[Diagnosis and management of type 2 diabetes].

In 2010, the American Diabetes Association included glycosylated hemoglobin (HbA1c) as a criterion for the diagnosis of diabetes with a cut-off point of ≥ 6.5%. However, there may be a substantial percentage of false negative results. Other scientific societies have accepted this approach but with slight differences. HbA1c complements, but does not substitute, basal glycemia as a screening and diagnostic test or the 2-hour oral glucose tolerance test. HbA1c should not be used for the diagnosis of gestational diabetes. Interpretation of HbA1c is limited in persons with anemia and hemoglobinopathies. Therefore, in addition to its sensitivity and specificity, its costs and the epidemiological situation should also be evaluated. An important question is whether almost normal HbA1c levels are safe in patients with type 2 diabetes. The results of the ACCORD, ADVANCE and VADT trials are contradictory and have aroused considerable controversy. However, the extensions of the UKPDS and STENO-2 studies have shown the benefits of good glycemic control in the long term. As a general rule, a target of HbA1c < 7% seems appropriate. In patients at low risk of hypoglycemic episodes, short disease duration and young persons, HbA1c < 6.5% can be considered. In patients with frequent hypoglycemic episodes, the elderly and persons with short life expectancy, values of more than 7% are acceptable. This target should be achieved through individualized, early, intensive and safe treatment, without risk of hypoglycemia, and should be integrated in an overall program of cardiovascular risk prevention.