I don't understand. Are you saying it can be manipulated in some way? Something along the lines of: If a perfectly healthy athlete munches on gummy bears all day it will result in a higher a1c?
HbA1C is an average (a “mean” in statistics). There is a companion piece of information used in statistics called a “standard deviation”, which indicates the variability of the data sample. The HbA1C does not provide any way to measure the standard deviation, which means that wild gyrations cannot be accounted for.
The average value for the data set {48, 49, 50, 51, 52} is 50.
Similarly, the average value for the data set {10, 20, 50, 80, 90} is 50, but the standard deviation is over 35 (as opposed to 1.58 for the other data set).
2/3 of all data values are expected to fall within one standard deviation of the mean. 98 or 99 percent of all values are expected to fall within two standard deviations of the mean. The variability of the second data set means that the average value is unreliable for predicting any particular instantaneous value.
Now, with respect to blood sugar, consider that non-diabetics are expected to have a low variability in their sugar readings. Therefore, the HbA1C’s ignoring the variability is of no real concern.
But a diabetic who varies wildly around a “normal” mean value will be shown by the same HbA1C test to fall within the expected range - thereby missing the diagnosis. Because of the variability, the instantaneous reading is a better option for diagnosing the issue.
That’s what I meant by “fooling” the test - the fact that highly variable data can invalidate the outcome. My recent 6.6% reading (approximately 158 mg/dl average) directly conflicts with the 205 mg/dl fasting sugar average (with a standard deviation of 45). In fact, the HbA1C fell below the range in which two-thirds of all of my sugar readings are expected to fall - and this was the fasting sugar.
That means that I was experiencing wild swings around a high but relatively benign sugar reading. With some of the numbers I was seeing, I had to be going seriously low at least once a day.
The bottom line to me is that an average without variability info is incomplete data. There may be mitigating factors due to the biology of the situation, but I’m not convinced this is the best route forward.
Of course, I’m just one guy expressing his opinion, too! ;-P