CGM vs A1c vs Fasting Glucose: Three Different Measurements
People arrive with a fasting glucose that was fine, an A1c that was borderline, and a two-week CGM trace from a wearable that looked alarming — and want to know which one is telling the truth. All three are. They are measuring different things over different windows, and the disagreement is the information.
This page is the comparison. Our explainers on A1c and fasting insulin cover each marker on its own.
Fasting glucose: one moment
A fasting glucose is the level in your blood at one moment, after a night without food. It is cheap, standardized and the basis of most diagnostic thresholds — prediabetes runs from 100 to 125 mg/dL fasting, and diabetes from 126.
Its limitation is exactly that it is one moment, and the least demanding moment of the day. Fasting glucose is the last thing to go wrong in developing insulin resistance: the liver's overnight output is well controlled long after the response to meals has deteriorated. A normal fasting glucose rules out very little. That is why we draw fasting insulin beside it — insulin rises years before glucose does — and why HOMA-IR exists.
A1c: three months, averaged
Hemoglobin A1c measures the fraction of red-cell hemoglobin that has had glucose attached to it. Because red cells live about three months, it reflects the average glucose over that period. Prediabetes runs from 5.7 to 6.4%, diabetes from 6.5.
Its strengths are stability — it does not care whether you fasted or what you ate yesterday — and the long track record behind its thresholds. Its limitations are less known. It is an average, so it cannot distinguish a steady moderate glucose from a pattern of large spikes and dips that average to the same figure. And it depends on red-cell lifespan: anemia, iron deficiency, kidney disease, some hemoglobin variants and even recent blood loss shift it, in both directions. A person with low ferritin can run an A1c that overstates their glucose.
CGM: two weeks, continuously
A monitor samples interstitial glucose every few minutes and reports the whole curve — every meal, every night, every workout. The standard summary metrics are time in range, which for the general adult is the proportion of readings between 70 and 180 mg/dL; mean glucose; and variability, the measure of how far and how often the line swings.
Its strength is the one thing the other two cannot see: pattern. A post-meal excursion that resolves in an hour, a 3am rise, a flat line that never moves. Its limitations are the ones the hub page sets out at length: a short window that may not represent the year; a normal range for non-diabetics that is still being defined; and a tendency, in a metabolically healthy wearer, to turn ordinary fluctuation into anxiety and unnecessary restriction. The 2026 JAMA Internal Medicine review is blunt about that last point, and so are we.
Where they disagree, and what it means
Normal fasting glucose, prediabetic A1c. Common, and usually real: the post-meal response has deteriorated while the overnight output holds. A CGM will typically show the excursions the fasting draw missed. This person has prediabetes.
Prediabetic A1c, normal CGM. Check the red cells. Iron deficiency or anemia can inflate A1c; a genuinely flat two-week trace plus a normal fasting insulin argues that the A1c is the outlier.
Normal A1c, alarming CGM. Usually not alarming. An A1c in the normal range with a trace showing meal peaks in the upper part of the normal range describes most healthy adults eating normally. This is the over-interpretation case, and it is the reason we do not hand out monitors without a reason to read one.
Everything normal, symptoms persist. The glucose axis is not the answer, and it is time to look elsewhere — thyroid, sleep, hormones, iron.
Which one you need
Everyone gets fasting glucose, fasting insulin and A1c on the baseline panel; together they cover the moment, the average and the underlying resistance. A CGM is added when there is a question those three cannot answer: what the pattern looks like in someone with prediabetes, why weight is not moving, what happens on or off a GLP-1. For a person over 45 with a family history and a rising fasting insulin, that question exists. For a person with three normal numbers and no symptoms, it usually does not.
Frequently asked questions
Neither is more accurate; they measure different things. A1c averages three months; a CGM shows two weeks of pattern. Disagreement between them is information, not error.
Yes, commonly. Fasting glucose is the last marker to deteriorate. Fasting insulin and A1c, or a CGM, usually reveal it.
Conventionally the proportion of readings between 70 and 180 mg/dL, though healthy adults spend almost all their time in a narrower band. The normal range for non-diabetics is still being defined.
Anemia, iron deficiency, kidney disease or a hemoglobin variant can inflate A1c independent of glucose. A flat trace and a normal fasting insulin point to the A1c being the outlier.
No. Meal peaks within the normal range are what a healthy trace looks like. A single reading is never a diagnosis.
Usually not. It adds value when there is a specific question the panel cannot answer.
Where this fits in your plan
Start with the three numbers on the baseline panel. The Continuous Glucose Monitor page covers when the fourth is worth adding.
We measure first. Then we act.
References
- American Diabetes Association. Standards of Care in Diabetes — 2026, Section 2: Diagnosis and Classification (thresholds); Section 7: Diabetes Technology (time in range).
- Battelino T et al. Clinical targets for continuous glucose monitoring data interpretation: recommendations from the International Consensus on Time in Range. Diabetes Care 2019;42:1593–1603.
- Klonoff DC et al. Continuous glucose monitoring in people without diabetes: a narrative review. JAMA Internal Medicine 2026.
- Shah VN et al. Continuous glucose monitoring profiles in healthy nondiabetic participants: a multicenter prospective study. JCEM 2019;104:4356–4364.
- Cavagnolli G et al. Effect of ethnicity, iron deficiency and other factors on HbA1c. PLoS One 2017;12:e0171315 — non-glycemic influences on A1c.
- Back toContinuous Glucose Monitor
- Using a CGM on a GLP-1A glucose monitor on a GLP-1 shows the drug working in real time — and, for most people, that is all it shows. When it is worth adding, when it is noise, and the one safety case.Read
- A1c levels explainedA1c estimates your average glucose over about three months. The diagnostic thresholds, what shifts a result without your glucose changing, and what it misses.Read
- HOMA-IR explainedHOMA-IR combines fasting glucose and fasting insulin into a single index. What it is good for, where it falls down, and why it is better for tracking than diagnosing.Read
How we write and review our content
ACT 2 Health provides clinician-led care. Treatments are available only to eligible patients following clinical evaluation and within applicable regulations. This content is educational and is not medical advice. Individual results vary.
Diagnostic testing does not diagnose or rule out disease on its own and is interpreted by a licensed provider alongside your history and examination.
All medical decisions are made solely by licensed healthcare professionals. Medications are prescribed only when medically necessary. GLP-1 medications are not suitable for everyone. Results may vary.
Care is delivered via telemedicine by healthcare professionals licensed in the state where the patient is located. Services are available only in states where our providers are licensed.