QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Monro, J. Search for Related Content PubMed PubMed Citation Articles by Monro, J.

---

Issues and Opinions

Redefining the Glycemic Index for Dietary Management of Postprandial Glycemia

New Zealand Institute for Crop and Food Research Limited, Palmerston North, New Zealand

¹To whom correspondence should be addressed. E-mail: monroj{at}crop.cri.nz

	ABSTRACT

TOP ABSTRACT The limitations of GI... Overcoming the limitations of... Current definition of GI... Proposed extended definition of... Implications of a food-based... LITERATURE CITED

 
The glycemic index (GI) is a component-referenced index, defined as the effect on blood glucose of glycemic carbohydrate in a food as a percentage of the effect of an equal amount of glucose. GI is not suitable for dietary management of postprandial glycemia because it refers to glycemic carbohydrate, not to food, it is static, i.e., unresponsive to intake, and it is relative in that it can be used appropriately only for equicarbohydrate comparisons. Food values for dietary management, in contrast, must be food-based, intake responsive and sensitive to variations in composition typical of diets. The limitations of GI could be overcome by expanding its definition beyond glycemic carbohydrate, to foods. GI could then be used as a food-referenced index, expressed as glycemic glucose equivalents (GGE)/100 g food, and extrapolated to GGE per common standard measure (CSM) as a value for the glycemic load exerted by familiar food quantities. GI_food and GGE/CSM would then be expressed as are nutrients in food composition databases, would lie on the same quantity-dependent continuum and could be used with nutrient information to provide a more complete nutritional profile of a food than is provided by nutrients alone. GI would then be a practical adjunct to food composition data, which was its intended role.

KEY WORDS: • glycemic index • foods • glycemic glucose equivalents • virtual food components • glycemic load

Otto et al. (1) were the first to propose food exchanges by glycemic effect, for control of postprandial glycemia. Later, Jenkins et al. (2) developed the glycemic index (GI)¹ as a datum for use in managing diabetes by carbohydrate exchange. GI is defined as the effect of glycemic carbohydrate in a food on blood glucose as a percentage of the effect of an equal amount of glucose. GI is the independent variable in much research on the relationship of available carbohydrates to health; it is being promoted vigorously as a guide to food choices for the control of postprandial glycemia (3).

The decision that GI should represent the relative glycemic potency of glycemic carbohydrate in food, rather than of food, was an important divergence from the food-based approach of Otto et al., and it placed a number of limitations on the use of GI.

	The limitations of GI in glycemia management

TOP ABSTRACT The limitations of GI... Overcoming the limitations of... Current definition of GI... Proposed extended definition of... Implications of a food-based... LITERATURE CITED

 

1. Food rankings by GI are not rankings by glycemic potency (4), and GI categories cannot accurately guide food choices unless equal amounts of glycemic carbohydrate are involved. Safety and liability issues when using GI-based rankings to promote foods are, therefore, of concern.
   
2. GI does not include a factor for carbohydrate content, whereas glycemic effect is strongly dependent on carbohydrate dose; thus, GI predicts relative food effects only for equal glycemic carbohydrate doses (4). The need to combine GI with intake of glycemic carbohydrate to obtain glycemic load (GL) complicates its usage.
   
3. GI does not change with food quantity; thus, it cannot be easily used to regulate food intakes. It does not behave like nutrients, which usually have dose-dependent effects.
   
4. Measuring GI requires glycemic carbohydrate and area under the blood glucose response curve to be determined and combined; thus, errors in GI are large, and exacerbated when GI is used with carbohydrate intakes to determine GL (5). Furthermore, analytically determined available carbohydrate and glycemic carbohydrate are not the same (6).
   

	Overcoming the limitations of GI

TOP ABSTRACT The limitations of GI... Overcoming the limitations of... Current definition of GI... Proposed extended definition of... Implications of a food-based... LITERATURE CITED

 
For practical glycemic comparisons of foods, a whole-food equivalent to GI is required and may be acquired simply by removing the equicarbohydrate requirement from the definition of glycemic index:

	Current definition of GI (7)

TOP ABSTRACT The limitations of GI... Overcoming the limitations of... Current definition of GI... Proposed extended definition of... Implications of a food-based... LITERATURE CITED

 

	Proposed extended definition of GI

TOP ABSTRACT The limitations of GI... Overcoming the limitations of... Current definition of GI... Proposed extended definition of... Implications of a food-based... LITERATURE CITED

 

A GI value based on response relative to glucose can be expressed as glycemic glucose equivalents (GGE)/100 g of food or of food component (4).

