7 × A1c − 46.7. LOINC, the system that names laboratory tests, records this in the test's own method field as "estimated from glycated hemoglobin". A blood sugar test measures glucose in a sample at one moment. This estimates an average across roughly two to three months, from a different measurement entirely. The two answer different questions and will routinely give different-looking numbers without either being wrong.

What A1c does my eAG correspond to?

Both numbers are the same result, so you can convert either direction with the equation your laboratory calculated it with. In mg/dL, eAG = 28.7 × A1c − 46.7; going backwards, A1c = (eAG + 46.7) ÷ 28.7. Some reference points: an A1c of 5.0% is an eAG of about 97 mg/dL; 5.7% is about 117; 6.5% is about 140; 7.0% is about 154; and 8.0% is about 183. The range stored on this record, 68 to 114 mg/dL, corresponds to an A1c of 4.0 to 5.6%. In practice you should not need to do the arithmetic — look for the A1c on the same report, since that is the number that was actually measured.

Why does my eAG not match my glucose meter or CGM average?

Partly because they are different estimates, and partly because they may be different quantities. Your eAG comes from an equation fitted to a study population; a sensor's mean glucose is an average of its own readings; and the glucose management indicator an app shows is a calculated stand-in for A1c rather than a mean glucose. The American Diabetes Association says it is unclear how generalizable the A1c-to-eAG conversion is to averages from modern monitors, and calls the glucose management indicator a calculated value approximating A1c that is "not always equivalent". A small difference is unremarkable. A large or persistent one deserves review with your clinician: it can reflect the periods compared, glucose that has been changing, limits of either measurement, or a stable non-glycemic factor affecting your A1c, which can lead to under- or overtreatment if it goes unrecognised.

My eAG is flagged high. Does that mean I have diabetes?

Not on its own. The stored range corresponds to an A1c of 4.0 to 5.6%, so the high flag begins near where ADA's prediabetes band starts — an A1c of 5.7%, an eAG around 117 mg/dL. ADA's diabetes threshold of 6.5% is an eAG around 140. But a diagnosis rests on clinical assessment and confirmatory testing, and an A1c and the eAG calculated from it are one result rather than two, so they cannot confirm each other. One exception matters: if you have symptoms of high blood sugar — heavy thirst, passing a lot of urine, blurred vision, unexplained weight loss — be seen promptly rather than waiting for a repeat. Otherwise, take it to the clinician who ordered it.

Is a low eAG good?

Not automatically, and the answer depends on whether you take medication that lowers glucose. If you do — insulin or a sulfonylurea, for instance — a low average deserves a prompt conversation with your prescriber, because an average cannot show the swings underneath it. ADA states that A1C "does not provide a measure of glycemic variability or hypoglycemia", and that severe or frequent hypoglycemia is an absolute indication for modifying a treatment plan. A low eAG does not prove you are having low readings; it simply cannot rule them out. If you take no glucose-lowering medication, an unexpectedly low value still needs interpreting against the A1c it came from, your symptoms, and anything affecting your red cells — no medication does not by itself make it benign.

What conditions make eAG unreliable?

Every condition that makes HbA1c unreliable, because eAG is A1c rescaled and carries no independent information. ADA states that factors affecting hemoglobin or red blood cells may affect A1c, and lists anemias, G6PD deficiency, recent blood transfusion, drugs that stimulate red cell production, end-stage kidney disease and pregnancy as able to interfere with its accuracy. It also says A1c cannot be measured at all in people with sickle cell disease or other homozygous hemoglobin variants. The study the eAG equation comes from excluded people with red cell disorders and with chronic kidney or liver disease, so the formula was never validated in them. Where A1c is unreliable, ADA recommends monitoring glucose directly instead.

How accurate is the eAG estimate?

It describes a population well and an individual less precisely. In the ADAG study the equation explained much of the variation between A1c and measured average glucose, with an R² of 0.84 across 507 participants, a study-wide prediction-error standard deviation of 15.7 mg/dL, and about 90% of samples within 15% of the calculated value. All of that describes how far your measured average glucose may sit from the estimate. It says nothing about how precisely the A1c itself was measured, and a trend across results is more informative than any single figure.

Where does the eAG formula come from?

