My aim here is to reduce the apparently widespread confusion about what the hemoglobin A1c test actually measures and why we want to measure it. Most explanations of the test that I have read were seemingly written by a very busy and distracted person who once knew what the test was for, but sort of lost track of it somewhere along the way.

For example, it is usually stated that the A1c test measures "average blood glucose". It doesn't really measure that, any more than a bathroom scale measures calories. We can work backwards from the A1c test result to estimate average blood glucose (and probably rather accurately, these days), but that's not the same thing as a measurement, and anyway it's not why the test is important.

If we think we should care about the A1c test result only because of what it reveals about average blood sugar, we are almost missing the point. In fact, that interpretation gets the whole thing a little backwards. It would be nearer the truth to say that we care about average blood sugar mainly because of the impact it has on what the A1c test actually does measure: a chemical reaction known as "glycation".

What it really measures

The A1c test examines hemoglobin, a protein found in red blood cells, to determine what percentage of it is "glycated" (in other words, what percentage of the hemoglobin has glucose bonded to it). Hemoglobin that is in this sugar-coated state is designated by the subtype "hemoglobin A1c", hence the puzzling name of the test. (It's yet one more illustration of the general rule that technical experts should never be allowed to choose the name of anything, ever.)

The result of an A1c test is a percentage. The percent sign is often omitted on lab reports, apparently for the sake of holding clarity to a minimum, but just pretend it's there. If your result says 6.2, that means 6.2% (in other words, 6.2% of your hemoglobin is glycated).

The glycation process (that is, the bonding of glucose onto proteins) occurs spontaneously whenever proteins (any proteins, not just hemoglobin) are exposed to glucose. Because blood always has glucose in it, a certain amount of glycation is always taking place throughout the body. Nobody ever got a result of 0% on an A1c test.

However, under normal conditions, glycation is a very slow and steady affair. The process accelerates whenever your glucose level rises, but if you stay within normal glucose limits it doesn't accelerate very much. Consequently, the fraction of hemoglobin that is glycated is small. In most healthy people, the fraction is in the range of 4 to 6%.

Surprising as it may seem, this fraction is stable over time; it normally doesn't grow. Obviously that fact needs some explaining. I mean, if there is always glucose in the bloodstream, and it keeps causing glycation, you would think the percentage of glycated hemoglobin would keep increasing over time, from birth to death. Why doesn't it?

The percentage doesn't grow because the body's maintenance processes work day and night to undo the glycation process. The body recycles its proteins over time, replacing old (and possibly glycated) proteins with fresh new ones. Because of this continuous renewal process, the percentage of protein that is glycated doesn't keep rising higher and higher over time. For a healthy, non-diabetic individual, the A1c result is not going to rise above 6%.

However, if the glycation process becomes accelerated, the body's maintenance processes will not be able to keep up with it. Proteins will become glycated faster than your body can replace them. If your glucose levels are routinely above the normal range, so that glycation is routinely accelerated, the result will be that the glycated fraction of your hemoglobin will become larger. For this reason, people with diabetes usually get a result greater than 6%, and sometimes greater than 10%. (14% is not unheard-of.) The size of this fraction makes a big difference in terms of long-term health; you want it to be in the normal range, or as close to the normal range as you can possibly get it.

Why hemoglobin?

The reason we measure hemoglobin (rather than one of the thousands of other proteins found in the human body) is not that hemoglobin is uniquely important. Hemoglobin just happens to be a convenient protein to choose when you're measuring glycation. It is an easily-collected, easily-measured protein which, when you find out how much of it glycated, gives a good indication of how much glycation in general (involving all proteins) is going on throughout the body.

Also, the results of a hemoglobin test reflect a comparatively narrow slice of recent history (about 3 months). 

Why 3 months?

Red blood cells only last about 3 months before the body replaces them. Therefore, whatever was going on in your blood more than 3 months ago, it won't be reflected in an A1c test result, because none of your hemoglobin dates back that far. In effect, the A1c test tells you how much glycation has been going on over the past 3 months, but it tells you nothing about what was going on before that.

Also worth mentioning: those three months, considered individually, do not have an equal influence on the test result. The reason is that the recycling process is continuous; it doesn't happen at scheduled intervals. By the time you take the test, a lot of the hemoglobin that was around 2 to 3 months earlier is already recycled, and the same is true of a fair amount of the hemoglobin that was around 1 to 2 months earlier.

