Does optimal HbA1c in diabetes differ according to drug treatment? An evaluation of national electronic database in Malta.

Background And Aims: A J-shaped relationship between HbA and mortality has been reported in subjects with type 2 diabetes. The postulated mechanism linking low HbA with increased mortality is increased hypoglycaemia risk. We tested this hypothesis by comparing the relationship between low HbA to mortality in patients on therapies with different hypoglycaemia risk.

Methods: We selected patients on any type of treatment for diabetes from a national electronic database (n = 25,743) and linked to other databases, including laboratory database and the national mortality register.

Results: We observed a J-shaped or U-shaped association between HbA and all-cause mortality in the whole type 2 diabetes patient cohort as well as in patients on metformin monotherapy and in those on metformin-sulphonylurea combination therapy, but not in subjects on sulphonylurea monotherapy or in those on insulin.

Conclusions: Our data confirm the J-shaped relationship between HbA and mortality in type 2 diabetes, but suggest that a low HbA is deleterious even in absence of hypoglycaemia and that subjects with type 2 diabetes might require a slightly elevated blood glucose for optimal outcome. Our data also suggest that the increased mortality associated with sulphonylureas cannot be mediated solely through increased hypoglycaemia risk.