The Diabetes Diet
Richard K. Bernstein, MD, FACE, FACN, FACCWS
Part 2 of Chapter 3
Essential Guidelines for the Diabetes Diet
What’s Wrong with the Glycemic Index
The glycemic index seeks to measure the impact that particular foods have on blood sugar. It’s my opinion that there is simply no way to determine objectively without measuring an individual’s blood sugars how any given food at any given time is going to behave in any given individual.
The notion of the glycemic index was first conceived of by Dr. David Jenkins around 1980 in attempting to find foods that were appropriate for diabetics. The index numbers you may have seen are derived from graphs of the averages that particular foods have on both diabetics and nondiabetics. Since glucose is the blood sugar, dietary glucose is the standard against which other foods are measured, and rates a GI of 100. Whole wheat bread rates a 70 on one scale, while it comes in at 99 on another. Whole grain pasta rates a 45 or a 53, depending again on the group who did the testing. The baseline 100 number represents the area under the curve of a graph depicting a rise in blood sugar over 2 hours from eating a particular number of grams of glucose. You take another carbohydrate food whole wheat bread or pasta, for example containing an equal amount of carbohydrate, then plot the 2-hour curve of the resulting blood sugar rise after eating it. The glycemic index number comes from dividing the area under the test food’s curve by the area under the glucose curve and then multiplying by 100 to get percent.
Dr. Jenkins defined the glycemic index of a carbohydrate food in terms of how its curve related to that of the glucose curve. If a food has a glycemic index of 50, then the area beneath its curve is 50 percent of the area beneath the glucose curve.
So what’s wrong with that?
Well, as it turns out, a lot of things. As attractive as it may seem, the glycemic index, unlike the measurement of calories, seeks to quantify something much less hard and fast. Instead of measuring the heat released by burning a food in a laboratory, the glycemic index plots the average of a considerably more variable series of chemical and physiologic reactions that occur within the human body.
When you eat, the digestive process begins pretty much as soon as food hits your saliva. How digestible a food may be depends not just on your body and its ability to produce enzymes and hormones, but also on how the food was processed (cooked, chopped, ground, pureed, etc.). How you react to foods depends on a host of factors, from your own physiology to the combinations of foods you are eating. Fats, for example, tend to slow down digestion. The only way to measure the real effect of foods on blood sugar is to feed people the foods, then frequently measure blood sugar for a given period of time, then average the results.
For a type 1 diabetic like me, it’s easy to see the effect that, say, a bagel has on my blood sugar. Since I make no insulin, I have no built-in mechanism to automatically bring down the ensuing blood sugar surge. The surge for a nondiabetic man of similar height, weight, and age is going to be much less. My experience with my patients has been that a food that can raise my blood sugar significantly might have a similar but not exact effect on the blood sugar of another type 1 diabetic. It will not have the same effect on a type 2 diabetic — nor will the effect be the same from one type 2 to another. Many things besides the foods we eat can affect blood sugar — everything from insulin resistance to adrenaline surges and exercise to infection to the quantity of what we eat (don’t forget the Chinese Restaurant Effect) — especially for the diabetic.
While the glycemic index reflects only the 2-hour period after eating a food, in reality it typically takes 5 hours to fully digest the carbohydrate portion of a meal (in the absence of gastroparesis). This discrepancy has been justified by the claim that “while digestion and absorption may take 5 hours or more, most of the blood glucose fluctuation takes place in the first 2 hours. That’s where big differences between foods become apparent and where damage might be done to the body. Small differences seen after 2–3 hours are not of major clinical significance.”*
On the surface this position doesn’t seem particularly unreasonable — but it reflects a basic ignorance of the impact that issues “not of major clinical significance” can have on diabetics. Having lived with this disease for nearly sixty years, I know intimately how things that seem of small significance to the nondiabetic clinician can in reality completely defeat efforts to normalize blood sugars. The same food that makes my blood sugar rise dramatically may have little or no effect on that of one of my patients who still makes considerable insulin.
*Jennie Brand-Miller, author of The New Glucose Revolution, responding to a question from medical writer David Mendosa on www.mendosa.com/bernstein.htm.
Because the GI is an average, true numbers vary considerably from one person to another, from one time to another, and from one study to another. But even if we accept it as a given that the glycemic index is an accurate reflection of the effect that carbohydrate foods have on blood sugar which is simply not the case — there is still the issue of interpretation.
What is low, medium, or high?
I would say that if a food is going to be classified as low, then it must not provoke a sharp increase in blood sugar levels. It should allow you to adhere to the Laws of Small Numbers and keep your need for insulin whether injected or made by your body to the minimum.
