What Is a Hemoglobin A1C Test and What Do Results Mean?

A hemoglobin A1C test measures the percentage of your hemoglobin (the oxygen-carrying protein in red blood cells) that has glucose stuck to it. Because red blood cells circulate for roughly two to three months before the body replaces them, the test gives a weighted snapshot of your average blood sugar over that window. Doctors use it both to diagnose diabetes and prediabetes and to track how well blood sugar is being managed over time. The result is reported as a simple percentage, with higher numbers meaning higher average blood sugar.

How Glucose Attaches to Hemoglobin

Glucose floating through your bloodstream naturally latches onto hemoglobin in a two-step chemical reaction. First, glucose loosely bonds to the hemoglobin molecule, forming what chemists call a Schiff base, an unstable intermediate. That weak bond can break apart. But if the glucose stays attached long enough, a second rearrangement locks it in place permanently, creating the stable form known as HbA1c.¹ This process does not require any enzymes; it happens passively whenever glucose and hemoglobin are in contact. The higher your blood sugar runs on an average day, the more hemoglobin gets permanently glycated, and the higher your A1C percentage climbs.

This discovery dates back to the 1960s, when researcher Samuel Rahbar noticed an unusual fast-moving hemoglobin band that appeared consistently in blood samples from people with diabetes. That “minor abnormal hemoglobin” turned out to be HbA1c, and identifying it eventually transformed diabetes care by giving clinicians a reliable way to judge long-term blood sugar control.²

What the Numbers Mean

In the United States and many other countries, results are reported as a percentage. The standard clinical thresholds look like this:

Below 5.7%: Considered normal. Your average blood sugar over the past few months has been in a healthy range.
5.7% to 6.4%: Prediabetes. Blood sugar has been running higher than optimal, and you’re at increased risk of developing type 2 diabetes.
6.5% or higher: Diabetes. This cutoff, when confirmed on a second test, is one of the criteria used for a diabetes diagnosis.

These thresholds are widely used, but they are not universally perfect. Research in different populations has found that optimal cutoffs can shift. A large study in Shanghai, for example, found that an A1C of 6.0% was the best diagnostic threshold for newly diagnosed diabetes in that population, and 5.6% for prediabetes, both slightly different from the standard American cutoffs.³ The thresholds are guidelines, not laws of nature, and your doctor may interpret your result in the context of other tests and risk factors.

Translating A1C Into an Average Blood Sugar

One of the more practical things about A1C is that it can be converted into an “estimated average glucose,” or eAG, which is expressed in the same units you’d see on a home glucose meter. A landmark study that correlated A1C values with continuous glucose measurements developed the formula that labs and clinics now use for this conversion.⁴ As a rough guide:

A1C of 5%: Average blood sugar around about 97 mg/dL
A1C of 6%: Around 126 mg/dL
A1C of 7%: Around 154 mg/dL
A1C of 8%: Around 183 mg/dL
A1C of 9%: Around 212 mg/dL

These translations help put the percentage into everyday terms. If your A1C is 7%, your blood sugar has been averaging roughly 154 mg/dL over the past couple of months. The conversion is not exact for every individual, but the correlation is strong enough that many lab reports now print the eAG alongside the A1C.

Why the Window Is About Three Months

Red blood cells have a finite lifespan. Once a mature red blood cell enters the bloodstream, it typically survives around 100 to 120 days before it is broken down and replaced. Since glycation accumulates over the life of each cell, the A1C reflects an average weighted toward more recent weeks. Blood sugar levels from the past 30 days contribute more to the result than levels from 90 days ago, because newer red blood cells have had less time to accumulate glucose on their hemoglobin while still being present in larger numbers.

This three-month window is why doctors usually repeat A1C tests every three months when monitoring diabetes management. It’s also why a single week of very high or very low blood sugar won’t dramatically swing the result the way a fasting glucose test might change from one morning to the next.

When the Test Can Be Misleading

A1C is reliable for most people, but certain conditions can push the number up or down independent of actual blood sugar levels. Recognizing these scenarios matters because a falsely high or low result can lead to the wrong treatment decision.

Conditions That Push A1C Higher Than Expected

Iron deficiency anemia is one of the most common culprits behind a falsely elevated A1C. In iron-deficient individuals, studies have found A1C readings that are meaningfully higher than what their blood sugar levels would predict. One study of diabetic individuals with controlled blood glucose found that those with iron deficiency had an average A1C around 6.8%, with even higher values in women and in those with borderline fasting glucose levels.⁵ The mechanism appears to involve changes to hemoglobin’s structure and increased oxidative stress that accelerate glycation.⁶ Other conditions that can inflate the result include vitamin B12 or folate deficiency, chronic kidney disease, and excessive alcohol use.

