Postscript: I leaving my original post as is, below. But actually, it boils down to this. "Diagnosing diabetes or prediabetes is almost entirely done by test results. If certain numbers are right, then it's 'right, your card is marked' for you! Only under arcane medical exceptions can you have these numbers disregarded. Therefore, a person can diagnose themselves by those numbers. And in my case, the European standards don't apply; use the American."
=====================================
Recalling that I have put some A1c tests in my signature, I saw the light: all I should need to do is elaborate on them. Wrong conclusions can be drawn if one hasn't learned the highlights of the theory of normal and abnormal of glucose metabolism.
The sequence of my A1c measurements is, in part: 6.0, 6.3, 5.5, 6.2, 5.9 (42, 45, 36.5, 44, 41). These were all done in clinic. The samples are intravenous. As a first step to interpreting the results, consider the test threshold. We all already know what that is, right? Although there's a pitfall here!
Next step: what are the ifs, ands, and buts? Now, personally I doubt the scientific validity of this test. Not for most people tested. For most people, I don't know, for 3/4, the test will be accurate. The other patients test too high or test too low, because there are interferences, which are physiological or genetic factors that can invalidate the measurement. Unfortunately, in regard to most of these interferences, they are not quantifiable and there are no official standards for adjusting the A1c result or discarding it on account of them. But if someone is asking official of me, the A1c's are official.
Among the HbA1c interferences are 4 or 5 genetic mutations in haemoglobin which summarily invalidate certain brands of A1c. "Brands"? You see, there are many manufacturers in the A1c market, and there is an agency (NGSP) that is charged with testing their tests to detect susceptibility to the haemoglobin mutations. Some manufacturers have multiple brands of A1c assay. The agency tests all brands to see if they give false results in individuals with, say, mutated HbE. The agency does this for all the haemoglobin mutations. The list of invalidities can be found posted online at the agency's Website.
However, these mutations occur almost exclusively in people with non-European ancestry. If your "HCP" does not advise you of the existence of these mutations and these A1c interferences, you will be ignorant that you could check your clinic's brand of A1c test against the invalid list. Then again, if you have no subcontinent, Far Eastern, Arab, or west African ancestry, there is no point in checking. It is the "HCP's" scientific responsibility to suss out whether a patient has one of these ancestries, and in case of yes, to consult NGSP's list and, erring on the side of caution, use only a brand that has proven insensitive to that mutation. (Just as most Europeans do not have dwarfism, most of the members of these racial communities will not have the haemoglobin mutations, so it would be a waste of money to test for them routinely.)
However, in case of any other interference, such as low red blood cell count, there is no possibility of disputing the A1c result. Although the authorities (including WHO) recognize that a low red cell count entails a falsely high A1c (Google on A1c interferences), this is purely a qualitative acknowledgement. If a patient with low RBC tests at 48, the medical power structure affords them no recourse to claim, "I'm really prediabetic because my RBC measures X".
Aside from actual beta cell disease, there are miscellaneous phenomena that can cause high BG. Examples are medication toxicity, or an exotic endocrine problem from
outside the beta cells. Unfortunately the burden is, in effect, on the patient to prove that one of these processes is present. Some possibilities of medication toxicity are not acknowledged by the "establishment". Then the establishment won't listen if the patient protests that their hyperglycaemia occurred fast on the heels of one of those.
In conclusion,
with the few, strict exceptions laid out above (eg, you prove you have a rare, non beta cell tumor that is making you seem to be a diabetic), two A1c blood test results (confirmation is mandatory, just as with HIV) summarily classify one as normal, prediabetic, or diabetic --
no qualified professional interpretation needed. Put another way, the clinical standards for diagnosing diabetes afford licenced HCP's little objective basis for disregarding the A1c result, although maybe: does their licence bestow on them total subjective discretion to do so? Therefore, a patient needs no validation by a licenced HCP to believe themselves pre-D or D. A patient needs validation by a licenced HCP only to
thwart a positive diagnosis. How many GP's do you suppose go looking for reasons to overrule the A1c result?
Back to the topic of what really is the diagnostic threshold of pre-D? Here's the last caveat. In Europe and Canada, prediabetes begins at 42. But in America, it begins at 38.7; and I'm in America. In the UK, you would be correct to say that I have rated as prediabetic in only 3 out of 5 test results. But in America, it has been 4 out of the 5, including the latest one, which is one month old. Moreover, even using the European threshold, it's more important that I've been objectively -- yes, not officially, but objectively -- pre-D for half of the last 3 years; and that I keep crossing the threshold.