HbA1c Versus Fasting Glucose for Diabetes Screening

HbA1c and fasting glucose offer different views of blood sugar regulation, so diabetes screening is not simply a choice between two interchangeable tests. Each has blind spots that the other can help expose. For a proactive patient, understanding those differences makes a strong case for reading both results together and investigating disagreement rather than accepting whichever result seems more reassuring.

What HbA1c and fasting plasma glucose each actually measure

HbA1c measures the percentage of hemoglobin coated with glucose. Duong et al. 2023 describes how glucose attaches to hemoglobin over the lifespan of a red blood cell, creating a record of past exposure. The result reflects average blood sugar over roughly two to three months rather than the concentration present when blood is drawn. Time is built into the measurement itself. Its meaning therefore depends on both glucose exposure and the cells that record it.

Fasting plasma glucose, or FPG, directly measures how much glucose is in plasma after an eight-to-twelve-hour fast. It captures the level at that particular moment. Unlike HbA1c, it does not use hemoglobin as an indirect record of blood sugar. A fasting result and a longer-term average can differ and still both be correct, because they answer different questions about the same person's glucose regulation.

Both tests were endorsed for diabetes diagnosis by the ADA and WHO, but HbA1c entered that role only in 2009–2011, after standardized assay methods became available. Before then, it was used for monitoring glycemic control. Its convenience is a genuine practical advantage. HbA1c requires no fasting and can be drawn at any time of day, while FPG generally requires an overnight fast and a morning appointment. The fasting and appointment requirements can prevent someone from completing screening at all.

Why HbA1c results can be wrong even when the lab is right

Because HbA1c depends on red blood cells, changes in their lifespan can change the result without any change in blood sugar. The distinction is between measuring accurately and interpreting correctly. If the cells have had unusually little or unusually long to accumulate glucose, their percentage of glycated hemoglobin no longer represents the usual window.

Conditions that shorten red blood cell lifespan, including hemolytic anemia, chronic blood loss and pregnancy, can produce falsely low HbA1c readings. Newer cells have had less time to accumulate glucose. Someone with uncontrolled blood sugar and hemolytic anemia may therefore appear to have a normal HbA1c. The reverse can occur with iron deficiency anemia, folate deficiency or B12 deficiency, which can extend the exposure window and falsely elevate the result. Iron deficiency can give someone with normal glucose an HbA1c result in the range linked to prediabetes or diabetes.

Hemoglobin structure introduces a separate source of uncertainty. Variants, including sickle cell trait and those associated with thalassemia, can make HbA1c readings unreliable, with the direction of error depending on the variant and assay method. These are conditions to consider actively, not dismiss as obscure laboratory caveats. Consider whether anything about your blood makes HbA1c harder to interpret. When red blood cell biology is in question, FPG is the appropriate anchor because it measures glucose independently of those cells, although illness, stress and daily variation can still affect it.

How often the two tests disagree — and what that reveals about each one's blind spots

Karnchanasorn et al. 2016 found that HbA1c ≥ 6.5% detected less than half of the people whose diabetes was confirmed by FPG. Sensitivity fell further when HbA1c was compared with the two-hour oral glucose tolerance test, or OGTT, criterion. Sensitivity describes how well a test identifies people who have the condition by the comparison standard. The choice of test changed who was identified.

The Duong analysis found that HbA1c and FPG had similar sensitivity against the OGTT reference standard, with both below the combined approach. FPG ranked highest on specificity and positive likelihood ratio. Detecting more true cases and avoiding false positives are related goals, but a test can perform better on one than the other.

The disagreement also changes across diagnostic boundaries. HbA1c tends to classify more people as prediabetic, while FPG is more conservative at that boundary. At the diabetes threshold, however, HbA1c misses cases that FPG catches. This shows that screening's reach depends on the biological dimension being measured and the cutoff applied. If neither test is sufficient alone, what does each reveal that the other cannot?

What each test uniquely contributes — and what it cannot tell you

FPG's strength is its direct view of current glucose. It can capture acute or rapidly changing blood sugar that a longer-term average obscures, without relying on red blood cell lifespan or hemoglobin structure. That immediacy is also its main limitation. Acute illness, poor sleep or high stress can elevate a single reading, while an unusually restrained eating period before the draw can depress it. Skipping the required fast can also change the result for reasons that have nothing to do with underlying metabolic health.

HbA1c smooths out much of that day-to-day noise. Because it covers a longer time, it helps track sustained glucose exposure and guide therapy while keeping one unusual morning from having too much weight in the assessment. But an average is not a timeline. If glucose has risen recently, the full change will not appear in HbA1c for weeks. The same averaging that makes the result steadier can delay recognition of a new problem.

