Food Product Design: Health/Nutrition - April 2005 - Low-Glycemic Foods — Ready for Prime Time?

April 2005

Low-Glycemic Foods -- Ready for Prime Time?

By Ronald C. DeisContributing Editor

With a 30-year career in food product and ingredient development devoted primarily to carbohydrates, I find it very exciting that we have entered a new era of consumer awareness, an awareness of the importance carbohydrates, along with some recognition that all carbohydrates are not the same. My Feb. 2005 article in Food Product Design, "The Low Down on Low-Carb," discusses the important contributions of the low-carb diet frenzy. One of the effects is an increased focus on low-glycemic foods and their relationship to health.

Chemistry versus nutrition As food scientists, we were all trained in proximate analysis -- determining fat, protein, water and ash by analysis. Carbohydrates were -- and still are -- determined by difference (subtracting the analyzed values from total weight), and they were all traditionally 4 kcal/gram.

Traditionally, carbohydrates have been identified by chemical means. Sugars are the mono- and disaccharides (1 to 2 degrees of polymerization, or DP 1 to 2). This class contains the recognized sugars, such as glucose, fructose, sucrose and lactose, as well as polyols, such as sorbitol, mannitol, maltitol and lactitol. From a labeling standpoint, however, polyols are not regarded as sugars. Oligosaccharides (DP 3 to 9) are the malto-oligosaccharides, such as starch degradation products, such as corn syrups and maltodextrins, and resistant dextrins, as well as other oligosaccharides, such as fructo-oligosaccharides and galacto-oligosaccharides. Polysaccharides are carbohydrates with a DP greater than 9, such as starches, cellulose, hemicellulose and pectins.

This subject was considered at length in the Food and Agriculture Organization of the United Nations (FAO) Food and Nutrition Paper 66, "Carbohydrates in Human Nutrition," published in 1998. This paper concluded that "a classification based purely on chemistry does not allow a ready translation into nutritional terms since each of the major classes of carbohydrates have a variety of physiological effects."

This lack of correlation between the chemistry and nutritional aspects of carbohydrates has created a great deal of confusion in the scientfic community, and even greater confusion in the eyes of the consumer. At the recent symposium, "The Nutritional Importance of Carbohydrate Quality in Cereal Foods," (Feb. 2005, Brussels, Belgium, AACC International, Minneapolis), Julie Miller Jones, Ph.D., professor of nutrition and foods, College of St. Catherine, St. Paul, MN, addressed the plight of the consumer: "Carbohydrates are complex. It is a word that comes directly from chemistry and means a hydrate of carbons, so right off the bat it is confusing to consumers. Then nutritionists added to the complexity by dividing carbohydrates into categories such as complex and simple. For a number of years, consumers were taught that complex carbohydrates were equated to starch and to foods such as potatoes, rice, and pasta, and that those foods delivered a steady supply of glucose and that all foods containing complex carbohydrates were the same. Consumers learned that simple carbohydrates were sugars and that these all entered the bloodstream rapidly, and all but fructose required insulin. To add to the complexity of simple sugars, there were sugars that occur in the bloodstream, those in human milk, and those in food, but not in all foods. Consumers know that carbohydrates are sweet and not sweet and that some sugars aren't sweet and don't enter the bloodstream rapidly. Then consumers tried to grapple with the fact that while the nutritional role of carbohydrates was to provide energy and spare protein, they learned that not all carbohydrates provide calories. Some are caloric and some are not. Some raise blood sugar and some do not. They raise insulin levels and they do not. Some are absorbed in the stomach immediately and some are absorbed slowly as they move through the intestine."

All of this confusion was compounded by terminology such as "zero carbs," "net carbs," "glycemic index," "glycemic load," "complex carbohydrates," "slow carbs," and so on. AACC International (the new name for AACC, the American Association of Cereal Chemists) convened a committee in late 2004 to address how to define carbohydrates for labeling purposes. The committee, headed by Jones, is comprised of representatives from North America, Europe and Australia, and has five academics, two government representatives and two industry representatives. After reaching consensus on definitions within this group, the committee has posted these definitions to www.aaccnet.org for international comment, which can be posted onsite. The definitions posted to date are as follows:

• Available (net) carbohydrate can be absorbed as monosaccharide and metabolized by the body.

