the statement about ref [21] is wrong, that study found that patients with A1C > 6.5% had lower mortality, not higher.