The study questions the way patients with type 1 diabetes determine how much insulin they should take before meals.
Jiansong Bao at the University of Sydney in Australia and colleagues took finger-prick blood samples from 10 healthy young people who ate a total of 120 different types of food - all with the same kilojoule content.
The glycaemia load repeatedly trumped the carb count in predicting the blood sugar and insulin rise after a meal.
Carbs and type 1 diabetes
"It suggests that the methods used to assess carbs in persons with type 1 diabetes might benefit from some rethinking," said Dr Edward J. Boyko, a diabetes expert at the University of Washington in Seattle who wasn't involved in the study.
But he said it wasn't certain the findings would hold up in people who aren't completely healthy.
The glycaemia load is calculated by multiplying the amount of carbs in grams per serving by the food's glycaemia index divided by 100. The glycaemia index for a variety of foods can be found at http://www.glycemicindex.com/.
Foods with a low glycaemia index cause the blood sugar to rise slowly.
Writing online in the American Journal of Clinical Nutrition, the researchers say their findings also suggest that eating foods with high glycaemia loads could be linked to chronic disease like type 2 diabetes and heart disease.
But that is not clear from the study, which only looked at blood sugar and insulin changes up to two hours after a meal, said Dr Boyko.
"It would just be speculation whether a dietary change like this would help people with type 2 diabetes," he told.
Long-term effects and other nutrients in the food might also be important for disease risk, for instance. And the most important problem remains pure and simple overeating, according to Dr Boyko.
"The excess weight is the main thing we ought to focus on," he said. "The simplest message would be, eat less."(Reuters Health/ March 2011)
The lowdown on carbo-loading