The concept sounds simple — only carbohydrates have more than minimal effect on blood glucose. The problem with understanding it is, however, that different carbohydrates affect blood glucose to different degrees. That’s the basis of the glycemic index, which is having more and more influence on low-carb diets like that of the late Dr. Robert C. Atkins.
Maltitol does increase blood glucose.
We call them carbohydrates because they are essentially hydrates of carbon. That means one carbon atom links one atom of water. Their composition is CxH2xOx.
We call the simple sugars — glucose, fructose, and galactose — monosaccharides. Their structural formula is C6H12O6.
What we call disaccharides have two sugar units bonded together. For example, common table sugar (sucrose) is a disaccharide that consists of a glucose unit bonded to a fructose unit.
Other carbohydrates are long chains of simple sugar units bonded together. That’s why we often refer to them as polysaccharides. Starch, a polymer of glucose, is the principal polysaccharide that plants use to store glucose for later use as energy.
Glycogen is another polymer of glucose. It is the polysaccharide that animals (including humans) use to store energy. Excess glucose bonds together to form glycogen molecules, which animals store in the liver and muscle tissue as a quick source of energy. Alpha cells of the pancreas secrete glucagon, which stimulates liver cells to break down glycogen and release glucose to the blood stream. We use it to treat hypoglycemia.
Cellulose is a third polymer of glucose. It’s different from starch and glycogen because it has hydrogen bonds holding together nearby polymers, which gives it added stability. Humans can’t digest cellulose, which we also know as plant fiber. Consequently, it passes through the digestive tract without being absorbed into the body.
When we talk about available carbohydrate, people have generally meant all carbohydrate except fiber, because we can’t digest it. Available carbohydrate is the carbohydrate that can be digested. Some people refer to it as “glycemic” or “usable” “net” carbohydrate, or “nutritive” carbohydrate. All of these terms refer to the same thing.
Many countries, including the U.S. and Canada, determine the amount of carbohydrate in foods indirectly, that is “by difference.” They measure the amount of protein, fat, water, and ash per 100 grams and subtract the sum of these from 100. In contrast, countries in Europe and Oceania analyze carbohydrate directly, so their carbohydrate figures do not contain unavailable carbohydrate (e.g. fiber), while values for the U.S. and Canada do.
As a result of this international difference, nutrition labels on packages imported to the U.S. from Europe and Oceania can be misinterpreted. For example, Bran-A-Crisp Fiber Bread, from Norway, is sold in the U.S. with a nutrition label that says it has 6 grams of total carbohydrates and 6 grams of fiber. It would be a mistake to conclude that this product contains no available carbohydrate. By comparison, either wheat or rye bread from Atkins Bakery made in America says on its nutrition label that it has 7 grams of total carbohydrate per serving and 4 grams of dietary fiber. Since it follows U.S. practice, the fiber is included in the carbohydrate, so this bread has 3 grams of available carbohydrate per serving.
What About Net Carbs?
Several manufacturers of low-carb products, including Atkins Nutritionals, Keto, and Biochem, say that carb counters should count only what they call net carbs or net impact carbs. Their definition of these terms is total carbohydrates less fiber, glycerin(e), the sugar alcohols, and polydextrose. They say that glycerin(e), the sugar alcohols, and polydextrose have “a negligible effect on blood glucose” or “a minimal impact on blood sugar.”
This is a fairly new development. The 1999 edition of Dr. Atkins’ New Diet Revolution, says that “Sweeteners such as sorbitol, mannitol and other hexitols (sugar alcohols) are not allowed….”
Then in 2002 Dr. Atkins published the revised and current edition of his bestseller, which for many is the bible of low-carb dieting. The book now says that you don’t count “non-blood sugar impacting carbs,” including polydextrose, glycerine, and sugar alcohol, as well as fiber, “when doing Atkins.” The Atkins Nutritionals website says, “We do use fiber and other carbohydrates, such as sugar alcohols, that have a minimal impact on blood sugar and thus fit the Atkins definition of a ‘non-digestible’ or net carb.”
What gives? The cynics say that it’s just business as usual. By 2002 Atkins Nutritionals had a growing product line with many products that included sugar alcohols among their ingredients.
To those who are less cynical it sounds like the Atkins people are now beginning to embrace the concept of the glycemic index. Indeed, the current edition of Dr. Atkins’ book called the glycemic index “A Beautiful Tool.”
This, however, is rather strange. After all, the glycemic index includes several others foods that have only a minimal impact on blood glucose. Nopal (prickly pear cactus) has a glycemic index of 7 (where glucose = 100). The mean of two studies of chana dal is 8. The mean of three studies of peanuts is 14. Yet no carb-counting diet that I am aware of excludes these fine foods.
It seems to me that there is a huge difference between “non-blood sugar impacting carbs” and those with “a minimal impact.” This is a difference that Atkins Nutritionals and others skirt over.
What Is the Impact of the Sugar Alcohols?
Sugar Alcohols — technically called polyols — are carbohydrates that we do not completely absorb. Of the eight sugar alcohols tested for their glycemic index, the most common ones are sorbitol, xylitol, mannitol, and maltitol.
If the sugar alcohols had no impact on our blood glucose, they would have a glycemic index of zero. With the December 2003 publication of Geoffrey Livesey’s amazing review of sugar alcohols, we now know a lot more about them than ever before. His article, “Health potential of polyols as sugar replacers, with emphasis on low glycemic properties,”; is in Nutrition Research Reviews 2003;16:163-91.