The bottom line is that the A1c test is "weighted" toward the most recent weeks of the 3 month period prior to the test. If your diabetic control was very good 2 to 3 months ago, but very bad over the past month, the test result will reflect your recent lapse more strongly than it will reflect your earlier success. Conveserly, if your control improved markedly in the past few weeks, the test result will look better than it would if it reflected the actual average over the past three months.

It would be pointless to test A1c every day, because the result is cumulative and it wouldn't change that fast. Usually the test is performed 1 to 4 times a year.

Sometimes doctors want to do a glycation test that looks at a more narrow span of time (to evaluate the effect of a new medication, for example). In that case, they use the "fructosamine" test, which looks at glycated albumin in the blood and indicates the glycation rate over a shorter time period (about 3 weeks).

Why glycation?

So why measure glycation in the first place? Because glycation is a large part of the reason that diabetes is bad for you. Glycation of protein is a permanent thing (the sugar doesn't just fall off the protein after a while), and unfortunately it is also a harmful thing.

Proteins are like tools: they have functions to perform within the body, and their molecular shapes are crucial to those functions. Encrusting your proteins with sugar is not a good way to make them work better; in fact, it's about as helpful as dipping your car keys in epoxy.

Over the long run, glycation of proteins is injurious to tissues and organs all over the body. Your blood, after all, reaches everywhere, and if your blood is too sugary, glycation accelerates everywhere, from your eyes to your ankles. Excessive glycation leads to a variety of health problems. Most of the famous "complications" associated with diabetes appear to be caused by glycation (or by secondary reactions which glycation promotes). The faster the rate of glycation, the greater the risk that these complications will develop.

For example, if your A1c test result held steady at just under 7% for a decade, your risk of developing the eye disease known as retinopathy during that time would be only 1 out of 20. The risk would shoot up to 20 out of 20 if your result held steady at 11% for the same period.

Therefore, the amount of glycation going on in your system makes a very big difference to your health, and the point of the A1c test is to find out how much glycation is going on. It's a direct measure of how your diabetes is affecting you. A blood glucose test measures a factor which we're afraid will lead to damage; an A1c test measures the damage itself (and in an early stage, when it's still possible to undo the damage if you know it's occurring).

The good news about glycation and the harm it does is that glycation doesn't produce serious medical consequences overnight. It takes a long time for glycation to result in "complications", and if you're testing your A1c (and working to drive it downward) you have a good window of opportunity to improve the situation.

Blood glucose testing is certainly important, and I don't mean to suggest otherwise, but it is mainly useful (as headlights are useful when you're driving at night) to help keep you steering properly during each little segment of your journey. It's the A1c test that, like a GPS system, tells you where you are.

Dabetes patients are generally advised to drive their A1c result as low as they can, and above all to try to keep it under 7%, because 7% is the threshold at which the risk of serious complications starts to climb significantly. 

How does A1c relate to average glucose?

A rough-and-ready formula was developed years ago for estimating average BG based on an A1c result; according to this formula, an A1c result of 6% was equivalent to an average BG of 135 mg/dl. However, more recent research, using modern diagnostic tools and a more continous measurement process, has concluded that the old formula over-estimated average BG. The ADA has published a new formula based on this research:

(A1c X 28.7) - 46.7 = eAG in mg/dl

By this formula, an A1c result of 6% indicates an average glucose level of 126 mg/dl, not 135. Here are some A1c percentages, and the average glucose values which they are believed to indicate:

A1c result in % eAG in mg/dl
6.0 126
6.5 140
7.0 154
7.5 169
8.0 183
8.5 197
9.0 212
9.5 226
10.0 240

The ADA has been promoting the idea that A1c test results should be converted to "eAG" (estimated average glucose), and reported to patients in that form. That is, you won't be told that your A1c was result was 6% glycated hemoglobin; you'll be told that the result was 126 mg/dl average glucose. The ADA thinks this will mean more to patients because they already do glucose testing in mg/dl on a daily basis.

I don't like this idea, and not just because of my reflexive distaste for the practice of dumbing-down scientific or medical information before sharing it with the public. It seems to me that an average glucose value is not as easy to relate to one's daily testing as the ADA thinks. Glucose values are volatile in the extreme, and we measure them at certain specific points in time (before the first meal of the day, an hour after eating, or whatever). An overall average which includes times when we never test (3 AM, say) is not going to mean much, any more than it would mean much to be told by a GPS system that one's averge speed on a trip which included both city traffic and open freeways was 46 miles per hour.

