The Normal A1C Level

by David Mendosa Patient Advocate

You want to control your diabetes as much as possible. You wouldn't be reading this if you didn't.

So you regularly check your A1C level. This is the best measurement of our blood glucose control that we have now. It tells us what percentage of our hemoglobin -- the protein in our red blood cells that carry oxygen -- has glucose sticking to it. The less glucose that remains in our bloodstream rather than going to work in the cells that need it the better we feel now and the better our health will continue to be. Less glucose in the bloodstream over time leads to lower A1C values.

As we are able to control our diabetes better and better, the reasonable goal is to bring our A1C levels down to normal -- the A1C level that people who don't have diabetes have. But before we can even set that goal, we have to know what the target is.

The trouble with setting that target is that different experts tell us that quite different A1C levels are "normal." They tell us that different levels are normal -- but I have never heard of actual studies of normal A1C levels among people without diabetes -- until now.

How are 'normal' A1C levels defined?

The major laboratories that test our levels often say that the normal range is 4.0 to 6.0. They base that range on an old standard chemistry text, Tietz Fundamentals of Clinical Chemistry. The Diabetes Control and Complications Trial or DCCT, one of the two largest and most important studies of people with diabetes, said that 6.0 was a normal level. But the other key study, the United Kingdom Prospective Diabetes Study or UKPDS, which compared conventional and intensive therapy in more than 5,000 newly diagnosed people with type 2 diabetes, said that 6.2 is the normal level.

Those levels, while unsubstantiated, are close. But then comes along one of my heroes, Dr. Richard K. Bernstein, the author of the key text of very low-carb eating for people with diabetes, Dr. Bernstein's Diabetes Solution. Dr. Bernstein himself developed type 1 diabetes in 1946 at the age of 12.

"For my patients...a truly normal HgbA1C ranges from 4.2 percent to 4.6 percent," he writes on page 54 of the third edition of that book. "Mine is consistently 4.5 percent." Then in his July 30, 2008, telecast he reiterated that as far as he has been able to determine, a normal A1C is 4.2 to 4.6.

What Dr. Bernstein says is normal is so at odds with the other experts that at least a year ago I determined to find scientific proof of what a normal A1C level actually is. It turned out to be a lot more difficult to find than I ever imagined.

My personal quest for a normal A1C level

When I learned in 1994 that I had diabetes and that my A1C level was 14.4, I was gradually able to bring it way down. Lately I have been doing everything I can think of to try to get my A1C down to normal. But in 2008 my level in nine separate A1C tests always ranged from 5.2 to 5.6. That's far from normal, according to Dr. Bernstein. My favorite Certified Diabetes Educator is also doing everything she can to get a normal A1C level. And she doesn't even have diabetes -- which she double-checked by taking a glucose tolerance test -- but her most recent A1C was 5.4.

What could we be doing that is so wrong? Each of us is thin, eats a very healthy diet, exercises a lot, takes care of our teeth and gums, which is a major source of infection. Could we have other infections or stresses that prevent us from getting our A1C levels down to "normal"?

It turns out that my favorite Certified Diabetes Educator and I have normal A1C levels after all. I learned this just yesterday when I finally tracked down actual research determining what normal levels are.

Setting the A1C standard

A friend suggested that I contact the people who run the standardization program for A1C testing. This organization affiliated with the University of Missouri is the NGSP. Those initials used to stand for the National Glycohemoglobin Standardization Program. But now that the NGSP is international, they changed the name.

So I called Curt Rohlfing, the NGSP data manager and technical writer/research analyst at the University of Missouri. And finally hit pay dirt in my quest for learning what a normal A1C is.

Curt told me that every three or four years his lab at the university studies a group of people who don't have diabetes to scientifically determine what a normal A1C level is. The results from one study to the next are always close, Curt told me. In their most recent study they tested 29 people who lived nearby in central Missouri.

I asked how they knew if the people they tested didn't have diabetes. "Because we did fasting glucose tests on them, they had no prior history of diabetes, and none of them were obese," Curt replied.

So what were their levels? They ranged from 4.5 to 6, Curt replied. That's at plus or minus 3 standard deviations. I am certainly no statistician. But Curt tells me that it includes about 99 percent of the values.

The range is narrower -- 4.7 to 5.7 -- at plus or minus 2 standard deviations. This includes about 95 percent of the values.

"The upper limit is the more important one," Curt explained further. "The lower limit doesn't convey as much meaning."

They also see "a little skew toward the high end of the range, a bit of tailing at the high side," Curt continued. In fact, levels below 4.5 are "quite unusual," and usually are only when people have anemia or other abnormalities of the red blood cells.

How to set a reasonable A1C target

Remember these are the ranges obtained by the people who set the standards for A1C tests. Sadly, however, not every laboratory or home test kit meets those standards. Maybe the lab that Dr. Bernstein uses doesn't. Does yours? Curt suggests that you ask your doctor if the lab running the test uses a method that is certified by the NGSP. The first conclusion of the research for me is that we need to shoot for a normal A1C level of no more than 6.0 instead of trying what may be impossible, a level of 4.2 to 4.6.

However, an A1C level of 6.0 can cause people who take insulin injections or one of the sulfonylureas to go hypo. That's why the American Diabetes Associations sets the goal conservatively at 7.0 and the American Academy of Clilnical Endocrinologists uses 6.5. People with diabetes whose A1C values creep over these cuttoff values are at increasesd risk of developing long-term complications, such as nerve and eye damage.

Still, a lower A1C level among people who take those medications is possible without hypos. Dr. Bernstein has amply shown that both in his own life and that of thousands of his patients.

And certainly, for those of us who don't take insulin injections or one of the sulfonylureas we can set our goal even lower. That's because we have to understand the difference between normal and optimal. For example, two-thirds of all American adults are overweight. Thus it has become normal in our culture to be overweight. Likewise, the average American gets little exercise, and that is also normal. We know that being a chubby couch potato isn't optimal.

"I'm going to aim to be in the lower end of the normal A1C range," my favorite CDE tells me, "because that is what I believe is optimal for human health." And now that I know my A1C is in the normal range I am still going to do my best to bring it down as much as possible. Are you?

David Mendosa
Meet Our Writer
David Mendosa

David Mendosa was a journalist who learned in 1994 that he had type 2 diabetes, which he wrote about exclusively. He died in May 2017 after a short illness unrelated to diabetes. He wrote thousands of diabetes articles, two books about it, created one of the first diabetes websites, and published a monthly newsletter, “Diabetes Update.” His very low-carbohydrate diet, A1C level of 5.3, and BMI of 19.8 kept his diabetes in remission without any drugs until his death.