IS HbA1c considered interval or ratio measurement?

Daytonite, BSN, RN · Mar 18, 2008

hbg-a1c, or glycosylated hemoglobin, measures the average combination of glucose and hemoglobin in the lifespan of an rbc. an rbc has a lifespan of 3 months or 120 days. the blood sample, however, will contain blood cells that will be from 1 to 120 days old, so the test will be an average of the glycosylated hemoglobin these cells contain. 4 to 5.5% is considered normal per my reference for hbg-a1c (page 707, davis's comprehensive handbook of laboratory and diagnostic tests with nursing implications, 2nd edition, by anne m. van leeuwen, todd r., kranpitz and lynette smith). to answer your question, it can never be zero. the amount of glycosylated hemoglobin depends on the amount of glucose that was available in the bloodstream over the lifespan of each individual rbc. that is why the value obtained is an average of all the rbcs in the sample that is tested. therefore, the more glucose the rbcs were exposed to is going to give you a higher result. the result could never be zero because there is always glucose in the blood or the patient would be dead dead. can anyone live with a blood glucose of o? i don't think so. they'd be dead from the seizures they'd be having in the first place.

you can find lab information in the weblinks on this thread if you don't have a lab reference book (bookmark this thread):

https://allnurses.com/forums/f205/medical-disease-information-treatment-procedures-test-reference-websites-258109.html - medical disease information/treatment/procedures/test reference websites

here are two websites with information about the hba1c:

llg, PhD, RN · Mar 18, 2008

To answer your question, it can never be zero. ... The result could never be zero because there is always glucose in the blood or the patient would be dead dead. Can anyone live with a blood glucose of O? I don't think so. They'd be dead from the seizures they'd be having in the first place.
]

Actually, for the purpose of determining the level of the measurement ... the fact that the person would be dead is irrelevant. What counts is the type of measurement being made. If there could be a theoretical zero ... and if the numbers obtained represent true quantities, then it is a ratio level measurement. For example, a result of 8% means there is actually twice as much of something as a result of 4%. A result of 6% actually corresponds to twice as much as indicated by a result of 3%.

That's why the category for that level of data is called "ratio level" data. The numbers represent true amounts of something and true relationships can be expressed mathematically as ratios.

The classic example of a measurment that appears to be ratio level that is NOT is standard temperature. On either a Fehrenheit or Celcius scale, the number 0 does not represent a "true absence of any temperature." Zero simply designates a chosen spot on a scale that goes both further down and futher up from that point. "Absolute zero" is actually -273 on the Celcisu scale -- far, far, below the 0 on a standard thermometer. (That's the point at which all molecular motions stops -- and heat is actually a measurement of energy associated with molecular motion. At absolute zero, all motion stops and there is actually 0 heat.) However on one of the more commonly used scales, 0 is simply a point on the contiuum chosen for convenience. On the Celcius scale, 0 is the point at which water freezes. Things can get a lot colder than that. Something with a temperature of 0 does not have zero heat/molecular motion ... it actually has more heat than a lot of things. Does that make sense?

So ... with the HA1C question and other times when you are deciding whether something is interval or ratio level data ... the question is not whether or not the patient would be alive or dead if they had a result of 0. The question is whether or not a score of 0 on the measuring scale would represent a true absence of the thing being measured -- which enables the numbers on the scale to represent true quantities of the thing being measured and thus, the expression of relationships as ratios. Does 8% actually mean twice as much as 4%, etc.

I hope that helps.