Sadly, it doesn't mean that high tests are wrong, but those around the edges of prediabetes are untrustworthy, the conclusion being that HbA1c results for prediabetes are neither "accurate nor specific", or in layman's terms the margin of error on this test is very large!
Food for thought.
I think this is quite well known which is why my GP ordered a number of different tests on diagnosis. The DSN still offers me a GTT every year which I refuse(YUK).
Could not see anything about the sickle cell trait in the study which can also skew diagnostic tests.