Only two of the sugar alcohols have a GI of zero, according to Livesey’s research. These are mannitol and erythritol. Several others have a very low GI, but two maltitol syrups have a GI greater than 50. This is a higher GI value than that of spaghetti, orange juice, or carrots.
Various articles about blood glucose control have incorrectly reported energy values of polyols as about 4 calories per gram and more recently on the American Diabetes Association website as about 2 calories per gram. In fact, Livesey reports that the energy values of sugar alcohols vary from 0.2 to 3.
|Maltitol syrup (intermediate)||53||3|
|Maltitol syrup (regular)||52||3|
|Maltitol syrup (high)||48||3|
|Polyglycitol (hydrogenated starch hydrolysate)||39||2.8|
|Maltitol syrup (high-polymer)||36||3|
Not all the low-carb gurus are on the polyol bandwagon. Dr. Richard K. Bernstein, a noted endocrinologist who wrote Dr. Bernstein’s Diabetes Solution (Boston, Little, Brown, revised edition 2003) says on page 139 that, “Some [sugars], such as sorbitol…, will raise blood sugar more slowly than glucose but still too much and too rapidly to prevent a postprandial blood sugar rise in people with diabetes.”
Confirmation of Dr. Bernstein’s position comes from a correspondent, Mary Lu Connolly. She wrote me in January that she has type 1 diabetes and has tried to reduce her carb intake by purchasing the low-carb foods now available. “What I have found is that these foods (especially breakfast bars) cause major rises in my blood sugars hours after eating. Can you explain what is happening?”
At the time she wrote I couldn’t explain it. Now, it’s clear that the culprit is probably maltitol or maltitol syrup. For example, Atkins Nutritionals Peanut Butter Cups have 11 grams of maltitol per serving. The “Net Akins Count” is 2 grams. Atkins Praline Sauce Duet has more maltitol syrup than anything else — 19 grams per serving. The net carbs count is 2. Or you could buy the Atkins Endulge Caramel Nut Chew Box, advertised as having 2 grams net carbs per serving. Yet a serving has 15 grams of maltitol.
Each of these examples come from the Atkins.com. None of them indicate that the glycemic index of one of their primary ingredients — maltitol — is higher than that of pearled barley or kidney beans.
Sugar alcohols do vary considerably in their glycemic indexes. It’s complicated, but they aren’t all created equal.
What Is the Impact of Glycerin?
Glycerin (or glycerine) is a liquid byproduct of making soap. It is wonderfully versatile and has been used as a solvent, antifreeze, plasticizer, drug medium, and in the manufacture of soaps, cosmetics, inks, lubricants, and dynamite. Now it is also used as a sweetener.
Atkins Nutritionals says that glycerine is another carbohydrate that has “a minimal impact on blood sugar.” Dr. Thomas Wolever, professor and acting chair of the department of nutritional sciences at the University of Toronto, confirms this in personal correspondence with me. He also heads a company, Glycaemic Index Testing Inc., which has ascertained the GI value of hundreds of foods.
“We did a study on glycerine at GI Testing, but the data don’t belong to me so I cannot publish it — except it was published in abstract form — and up to 75g glycerine had a negligible effect on blood glucose and insulin in normal subjects.’ He cites his article, “Oral glycerine has a negligible effect on plasma glucose and insulin in normal subjects” in Diabetes 2002;51(Supplement 2):A602. Some others believe, however, that it might have a greater impact on people with type 2 diabetes who have overactive livers.
What Is the Impact of Polydextrose?
Polydextrose is another carbohydrate. It is used primarily as a bulking agent for the preparation of calorie-reduced foods. Atkins Nutritionals says that polydextrose has “a minimal impact on blood sugar.”
Again, Dr. Wolever can confirm the Atkins claim. ”I don’t think polydextrose is available in the small intestine at all,” Dr. Wolever tells me. “If that is so, it has no effect on blood glucose.”
A recent study lead by Zhong Jie of Rui Jin Hospital in Shanghai, “Studies on the effects of polydextrose intake on physiologic functions in Chinese people,” confirms Dr. Wolever’s belief. This study, reported in the American Journal of Clinical Nutrition, Vol. 72, No. 6, 1503-1509, December 2000, concluded that “polydextrose had no significant effect on blood biochemistry indexes” include the glycemic index. Their study confirmed “that polydextrose is nonglycemic.”
Dr. Atkins and the vendors of low-carb products are correct that not only fiber but also glycerin and polydextrose have little or no effect on blood glucose. The story with sugar alcohols, however, is different. One of the most commonly used sugar alcohols, maltitol and its syrups, does have a considerable effect on blood glucose. Two sugar alcohols, erythritol and mannitol, have no effect, and four others have some effect.
This article originally appeared on mendosa.com on February 13, 2004.
Update: Response from Atkins
It took the Atkins people seven months to recognize this article and respond to it. And when they did write, it became clear that they didn’t like it at all.
My article “makes an unfounded claim and does a considerable disservice to the millions of Americans who lead healthier lives by following the Atkins Nutritional Approach,” writes Colette Heimowitz, vice president, education and research, Atkins Health and Medical Information Services. Actually, she enumerated three unfounded claims that I supposedly made in this article: Basing my article on the glycemic index is “problematic” for several reasons:
Last modified: May 13, 2005
Go back to Home Page
Go back to Diabetes Directory