I seem to be in the minority here — not an unusual position for me. A lot of well-meaning dietitians and diabetes educators now use the glycemic index to recommend foods that are “low” but in reality are completely inappropriate in a sensible blood sugar normalization regimen. According to www.glycemicindex.com, low, medium, and high glycemic indices are as follows:
• Low = 55 or lower
• Medium = 56–69
• High = 70 or above
My own experience is that a vast number of foods that are rated low by this definition will cause considerable blood sugar elevations in diabetics. In addition, these foods will impede blood sugar normalization in diabetics and contribute to weight gain or prevent weight loss in nondiabetics.
The carbohydrate foods that I recommend in the ensuing pages — salads and selected vegetables typically have glycemic indices less than one-quarter that of the “low” threshold above.
I recommend that you ignore the glycemic index, because in the final analysis it is no more helpful than the old calorie system of weight loss. Calorie counts give us a rough idea of the energy available from a particular food, and the old idea was that you could lose weight if you ate less and burned more calories. So you would guesstimate your daily caloric burn based on your height, weight, age, and so forth. Then you could count calories and eat fewer than you burned. That always sounded good on paper, but my experience is that it never really worked.*
*The beta oxidation (or “burning”) of fat by the body requires the action of an enzyme called insulin-sensitive lipase. This enzyme is turned off by insulin. Eating carbohydrate obliges the body of a nondiabetic to make insulin in proportion to the amount consumed and obliges many diabetics to inject insulin to prevent blood sugar elevation. When insulin levels go up, fat oxidation therefore goes down, and since insulin is also the fat-storage hormone, dietary fat is stored. Furthermore, insulin signals the liver to convert the carbon backbone of carbohydrates (glucose) to saturated fat, which then appears in the blood as triglycerides, which are subsequently stored. So calories of fat are handled much differently on a low-carbohydrate diet than on a high-carbohydrate diet. Recent studies on humans eating equivalent amounts of fat show that those eating more carbs store more fat.
Using the glycemic index, by comparison, tells us what? Its original intent was to find foods appropriate for diabetics. Again, however, there is the question of interpretation.
The apparent conclusion of the popular press on low-carbohydrate diets is that you can eat your fill of low-GI foods and still lose weight. As long as we eat foods with a GI of 55 or lower, the thinking goes, the pounds will just melt away, diabetes will fall into line, and all will be well with the health of the population. I wish it were so. And now, added to the concept of the glycemic index is the “glycemic load,” which was developed by Dr. Walter Willett. The idea behind the glycemic load (GL) is that the glycemic index of a food doesn’t tell you enough. (Which I would support.) The glycemic index tells you how fast the carbohydrate in a particular food turns to sugar, but it doesn’t tell you how much carbohydrate is actually in a serving of that food. The blood sugar effect of a “serving” of a very dense food like a potato is likely to differ considerably from the effect of a serving of a less dense food such as popcorn. The glycemic load, which takes carbohydrate density and serving size into account, is supposed to give you a better view of the overall burden the food you eat places on your body.
Unless you’re very clever and keep a calculator and a book of indices with you at all times, this seems to me unnecessarily complex and labor intensive. Are you going to sit down at a restaurant with your book of the glycemic index and have the waiter go over the gram count of the foods on the menu? Will you do the calculations when you open
up a recipe book?
I think it’s a great idea to know what you eat. But I also think that all this calculation is too much like the old system of “exchanges” — “Well, since a slice of bread has a glycemic load of X and this other food has a glycemic load of X and a half, if I just eat two-thirds of the slice of bread — no, wait a minute. . . .”
In addition, the glycemic load theory assigns foods different numbers than the glycemic index, so there is always the potential for confusing them.
Too Be Continued in Part 3 Keeping It Simple
There are no complicated formulas or calculations in the Diabetes Diet. In the following pages I will show you the foods that you ought to emphasize in your diet and the foods you ought to eliminate.
We would like to thank the publisher Little Brown and Company and Dr. Richard K. Bernstein, for allowing us to provide excerpts from The Diabetes Diet.
Copyright © 2005 by Richard K. Bernstein, M.D. All rights reserved. No part of this book may be reproduced in any form or by any electronic or mechanical means, including information storage and retrieval systems, without permission in writing from the publisher, except by a reviewer who may quote brief passages in a review.
This book is not intended as a substitute for professional medical care. The reader should regularly consult a physician for all health-related problems and routine care.
For more information on Dr. Bernstein’s and to purchase his books, CD’s or get access to his free monthly webinars, visit his website at DiabetesBook.com.