Glucose, A1C and insulin: what does each test actually show?
A moment, a longer-term pattern, and a hormone response. Related information—but not interchangeable diagnoses.
The quick answer
The three tests look at different parts of glucose regulation.
Plasma glucose measures glucose at one point in time. A1C estimates longer-term glycemic exposure through hemoglobin glycation. Insulin measures the hormone circulating at the moment of collection. Fasting glucose and A1C have validated screening and diagnostic roles; fasting insulin and HOMA-IR do not have universal diagnostic cutoffs and should not be treated as stand-alone diagnoses.
Fasting glucose — the snapshot
Shows plasma glucose after a defined fasting period. It is accessible and validated, but illness, stress, sleep, medications, and day-to-day biological variation can affect it.
A1C — the longer view
Reflects glycemic exposure over roughly the prior two to three months, with greater influence from recent weeks. It does not require fasting, but red-cell biology and assay interference matter.
Insulin — the response signal
Shows circulating insulin at one moment. The result depends on glucose, fasting duration, timing, medications, assay, and physiology; by itself, it does not diagnose prediabetes, diabetes, or insulin resistance.
The central distinctionGlucose asks: what is the concentration now? A1C asks: what has glycemic exposure looked like over time? Insulin asks: how much circulating hormone is present at this moment?
Side by side
Four measurements. Four different questions.
MERHI ONE Evidence Map
What the evidence supports—and where interpretation often goes too far.
Each rating applies to the exact claim. A test can be biologically interesting without being validated as a universal screening tool, diagnostic cutoff, or treatment target.
Fasting plasma glucose, A1C, and the oral glucose tolerance test are validated tests for screening and diagnosing diabetes.1,2,3,4
Current U.S. and Brazilian guidance recognizes these tests, with method-specific criteria. They do not identify exactly the same people because each samples a different aspect of glycemia.
An abnormal diagnostic result usually requires confirmation when unequivocal hyperglycemia is absent.1,2,4
A1C ≥6.5%, fasting plasma glucose ≥126 mg/dL, or 2-hour glucose ≥200 mg/dL during a 75-g oral glucose tolerance test meets a diagnostic threshold. Without classic symptoms or a hyperglycemic crisis, a second abnormal result is generally required.
A1C reflects longer-term glycemic exposure, not every glucose peak or day-to-day fluctuation.1,5,7
A1C integrates glycation over the red-cell lifespan and is weighted toward more recent weeks. The estimated average glucose derived from it is an approximation, not a direct record of every meal or glucose excursion.
Red-cell disorders and assay interference can make A1C misleading.4,5,6,8
Hemolysis or recent blood loss can lower A1C; iron deficiency may raise it. Transfusion, hemoglobin variants, kidney failure, liver disease, pregnancy, and some assay methods can also alter interpretation.
Routine insulin or proinsulin testing is recommended for most adults at risk for diabetes or cardiovascular disease.4
Major laboratory guidance specifically does not recommend routine insulin or proinsulin measurement for most people with diabetes or at risk for diabetes or cardiovascular disease. Clinical decisions are better validated against glucose-based measures, A1C, risk factors, and the relevant diagnosis.
One HOMA-IR threshold can diagnose insulin resistance in every adult.2,4
HOMA-IR combines fasting insulin and fasting glucose, commonly as insulin × glucose in mg/dL ÷ 405. Its value changes with the insulin assay, population, physiology, and study definition, and guidelines do not establish one universal diagnostic cutoff for routine care.
Discordance between fasting glucose and A1C can contain useful information.1,2,5,7
One test may cross a threshold while another does not because they reflect different time windows, biological variation, or interference. Repeat testing, an oral glucose tolerance test, or an alternative marker may clarify selected cases.
Insulin and C-peptide can answer selected diagnostic questions.1,4
Paired glucose, insulin, C-peptide, and related measurements may be important during a documented hypoglycemic episode. C-peptide can also help assess endogenous insulin secretion in selected diabetes-classification questions, especially when insulin treatment complicates interpretation.
