Glycaemic index is quoted as a fixed property of a food, like its fibre content. It is actually the output of a testing procedure involving human volunteers, and knowing the procedure explains what the number can and cannot tell you.
The test standardises carbohydrate, not portion size
Volunteers are given a portion of the test food containing a fixed amount of available carbohydrate, commonly fifty grams, with fibre excluded from that count.
Because foods differ in carbohydrate density, the physical portions differ enormously. Reaching that fixed quantity from a low-carbohydrate vegetable may require an amount nobody would eat at a sitting.
The index therefore compares carbohydrate quality, holding quantity constant, which is a narrower question than how a meal behaves.
Blood glucose is tracked as a curve
After the food is eaten, blood glucose is sampled repeatedly over about two hours, producing a rise-and-fall curve rather than a single reading.
The area under that curve, above the starting value, is calculated. A single peak reading would miss whether the rise was brief or sustained.
The same person then repeats the procedure with a reference carbohydrate, usually pure glucose, on a separate day.
The number is a ratio and an average
The test food's area under the curve is expressed as a proportion of the reference food's, which is why the scale is anchored with glucose at one hundred.
The procedure is repeated across a group of volunteers and the results averaged, because individual glucose responses to the same food vary considerably.
A published index value is therefore a group mean, not a prediction for any particular person.
Real meals break several of the assumptions
Foods are tested alone. Fat, protein and acidity slow gastric emptying, so the same carbohydrate eaten within a mixed meal produces a flatter curve.
Processing matters too: grinding, cooking time and cooling all change how accessible the starch is to digestive enzymes, and two versions of the same ingredient can test differently.
Glycaemic load was added to restore portion
Because the index deliberately fixes carbohydrate quantity, a second measure multiplies the index by the carbohydrate in a realistic serving.
That adjustment brings portion size back into the picture and is why some high-index foods carry a modest load.
Anyone managing blood glucose for a diagnosed condition needs individual guidance, since these population averages are not a substitute for measured personal response.