Editorial The New England Journal of Medicine -- May 11, 2000 -- Vol. 342, No. 19

In the past decade, many new drugs have become available for patients with diabetes mellitus. The sulfonylurea drugs and metformin have been joined by disaccharidase inhibitors, meglitinides, and thiazolidinediones. Several new insulin preparations have been developed, with a wide spectrum of times of peak action and durations of action. Insulin lispro is the first of these to reach the market. In the face of this impressive progress in drug treatment of diabetes, the article by Chandalia et al. in this issue of the Journal (1) serves as a timely reminder that, until the advent of insulin in the 1920s, the only treatment for diabetes was dietary.

Increasing consumption of soluble fiber is one effective dietary treatment. Its benefits in terms of serum cholesterol concentrations and colonic function are well established, but its effects on hyperglycemia have been less well accepted. In a crossover study, Chandalia et al. treated a group of 13 obese patients with type 2 diabetes with a standard American Diabetes Association (ADA)-recommended diet containing about 24 g of fiber per day, of which 8 g was soluble fiber, and a diet containing about twice as much total fiber, with three times the soluble fiber. Each diet was followed for six weeks. At the end of the high-fiber diet, the patients' 24-hour plasma glucose concentration (i.e., the area under the curve) was 10 percent lower and the plasma insulin concentration was 12 percent lower than at the end of the ADA diet. The decrease in glycosylated hemoglobin values was smaller, because of the short duration of treatment. The overall decrease in plasma glucose concentrations was similar to that typically achieved with an oral hypoglycemic drug. The absence of weight loss during consumption of the high-fiber diet indicates that the improvement was not simply due to caloric restriction.

This new evidence that high dietary fiber intake is beneficial in type 2 diabetes confirms previously published work. Increasing fiber intake as a treatment for diabetes is actually less controversial than the related concept of using foods with a low glycemic index. The glycemic index refers to the increase in blood glucose concentrations in the three hours after the consumption of a test food containing 50 g of available carbohydrate. The index is calculated as a percentage of the increase in the blood glucose concentration produced by a reference food, typically white bread, with equivalent carbohydrate content. The glycemic index of a food is minimally affected by its protein and fat content. Although foods with high fiber content typically have a low glycemic index, the two concepts are independent. Foods with a low glycemic index and high fiber content typically raise postprandial blood glucose concentrations less than foods that have the same fiber content but higher values on the glycemic index.

The glycemic index has been calculated for a large number of foods. Although the use of this index to plan carbohydrate intake is complicated, diets that incorporate foods with a low glycemic index reduce hyperglycemia in type 2 diabetes. Furthermore, in type 1 diabetes, as well, the variation in blood glucose concentrations after meals is affected primarily by the carbohydrate load. Patients who are dependent on insulin have popularized the technique of "carbohydrate counting" to vary the dose of short-acting insulin before a meal according to the carbohydrate content of that meal.

A straightforward explanation of the benefit of both high-fiber foods and foods with a low glycemic index in reducing postprandial hyperglycemia relates to the timing of disposal of the dietary glucose intake. In patients with type 2 diabetes, early-phase insulin release is deficient. Even patients with good dietary control of diabetes have diminished release of insulin in the first half-hour after a meal. Although substantial insulin release may occur later, the decreased early-phase insulin secretion leads to postprandial hyperglycemia. Carbohydrates with high fiber content and a low glycemic index may delay the absorption of glucose, thereby permitting a better match between the timing of insulin release and peak blood glucose concentrations.

In addition, the production of free fatty acids is lower after a meal with a low glycemic index. Fatty acids promote insulin resistance. Lower insulin resistance after a meal translates into lower blood glucose concentrations.   Although carbohydrates with a low glycemic index appear to be desirable, elimination of carbohydrates from the diet is not beneficial. On the contrary, meals with a low carbohydrate content, even when the glycemic index is low, are associated with higher serum concentrations of free fatty acids.

It is clear that dietary manipulation produces dividends in patients with diabetes. Combining a high-fiber diet with the use of foods with a low glycemic index effectively lowers blood glucose concentrations synergistically. The use of "natural foods" -- that is, not supplements or synthetic products -- is considered an advantage by many.   However, one should not discourage the testing of synthetically derived food products in the search for improved diets for diabetic patients. The key element in dietary change is the acceptability of the food choices. If the dietary modifications are attractive to patients, they will be successful. Thus, the use of plant-derived sterols in spreads and cooking oils has been well received and has proved beneficial in lowering serum cholesterol concentrations.   Fat substitutes have also been accepted in the marketplace. Similar manipulations of food to lower blood glucose concentrations would be welcomed.

Chandalia et al. note with regret that, despite the evidence of benefit, dietary fiber ingestion by persons with diabetes has not increased. In part, this failure is due to the lack of concerted educational campaigns like those that led to recognition of the role of unsaturated fats in reducing serum cholesterol concentrations. In addition, there is insufficient awareness on the part of physicians of the benefits of dietary treatment. Although dietitians do have an important role in patient care, the key to dietary change is the repetition of dietary education by the primary care physician at each visit. In an era in which physicians spend too much of their time documenting their activities in order to secure reimbursement, it is difficult to invest the considerable effort required for dietary counseling. Nor do the reimbursing authorities make it easy for physicians to obtain compensation for such efforts.

Yet the benefits of dietary intervention in patients with diabetes are clear and important. The decrease in the degree of hyperglycemia achieved in the study by Chandalia et al. by increasing patients' fiber intake is similar to that typically obtained by the addition of another oral hypoglycemic drug to the therapeutic regimen. Such escalating combinations of oral hypoglycemic drugs are usually necessary to lower blood glucose concentrations to current target values. It would be more desirable first to take full advantage of the improvements that can be attained by dietary modification.

Marc Rendell, M.D.
Creighton Diabetes Center .  Omaha, NE 68131