The current criteria for the diagnosis of diabetes used in France are now based on those published by the American Diabetes Association in 1997: fasting plasma glucose >/= 7.0 mmol/l (126 mg/dl) (previously >/= 7.8 mmol/l (140 mg/dl)), 2-hour glucose >/= 11.1 mmol/l (200 mg/dl) following a 75g oral glucose tolerance test. However, while the American Diabetes Association recommended that the post charge test not be used, both the World Health Organisation and the French Language Association for the study of Diabetes and Metabolic diseases (ALFEDIAM) retained this test. The DECODE (Diabetes Epidemiology: Collaborative analysis of Diagnostic criteria in Europe) study analysed the effect of these changes on the prevalence of diabetes, and whether the changes were justified by the mortality in the various glycaemic groups, using epidemiological data on close to 30,000 subjects from twenty European epidemiological studies. The prevalence of diabetes, using fasting rather than the 2-hour glucose concentrations (as had previously been recommended for epidemiological studies) resulted in changes in the prevalence of diabetes, an increase or a decrease, depending on the population studied. The fasting criteria tended to diagnose younger and more obese subjects than the 2-hour criteria. The subjects who would now be diabetic with the new fasting diagnostic criteria suffered a high mortality, similar to that of other diabetic subjects, thus the new criteria can be justified. However, the diabetic subjects who only have a post-charge diabetic hyperglycaemia (>/= 11.1 mmol/l (200 mg/dl)), are now even less likely to be screened as diabetic, despite the fact that they have a risk of premature death of the same order as other diabetic subjects.