FPG can be normal even when post-meal glucose is high, while HbA1c cannot show when glucose rose or how sharply. The OGTT measures fasting glucose and glucose two hours after a glucose load, exposing impaired regulation that the other tests can miss. This is why it remains the reference standard.

Which patients are most likely to be misclassified by relying on only one test

The Karnchanasorn study found that people with diabetes by the two-hour glucose criterion but HbA1c below the diabetes threshold were more likely to be older, female and leaner than those identified by HbA1c. HbA1c at the current cutoff captured only 28.1% of those classified as diabetic by the two-hour plasma glucose criterion. Those characteristics describe a study-level pattern. A reassuring HbA1c result does not necessarily mean that a person's response to glucose is normal.

Other vulnerabilities follow from the measurement mechanisms themselves. With hemolytic anemia, pregnancy or chronic blood loss, a low HbA1c may be partly due to red blood cells that die sooner, even when glucose exposure is not low. With iron deficiency anemia, a high HbA1c may be partly due to cells having more time to take up glucose. These errors point in different directions. One can conceal high blood sugar, while the other can create an abnormal signal that prompts unnecessary intervention.

A known hemoglobin variant creates another reason to question HbA1c-only screening, because the result may depend on the assay used. Ancestry can inform a discussion about hemoglobin variants, but it does not establish whether an individual has one. FPG has a different vulnerable group. Someone with significant post-meal glucose elevation may have a normal fasting result, and an atypically low fasting day can also obscure a longer-term problem. For a patient who recognizes any of these circumstances, that self-knowledge is a reason to request both tests or discuss glucose-based follow-up.

Why running both tests together gives a more complete metabolic picture

For a proactive patient seeking a fuller assessment, requesting HbA1c and FPG together is a better default than treating them as substitutes. The combined strategy in the Duong analysis ranked highest on sensitivity when either HbA1c ≥ 6.5% or FPG ≥ 126 mg/dL counted as a positive screen. It found more true cases at a modest cost to specificity. Finding more cases also means some positive screens will need more checking before they can be accepted as a firm diagnosis.

Agreement supports greater confidence in the classification, while disagreement directs attention toward timing, fasting conditions, red blood cell biology or glucose regulation that neither test fully captures. An OGTT is the appropriate next step when results disagree, rather than choosing one test's verdict by preference. Even agreement, however, does not turn these two measurements into a complete account of glucose metabolism.

For patients who want to know about earlier changes, adding fasting insulin can help them look beyond glucose thresholds. Insulin resistance can be present and rising for years before either HbA1c or FPG crosses a diagnostic boundary. It addresses a different question from established diabetes screening. Its potential role is as an early signal of metabolic change.

Practical barriers are lower than two test names suggest. Both can be obtained from the same blood draw without an additional visit, although fasting is still required for FPG. The 2025 decision-model cost-effectiveness analysis published in BMC Health Services Research found that point-of-care HbA1c was cost-effective relative to FPG in a low-resource setting. That finding gives weight to how easy a test is to get, but it does not show that running both tests together is cost-effective in every setting.

How to read your own results and decide when to go further

The first questions to ask are which test was used and whether anything about your health could distort it. Look more closely at an HbA1c result if you have iron deficiency, a hemoglobin variant or a condition that affects red blood cell turnover. In that situation, request FPG or discuss an OGTT with your clinician as a check. For FPG, consider whether you completed the fast and whether illness or stress could have affected the draw. The number needs its biological and practical context.

At the diabetes thresholds discussed here, HbA1c ≥ 6.5% and FPG ≥ 126 mg/dL are positive screening signals even if you feel well or the other test is lower. When both results fall in the same category, whether normal, prediabetic or diabetic, the classification is more reliable. That agreement supports acting with greater confidence in consultation with your clinician. It does not remove either test's limits. Normal fasting glucose and a reassuring average still do not directly test the response to a glucose load.

When the results disagree, avoid choosing the answer that makes you feel more at ease. One normal result does not cancel an abnormal one, and a difference between them alone does not show that the lab made an error. Discuss an OGTT to investigate the gap. If sickle cell trait, thalassemia or another hemoglobin variant is known or suspected, also ask whether hemoglobin screening or an alternative glucose test is appropriate before treating HbA1c as definitive.

A result in the prediabetes range calls for attention. Disagreement is particularly pronounced at this boundary, and prediabetes is an established risk factor for progression to diabetes and for cardiovascular disease. A prediabetes-range result from either test is a reason to request the other before the next annual physical.