How accurate are the hemoglobin A1c tests (which I'll call just A1c) that most of us with diabetes get about 4 times a year? The A1c is supposed to tell us what our average blood glucose (BG) level has been over the previous several months, and it has provided information that was lacking in the early days of diabetes treatment.
However, researchers have now begun to dig deeper into the meaning of the A1c test to try to explain why some people's A1c doesn't seem to agree with their daily BG measurements.
We should have an approximate idea of how well we're controlling just by measuring our BG levels with our home meters. But the meters only tell us what the BG level is at the time we test. Even if we test 8 or 10 times a day, there are times when we're not testing, especially overnight.
And not everyone can afford to test even eight times a day. For instance, Medicare thinks that people on insulin need only 100 strips a month, enough to test about three times a day. For people not on insulin, they think 100 strips every three months should be just hunky dory.
Some people are using continuous glucose monitors, which measure BG levels every 5 minutes, tell you what your BG is right now, and produce graphs telling you what your BG was in the past. This sounds wonderful, and it can be, but the meters are expensive (up to $1000 for a starter kit), the sensors cost about $9 or $10 a day, and it's difficult to get insurance to cover them.
So until the prices come down, most of us will have to rely on the A1c to let us know how we're doing overall.
The A1c is based on the nonenzymatic glycation (addition of glucose) of hemoglobin, which is the pigment that makes your blood red and carries oxygen from the lungs to the tissues that need it. The higher the concentration of glucose in your blood, the faster your hemoglobin becomes glycated. And when glucose adds onto the hemoglobin, the reaction is essentially irreversible, so it stays there for the lifetime of the hemoglobin molecule.
When blood is analyzed, scientists can separate it into different fractions, and the A1c is the fraction that contains most of the glucose, which is why it was chosen to indicate your BG control. When your control is poor and your BG levels are high, more hemoglobin becomes glycated, and the A1c is higher.
Hemoglobin is contained within red blood cells (RBCs), and eventually the body gets rid of old RBCs and replaces them with new ones. The average lifetime of a RBC is estimated to be about 120 days, so it is said that the A1c reflects your BG control over the previous 120 days.
In fact, for reasons I won't try to explain, the A1c is heavily weighted toward your recent control, say in the past several weeks, although a few RBCs are the older ones.
But how accurate is the A1c in estimating your real BG control? Many people have complained that their BG control is good but their A1c doesn't reflect that control. Some have fasting levels of 150 or more and go even higher after meals and yet have excellent A1cs. Others have much much lower fasting levels, eat very few carbohydrates so they don't go high after meals, and have much higher A1cs.
There are many variables that can contribute to the A1c value. For instance, people who see-saw between very high BG levels and very low BG levels might come out with intermediate A1c levels, because the A1c reflects the average that your BG has been throughout the previous weeks and months, and the highs balance the lows.
Abnormal hemoglobin varieties can also cause aberrations in the A1c. Some unusual hemoglobins are similar to A1c in chemical properties, and depending on the particular method used to measure the A1c, may show up along with the A1c even when not glycated and give erroneously high results. Other types might not show up with the A1c when they're glycated and give erroneously low results. If you suspect you may have an unusual hemoglobin type, you can have another test, called the fructosamine test, which measures the glycation of another protein in the blood and shouldn't be affected by the hemoglobin type.
Because the lifetime of the RBC is assumed to be 120 days, anything that changes the average lifetime of the cells will affect the A1c. If the cells live a shorter time, the A1c will be lower than expected. If they live a longer time, the A1c will be higher than expected.
One study by a group of researchers at the University of Cincinnati led by Robert M. Cohen found that the average age of RBCs varied from 38.4 to 59.5 days. The researchers estimated that patients with identical BG levels but mean RBC lifetimes of 38.4 and 59.5 days could have A1cs of 7.5 or 9.9, respectively.
"An ostensibly acceptable HbA1c that would not prompt a change in management may actually be unacceptably high if a shorter RBC survival were taken into account, and at the other extreme, the measured HbA1c may correct to one much lower if a longer RBC lifespan is considered. In the latter case, inappropriately aggressive therapy might be instituted, with increased risk for hypoglycemia," wrote the authors.
This study found that the patients with the poorer BG control had longer RBC lifetimes. Two other studies found the same thing, but others found just the opposite, that poor BG control resulted in shorter RBC lifetimes.
Factors that reduce the RBC lifetime include blood loss (for example, internal bleeding or donating blood) and certain types of anemia. Factors that increase the RBC lifetime include damage to the spleen, which is the organ that participates in removing old RBCs.
