The New GI Tracts

For preventing heart disease, diets that control insulin are all the buzz

It’s hard to meander down the health-and-nutrition aisles of book megastores without encountering displays for competing diet guides. The books target different parts of the population—people struggling with obesity, lack of pep, or various diseases. Recently, there’s been an explosion of volumes for a condition that some authors claim affects one-tenth to one-third of people in the United States.

Glycemic index: bananas, 55.

Glycemic index: apple, 38.

Glycemic index: cherries, 22.

Glycemic index: pear, 38.

Glycemic index: grapefruit, 25.

Glycemic index: peaches, 42.

Glycemic index: watermelon, 72.

People usually don’t know whether they have this characteristic, called insulin resistance, but if they do, they face an unusually high risk of developing heart disease.

Two new tomes offering ways to diminish health complications of insulin resistance share the same titleSyndrome X—the medical name for a constellation of abnormalities that often develop in insulin-resistant individuals (SN: 9/16/89, p. 184). These include increases in blood pressure (SN: 10/14/89, p. 254), the length of time fats circulate in the blood, and the share of cholesterol residing in especially small, low-density lipoproteins (LDLs), which pose a potent heart-attack risk (SN: 9/21/96, p. 182: Syndrome X differs from diabetes in that the body can still produce enough insulin to cope with its elevated blood sugar.

Both books move away from the low-fat mantra that has pervaded major health guidelines for the general population during the past decade and campaign instead for limits on carbohydrates—sugary and starchy fare. The approaches to managing carbs, however, differ dramatically, reflecting a sharp dichotomy in the research community.

One of the Syndrome X books, by Jack Challem (2000, John Wiley & Sons), recommends that people with this problem select fare with a low glycemic index (GI). In these foods, carbohydrates break down slowly in the digestive system to provide the blood glucose that powers cells. Challem and his coauthors reason that this sluggish digestion will keep blood-glucose concentrations fairly low, enabling the body to metabolize this sugar via a stable, moderate production of insulin.

In the other new book with the same title (2000, Simon & Schuster), Stanford Medical School endocrinologist Gerald Reaven—who first defined syndrome X in 1988—asserts that the easiest way to maintain steady, moderate insulin concentrations is to keep body weight low, avoid bingeing on food, and steer clear of low-fat diets while limiting saturated fats. Indeed, he told Science News, “the last thing I’d worry about is glycemic index.”

Yet the term glycemic index is entering the public lexicon, not to mention medical symposia, health guidelines, and nutritional prescriptions. Fueling this interest is a rapidly growing body of research reporting tantalizing prospects for diets guided by the rate at which their carbohydrates are digested. Some researchers suggest that such diets can also foster weight loss (see below: “Glycemic index for weight loss?”) and perhaps even slow or prevent the development of type II diabetes.

Although the authors of these studies are quick to acknowledge limitations in their data—often owing to small test groups or short trials of diets—many are enthusiastic about the potential health payoff from managing insulin via the diet, especially in people with no apparent risk factors for diabetes.

In fact, to nutritional epidemiologist Simin Liu of the Harvard Medical School in Boston, the work on glycemic index “is the most fascinating and promising area in nutrition research today.”

Getting research rolling

Reaven takes credit for getting research on the glycemic index rolling some 30 years ago. That’s when his team at Stanford first began investigating the rates at which different starches break down.

Starches, a family of carbohydrates, consist of linked glucose molecules. The prevailing view had been that once in the digestive system, the glucose units would uncouple at some common rate and enter the blood. For any given quantity of starch eaten, then, the rise and fall of blood-glucose concentrations—and insulin—should be identical.

In fact, Reaven showed, starchy foods can release their glucose at very different rates. He found that baked potatoes and bread release their glucose quickly, and rice and cooked corn parcel it out more slowly. One study by his team showed that the gut even digests a single type of rice at different rates depending on how it has been milled. “This really emphasized the importance of the physical properties of a food,” he recalls.