	Implications of a food-based redefinition of GI

TOP ABSTRACT The limitations of GI... Overcoming the limitations of... Current definition of GI... Proposed extended definition of... Implications of a food-based... LITERATURE CITED

 

1. The extended definition of GI would allow it to function either as a component-referenced index (GI_carb), as at present, or as a food-referenced index (GI_food) (8). One need only specify what is being compared with glucose.
   
2. Extending GI to foods would allow food-based dietary management of glycemia without the need for new terms that may add confusion.
   
3. A food basis for GI would improve its relevance because diets are managed with foods, and people consume foods, not the carbohydrate component in isolation.
   
4. Food classifications based on GI_food would accurately reflect relative glycemic potency; thus, unhelpful food choices and incorrect advice arising from current GI_carb-based food classifications (4,9) would be reduced.
   
5. Food comparisons would no longer have to be limited to foods or food quantities containing equal amounts of glycemic carbohydrate, but would relate to a GGE continuum within which GI_food (GGE/100 g food) would be a comparison point at 100 g of food.
   
6. If GI_carb is used as an independent variable in research, carbohydrate digestion must be measured to avoid confounding carbohydrate availability with GI_carb per se. This problem would be eliminated if GI were food based.
   
7. Directly measuring GI_food would greatly improve accuracy because separate measurement of glycemic carbohydrate and blood glucose response to determine GI, and use of the carbohydrate value again with GI to determine GL, with compounding errors, would be avoided.
   
8. Measuring GI (GI_food) directly as a relative food effect would remove uncertainties in measurement and meaning of "glycemic carbohydrate" as a carbohydrate value would not be required.
   
9. A GI_food/GGE-based system of choosing foods would make more sense than the current GI_carb because GGE, as a virtual food component (VFC) (8), responds to changes in food intake in the same way as a nutrient (4).
   
10. A GI_food, or GGE content, or classification based on GI_food, could do all that the current GI_carb does to guide food choices, but accurately and without anomalies.
    

Current tables of glycemic index and glycemic load.

With GI extended to represent the relative glycemic effect of food, GI values in tables of GI and GL (10) could all be retained. Current GI (GI_carb) values need only be identified as referring to carbohydrates in the foods. GI_food values could be estimated from GI_carb and the carbohydrate value used in the original GI_carb calculation. The result would be GGE/100 g food, and the serving weight could then be used to determine GL of a serving as GGE/serving. Derived values would combine errors in GI and carbohydrate values, but could be replaced in time with accurate GI_food values, measured directly as response to relevant amounts of foods. At present, GL has no units. Providing it with GGE units confers meaning in terms of glycemic control and makes clear its basis.

Confounding of GI_carb and available carbohydrate.

Defining GI in terms of glycemic carbohydrate requires that "glycemic carbohydrate" be measured when using GI as an independent variable. Carbohydrate that is glycemic under a particular set of experimental conditions is, however, almost never measured; thus, it is usually impossible to know whether to attribute effects to the amount of carbohydrate available, or to the GI of the carbohydrate that was available. Thus, GI is often confounded with bioavailability.

For instance, studies of high GI vs. low GI starch have compared amylopectin with amylose, without adjusting for the extent of digestion (11). But amylose is a partially resistant starch, and nonglycemic by definition. Therefore, whether observed effects were the result of difference in amount of carbohydrate digested, or difference in GI of the digested carbohydrate, remains unknown.

A clear definition of "glycemic" and "available" carbohydrate is required. "Available carbohydrate" has been the sum of sugars plus starch in food tables, the sum of sugars plus starch digestible under analytical conditions, or carbohydrate by difference, none of which are the same as "glycemic carbohydrate." Does "glycemic carbohydrate" mean carbohydrate that would be glycemic if digestive enzymes could gain access to it and there were no constraints on digesta diffusion? Then, where is the boundary between "glycemic carbohydrate" and dietary fiber? If the distinction is to be physiologic, glycemic carbohydrate will have to be measured in food in its state of existence in the gut, under experimental conditions relevant to the study being undertaken.

