The Fructose Index is the New Glycemic Index

I stumbled upon an interesting editorial recently in the American Journal of Clinical Nutrition from Dr. Richard Johnson's group, entitled "How Safe is Fructose for Persons With or Without Diabetes?" It was a response to a meta-analysis in the same journal pronouncing fructose safe up to 90 grams per day. That's the amount in eight apples or four cans of soda. Not quite what our hunter-gatherer ancestors were eating! The editorial outlined the case against excessive fructose, which I feel is quite strong. That led me to another, more comprehensive paper from Dr. Johnson's group, which argues that the amount of fructose found in a food, which they call the "fructose index", is more relevant to health than the food's glycemic index.

The glycemic index is a measure of the blood sugar response to a fixed amount of carbohydrate from a particular food. For example, white bread has a high glycemic index because it raises blood sugar more than another food containing the same amount of carbohydrate, say, lentils. Since chronically elevated blood sugar and its natural partner, insulin resistance, are part of the metabolic syndrome, it made sense that the glycemic index would be a good predictor of the metabolic effect of a food. I believed this myself for a long time.

My faith in the concept began to erode when I learned more about the diets of healthy traditional cultures. For example, the Kitavans get 69% of their calories from high-glycemic index carbohydrates (mostly starchy root vegetables), with little added fat-- that's a lot of fast-digesting carbohydrate! Overweight, elevated insulin and other symptoms of the metabolic syndrome are essentially nonexistent. Throughout Africa, healthy cultures make dishes from grains or starchy tubers that are soaked, pounded, fermented and then cooked. The result is a pile of mush that is very easily absorbed by the digestive tract, which is exactly the point of going through all the trouble.

The more I thought about the glycemic index and its relationship to insulin resistance and the metabolic syndrome, the more I realized there is a disconnect in the logic: elevated post-meal glucose and insulin do not necessarily lead to chronically elevated glucose and insulin. Here's what Dr. Mark Segal from Dr. Johnson's group had to say:

We suggest that the [glycemic index] is better aimed at identifying foods that stimulate insulin secretion rather than foods that stimulate insulin resistance. The underlying concept is based on the principle that it is the ingestion of foods that induce insulin resistance that carries the increased risk for obesity and cardiovascular disease and not eating foods that stimulate insulin secretion.

Well said! I decided to take a look through the literature to see if there had been any trials on the relationship between a diet's glycemic index and its ability to cause satiety (fullness) and affect weight. I found a meta-analysis from 2007. Two things are clear from the paper: 1) in the short term, given an equal amount of carbohydrate, a diet with a low glycemic index is more satiating (filling) than one with a high glycemic index, leading to a lower intake of calories. 2) this effect disappears in the long-term, and the three trials (1, 2, 3) lasting 10 weeks or longer found no consistent effect on caloric intake or weight*. As a matter of fact, the only statistically significant (p less than 0.001) weight difference was a greater weight loss in one of the high-glycemic index groups!

As I've said many times, the body has mechanisms for maintaining weight and caloric intake where they should be in the long term. As long as those mechanisms are working properly, weight and caloric intake will be appropriate. The big question is, how does the modern lifestyle derail those mechanisms?

Dr. Johnson believes fructose is a major contributor. Table sugar, fruit, high-fructose corn syrup and honey are all roughly 50% fructose by calories. Total fructose consumption has increased about 19% in the U.S. since 1970, currently accounting for almost one eighth of our total calorie intake (total sugars account for one quarter!). That's the average, so many people actually consume more.

Fructose, but not starch or its component sugar glucose, causes insulin resistance, elevated serum uric acid (think gout and kidney stones), poorer blood glucose control, increased triglycerides and LDL cholesterol in animal studies and controlled human trials. All of these effects relate to the liver, which clearly does not like excessive fructose (or omega-6 oils). Some of these trials were conducted using doses that are near the average U.S. intake. The effect seems to compound over time both in humans and animals. The overweight, the elderly and the physically unfit are particularly vulnerable. I find this pretty damning.

Drs. Johnson and Segal recommend limiting fructose to 15-40 grams per day, which is the equivalent of about two apples or one soda (choose the apples!). They also recommend temporarily eliminating fructose for two weeks, to allow the body to recover from the negative long-term metabolic adaptation that can persist even when intake is low. I think this makes good sense.

The glycemic index may still be a useful tool for people with poor glucose control, like type II diabetics, but I'm not sure how much it adds to simply restricting carbohydrate. Reducing fructose may be a more effective way to address insulin resistance than eating a low glycemic index diet.


*Here was the author's way of putting it in the abstract: "Because of the increasing number of confounding variables in the available long-term studies, it is not possible to conclude that low-glycaemic diets mediate a health benefit based on body weight regulation. The difficulty of demonstrating the long-term health benefit of a satietogenic food or diet may constitute an obstacle to the recognition of associated claims." In other words, the data not supporting our favorite hypothesis is an obstacle to its recognition. You don't say?