Does your hemoglobin A1c level not appear to agree with the average meter readings you get at home? You're not alone.

There are numerous reasons your A1c might appear to be higher or lower than what you were expecting. The most common reason is related to the fact that your A1c reflects an average blood glucose (BG) level. You can have a lot of highs but also a lot of lows and end up with a relatively normal A1c, the same as you'd have if you kept your BG levels normal all the time.

But this isn't the only reason for variation.

The A1c depends on glycation of the hemoglobin in your red blood cells (RBCs). Glycation means adding glucose, and the higher your BGs are, the more glucose you'll add to the hemoglobin.

Anything that affects the lifetime of your RBCs, which are assumed to live 120 days, will affect the A1c. If you give blood or have some kind of internal bleeding, or if you have a hemolytic anemia, you will lose some of the older RBCs cells with a lot of glycated hemoglobin, and your body will make new RBCs with unglycated hemoglobin, so the percentage of glycated hemoglobin will be lower, and your A1c will be lower.

Conversely, if you have spleen damage or no spleen at all, your body will take longer to remove old RBCs from your body, because the spleen is where this housekeeping chore normally happens.

In addition, individuals may have average RBC lifetimes that are different from normal.

Finally, different people may glycate hemoglobin at different rates because of individual variation in the enzymes involved.

Numerous studies throughout the years have investigated this problem and shown that the A1c does not always match average BG levels. Recently a  report, titled Does A1c consistently reflect mean plasma glucose? was published in the Journal of Diabetes.

They had type 2 patients treated with insulin measure BG 7 times a day: before and 2 hours after breakfast, lunch, and dinner, and again at 3 a.m. They said previous studies showed that this was enough to determine average BG. The study would have been better if they'd wired the patients up with continuous monitors, but they didn't do that.

What they found was that the average BGs in the groups matched the A1cs pretty well. (Formulas for converting BG and A1c differ slightly depending on the formula you use. There's a handy converter here that will give you results on the basis of 2007 ADAG [A1c-Derived Average Glucose] study group recommendations.)

But there was a lot of variation among the patients. For example, in the group that had A1cs between  6.5 and 7.5, the average home-tested BG among  260 patients was 142. With the A1c calculator mentioned, the range should have been 140 to 169.

Patients in the 90^th percentile of that group had an average BG of 171, which agrees well with the top of the range you get with the calculator. But patients in the 10^th percentile had an average BG of only 115.