Nutritionists quickly translated such findings into new recommendations for people with diabetes. Many began advising patients who need to control their blood-sugar concentrations to eat less-processed forms of a food—such as whole carrots instead of grated ones or firm instead of overcooked pasta.

To Reaven and several others, however, the greatest significance of those early studies was their implications for people who hadn’t been diagnosed with diabetes. “The first actual measurements of insulin resistance probably took place around 1970, with techniques that we introduced,” Reaven recalls. These and later studies would show that although much of the nondiabetic population makes what should be adequate amounts of insulin, some people have trouble using the hormone.

Certain tissues—notably, muscle and fat—can’t absorb glucose from blood unless it’s shepherded by insulin. In nondiabetic people with insulin resistance, normal concentrations of the hormone are insufficient, for reasons that remain unknown. The body compensates by pumping out increasingly larger amounts of the hormone.

Genetic predisposition and obesity can both contribute to insulin resistance. The more insulin-resistant a person becomes, the more insulin that person makes. In some cases, resistance eventually leads to type II diabetes, in which the body can’t reliably make enough insulin to compensate for resistance. Reaven observes that most people, however, just rev up their insulin production to increasing levels, as needed—an adaptation that represents a mixed blessing.

The resulting overproduction, called hyperinsulinemia, can damage blood vessels and prompt the liver to increase its output of very-low-density lipoproteins (VLDLs). These compounds have been linked to increased risk of heart disease. Also, as they circulate in the blood, VLDLs trigger a fall in high-density-lipoprotein (HDL) cholesterol, the so-called good cholesterol (SN: 9/9/89, p. 171). The other changes seen in syndrome X, Reaven adds, may also be the result of hyperinsulinemia trying to overcome insulin resistance.

Reaven contends that for people with insulin resistance, carbohydrate-rich diets are at least as dangerous as those laden with saturated fats.

Breakdown rates

Some 20 years ago, David J.A. Jenkins of the University of Toronto also began worrying about the dangers that certain carbohydrates might pose to insulin-resistant people. To investigate these risks, he needed to know the breakdown rates of comestibles containing carbs. So his team began feeding a battery of foods to healthy individuals and ranking the foods’ glycemic indexes on a 100-point scale, where 100 is pure glucose. The higher the GI values, the more rapidly carbohydrates turn into blood sugar.

“These studies are quite expensive,” which explains their slow accumulation in the medical literature, says Jennie Brand-Miller of the University of Sydney, Australia, who has done many of her own. But such studies prove essential, she adds, because GI rankings often turn out to be counterintuitive.

For instance, the GIs for ice cream and chocolate milk tend to fall under 50, she’s found, while that for baked potatoes can range from 75 to 100, depending on the type of potato.

Once GIs were measured for a range of foods, Jenkins’ group began giving volunteers diets with a low or high rating. “And we showed that diets based on low-GI carbohydrates improved [serum] triglyceride levels, total cholesterol levels, and the ratio of LDLs to HDLs,” Jenkins says.

His crew also observed signs of other advantages, he says, such as better removal of urea wastes from the blood and inklings that limiting hyperinsulinemia via the diet might limit cancer growth. A just-completed follow-up study, he says, lends strong support to the idea that a low-GI diet can fight cancer.

However, Jenkins points out, the quantity and quality of a food’s fiber can prove pivotal to its GI. “It was therefore hard to tell whether changes were due simply to low-GI intakes, to fiber, or both,” he says.

“Fiber and glycemic index are intimately mixed up in a way that most people don’t understand,” Brand-Miller explains. Fiber can slow the digestion of associated carbohydrates—but only if the fiber remains an intact part of the original food.

For instance, she notes, whole grains tend to have a GI of 50 or so. But mill those grains, and their fiber will be crushed into small shards that separate from the carbohydrates they had been protecting. Whole-grain flours and refined white flour both usually yield breads with a GI of about 70, she notes.

Vitriolic debates

About 15 years ago, Jenkins says, “vitriolic” debates broke out between factions of the nutrition community that were warming to the value of GI-tailored diets and that thought them a waste of time. Eventually tiring of the politics “and venom,” Jenkins says, he switched his research focus.