Conditions That Push A1C Lower Than Expected

Anything that shortens the lifespan of red blood cells will lower A1C, because the hemoglobin has less time to accumulate glucose. This includes hemolytic anemias, recent major blood loss, blood transfusions, and conditions like sickle cell disease that destroy red blood cells prematurely. A shorter red blood cell lifespan means less total exposure of hemoglobin to glucose, leading to a falsely reassuring result.⁷ Pregnancy also lowers A1C, partly because of increased red blood cell turnover and a naturally expanded blood volume. Research on using A1C to diagnose gestational diabetes has found poor sensitivity, with one study reporting that an A1C cutoff of 5.8% caught only about 17% of gestational diabetes cases, even though the specificity was high.

Even among people without any diagnosed blood disorder, natural person-to-person variation in red blood cell lifespan is wide enough to shift A1C results in clinically meaningful ways.⁸ Two people with the same average blood sugar can have noticeably different A1C results simply because one person’s red blood cells live a bit longer than the other’s.

Hemoglobin Variants and Lab Methods

Hemoglobin comes in different genetic forms. The most common are HbS (associated with sickle cell trait), HbC, HbD, and HbE. These variants can interfere with certain laboratory techniques used to measure A1C, either inflating or deflating the result depending on the assay. Studies comparing lab methods have found that some commonly used systems produced clinically significant errors in the presence of these variants.⁹ Other methods, particularly certain immunoassay-based tests, are less affected.¹⁰

This is not a niche concern. Hemoglobin variants are common in people of African, Mediterranean, and Southeast Asian descent. If you carry a variant and your A1C results seem inconsistent with your home glucose readings or your symptoms, it’s worth asking your doctor whether the lab method used could be affected. Labs that are aware of a variant can often switch to an assay that gives accurate results, or your doctor may rely on alternative markers instead.

How Age and Ethnicity Affect Results

Even after accounting for actual blood sugar levels, A1C tends to be slightly higher in older adults and in certain racial and ethnic groups. One analysis found that A1C increases by about 0.10% per decade of age in people with normal glucose tolerance, independent of glucose levels.¹¹ Non-Hispanic Black individuals showed A1C levels roughly 0.21% higher than non-Hispanic white individuals at the same glucose levels, and the gap widened further in those with prediabetes.

A separate large study similarly found that at the same fasting glucose level, Black individuals had a higher A1C than white individuals, with a difference of about 0.2 percentage points.¹² These differences are not thought to reflect worse glucose control. Instead, they likely arise from biological differences in glycation rates, red blood cell turnover, or other non-glycemic factors. The practical implication is real: using the same 6.5% cutoff across all populations may over-diagnose diabetes in some groups and under-diagnose it in others.

Why A1C Targets Are Not One-Size-Fits-All

For most adults with diabetes, guidelines suggest aiming for an A1C below 7%. But that target is deliberately flexible. In younger adults who were recently diagnosed and have no major complications, the goal might be tighter, perhaps below 6.5%. In older adults with multiple chronic conditions, frailty, or a history of dangerous low blood sugar episodes, guidelines recommend relaxed targets to reduce the risk of hypoglycemia and other adverse events.¹³

The reason is that pushing A1C very low with aggressive medication comes with tradeoffs. While lower A1C clearly reduces the risk of microvascular complications like eye disease and kidney damage, the relationship with heart disease and mortality is more complex. Research suggests that every 1% increase in A1C above a threshold is associated with roughly a 38-40% higher risk of both large- and small-vessel events and death.¹⁴ But other evidence shows that lowering A1C reduces microvascular complications more consistently than it reduces cardiovascular events.¹⁵ For someone who is 80 years old and taking multiple medications, chasing a very low A1C could mean frequent dangerous blood sugar lows without a clear payoff for heart protection.

How Lifestyle and Medication Move the Number

If your A1C comes back in the prediabetes range, the first-line recommendation is almost always lifestyle change: more physical activity, a diet lower in refined carbohydrates, and weight loss if needed. These interventions can bring A1C down meaningfully. Adding metformin on top of lifestyle changes provides an additional modest benefit. A meta-analysis of randomized trials found that combining metformin with lifestyle modifications produced a statistically significant drop in A1C compared with lifestyle changes alone at three and six months, though the difference was no longer significant at 12 months, suggesting the two strategies converge over time.¹⁶

For people with newly diagnosed type 2 diabetes, metformin typically produces a larger A1C drop. A comparative study found that patients started on metformin saw their A1C fall by about 1.5 percentage points, while those using lifestyle changes alone saw a drop of about 1.0 percentage point.¹⁷ Other diabetes medications, including sulfonylureas, GLP-1 receptor agonists, and insulin, have their own expected A1C-lowering effects, which is why doctors choose among them partly based on how far the A1C needs to come down.

