One of the reasons, I think, for switching units was to make more separation between fingerstick BG numbers and HbA1c, which are related, but can’t be directly converted one to the other.
Sort of a bit, but not really. It was a conscious decision to avoid confusion with HbA1c numbers!
Despite HbA1c methods being used since the later 1970s, the Diabetes Control and Complications Trial in the U.S. only reported in 1993. This lead to standardized testing in the U.S. aligned to its results, whi h wa also what was adopted in the U.K.
Other countries were establishing their own standardization too, notably Japan and Sweden. But while these lead to comparable results between laboratories, difference still existed between the national standards. So in 1995 the International Federation of Clinical Chemistry and Laboratory Medicine formed a working group to establish an international standard. The reference method they produced was approved in 2002.
However the various national standards gave different results to that reference. This was mainly due to the I.F.C.C. more specifically ignoring other forms of haemoglobin that affected the results of the other standards. That said, those standards were still reliable as indicators, but the percentage stated was not a direct correlation to the actual percentage of adult haemoglobin that was made up of HbA1c. For the D.C.C.T.-aligned scale which we used, the results were around 1.5% higher than they should be.
The IFCC-IUPAC Committee on Nomenclature, Properties and Units (IUPAC being the International Union of Pure and Applied Chemistry) made a number of proposals, including renaming HbA1c to "Haemoglobin beta chain(Blood)—N-(1-deoxyfructos-1-yl)haemoglobin beta chain"! They clearly have a preference for long detailed names, though they did say it could be call DOF Haemoglobin for short.
Fortunately for us, that was rejected by the working group, but they did accept the need to change the system of units to avoid confusion. And there were already three slightly different percentage scales. But someone who was tested as 8.0% on the D.C.C.T.-aligned scale would suddenly find themselves "dropping" to 6.4% when tested on the I.F.C.C. scale, for example. As well as patient confusion with it falsely implying an improvement, it is also causes a big problem in research when comparing data across the two systems if the values suddenly change meaning.
Though the comittee did also include the argument that the similarity of numbers for a glucose level in mmol/L and an HbA1c expressed as a percentage could risk confusion. Although there was no evidence for that. And many countries give glucose levels in mg/dL, so the change in would create a new potential confusion where those scales overlap instead. And as seen on this site, plenty of people still get confused by different scales and they have a similar appearance with both being in "mmol".
The main argument, though, is that the percentage scale was not scientific. A percentage is seen as ambiguous and not encouraged and in scientific terms a percentage is a unit of 0.01. Expressed properly as a fraction of a substance, 6.4% should be a very unfriendly 0.064 and clearly outside the preference for units to fall within a range between 0.1 and 999.
Although some countries started using the new reference method shortly after its approval, in 2007 it was agreed that the I.F.C.C. method would be adopted worldwide. Equations fixed the three main national scales to it and it was agreed both the original and new units would be reported together. That was when the U.K. started using it, but we officially switched to reporting only the I.F.C.C. system a few years later in October 2011. Despite that, clinician and patient familiarity has meant that in practice both systems continued to be used. I was only given percentage units when diagnosed in 2014, and the my local health trust's lab still dual reports, with separate items given using each system.
I believe I have all that right, a combination or memory and looking up the details. I certainly do not remember that deoxyfructose haemoglobin name we escaped needing to remember.
Presumably this partic Hb test has been designated the A1c test.
If you were not put to sleep by history, now it is time for science! Done properly the "1c" part of HbA1c is a subscript, so it breaks down as:
Hb — Haemoglobin.
HbA — Adult haemoglobin, the most dominant form making up around 97% of the haemoglobin in a healthy adult and which has two alpha and two beta chains, others varients include:
HbA2 — A variant with two alpha and two delta chains.
HbF — Fetal haemoglobin with two alpha and two gamma chains.
HbA2 and HbF make up the remainder of haemoglobin in a healthy adult, but there many other less common variants, such as HbS which are sickle cell shaped due to sickle cell disease.
HbA
0 — Haemoglobin that has not been glycated, sometimes called pure haemoglobin.
HbA1 — Glycated haemoglobin, of which:
HbA1a1 — Haemoglobin glycated with fructose-1,6 diphosphate.
HbA1a2 — Haemoglobin glycated with glucose-6 phosphate.
HbA1b — Haemoglobin glycated with pyruvic acid adducts.
HbA1c — Haemoglobin glycated with d-glucose (a.k.a. glucose of dextrose)
HbA
1c makes up around 80–90% of total HbA
1. The result of an HbA
1c test when given as a percentage is (roughly, as explained above) how much of it makes up your total amount of HbA.
Very simplified, an average HbA
1c of 6% for a healthy person means that 6% of their adult haemoglobin has glucose attached to it (which would normally be around 5.8% of all red blood cells).
Its equivalent of 42 mmol/mol means that for every thousand moles of adult haemoglobin, 42 of them will have glucose attached.
And to tie it into the above, if you divide the mmol/mol value by ten then you have what the percentage value that it would have been had the I.F.C.C. not changed units. i.e. were the D.C.C.T.-aligned scale said someone had an HbA
1c of 6.0%, actually only 4.2% of their HbA had glucose attached.
Sometimes my need for things to make sense does seem to go too far. Though not far enough that I have tried to read about N terminals, where the glucose attaches to the beta chain.