Others, however, continued to probe the value of controlling insulin through diet. While at the Institute for Research in Extramural Medicine in Amsterdam, Johannas B. Ruige reviewed 17 studies that had looked at cardiovascular risks from hyperinsulinemia. His analysis noted that it was hard to compare the studies because of differences in their designs. Also, most were small; seven contained fewer than 20 subjects each.

Still, he told Science News, “we saw some increasing risk with increasing insulin levels.” Overall, “it was a very small effect,” perhaps an increase in cardiovascular disease risk of 15 or 20 percent, he says.

Gary Frost of Hammersmith Hospitals in London has also been homing in on the ways that food choices affect insulin. By changing the diet of heart-disease patients, for instance, “we showed that a low-GI diet improves insulin sensitivity,” he says. Recently, his team analyzed diets in a cross-section of British adults—1,420 men and women participating in a study of heart risks. Searching for links between HDLs and various facets of the diet, such as fats and carbohydrate consumption, only one strong correlating factor emerged: glycemic index.

The 20 percent of the volunteers eating foods having the lowest GI possessed substantially higher HDL concentrations than the 20 percent eating the highest-GI foods. In women, the HDL difference between groups came to 9.7 milligrams per deciliter of blood. That should translate to a 29 percent reduction in coronary heart disease mortality for the low-GI group, Frost’s team concluded in the March 27, 1999 Lancet. Men eating low-GI fare showed a similar though smaller HDL boost.

“We believe that what’s going on is that these low-GI foods attenuated the insulin response of the diet,” Frost told Science News.

Using 10 years of data collected from more than 75,000 women in the long-running Nurses Health Study, Liu says his team at Harvard “basically confirmed what Frost found.” The Harvard researchers showed that average daily glycemic load—the amount of carbohydrates downed, adjusted for the GI values of the food typically eaten—”was directly associated with risk of coronary heart disease,” Liu says. Even after taking into account other potentially confounding risk factors, people who routinely ate the lowest glycemic load had only about half the incidence of heart disease as those eating the highest load.

The trend was strongest among the heaviest women, Liu notes. Since obesity tends to increase an individual’s risk of insulin resistance, he notes, this study again finds evidence that hyperinsulinemia could explain the link between carbohydrate consumption and the increased incidence of heart disease.

Nutritionist James J. Kenney argues that many studies that purport to find advantages to low-GI diets have been flawed. They tend to offer participants equal-calorie portions of high- and low-GI fare. Because high-carbohydrate foods are usually not as energy dense as high-fat alternatives, it becomes hard to down excessive calories on high-carb diets, Kenney asserts. In people who don’t overeat, he says, high-carb diets don’t seem to lead to the worrisome metabolic features of syndrome X. Kenney works in Santa Monica, Calif., at the Pritikin Longevity Center, which promotes diets of low-fat foods with low caloric densities.

Dietary options

Other dietary options may be as beneficial as controlling carbohydrates—and more palatable, according to Manny Noakes of the Commonwealth Scientific and Industrial Research Organization in Adelaide, Australia. Last year, she and her colleagues investigated the effects of glycemic index on HDLs in 21 men and women with type II diabetes.

The researchers assigned each person to a succession of 4-week diets. In one phase, all the volunteers ate high-carbohydrate, high-GI meals. In another segment, all consumed high-carbohydrate, low-GI foods. In a third phase, they downed a high-fat diet, in which most of the fats were heart-healthy monounsaturates (SN: 11/21/98, p. 328: Any carbohydrates in those meals had high GIs.

As in Frost’s study, low-GI meals elevated the volunteers’ good HDLs, although the magnitude of that increase was small, just 6 percent. More surprising, the Australian team reported in the June 1999 European Journal of Clinical Nutrition, was a comparable HDL elevation when the participants downed high-fat, high-GI meals.

To Noakes, this suggests that the benefits of heart-healthy mono fats may compensate for any detrimental effects of high-GI foods—provided the total share of carbohydrates in the diet is kept relatively low. That could prove important, she points out, since high-fat fare tends to be tastier, making diets including it easier to stick with.

