The employee health newsletter I get at work recently had an article about the glycemic index (GI), "a food rating system that tells you how fast a particular carbohydrate-containing food enters the bloodstream as glucose (sugar)." Low-GI foods, the article said, "promote a slow, healthy rise in blood sugar," while high-GI ones cause a sharp spike accompanied by a rapid rise in insulin production. "If this happens repeatedly over time, you could be at risk for type 2 diabetes and heart disease," it went on. At first I thought the glycemic index was just a fancy way of getting you to skip simple carbohydrates (sugar, processed foods) in favor of complex carbohydrates (pasta is the one I remember, but I suppose green vegetables and that kind of thing are in there too). Imagine my surprise when I turned to the list of GI ratings and found that potato chips and low-fat ice cream were considered low-GI foods, Coke and table sugar were intermediate, and bran flakes, rice cakes, and dried dates were considered high GI! What's the story? Should I have been pigging on chips and ice cream all those years when I was virtuously eating rice cakes instead?
Illustration by Slug Signorino
Doesn’t that just suck? You try to eat the right stuff, but they keep changing what the right stuff is. I agree with Woody Allen: someday they’ll decide the real health foods are tobacco, steak, and beer.
The glycemic index was introduced in 1981 as a way to help diabetics control their blood-sugar levels. Its use in diabetes management remains controversial — the American Diabetes Association still won’t endorse it. But it seems to have gained fairly wide acceptance in the medical community.
Now some are promoting the glycemic index as a dietary tool for the general public. One of the more vocal advocates is University of Sydney nutritionist Jennie Brand-Miller, whose book The Glucose Revolution appeared in 1999. The University of Sydney, in cooperation with the Australian diabetes associations, is now working with manufacturers to publish GI ratings on food labels in Australia and North America (look for a circled G near the nutritional info).
In her book Brand-Miller emphasizes that the GI isn’t just a matter of simple versus complex carbohydrates, which is based on molecular structure. Instead, she says, it uses “real people and real foods.” GI is calculated by feeding a standard amount of a food to test subjects and periodically measuring their blood sugar. The food is then ranked from 0 to 100, with 100 the equivalent of eating pure glucose, the form of sugar most directly absorbed by cells. A long list of factors can cause GI to vary: amount of processing, fiber content, “degree of starch gelatinization,” and so on. Interestingly, sugar and fat tend to lower GI, the latter because it “slows down the rate of stomach emptying, thereby slowing the digestion of starch.”
There’s no simple divide between high-GI and low-GI foods. Kellogg’s Bran Buds With Psyllium is 45, but Raisin Bran is 73. Rice can vary from 40 to 90, depending on the type. Glucose response can vary among individuals, and different combinations of foods can produce unexpected results. There’s disagreement about the best way to compute GI rankings (some say white bread offers a better baseline than sucrose). When you compare published GI lists, sometimes you find widely varying numbers for what seem to be the same food. All of this makes you wonder whether GI has any practical value when making dietary choices.
GI numbers can be misleading. Carrots, for example, have a published GI of 71, which is high. Does that mean you should avoid carrots? Not really. To get the standard amount of carbohydrate (50 grams) the test subjects had to eat a pound and a half of carrots. Nobody does that except Bugs Bunny. To get around this problem, researchers at Harvard have proposed an adaptation of the glycemic index called the glycemic load (GL), in which you multiply the GI for a food by the amount you actually eat. In a recent study they claimed that women with a high GL have a much greater risk of heart disease (for more, see www.health.harvard.edu/newsletters/giload.shtml). [Update 2013: Now behind a paywall/not available at this url.]
Confused? Believe me, I’ve barely scratched the surface. I’m not saying the GI is useless — on the contrary, it can be a valuable analytical tool. There’s little question that frequent wild swings in blood-sugar level don’t do your health any good. But I doubt the GI will be much help when you’re planning your next meal.
Send questions to Cecil via firstname.lastname@example.org.