Internet charts define a universal ‘optimal fasting insulin’ that should be treated to a target.4
There is no harmonized fasting-insulin assay, universal reference target, or outcomes-based treatment threshold endorsed for the general population. A lower or higher result cannot be interpreted independently of glucose and the clinical setting.
Two native pathways
The diagnostic biology is shared. Screening recommendations are localized.
United States — ADA 2026 and USPSTF
- ADA diagnostic thresholds include A1C ≥6.5%, fasting plasma glucose ≥126 mg/dL, or 2-hour glucose ≥200 mg/dL during a 75-g oral glucose tolerance test; confirmation is required without unequivocal hyperglycemia.
- ADA recommends risk-based testing at any age and general adult screening beginning no later than age 35.
- USPSTF gives a Grade B recommendation for asymptomatic adults ages 35–70 with overweight or obesity—a narrower preventive-services population than ADA guidance.
- USPSTF accepts fasting glucose, A1C, or oral glucose tolerance testing and notes that a 3-year interval may be reasonable after a normal result, although the optimal interval is uncertain.
Brazil — SBD 2026
- Diabetes thresholds include fasting glucose ≥126 mg/dL, A1C ≥6.5%, 1-hour glucose ≥209 mg/dL, or 2-hour glucose ≥200 mg/dL during oral glucose tolerance testing.
- Prediabetes ranges include fasting glucose 100–125 mg/dL, A1C 5.7–6.4%, 1-hour glucose 155–208 mg/dL, or 2-hour glucose 140–199 mg/dL.
- The SBD recommends universal type 2 diabetes screening from age 35, with earlier testing in adults with overweight/obesity and risk factors.
- Fasting glucose and/or A1C are initial options; oral glucose tolerance testing can detect abnormalities missed by either test and is used selectively.
Reading combinations
When the results disagree, ask what could explain the pattern.
Fasting glucose normal + A1C elevated
Consider repeat confirmation, post-meal hyperglycemia, biological variation, iron deficiency, kidney/liver disease, hemoglobin-related issues, and assay interference. One explanation does not fit every person.
Fasting glucose elevated + A1C below threshold
A transient stressor, medication, fasting/sample conditions, emerging dysglycemia, or ordinary variation may contribute. Repeat or alternative testing may be appropriate.
Both in the prediabetes range
The pattern supports increased future diabetes and cardiovascular risk, but not inevitable progression. Overall risk and evidence-based lifestyle or prevention options matter more than one label.
Both reach a diabetes threshold
Concordant abnormal results strongly support the diagnosis. Clinical review should define diabetes type, urgency, symptoms, complications, and a treatment plan.
Insulin high with glucose below a diabetes threshold
This may reflect compensatory insulin secretion, but it is not a stand-alone diagnosis. Timing, assay, body composition, medications, puberty, pregnancy, liver/kidney function, and other factors can influence it.
Insulin low or ‘normal’
The number does not exclude dysglycemia, beta-cell dysfunction, or diabetes. Interpretation depends on the simultaneous glucose and the diagnostic question.
Common interpretation errors
A precise number can still create a false conclusion.
The most common error is assigning more diagnostic certainty to insulin or A1C than the method and context can support.
‘One high result means I have diabetes’
Without unequivocal hyperglycemia, diagnosis usually requires confirmation. Symptoms, acute illness, medications, and sample reliability matter.
‘My A1C is normal, so every glucose peak is normal’
A1C is an integrated average and can miss variability or post-meal abnormalities. It is also vulnerable to red-cell and assay effects.
‘HOMA-IR above an online cutoff proves insulin resistance’
Cutoffs vary by assay, population, and research method. No single value is a universal routine-care diagnosis.
‘Fasting insulin has one ideal target for everyone’
No outcomes-based universal target is established. The value must be paired with glucose and a defined clinical question.
‘A1C is exact average glucose’
Estimated average glucose is a population-derived approximation. Two people with the same A1C can have different glucose patterns.