Such questions have plagued dietary fiber research for years (12,13), and they cloud the important issue, the effect of food on glycemia. They would be avoided if GI were a food-referenced index.

The properties of GI_carb as a food value.

Food values for self-care require a number of properties embodied in the concept of the "virtual food component" (8). Several criteria used to establish VFC are applied to GI_carb and GI_food in Table 1, and they show the inadequacy of GI_carb for dietary management compared with GI_food.

A food-based glycemic index (GI_food) could be incorporated into food composition databases alongside other food data because it is a value that can be expressed, as in the case of other nutrients, per 100 g food (GI_food = GGE/100 g food), and as GGE/CSM. GI_food values would be an equiquantity (100 g) reference point for comparing foods by relative glycemic potency, within a quantity-related continuum.

Because GGE behaves as a nutrient, it can be used with nutrient information as a VFC, to show food effects concurrently with nutrient intakes (14). To provide a more complete nutritional profile of a food or meal than is presently available for managing complex syndromes, such as diabetes, in which effects that depend on both nutrients and properties of food are important (15). GGE has been shown to predict glycemic response well within individuals (16).

GI_carb values are not appropriate for food composition databases because such databases deal with food values, and GI_carb, unlike GI_food and other food components, does not lie within the intake-dependent continuum of effect.

Personalized management of postprandial glycemia.

Individual responsiveness per GGE is easily established by monitoring responses to foods of known GGE content, and is quite reliable (16). Individuals who have determined their GGE responsiveness by self-calibration could use GI_food values to gauge their likely glycemic response to a food because GI_food and GGE dose are on the same dose-response continuum. The current GI_carb value is more difficult because it must be multiplied by item size and glycemic carbohydrate content to obtain a value for GL.

Options for the future.

The highest quality GI_food data require direct measurement of the glycemic effect of foods relative to glucose, but until such data are available, GI_food values will have to be calculated from GI_carb or estimated in vitro to create interim food data sets, and the errors tolerated.

Alternatively, dietary management could be based more on GGE as a VFC, with GGE/100 g and GGE/serving as reference points, making GI superfluous.

If GI is retained exclusively as a glycemic carbohydrate-based index, another variable may be required, such as relative glycemic potency, which was intended to be a food-based GI (17), and from which GGE were originally derived (14). However, now that the idea of GI is becoming established, it would be better to extend it to food rather than introduce another term of similar meaning.

The current use of carbohydrate-based GI (GI_carb) values to control food intakes is problematic for people with diabetes because they are trying to use a component-referenced index to manage whole foods. In the long term, allowing GI to a be food-referenced index would be a valuable step as the world faces an increasing need for self-management of glycemia.

In conclusion, the following food variables are proposed for use in dietary management of postprandial glycemia:

Glycemic index (GI).

The incremental blood glucose response to specified ingested material, food or food component, as a percentage of the response to an equal weight of glucose. Where necessary, abbreviations such as GI_carb and GI_food could be used to identify the GI as component-referenced vs. food-referenced. GI_food (= GGE/100 g food) is for use in dietary management of glycemia based on foods.

Glycemic glucose equivalent (GGE).

The amount of glucose equivalent in its glycemic effect to a given weight of ingested material, and a VFC that represents glycemic effect, useable with other food composition data.

Glycemic load (GL).

The GGE dose delivered by a given quantity of ingested material.

	FOOTNOTES

 
² Abbreviations used: CSM, common standard measure; GGE, glycemic glucose equivalents; GI, glycemic index; GL, glycemic load; VFC, virtual good component.

Manuscript received 24 August 2003.

	LITERATURE CITED

TOP ABSTRACT The limitations of GI... Overcoming the limitations of... Current definition of GI... Proposed extended definition of... Implications of a food-based... LITERATURE CITED

 

1. Otto, H., Bleyer, G., Pennatz, M., Sabin, G., Schauberger, G. & Spaethe, K. (1973) Kohlenhydrataustauch nach biologischen aquivalanten. (Carbohydrate exchange according to biological equivalents.). Diateik Bei Diabetes Mellitus 1973:41-50 Huber Bern, Switzerland.