Genetic factors are most likely another element affecting RBC lifetime. Just as not every man weighs exactly 70 kilograms, has a blood pressure of 120/80 and a temperature of exactly 98.6 degrees F (the statistics for the "standard man" used in calculating a lot of physiological parameters), not everyone's RBCs will live exactly 120 days.
Another factor is thought to be the rate at which people glycate hemoglobin and other proteins. "High glycators" are thought to add glucose much more than "low glycators." This would mean that two people with normal hemoglobin and a RBC lifetime of exactly 120 days could have an average BG level of 110, but one would have a higher A1c than the other.
Now yet-another factor has been proposed by the University of Cincinnati researchers.
According to this research, RBCs can differ in how well glucose crosses the RBC membrane. If the RBCs were very good at taking glucose out of the blood, then the concentration of glucose would be higher than normal within the RBC. And because the amount of glycation of the hemoglobin molecule depends on the local concentration of glucose, the amount of glycation would be increased.
If, on the other hand, the RBCs were not as good at taking glucose out of the blood, then the amount of glycation would be reduced. The researchers hypothesized that the variations in the ability of glucose to cross the membrane in different RBCs could result in approximately 25% variation in the glycation of hemoglobin, or a difference of 1.5 to 2 A1c points for the same BG control.
"We speculate that the inter-individual variation in [glucose uptake by the RBC] may contribute to the observed genetic variation in the incidence of diabetic complications," they wrote.
Does this mean that people whose RBCs are very good at taking up glucose, so the A1c is higher than expected, would be more or less likely to have complications? This is an important question, and unfortunately, no one knows the answer yet. It would depend on whether the cells responsible for various complications, for example, the endothelial cells lining your arteries, reacted in the same way as the red blood cells. This has not yet been tested experimentally.
So what does all of this mean for you? It means that interpreting your A1c value is very complex. Many things can affect the results. You could have RBCs that took up a lot of glucose and endothelial cells that didn't, so a higher-than-expected A1c wouldn't predict complications for you. Or the converse could be true. You might have RBCs that didn't take up a lot of glucose, so your A1c would be lower than expected, but your BG levels could be high and could cause complications.
You could have RBCs that took up a lot of glucose but you could be a "low glycator," so one effect would cancel out the other and you'd have the A1c you'd predict from your BG meter readings. Numerous other permutations and combinations are possible. I'll let you think of them all.
So should we just throw out the A1c test and use our meters instead? Not at all. Numerous studies have shown that on average, or in population studies, the A1c is able to predict complications, especially microvascular complications. And even in people without diagnosed diabetes, on average, people with higher A1c values, even in the so-called normal ranges, have more heart attacks than people with lower A1c values.
The A1c is still a valuable test.
We don't know yet which is more important: the A1c or average BG levels as measured by BG meters. If the glycation rate of hemoglobin corresponds to complication rates, then it's actually more important than your actual BG levels. But if you were a "high glycator" whose RBCs took up a lot of glucose and lived longer than average, so you had a very high A1c compared with your average BG levels, trying to bring your A1c down to a better level might result in too much hypoglycemia.
On average, people with lower A1cs get fewer complications. But there are always exceptions to these averages. Some people have excellent A1cs and get complications. Other people have lousy A1cs and don't. Some people have recently suggested that variations in BG levels (going very high and then very low, resulting in intermediate A1cs) may be more important than steady higher BG levels, although this idea is still controversial.
So there's a lot we don't know. "People make a lot of assumptions and inferences that seem plausible to them, but when you look more carefully, you really have to see what has been proven and what not," Cohen told me. His group estimates that even if the A1c matches average BG levels in 85% of the population, the "one size fits all" concept would not apply to about 3 million people in the United States.
The A1c is not perfect. But for now it's pretty much all we've got.
However, if your A1c doesn't match your home BG readings and your doctor thinks you're cheating or just not testing when you're high, discuss the other possibilities with the doctor. If the doctor doesn't listen, find another doctor.
In general, except during illnesses that might affect the RBC lifetime, all these factors should be constant within one individual. So what you want to monitor is how your A1c is changing, or not changing, with time or with different treatments.
Don't compare your A1c with someone else's. They might have different RBC lifetimes, or their RBCs might take up more or less glucose than yours, or they might be higher or lower glycators. And if your A1c is higher or lower than you expected on the basis of your BG readings (assuming you're measuring both before and after meals), accept that something in your physiology is a little different from the average. Learn what's normal for you when you have good BG control and make sure that number doesn't go up.
Even better, see if you can get it to go down without hypoglycemia.