‘More frequent insulin testing means better prevention’
Routine insulin testing has not been validated as a superior universal prevention strategy. Risk-based follow-up uses established glycemic tests and clinical factors.
What a wrong interpretation can cause
- Premature or missed diagnosis
- Unnecessary anxiety and repeated testing
- False reassurance from one normal number
- Unvalidated diets, supplements, or self-medication
- Failure to recognize A1C interference
- Delay in confirming clinically important hyperglycemia
A practical next step
Turn the numbers into a reliable glycemic assessment.
- 01
Define the question: screening, diagnosis, monitoring, pregnancy, symptoms, medication effects, or documented hypoglycemia are different pathways.
- 02
Confirm sample conditions, fasting duration, acute illness, recent blood loss or transfusion, anemia, medications, and prior results.
- 03
Use fasting plasma glucose and/or A1C for routine screening when appropriate; choose an oral glucose tolerance test when a more sensitive or confirmatory assessment is indicated.
- 04
Confirm an abnormal diagnostic result when there is no unequivocal hyperglycemia, following the relevant guideline.
- 05
Investigate meaningful A1C–glucose discordance before making a treatment decision.
- 06
Order insulin, C-peptide, or HOMA-IR only when the result answers a defined question and its limitations are explicit.
Frequently asked
Direct answers to the questions people actually ask.
Which is better: fasting glucose or A1C?+
Neither is universally better. Fasting glucose is a point-in-time measurement; A1C reflects longer exposure and needs no fasting. Their limitations differ, and selected people benefit from both or from an oral glucose tolerance test.
Do I need to fast for A1C?+
No. A1C itself does not require fasting. A simultaneous fasting glucose or lipid panel may, so follow the laboratory order and instructions.
What do 5.7% and 6.5% mean?+
For nonpregnant adults, A1C 5.7–6.4% is a prediabetes range and ≥6.5% is a diabetes diagnostic threshold when measured appropriately. In the absence of unequivocal hyperglycemia, confirm the diagnosis.
Is fasting insulin useful?+
Sometimes, for a clearly defined question. It is not required for routine diabetes diagnosis and does not have one universal cutoff for insulin resistance.
What is HOMA-IR?+
It is a surrogate calculated from fasting glucose and fasting insulin. It is useful in research and selected contexts, but assay and population differences prevent one universal clinical cutoff.
Can anemia change A1C?+
Yes. Iron deficiency can raise A1C, while hemolysis or shortened red-cell survival can lower it. The direction depends on the condition, and meaningful discordance deserves review.
Can a glucose meter or CGM diagnose diabetes?+
Home meters and continuous glucose monitors are valuable in management, but ordinary diagnosis should use standardized laboratory criteria and confirmation rather than a consumer-device reading alone.
Claim-level foundation
References
Priority was given to current national standards, Brazilian guidance, preventive-services recommendations, laboratory guidance, and official assay resources. Citations beside each claim point to its supporting sources.
- 01Open source ↗
American Diabetes Association · 2026
Standards of Care in Diabetes — Diagnosis and Classification of Diabetes
- 02Open source ↗
Sociedade Brasileira de Diabetes · 2026
Diagnosis of diabetes mellitus — Brazilian guideline
- 03Open source ↗
U.S. Preventive Services Task Force · 2021; current
Screening for Prediabetes and Type 2 Diabetes
- 04Open source ↗
ADA / Association for Diagnostics & Laboratory Medicine · 2023
Guidelines and Recommendations for Laboratory Analysis in the Diagnosis and Management of Diabetes Mellitus
- 05Open source ↗
National Institute of Diabetes and Digestive and Kidney Diseases · Current resource
The A1C Test & Diabetes
- 06Open source ↗
National Glycohemoglobin Standardization Program · Updated 2024
Factors that Interfere with HbA1c Test Results
- 07Open source ↗
National Institute of Diabetes and Digestive and Kidney Diseases · Current resource
Diabetes & Prediabetes Tests
- 08Open source ↗
National Glycohemoglobin Standardization Program · Current resource
Clinical Use of HbA1c
Editorial record