• Glycemic carbohydrate is the portion of available carbohydrate that can elicit a blood-glucose response expressed as equivalent grams of glucose per serving, or per 100 grams of food.

• Glycemic response is the change in blood-glucose concentration induced by ingested food.

As a next step, the committee will reach consensus on appropriate methods of analysis. At the Feb. AACC Symposium, Nils-Georg Asp, professor, applied nutrition, Lund University, and director, SNF Swedish Nutrition Foundation, Lund, Sweden, described the wide range of definitions for carbohydrates currently used globally on a governmental basis. For example, Europe defines metabolizable carbohydrates to include polyols, and does not yet have a common definition for dietary fiber. (Note that a dietary fiber definition was a past effort of AACC International, and it, as well as a definition for whole grains, is posted on its website.)

  In the United States, carbohydrates are not yet defined, and there is no accepted definition for dietary fibers. Carbohydrate labeling is an "A" priority for FDA in 2005. Terms such as "digestible and undigestible," "available and unavailable," "dietary fiber," "non-starch polysaccharides," "complex and simple," "intrinsic and extrinsic," and "soluble and insoluble fibers" have been used at one time or another over the years. Because of the expertise of its members, AACC International has been able to establish highly regarded definitions for other components, such as dietary fiber and whole grains, and has taken the lead in establishing accepted methodologies. Comments are a valuable part of this process.

Enter GI One concept that has been at the center of discussion and controversy in the wake of the low-carb frenzy is the concept of glycemic index (GI). At the core of many of these discussions is the experience in Australia. Gareth Hughes, business manager, Glycemic Index, Ltd., Australia, discussed that experience at the Feb. 2005 AACC Symposium. He described the GI Symbol program, which is run by his organization -- a not-for-profit group formed by the University of Sydney, Diabetes Australia, and the Juvenile Diabetes Research Foundation.

At the core of the program is the GI-tested certification mark, which indicates that a certified clinical laboratory has tested a food   according to strict nutritional criteria. Australia's program was started in 2000, and is recognized by 67% of main grocery buyers in Australia. To date, the program has 19 licensees, seven of which are large multinational companies, and the program is seeking to establish partnerships in New Zealand, the European Union and North America.

The primary products seen in Australia that use the system at this time are bread and yogurt. Sweden, South Africa and Japan have also seen use of the GI as a tool, and Tesco grocery stores in the United Kingdom adopted a program in 2004. Tesco is actively trying to educate the public through brochures and recipes.

Defining GI What is the GI, and what are the controversies surrounding it? The GI was introduced in 1981 by David Jenkins, a speaker at the AACC Symposium, and professor, department of nutritional sciences, faculty of medicine, University of Toronto. Jenkins defined GI as "the incremental area under the blood-glucose response curve elicited by a 50-gram available carbohydrate portion of a test food expressed as a percentage of the response elicited by the same amount of carbohydrate from a reference food taken by the same subject."

Jenkins further noted that the reference food "should be taken on three occasions to reduce variability." Carbohydrates that are digested and absorbed quickly have a high GI, but those that are digested or absorbed slowly have a low GI. GI serves, then, as a scale to rank available carbohydrate according to its effect on blood glucose. GI related to foods has been grouped into three generally accepted categories: high GI, 70 or higher; intermediate GI, 56 to 69; and low GI, 55 or lower.

This information, as well as an extensive database, can be found on the University of Sydney's website, www. glycemicindex.com. Although Europe, Sweden, the United Kingdom and South Africa have seen acceptance of this concept, U.S. organizations and groups, such as the American Diabetes Association, the 2005 Dietary Guidelines committee and The American Institute for Cancer Research (AICR), have been slower to fully endorse the concept. The AICR states that "due to insufficient evidence of clinical efficiency and persistent concerns regarding how glycemic index values are determined, AICR cautions the public not to make dietary changes based solely on this interesting, but still unproven, concept."

