ELLIOT JOSLIN, a pioneering American researcher, argued vociferously until his death in 1962 that controlling the level of glucose in a person's bloodstream was the key to managing type 2 diabetes (the variant of the disease that appears later in life). Since the defining symptom of all types of diabetes is that the body cannot do this properly by itself, that made intuitive sense. It also seemed to make practical sense. His approach, which involved a combination of insulin treatment (insulin is the hormone most involved in regulating blood-sugar levels), exercise and a diet low in carbohydrates, showed promising results in the patients he treated at his clinic in Boston. Several big studies since Joslin's death appeared to vindicate him. One, published in 1993 in the New England Journal of Medicine, confirmed that carefully managing glucose levels delays the onset of complications. Another, the United Kingdom Prospective Diabetes Study, published in 1998, looked at levels of a substance called glycated haemoglobin A1C (a reliable indicator of blood-glucose levels). Healthy people usually have A1C levels of 4-6%. Any level above 9.5% is considered extremely dangerous. The study found that those whose A1C levels were reduced by treatment to around 7% suffered fewer heart attacks and strokes than those whose levels were held at around 8%. Despite this evidence, the glucose-control hypothesis has always had its sceptics. The core of their doubt is that what is being treated is a symptom