At the end of the 1970s confusion reigned both with regard to the classification of diabetes and to the appropriate diagnostic test. There was enormous variation in diagnostic cut-off values both for the fasting glucose level and after oral glucose loading. The size of the glucose load varied between 50 g and 100 g or was body weight related. Similarly the types of diabetes were loosely divided into 'juvenile onset' and 'maturity onset', with secondary diabetes, chemical diabetes, borderline diabetes and prediabetes all used in ill-defined ways. In 1979 and 1980 the National Diabetes Data Group (NDDG) in the USA 1 and the World Health Organization (WHO) Second Expert Committee on Diabetes 2 made valiant efforts to create order out of chaos. Types 1 and 2 diabetes, Impaired Glucose Tolerance, Gestational Diabetes and 'other types' took centre stage whilst the 75 g oral glucose tolerance test (OGTT) became the gold standard with fasting and 2-hour values defined. There were nonetheless some differences between WHO and NDDG which were resolved in part by a further WHO report in 1985 3 when Malnutrition Related Diabetes Mellitus (MRDM) was added as a major category.
There has been a growing feeling over the intervening period that further revision was necessary. Whilst discussions were taking place at WHO, independently the American Diabetes Association established an Expert Committee to re-examine diagnostic criteria and classification. It was felt, in particular, that with the burgeoning of pathophysiological knowledge about diabetes, a more aetiology-based system could be used. The WHO Consultation which took place in late 1996 was additionally charged with the more daunting task of making recommendations about the diagnosis and classification of diabetic complications.
The ADA Expert Committee reported in 1997 4 whilst Part I of the WHO Consultation appears in the current issue as a consultation document. 5 There was crossrepresentation between the groups which has certainly helped achieve some consistency in the recommendations. A new fasting plasma glucose level of 7.0 mmol l -1 or above is suggested for the diagnosis of diabetes compared with the previous 7.8 mmol l -1 . This is based both on equivalence with the 2-hour value of 11.1 mmol l -1 and on predictive power for microvascular complications, albeit in cross-sectional studies. So far so good! The ADA, however, makes a strong recommendation that the fasting plasma glucose (FPG) can be used on its own and that in general the OGTT need not be used. The WHO Consultation by contrast argues strongly for the retention of the OGTT and suggests using the 535