It's strictly a question of his likelihood P for developing diabetes. If his P is the same as that of other people with lower A1C, then "prediabetes" is nonsense.