A1C vs. Continuous Glucose Monitors

Continuous glucose monitors (CGMs) have become increasingly popular, and they report a metric called the Glucose Management Indicator (GMI), which estimates what your A1C “should” be based on your sensor glucose readings over 14 or more days. In theory, GMI and lab-measured A1C should agree. In practice, they frequently don’t. Research has found that the two values differ by more than 0.3 percentage points in about half of people, and by 0.5% or more in a large share of users.¹⁸

A real-world analysis confirmed the scale of this mismatch: only about 11% of patients had a discordance smaller than 0.1%, while roughly half had a gap of 0.5% or more. The discordance was even larger in people with advanced kidney disease.¹⁹ A smaller clinic-based study found that lab A1C was, on average, about 0.34% higher than the CGM-derived GMI.²⁰

If your CGM says your GMI is 6.5% but your lab A1C comes back at 7.2%, the difference is not necessarily an error on either side. The CGM measures glucose in the interstitial fluid over a short window. The lab A1C reflects a different biological process: cumulative glycation over the full life of your red blood cells. Factors like individual glycation rate, anemia, and hemoglobin variants all influence the lab value but not the CGM reading. Rather than treating one as “right,” it is better to use both numbers together. A persistently large gap between your GMI and your lab A1C may itself be useful clinical information, suggesting that one of the confounding factors discussed earlier is at play.

Alternative Blood Sugar Markers

When A1C is unreliable due to anemia, hemoglobin variants, or other interference, doctors can turn to alternative markers. The most widely used are fructosamine and glycated albumin. Both work on the same general principle as A1C: blood proteins get glucose stuck to them in proportion to blood sugar levels. But because albumin and other serum proteins turn over faster than red blood cells, these markers reflect a shorter window, roughly two to three weeks rather than two to three months.²¹

Another marker, 1,5-anhydroglucitol (often abbreviated 1,5-AG), works differently. It is a naturally occurring sugar whose blood level drops when glucose spills into the urine, making it a good indicator of recent glucose spikes rather than average glucose. These alternatives are not interchangeable with A1C. They each have their own quirks and conditions that interfere with accuracy. But they can be invaluable when A1C alone cannot be trusted, and they sometimes reveal information that A1C misses, like frequent post-meal glucose spikes that get hidden in a normal-looking average.

Point-of-Care Testing

Increasingly, A1C can be measured in the doctor’s office using a rapid point-of-care device rather than sent to an outside lab. These devices use a fingerstick blood sample and return results in minutes, which can make the conversation about treatment changes happen in real time during a single visit rather than requiring a follow-up after lab results come back. Accuracy has been a fair question, but testing of modern point-of-care systems shows results that are closely aligned with reference laboratory methods. One study of a fingerstick-based device found a correlation of 0.99 with the gold standard, with a mean bias of only about −0.02 percentage points.

Lab standardization in general has come a long way. International efforts have brought A1C measurements across different labs and methods into tight agreement. A comparison study evaluating an improved reference method found deviations of only 0.04 to 0.16 percentage points from the international reference.²² For practical purposes, this means a 7.0% from one accredited lab should be very close to a 7.0% from another.

A1C and Brain Health

Beyond the well-known effects on eyes, kidneys, and nerves, higher A1C levels appear to be linked to cognitive decline, and not only in people with diabetes. A large analysis using data from the Health and Retirement Study found that even among people whose A1C was below the diabetes threshold of 6.5%, higher values were associated with steeper memory decline over time.²³

Brain imaging studies tell a similar story. Prediabetes was associated with about a 42% higher odds of cognitive decline, and both prediabetes and diabetes were linked to greater volumes of white matter damage in the brain and lower hippocampal volume, a brain region central to memory formation. Interestingly, people with low-normal A1C showed slightly less white matter damage and slightly larger hippocampal volumes than those in the middle of the normal range.²⁴ The association between A1C and brain health appears to exist on a continuum, not just above or below a diagnostic cutoff. This is one of the more compelling reasons to pay attention to your A1C even if you’ve never been told you have diabetes.