An A1C test measures the percentage of your hemoglobin (the oxygen-carrying protein in red blood cells) that has glucose permanently attached to it. The higher your blood sugar has been over the past two to three months, the higher your A1C percentage. A normal result is below 5.7%, prediabetes falls between 5.7% and 6.4%, and diabetes is diagnosed at 6.5% or above.
How Glucose Attaches to Hemoglobin
The process behind an A1C reading is called glycation, and it happens continuously in your bloodstream without any help from enzymes. Glucose floating in your blood bumps into hemoglobin molecules and bonds to a specific spot on the protein’s surface, forming an unstable connection first. That unstable link then rearranges into a permanent bond that lasts for the entire life of the red blood cell. Once glucose locks on, it doesn’t let go.
This is why A1C works as a long-term average rather than a snapshot. A single blood sugar reading tells you what’s happening right now. A1C tells you how much glucose has been sticking to your hemoglobin over weeks and months, because every red blood cell in your sample carries a chemical record of the sugar it was exposed to during its lifetime.
Why It Reflects Two to Three Months
Red blood cells live about 106 days on average, with a mean age of roughly 53 days at any given time. When a lab draws your blood, the sample contains red blood cells of all ages: brand-new ones with very little glucose attached and older ones that have been accumulating glucose for months. The A1C result is essentially a weighted average across all of those cells.
Because younger red blood cells are more numerous (your body constantly produces new ones), more recent blood sugar levels have a slightly stronger influence on the result than levels from three months ago. Roughly half the A1C value reflects the most recent 30 days, with the remainder shaped by the weeks before that. So a big change in your diet or medication over the past month will start to move the number, but it won’t fully show up until the next test.
What Happens in the Lab
Several laboratory methods can measure A1C, and they approach the problem differently. The gold standard for many years has been high-performance liquid chromatography, or HPLC. This technique separates hemoglobin molecules by their electrical charge: glycated hemoglobin has a slightly different charge than non-glycated hemoglobin, so the machine can sort them apart and calculate the percentage that carries glucose.
Other common methods include immunoassays, which use antibodies designed to recognize the specific spot where glucose attaches, and enzymatic assays, which trigger a chemical reaction with the glycated portion and measure the result. A fourth approach, boronate affinity, exploits the fact that a certain compound binds specifically to sugar molecules on hemoglobin, pulling out only the glycated fraction.
All of these methods should give you essentially the same number, thanks to a global standardization system. In the U.S., the National Glycohemoglobin Standardization Program (NGSP), established in 1996, ensures that routine lab results are traceable to the landmark clinical trials that defined what A1C levels mean for diabetes risk. Internationally, a reference system developed by the IFCC serves as the anchor for worldwide calibration. Results reported in the familiar percentage format (the NGSP/DCCT scale) can be converted to the international mmol/mol scale using a standard equation.
Finger-Stick Tests vs. Lab Draws
Point-of-care A1C devices, the kind used in some doctor’s offices or pharmacies for a quick finger-stick result, are convenient but less reliable than a full laboratory test. A 2017 review of 61 studies comparing 13 different point-of-care devices found that every single one showed some degree of bias. Nine of the thirteen tended to read lower than the lab result, and four read higher. The variability was especially poor at lower A1C concentrations, right around the cutoffs that matter most for diagnosis.
For this reason, the American Diabetes Association and other organizations recommend that A1C used to diagnose diabetes should come from an accredited laboratory, not a point-of-care device. The World Health Organization says point-of-care testing should only be used for diagnosis where it’s the only option available or where a rigorous quality program is in place. If your diagnosis or treatment decision hinges on a borderline result from a finger-stick test, a follow-up lab draw is a good idea.
Conditions That Can Skew Results
Because A1C depends on both blood sugar levels and red blood cell lifespan, anything that changes how long your red blood cells survive can throw off the reading. Iron-deficiency anemia, for example, tends to push A1C higher because red blood cells live longer and accumulate more glucose. Conditions that destroy red blood cells faster than normal, like sickle cell disease or significant kidney disease, can pull A1C artificially low because the cells don’t live long enough to collect their usual share of glucose.
Hemoglobin variants also matter. Hundreds exist, but the four most common are hemoglobin S (sickle cell trait), hemoglobin E, hemoglobin C, and hemoglobin D. Depending on which lab method is used, these variants can interfere with the measurement itself, producing falsely high or falsely low numbers. If you carry a hemoglobin variant, your doctor may need to use a specific testing method known to be unaffected, or rely on alternative markers like fructosamine to track your blood sugar.
Liver failure, heavy bleeding, recent blood transfusions, and pregnancy can also alter red blood cell turnover enough to make A1C unreliable.
Translating A1C Into Average Blood Sugar
One of the most useful things about A1C is that it can be converted into an estimated average glucose (eAG), the number you’d see on a blood sugar meter. The conversion formula is straightforward: multiply your A1C by 28.7, then subtract 46.7. The result is your estimated average in mg/dL.
- A1C of 5.7%: estimated average glucose of about 117 mg/dL
- A1C of 6.5%: estimated average glucose of about 140 mg/dL
- A1C of 7.0%: estimated average glucose of about 154 mg/dL
- A1C of 8.0%: estimated average glucose of about 183 mg/dL
This conversion helps bridge the gap between the percentage your lab reports and the daily numbers you might check at home. Keep in mind that “average” means some readings were higher and some lower, so two people with the same A1C can have very different daily glucose patterns.
How Often to Get Tested
If you have diabetes and your blood sugar is well controlled, testing twice a year is typical. If your levels aren’t at target or you’ve recently changed medications, your doctor may check it every three months. Since the test reflects roughly a three-month window, testing more frequently than that won’t give you meaningfully new information. For people without diabetes, A1C is commonly part of routine screening, especially after age 45 or if other risk factors are present.