The idea that the diets of diabetics should be low in carbohydrates dates back nearly 200 years. Even 40 years after the introduction of insulin allowed them to control their blood sugars, diabetics were advised to restrict carbohydrate intake to between 30 and 40 percent of the day's calories.

In recent times, nutritionists and physicians have taken a different tack. They have begun to recommend that carbohydrates, primarily in complex form, be increased to between 45 and 55 percent. This regimen seems to keep a tighter rein on blood sugar, as well as reducing the level of fat in the diet of a group already at higher risk for developing coronary heart disease.

But as every diabetic knows, the diet is restrictive. Studies show compliance at less than 50 percent. According to some researchers, it is time to expand the dietary guidelines to include, on occasion, foods that used to be off limits. They point to evidence suggesting that moderate amounts of sweets added to a mixed meal may not be detrimental.

As a result, the previously held theories about the effects of different types of carbohydrates on blood sugar levels -- theories on which two centuries of dietary recommendations were built -- are being reevaluated.

Carbohydates can be divided into two types: simple and complex. It has always been assumed that the consumption of concentrated doses of simple sugars like glucose and sucrose (a combination of glucose and fructose) would bring a rapid rise in blood sugar because they require little or no digestion prior to absorption. Complex carbohydrates, such as the starch found in rice, potatoes and legumes, must first be broken down to their component simple sugars. Theoretically, this would take longer and thus help blunt the eventual effect of the absorbed sugar on blood glucose levels.

But in the 1970s, Phyllis Crapo, a nutritionist at the University of Colorado, questioned this reasoning. She called attention to evidence that the rate-limiting step is absorption rather than digestion. In other words, the dismantling of complex carbohydates can occur more rapidly than the intestine can absorb the simple sugar units. With her colleagues, Crapo began testing the effects of various carbohydrates on blood sugar levels. Their research, and that of others, confirms that blood sugar response cannot be predicted solely on the basis of the type of carbohydrate in a particular food.

Unfortunately, putting this new information into practice is not easy. It has been suggested that a "glycemic index" be developed to compare the blood glucose response of a particular food to that of the same amount of pure glucose. This could replace the currently used food-exchange lists that give sample sizes of bread, rice and other types of food providing equivalent amounts of carbohydates.

The problem with a glycemic index, or with any method devised to assist the diabetic in selecting foods, is that many other factors come into play in predicting blood glucose response. Besides the type of carbohydrate, much depends on how the food is prepared. It is the cooking of the potato that causes a rapid rise in blood sugar similar to pure glucose -- a much-publicized finding of Crapo's studies. Wheat as bread gives a lower response than does wheat as pasta. The same is true for whole rice versus rice flour.

Another stumbling block lies in the effects of other food components on carbohydrates when eaten in combination. Protein stimulates insulin release, increasing glucose uptake from the blood. Fat delays stomach emptying, thereby slowing absorption. Few studies have tackled these problems. Instead, most have used large doses of a single food administered alone.

Still another complicating factor is the disease itself, which can take different forms. Some diabetics need insulin; others do not. How this affects an individual's response to various carbohydrates has yet to be tested. Many studies have employed only nondiabetic subjects.

So even if the old theories about diet and diabetes are in line for revision, more work must be done before major changes in diet therapy can emerge.

Where does sugar fit into the picture? New research found that when added to a mixed meal, it did not increase blood glucose levels to any great extent, as compared to other carbohydrate sources -- at least in adult-onset diabetics. The effect on those with juvenile-onset diabetes was more variable and therefore less clear-cut. Some experts believe that allowing occasional sugar consumption would greatly improve overall compliance with diet therapy. But important as palatability is, weight loss and subsequent maintenance of ideal weight are prime goals of treatment for many diabetics. Usually, nutritionally adequate reducing diets leave little room for empty calories.

We emphasize that any change in diet should be made only after consulting a physician and/or professionally qualified registered dietitian or nutritionist.