From the A1c-Derived Average Glucose study, published in Diabetes Care in 2008. It compared A1c against continuous glucose monitoring and frequent fingerstick testing in 507 people — 268 with type 1 diabetes, 159 with type 2, and 80 without diabetes — and fitted the regression laboratories now use: eAG in mg/dL = 28.7 × A1c − 46.7. The authors proposed reporting A1c alongside this translation for the same reason creatinine is reported with an estimated GFR: the derived figure is easier to reason about. They were also explicit that their results applied directly to the population studied, and that children and pregnant women needed further data.

Can my eAG be good while my diabetes control is not?

Yes, and this is the main limitation of any average. The same midpoint can come from steady readings or from frequent highs and lows that cancel each other out, and the second pattern can include hypoglycemia, which matters clinically and is invisible here. ADA states plainly that A1C "does not provide a measure of glycemic variability or hypoglycemia". This is why time-in-range from a continuous monitor has become a separate target rather than a restatement of A1c: it describes the shape of the days, where eAG describes only their center. A reassuring eAG with troubling daily readings is a real and recognized combination, and the daily readings are the ones to raise.

Clinical Sources & References

All our content is backed by peer-reviewed studies, academic research, and trusted medical sources. We’re committed to accuracy and transparency — see our editorial policy for details.

Glycemic Goals and Hypoglycemia: Standards of Care in Diabetes—2025

American Diabetes Association Professional Practice Committee, Diabetes Care, volume 48 supplement 1. Accessed 13 October 2026 · 2025 View source ↗
Translating the A1C Assay Into Estimated Average Glucose Values

The A1c-Derived Average Glucose (ADAG) Study Group, Diabetes Care. DOI 10.2337/dc08-0545. Accessed 13 October 2026 · 2008 View source ↗
Diagnosis and Classification of Diabetes: Standards of Care in Diabetes—2026

American Diabetes Association Professional Practice Committee, Diabetes Care, volume 49 supplement 1, page S27 onward. Accessed 15 October 2026 · 2026 View source ↗
LOINC 27353-2 — Glucose mean value [Mass/volume] in Blood Estimated from glycated hemoglobin

Regenstrief Institute, Inc. Accessed 13 October 2026 · 2026 View source ↗
Estimating Glycemia From HbA1c and CGM: Analysis of Accuracy and Sources of Discrepancy

Diabetes Care, volume 47 issue 3. Accessed 15 October 2026 · 2024 View source ↗

Article Review Log

Last reviewed October 15, 2026

Originally published October 15, 2026

Reviewed by HealthMatters Editorial Team

Written by Healthmatters Editorial Team

Organize and understand all your lab results

Start organizing — $15/mo →

For Health Professionals Healthmatters Pro Multiple clients’ results in one dashboard. See Pro plans →

Healthmatters Lifetime Plan

Pay $250. Track your health forever

One payment. No renewals. Yours for life.

10,000+ biomarkers explained · trends over time

See other plans →

Secure checkout via Stripe · GDPR & HIPAA

We accept results from any lab

LabCorp Quest Diagnostics Genova Diagnostics DUTCH Test Doctor's Data Vibrant America ZRT Laboratory Cyrex Labs + any other lab

What members are saying

★★★★★ 10,000+ happy members

★★★★★

"Watching my numbers move over the years — instead of forgetting last time's results — completely changed how I manage my health."

Daniel

Member since 2022

★★★★★

"Living across several countries made keeping my lab results organized a real challenge. Healthmatters brought my results from the USA, Spain, and Germany together — 20 years of history, with long-term trends instead of scattered reports."

Ben

Member since 2021 · tracks across countries

★★★★★

"Standard lab ranges are too wide for how I practice — 'normal' isn't optimal. I set my own bands, so a client sees yellow before a marker goes officially out of range. That's the conversation I want to be having."

Dr. Alexandra

Functional medicine · 6 yrs on Pro

Estimated Average Glucose (eAG): What Your Number Means, and Why It Is Really Your A1c

At a Glance

How to read your result

What the number is good for

The estimate has a width, and it is about the wrong quantity to read precisely

When A1c — and therefore eAG — may be misleading

If you wear a continuous glucose monitor

If you remember only one thing

Two people, same result

What this result cannot tell you

Common interpretation mistakes

Questions your doctor may ask

Read together with

Clinical pearls

Clinical Takeaway

In one sentence

Bottom line

Organize and understand all your lab results

Every lab result you’ve ever had, charted and explained.

Frequently Asked Questions

Related Biomarkers

Related Health Conditions

Pay $250. Track your health forever

What members are saying