And that, Reaven contends, is his big argument against using glycemic index as a guide to food selection.

He’s found that “getting patients to change their diet is very difficult.” Most want the high-fat taste without its risks. A relatively simple change—substituting monounsaturates for saturated fats—can achieve that a lot more easily, he argues, than making people look up food choices in GI tables before they sit down to eat.

“There’s nothing wrong with [adjusting diets for] glycemic index,” he says, except that it’s more cumbersome than other options and doesn’t deliver the same magnitude of benefit that losing weight or cutting overall consumption of carbohydrates will offer.

Brand-Miller—whose GI tables appear in her popular book The Glucose Revolution (1999, Marlowe & Co.)—argues that this criticism just reflects an “American” bias. Elsewhere, especially in Australia, South Africa, and parts of Europe, she says, the use of GI tables to fashion healthier diets “has been well accepted for more than 10 years.” She notes that low-GI diets have been linked to reduced appetite and weight loss, and she expects that many North American holdouts will “come around.”

The important thing, she points out, is to realize that as long as a person doesn’t binge on high-GI offerings, no food needs to be ruled out of carb-managed diets—”even chocolate.”

Glycemic index for weight loss?

Studies of glycemic index (GI) were originally intended to help people suffering from an unusually low response to their own insulin, but some researchers now suggest that the diet they recommend for this so-called syndrome X can benefit nearly anyone. In The Glucose Revolution (1999, Marlowe & Co.), Jennie Brand-Miller of the University of Sydney, Australia and her coauthors propose that a low-GI diet can melt off pounds and even boost athletic performance.

Because foods ranking low on the GI scale tend to be rich in fiber, it’s difficult to sort out which aspect of diet is more important. The findings of a study published in the Oct. 27, 1999 Journal of the American Medical Association, for example, might be attributed to either factor.

Fiber emerged as the leading dietary constituent linked to weight control in this 10-year study of nearly 3,000 men and women. The more fiber an individual ate, the less likely he or she was to gain weight. Moreover, fiber consumption “had a major [beneficial] effect on blood pressure, cholesterol levels, and fibrinogen, which is a measure of the tendency for excessive blood clotting,” notes David S. Ludwig of Children’s Hospital in Boston, the study’s lead author.

Although the glycemic indexes of the foods eaten weren’t calculated in this trial, Ludwig says that fiber appears to exert its effects “by lowering levels of the hormone insulin.” That’s not surprising, he adds, because insulin can program the body to store calories. Indeed, he notes, “one complication of insulin treatment for diabetics can be excessive weight gain.”

Ludwig saw similar effects in a study of 12 overweight boys. The teens experienced greater and earlier hunger after a high-GI breakfast and lunch than after a breakfast and lunch of low-GI fare—even though all the meals contained the same number of calories. More importantly, his team reported in the March 1999 online version of Pediatrics, the youths snacked on 81 percent more calories after two high-GI meals of instant oatmeal than after two low-GI meals of vegetable omelets and fruit.

Before giving the keynote address at the meeting of the North American Association for the Study of Obesity last year, Susan B. Roberts of the Department of Agriculture’s Human Nutrition Research Center on Aging at Tufts University in Boston reviewed the 20 studies that were available on glycemic index and obesity. Though many had design flaws, six looked pretty strong, she says. “And five of those six, including the one I coauthored [with Ludwig, in Pediatrics], showed that low-GI foods reduced hunger,” she says.

High-GI foods foster a large increase in insulin within about 30 minutes, she notes. An hour later, however, blood concentrations of this hormone plummet, which the body translates into a feeling “that you’re ravenously hungry,” she observes.

“So, I have developed a personal rule of thumb,” Roberts says. “If you’re hungry 2 hours after a meal, you’re eating the wrong foods.”

Janet Raloff

Janet Raloff is the editor of Science News for Students, a daily online magazine for middle school students. She started at Science News in 1977 as the environment and policy writer.

More Stories from Science News on Health & Medicine

From the Nature Index

Paid Content