The oral glucose tolerance test, revisited.

Hyperglycaemia is necessary for the diagnosis of diabetes. For more than 35 years there has been a spirited controversy about the diagnostic criteria for Type 2 diabetes, in particular about the oral glucose tolerance test (OGTT). In a 1975 paper Marvin Siperstein referred to the OGTT as ‘a pitfall in the diagnosis of diabetes mellitus’. Now the controversy has returned in the wake of the 1997 recommendations by the American Diabetes Association to exclude this test from routine diagnostic and screening efforts. This discussion has expanded beyond the best method to determine the level of glycaemia that defines diabetes, and now embraces questions of the best glucose level to predict an increased risk of future diabetes or future cardiovascular disease. The paper by Qiao et al. in this issue is one of several recent publications that re-examine the OGTT question. The history of the OGTT is instructive. By the 1960s it was recognized that a fasting glucose test would miss many persons who had obvious diabetes when their glucose was measured after a test meal. This led to the development of at least six different recommendations for oral glucose loads ranging from 50 to 100 g without regard to body size, or based on ideal body weight or body surface. Most of these early tests for diabetes required both 1 and 2 h post-challenge blood draws, and the level of glycaemia required for a diagnosis of diabetes varied depending on the protocol. Only one of these early protocols included a fasting blood glucose, and this protocol (with the diagnosis of diabetes based on a point score) required four blood tests, one at 0, 1, 2 and 3 h after an OGTT. Several reasons for not testing post-challenge glycaemia were recognized early, including the poor reproducibility of post-challenge (compared to fasting) glucose levels and the striking effect of age on post-challenge (compared to fasting) glucose. In fact, the high prevalence of diabetes in older adults when the diagnosis was based on post-challenge glucose led to a reluctance to accept the OGTT — on the grounds that so many people could not possibly