The AICR cites several methodological concerns, specifically "standardizing how GI values are tested and measured, and using one reference food -- not two reference foods -- as the standard for calculating GI values to avoid inconsistencies in GI tables. Additionally, the variability in GI, whether it's due to the physical structure of the carbohydrate, the inclusion of carbohydrates in the mixed meal, or the variation in an individual's blood sugar response to a carbohydrate, needs to be clarified so that the appropriate dietary guidance on GI can be provided."

Mixed messages It also should be noted that GI is based on available carbohydrate, and the accurate calculation of this is still open to question. As an example, there are within-subject sources of variation (coefficient of variability of about 25% in healthy individuals as reported by Jenkins), source variations (venous versus capillary blood), as well as between-subject variations (an average of at least 10 people is required for the test). In addition to the testing method, many factors can affect the GI of a food, including:

• Type of starch in the food. Amylose is harder to digest than amylopectin, so higher-amylose foods have a lower GI.

• Method of cooking. Gelatinized starch granules are easier to digest. For example, white spaghetti boiled 5 minutes has a GI of 34, while the same spaghetti boiled 10 to 15 minutes has a GI of 40.

• Method of processing. The more the food has been processed, the higher the GI will be. An example of this is finely milled flour (high GI) versus stone-ground flour (low GI). Pumpernickel bread has a GI of 46, while white bread has a GI of 73.

• Fat content. Fat slows gastric emptying, slowing the rate at which carbohydrates are absorbed. This is an appropriate place to point out that GI is not the only factor to consider when judging whether a food is "healthy" -- the GI of French fries is 75, while a baked potato has a GI of 85. If the french fries were fried in saturated fat, which would be more healthful? Thus, GI is only one indicator determining whether a food is healthful.

• Acidity. Acid also slows gastric emptying, with the same effect of decreasing the absorption rate. Vinegar, lemon juice and sourdough breads have shown this benefit. A salad topped with a vinaigrette dressing is a good addition to a meal.

• Protein. Protein also delays gastric emptying.

• Fiber. Viscous, soluble fiber thickens gastric content, slowing digestion and decreasing absorption.

• Sugars. The type of sugar in a food can lower its GI. Glucose has a GI of 100; sucrose (glucose/fructose) is 65. All polyols are low GI, so use of polyols to partially replace sucrose and corn syrups (GI approximately 100) will effectively lower GI.

• Mixed meals. Consumption of a low-GI food with another will lower effective GI, and consumption of a low-GI meal will have the effect of lowering the GI impact of the subsequent meal.

A frequent criticism of GI is that it does not take serving size into account. An example is carrots. To consume 50 grams of available carbohydrates from carrots, the portion would have to be extremely large. The reported GI of carrots, then, is higher than the effect of a normal serving. Researchers at Harvard University, Cambridge, MA, developed the concept of glycemic load (GL) in 1997 to express the amount of available carbohydrate per serving of food. GL, however, is still linked to GI, because GL = GI x grams carbohydrate per serving, so available carbohydrate still must be determined as a part of the method. Since there is still global disagreement on how to calculate available carbohydrate, and because there are no accurate tables or methods to determine this, there is   some potential in both of these concepts for error.

Another recent method worth mentioning is expressed in U.S. Patent Application US2004/0043106A1, "Methods and Systems for Determining and Controlling Glycemic Responses," published March 4, 2004. This is a method proposed by Atkins Nutritionals, Ronkonkoma, NY, and is the basis for its "Net Atkins Count," detailed on its website. In the patent application, the method is termed "equivalent glycemic load," EGL. In this method, a glycemic standard curve is established within a standard comestible (in this case, white bread) by determining glycemic responses of the standard with set glycemic loads. The glycemic response for test foods can be compared to the standard curve. This method correlates glycemic responses at loads below 50 grams extremely well, and allows the consumer to evaluate the product against other foods more directly. Another advantage is that glycemic response is measured directly, without the need to calculate available carbohydrate.

It is clear that, although GI has been accepted in several countries around the globe, methodologies and appropriate terms still are not ready for prime time -- the consumer. Low-carb diets have had a major positive effect, however, as debate within the scientific community is now more open and accelerated.

