His interest was sparked by a paper published in 1979 by Dr B. Jeanrenaud, who found that all obese people suffered from 'a form of hyperinsulinemia' (Diabetologia, 1979; 17: 133-8). Hyperinsulinaemia is when the pancreas overreacts to excess glucose by producing too much insulin. At the time, nutritionists looking at Jeanrenaud’s data concluded that hyperinsulinaemia was the cause of obesity - but Montignac wasn’t so sure.

Taking a sideways look at the theory, Montignac proposed that the overproduction of insulin is not caused by obesity, but by a faulty pancreas - which itself causes obesity.

Montignac’s theory remained in the wilderness until 1988, when two other French scientists discovered that hyperinsulinaemia caused 'abnormally large amounts of recently consumed fat' to be laid down in the body (Creff EF, Obésités, Paris: Masson, 1988).

Here, then, was a plausible mechanism for obesity, one that would help explain why some people become fat while others do not, even with exactly the same food intake. Obese people, the new findings suggested, are not secret guzzlers - they have a pancreas that produces too much insulin.

What were the implications of this for diabetes? Montignac theorised that the reason most people with adult-onset (type 2) diabetes are obese is not, as most people thought, because obesity causes diabetes. Rather, obesity and diabetes are two sides of the same coin - both are manifestations of the same underlying problem: hyperinsulinaemia resulting from a faulty pancreas.

So, does too much insulin cause diabetes? No, said Montignac, it’s a symptom of an overworked pancreas. It’s the end result of a state of almost continuous hyperglycaemia (excess glucose in the blood), which places too high a demand on the pancreas to produce insulin. Eventually, the system goes into overdrive, creating a near-constant surplus of insulin. This, in turn, causes ‘insulin resistance’, where insulin ‘switch-off’ cells overload and cease to respond, further tightening the vicious circle.

Diabetes is, therefore, due to pancreas exhaustion, says Montignac, caused by a diet containing too many carbohydrates with a high glycaemic index. He puts the blame fairly and squarely on the modern Western diet, with its high content of refined carbohydrates and sugar.

When Montignac first proposed the theory, few experts supported him. Indeed, the prevailing theory was - and still is - that sugar does not cause diabetes. However, the evidence has been mounting that he may be right.

In 1997, researchers at the Harvard School of Public Health reported the results of a study of 65,000 middle-aged women followed for over six years, by which time, over 900 of them were diabetic.

What marked those women who developed diabetes out from the others who hadn’t developed diabetes was their diet, which mostly consisted of low-fibre foods and refined carbohydrates.

Looking at individual foods, the chief culprits were found to be 'cola beverages, white bread, white rice, french fried potatoes, and cooked potatoes' - the very foods listed by Montignac as having a high GI. These foods more than doubled the risk of diabetes.

The Harvard researchers unequivocally supported Montignac’s theory: 'hyperinsulinemia . . . is one of the best predictors of [type 2 diabetes]', they wrote. 'Our findings support the hypothesis that a diet with high glycemic load and a low cereal fiber content increases risk of [type 2 diabetes]' (JAMA, 1997; 277: 472-7).