Medical education reviewed by Domenico Savatta, MD, FACS on July 23, 2026; educational use only. Care routing remains closed.
Modern Sleep and Metabolic InstituteExplore the health library
Menu

Blood glucose test comparison

A1C versus fasting glucose: what does each test tell you?

A1C estimates glucose exposure over roughly two to three months, while fasting plasma glucose measures the concentration at one properly prepared point in time. They can disagree without either report being automatically wrong. The useful question is whether the sample and A1C are reliable and whether repeat or alternative testing changes the diagnosis.

Medically reviewed July 23, 2026 • Clinical reviewer: Domenico Savatta, MD, FACS • Educational use only

Read the guideMedication questions

Direct answer

Which is more accurate, A1C or fasting glucose?

Neither test is universally better. A1C is convenient and reflects a longer period, but conditions affecting red blood cells can distort it. Fasting plasma glucose is a direct point measurement, but preparation, illness, stress, sleep, and day-to-day variation matter. Clinicians select, repeat, or combine tests based on the person and the diagnostic question.

  • Use venous laboratory plasma glucose for diagnostic decisions unless a clinician explicitly directs another method.
  • Home meters and wearable glucose estimates can reveal patterns but do not replace standard diagnostic testing.
  • Do not diagnose diabetes from one borderline or unexpected result without the confirmation plan appropriate to the clinical situation.

At a glance

  • NIDDK lists A1C 5.7%–6.4% and fasting plasma glucose 100–125 mg/dL as prediabetes ranges.
  • A1C 6.5% or higher or fasting plasma glucose 126 mg/dL or higher is in a diabetes range, subject to clinical confirmation rules.
  • A1C does not require fasting; fasting glucose requires at least eight hours without caloric intake under standard instructions.
  • Pregnancy, anemia, transfusion, kidney or liver disease, and hemoglobin variants can affect which test is appropriate.
  • Discordance should trigger context and follow-up, not cherry-picking.

A1C measures a longer exposure window

The test uses glycated hemoglobin as an estimate of recent average glucose.

Because red blood cells circulate for weeks, A1C reflects more than the morning of the blood draw and does not require fasting. Recent weeks influence the result more than distant weeks, so it is not a perfect three-month average. A1C can miss short spikes and does not reveal when highs or lows occurred. The percentage should be interpreted from a certified laboratory method in context, not converted into a diagnosis from a consumer-device estimate.

Fasting plasma glucose is a point measurement

It answers what the plasma glucose was after the required fast at that moment.

NIDDK describes fasting as at least eight hours without eating, with only sips of water. Acute illness, stress, steroid treatment, sleep loss, timing, and preparation can influence a value. A correct fasting sample can still differ from another day because glucose regulation varies. A finger-stick meter is valuable for management when prescribed, but diagnostic thresholds were established for laboratory plasma testing and should not be applied casually to every device reading.

Know the standard categories without self-diagnosing

Ranges organize follow-up; the clinician establishes the diagnosis.

NIDDK lists normal A1C below 5.7%, prediabetes 5.7% through 6.4%, and diabetes 6.5% or higher. It lists normal fasting plasma glucose at 99 mg/dL or below, prediabetes from 100 through 125 mg/dL, and diabetes at 126 mg/dL or higher. In an asymptomatic person, a result in a diabetes range generally requires confirmation. Symptoms, pregnancy, laboratory quality, and acute illness can change the testing pathway.

Explain disagreement before repeating blindly

The two tests can identify different people because they measure different biology.

Review whether fasting instructions were followed and whether A1C may be altered by anemia, blood loss, transfusion, kidney or liver disease, pregnancy, or a hemoglobin variant. The clinician may repeat one test, obtain the other, or use an oral glucose tolerance test. A disagreement is clinically useful information; it can show that a single method does not capture the full pattern. Do not average the results or choose the one that supports a preferred conclusion.

Connect sleep evidence carefully

Sleep loss and sleep apnea can relate to glucose risk but cannot explain away an abnormal result.

Bring sleep duration, shift work, insomnia, snoring, breathing pauses, PAP use, medicines, and recent illness to the interpretation. If obstructive sleep apnea is suspected, evaluate it on its own track. Treating sleep may support overall health, but glucose still needs repeat laboratory measurement and a prevention or treatment plan. Mark who owns follow-up, which test will be repeated, the date, and what symptoms warrant earlier care.

Appointment checklist

Reconcile two glucose results

Use the original reports and answer the preparation and accuracy questions before comparing categories.

  1. 1

    A1C details

    Record percentage, date, laboratory, prior trend, and anemia, hemoglobin variant, kidney/liver disease, pregnancy, bleeding, or transfusion context.

  2. 2

    Fasting details

    Record mg/dL, fasting duration, water or calories consumed, illness, steroids or other medicines, sleep loss, stress, and collection time.

  3. 3

    Other evidence

    Bring symptoms, home readings only if requested, family and pregnancy history, weight trend, blood pressure, lipids, and relevant sleep pattern.

  4. 4

    Follow-up decision

    Ask which result is considered reliable, whether confirmation or an oral glucose tolerance test is needed, and when the next result is due.

Common questions

Questions patients ask first

Can fasting glucose be normal when A1C is high?

Yes. The tests measure different time windows, and biological or measurement factors can create discordance. Review A1C reliability, fasting preparation, medicines, and whether repeat or alternative testing is needed.

What A1C is prediabetes?

NIDDK lists 5.7% through 6.4% as the prediabetes range. A clinician should interpret the result and consider whether conditions affecting red blood cells make another test more reliable.

Can poor sleep raise fasting glucose?

Short or disrupted sleep can affect glucose regulation, but one abnormal result cannot be attributed to sleep without evaluation. Repeat and interpret the laboratory result appropriately while assessing suspected sleep problems separately.

Do I need to fast for an A1C?

No fasting is required for A1C. Other tests ordered at the same draw may require fasting, so follow the laboratory and clinician’s instructions for the complete order.

Authoritative sources

Review the public guidance

Medically reviewed by Domenico Savatta, MD, FACS on July 23, 2026. Source links support education, not a personal recommendation.