2. Jenkins, D.J.A., Wolever, T.M.S., Taylor, R. H., Barker, H., Fielden, H., Baldwin, J. M., Bowling, A. C., Newman, H. C., Jenkins, A. L. & Goff, D. V. (1981) Glycemic index of foods: a physiological basis for carbohydrate exchange. Am. J. Clin. Nutr. 34:362-366.[Abstract/Free Full Text]

3. Brand-Miller, J. C., Wolever, T.M.S., Foster-Powell, K. & Colagiuri, S. (2003) The New Glucose Revolution 2003 The Authoritative Guide to the Glycaemic Index. Marlow and Co New York, NY.

4. Monro, J. A. (2002) Glycaemic glucose equivalent: combining carbohydrate content, quantity and glycaemic index of foods for precision in glycaemia management. Asia Pac. J. Clin. Nutr. 11:217-225.[Medline]

5. Pi-Sunyer, F. X. (2003) Glycemic index and disease. Am. J. Clin. Nutr. 76:290S-298S.

6. Monro, J. A. (2000) Evidence-based food choice: the need for new measures of food effects. Trends Food Sci. Technol. 11:136-144.

7. FAO/WHO (1997) Carbohydrates in Human Nutrition: Report of a Joint FAO/WHO Expert Consultation, 14–18 April 1997 Rome, Italy.

8. Monro, J. A. (2003) Virtual food components: functional food effects expressed as food components. Eur. J. Clin. Nutr. (in press).

9. Monro, J. A. (2001) Glycaemic index–based classifications do not reflect the relative glycemic impact of foods. Proc. Nutr. Soc. N.Z. 24:96-103.

10. Foster-Powell, K., Holt, S. H. A. & Brand-Miller, J. C. (2002) International table of glycemic index and glycemic load values: 2002. Am. J. Clin. Nutr. 76:5-56.[Abstract/Free Full Text]

11. Pawlak, D. B., Bryson, J. M., Denyer, G. S. & Brand-Miller, J. (2001) High glycemic index starch promotes hypersecretion of insulin and higher body fat in rats without affecting insulin sensitivity. J. Nutr. 131:99-104.[Abstract/Free Full Text]

12. Monro, J. A. (2003) Dietary fibre. Nollet, L.M.L. eds. Handbook of Food Analysis 2nd ed. 2003 Marcel Dekker New York, NY (in press).

13. Donnelly, B. (2003) National Academies of Science definitions relating to food fibre only add confusion. Br. J. Nutr. 90:481-483.[Medline]

14. Monro, J. A. & Williams, M. (2000) Concurrent management of postprandial glycaemia and nutrient intake using glycaemic glucose equivalents, food composition data, and computer-assisted meal design. Asia Pac. J. Clin. Nutr. 9:67-73.

15. Ha, T.K.K. & Lean, M.E.J. (1998) Recommendations for the nutritional management of patients with diabetes mellitus. Eur. J. Clin. Nutr. 52:467-481.[Medline]

16. Liu, P., Perry, T. & Monro, J. A. (2003) Glycaemic glucose equivalent: validation as a predictor of the relative glycaemic effect of foods. Eur. J. Clin. Nutr. 57:1141-1149.[Medline]

17. Monro, J. A. (1999) Available carbohydrate data and glycaemic index combined in new data sets for managing glycaemia and diabetes. J. Food. Comp. Anal. 12:71-82.

This article has been cited by other articles:

				H. L Mitchell The glycemic index concept in action Am. J. Clinical Nutrition, January 1, 2008; 87(1): 244S - 246S. [Abstract] [Full Text] [PDF]

				B. J. Venn, A. J. Wallace, J. A. Monro, T. Perry, R. Brown, C. Frampton, and T. J. Green The Glycemic Load Estimated from the Glycemic Index Does Not Differ Greatly from That Measured Using a Standard Curve in Healthy Volunteers J. Nutr., May 1, 2006; 136(5): 1377 - 1381. [Abstract] [Full Text] [PDF]

				D. S. Kronfeld, K. H. Treiber, T. M. Hess, and R. C. Boston Insulin resistance in the horse: Definition, detection, and dietetics J Anim Sci, June 1, 2005; 83(13_suppl): E22 - 31. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Monro, J. Search for Related Content PubMed PubMed Citation Articles by Monro, J.