A healthy debate Another subject of controversy has been the effect of low GI on disease prevention. At the Feb. 2005 AACC Symposium, Janet Warren, research dietician, School of Biological and Molecular Sciences, Oxford, Brookes University, Oxford, England, summarized the impact of low GI on obesity:

• "Further research on GI and energy regulation is required, with an emphasis on long-term well-controlled study designs.

• "Studies need to be done in wider population groups, including ethnic groups.

• "Regular substitution of low-GI foods for high-GI refined products may help prevent excess weight gain.

• "Inclusion of low GI-foods in the diet fits in with current healthy eating recommendations for a diet high in fiber and low in fat."

In terms of effects on diabetes, Warren stated that "epidemiological data suggests a link between the GI of the diet and risk for diabetes," but longer-term studies with more subjects are needed. "Internationally, the recommendation of a low-GI diet for diabetes is controversial," Warren remarked, but "low-GI foods fit with accepted nutritional guidelines for diabetic patients."

Probably the best summation of this point of view was stated by the Food and Nutrition Board of the Institute of Medicine, U.S. National Academy of Sciences. According to the board: "Not all studies of low-GI or low-glycemic -load diets have resulted in beneficial effects; however, none have shown negative effects. There are also theoretical reasons at a time when populations are increasingly obese, inactive, and prone to insulin resistance that dietary interventions that reduce insulin demand may have advantages. In this section of the population, it is likely that more slowly absorbed carbohydrate foods and low-glycemic-load diets will have the greatest advantage." The report concluded that this principle of slow absorption might be a potentially important principle requiring further study.

It is clear from these discussions that, at least on a global basis, much work is needed to establish terminology and methodology. According to Jennie Brand- Miller, professor of human nutrition, School of Molecular and Microbial Biosciences, University of Sidney, Australia: "In Australia and United Kingdom, interest is at an all-time high. In the United States, it's an undercurrent that has been steadily building. The big companies are laying foundations now."

This view was shared by Diane Carnell, R&D director, sweet ingredients and functional bars, Kerry Americas, New Century, KS, who says: "An undercurrent is starting. I don't find the all-consuming 'must have' of low-carb, but more articles and customer requests are starting to filter through." Carnell sees "a leaning toward GI, perhaps not in the scientific meaning of the term, but more of an understanding that, for example, whole grains are better than refined sugars ... satiety will become more important -- this will link into the diet side of the market."

Seeking better strategies How can the GI, or glycemic response, of a food be transformed? Product developers will need to develop a better knowledge of the types of carbohydrates and how they impact glycemic response.

As discussed earlier, product designers can manipulate glycemic response through types of starch in the product, degree of starch gelatinization, level of processing, fat level (although the type of fat is important to the overall health benefit of the food), food acids and proteins. These are all variables that can be manipulated in a formulation, but their manipulation will require clinical evaluation to determine their benefit in a particular food item. Developers also must keep in mind that the goal is to improve diet or food without negatively impacting texture or taste. Therefore, product designers must make changes carefully, and must evaluate the impact of these changes on product shelf life. That has been the failure of the no-fat, high-fiber, no-carb fads of the past -- a healthy product cannot look and taste like cardboard. The goal is not to eliminate sugar from the diet, or to consume all low-GI foods!

According to Brand-Miller, most of the initial interest in low-glycemic formulating has been in breads, rice, confectionery and dairy applications. Most of the confectionery interest has occurred within nutritional bars, and much of the dairy interest has been in yogurts. Various types of rice show a wide difference in GI, based on the availability of starch -- the soft rice in many Chinese dishes has a high GI, while the GI of basmati rice is low. In general, Brand-Miller recommends avoiding corn syrups, glucose and maltodextrins as sweeteners, and using acidic fruits where possible. In breads and other bakery products, oats, whole grains, and viscous fibers such as psyllium, guar and acacia gums should be used whenever possible.