Anyway, when you make a measurement, the honest thing to do is to report the result that you actually measured, not the estimate of something else which it inspired you to make. The new formula for estimating average glucose from A1c results may be better than what we were using before, but that doesn't mean it's perfect; the formula changed once, and might change again. Rather than report an estimate (as if you measured it) which you might later decide was a mistake, why not just say what you measured?

What I like about the A1c result as we know it today is that its literal meaning can be stated in a very concrete and easily-visualized way ("7.3% of your hemoglobin is encrusted with sugar"). If you don't realize the implications of that, they are fairly easy to discover ("a result of 7% or higher puts you at risk for diabetic complications, so try hard to reduce it").

What the test won't tell you

I should point out that, in one important area, the A1c test does not give diabetes patients much help in figuring out how well they're doing. The heightened risk of a "macrovascular accident" (that is, an arterial blockage, usually by a blood clot, which can trigger a heart attack or stroke) is a diabetic complication which doesn't seem to have much to do with glycation, and doesn't seem to be much helped by a decline in the glycation rate. Bringing down your glucose level and your A1c test result is highly beneficial in other ways, but it provides little or no relief from the cardiovascular risk associated with being diabetic.

Strange as it may seem, heart attacks and strokes, which correlate very strongly with the status of being diabetic, don't seem to correlate with A1c results. In other words, becoming diabetic boosts the risk, but getting your diabetes under control doesn't reduce it, or at least doesn't reduce it much. 

Therefore, there must be something else about being diabetic (entirely apart from elevated glucose and the glycation it causes) that leads to dangerous problems in the arteries. Cardiologists are fairly sure they know what that something is: inflammation.

Chronic inflammation anywhere in the body seems to trigger a lot of other health problems, and inflammation in the arteries is especially dangerous. When the arteries are inflamed, they are highly vulnerable to being wounded (as, for example, when a fragment of cholesterol placque tears loose from the arterial wall), and a wounded artery can easily form a blood clot large enough to block the flow of blood to brain tissue or heart muscle.

Obviously, if you have diabetes, you need to do whatever you can to reduce or avoid inflammation of the arteries. But what is it about being diabetic that causes inflammation? It's a complicated question; maybe it would be better to simply list some important causes of arterial inflammation, and the apparent connection between these things and diabetes.

Cause of Inflammation Relevance to Diabetes
Chronic infection. When a tissue anywhere in the body is is constantly fighting a low-grade infection, inflammation is the result. A typical example is periodontal disease, which is strongly associated with Type 2 diabetes.
The human stress response. Any kind of stress (including time-pressure at work) causes a release of the "stress hormones" adrenaline and cortisol, which promote inflammation. The frustrations and fears of dealing with diabetes could certainly raise a patient's stress level.
High blood pressure. Excessive blood pressure places a constant stress on the arterial walls which result in inflammation. High blood pressure is strongly associated with Type 2 diabetes.
Excess insulin in the bloodstream. "Compensatory hyperinsulinemia" is a tactic the body uses to overcome insulin resistance; unfortunately, insulin (which is chemically similar to the stress hormone adrenaline) has a similar inflammatory effect. This perpetual insulin overdose may help reduce your glucose level, but it inflames the arteries.
All causes... Inflammation releases chemicals which seem to trigger insulin resistance. Type 2 diabetes is strongly associated with various conditions which cause chronic inflammation. It seems to be a vicious circle: inflammation promotes Type 2 (by increasing insulin resistance), and Type 2 promotes inflammation (by increasing insulin output).

Because reducing your A1c result doesn't take care of the inflammation problem, and therefore doesn't take care of the increased cardiovascular risk associated with the disease, you have to address this issue separately, and do something about all the contributors to diabetes listed in the table above.

  • Get adequate dental care or medical care to eliminate any sources of chronic infection.
  • Do whatever you can to reduce stress (exercise, deep-breathing, yoga, or whatever helps you).
  • Do whatever you can to reduce high blood pressure (exercise, weight loss, medication if necessary).
  • Do whatever you can to reduce excess insulin in your bloodstream (by increasing your insulin sensitivity through exercise, weight loss, and medication if necessary).