Carnell says: "We have a range of proteins, which can give a protein bar texture, but will not harden over time, while giving a good eating experience and still providing excellent levels of protein to the core. Fiber will be one of the major ingredients over the foreseeable future -- this may be in the form of inulin, FOS, resistant starch, polydextrose or oat fibers, to name a few."

The Calorie Control Council has a position paper titled "Glycemic Index: Summary of Current Status and Future Prospects," posted online at www.caloriecontrol.org/glycemicindex. html. This paper notes: "Sugar alcohols (polyols) such as lactitol, xylitol, isomalt and maltitol have a low glycemic effect, as do fructose, polydextrose and some resistant starches. These ingredients are used extensively to completely or partially replace sucrose, glucose and high-GI polysaccharides, such as starch and maltodextrin, in a wide range of processed foods such as dairy products, baked goods and confectionery. Specialty carbohydrates can have a useful role in reducing the overall glycemic challenge of the diet, and in so doing, may help to reduce the risk of a variety of 'lifestyle' related diseases."

What about some more-specific examples of formulating for low GI? Karen McPhee, senior product development specialist, Guelph Food Technology Centre, Ontario, writes in the Dec. 2004 Bakers Journal that "formulating food products with fibres can be challenging since they impart flavours and also tend to absorb a lot of water, which can cause dryness at high levels. Another possibility might be non-digestible carbohydrates that behave as fibre such as inulin, polydextrose, and resistant starch... Replacing the simple sugars such as sucrose and glucose with alternative sweeteners is another option to reduce the GI of baked products."

Sugar alcohols, such as maltitol, maltitol syrups, isomalt and sorbitol, are often used to replace the bulk of sugar and corn syrups, and also to provide humectancy for improved shelf life. High-potency sweeteners, such as sucralose, aspartame, acesulfame-K and neotame, can boost the level of sweetness. Levels of sugar alcohols should be maintained at 15 or less grams for disaccharides, and 20 or less grams for polysaccharides, to avoid potential single-dose laxative effects that can occur if consumers eat more than one serving of the product, as often happens. McPhee also recommends the addition of whole grains, seeds and nuts to baked goods such as breads, muffins and cookies. Where possible, fruit inclusions also reduce GI. In breads, sourdough fermentation, or the addition of food acids in some other way, also is a plus.

What about proteins? G. Harvey Anderson, Ph.D., professor, nutritional sciences and physiology, University of Toronto, notes, "food intake reduction can be achieved with food proteins and encrypted peptides that stimulate hormones." Anderson discovered that proteins suppress appetite more than carbohydrates and fat, but all proteins are not equal. Studies have shown that people whose diets include dairy products generally have more regulated body weight. According to Anderson, "whey casein holds food intake suppression capacities similar to that of a complete protein. Soybean beta-conglycine peptone (soy protein) also has a strong effect on CCK (a satiety hormone). This offers strong evidence that it is a matter of time before we can produce a functional food that will both enhance satiety and reduce food intake." Satiety is another challenging area of research, which is evolving at a more-rapid pace than in past years.

As has been stressed, we are nearer to defining carbohydrates for labeling than ever before, but time will tell -- reaching scientific agreement is a long process, as is educating the consumer. Methodologies also must be established. In the United States, many companies are improving products by following the recommendations of the 2005 Dietary Guidelines for Americans by reducing sugar and adding fiber -- and these are certainly steps in the right direction. Any glycemic claims will probably require some sort of in vivo clinical validation -- there are no valid in vitro methods established at this time. Users of the Australian GI symbol must conform to tight nutritional guidelines and be clinically validated. GI numbers established for Tesco, in the United Kingdom, also are clinically validated, as are the "Net Atkins Count" symbols on Atkins products. This has not happened elsewhere in the United States. Should it? Would the consumer understand? Stay tuned.

Ronald C. Deis, Ph.D., is the vice president of technology at SPI Polyols, Inc., New Castle, DE. He has 25 years of experience in the food industry, both in food-ingredient and consumer-product companies, and is an active member in a number of trade associations. He has been a short-course speaker, worked as a freelance writer covering a number of food-science-related subjects in food journals, and contributed chapters on sweeteners and fat replacers for several books.

Back to top