AJCN EB Program 2010
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Am J Clin Nutr 89: 97-105, 2009. First published December 3, 2008; doi:10.3945/ajcn.2008.26354
 American Journal of Clinical Nutrition, doi:10.3945/ajcn.2008.26354
Vol. 89, No. 1, 97-105, January 2009
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
89/1/97    most recent
ajcn.2008.26354v1
Right arrow Purchase Article
Right arrow View Shopping Cart
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Brand-Miller, J. C
Right arrow Articles by Denyer, G.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Brand-Miller, J. C
Right arrow Articles by Denyer, G.
Agricola
Right arrow Articles by Brand-Miller, J. C
Right arrow Articles by Denyer, G.
© 2009 American Society for Clinical Nutrition

ORIGINAL RESEARCH COMMUNICATION

Glycemic index, postprandial glycemia, and the shape of the curve in healthy subjects: analysis of a database of more than 1000 foods1,2

Jennie C Brand-Miller, Karola Stockmann, Fiona Atkinson, Peter Petocz and Gareth Denyer

1 From the Institute of Obesity, Nutrition and Exercise (JCB-M), and the School of Molecular and Microbial Biosciences (JCB-M, KS, FA, and GD), The University of Sydney, Sydney, Australia, and the Department of Statistics, Macquarie University, Sydney, Australia (PP).

2 Reprints not available. Address correspondence to JC Brand-Miller, Human Nutrition Unit, G08, The University of Sydney, Sydney, NSW, 2006. E-mail: j.brandmiller{at}mmb.usyd.edu.au.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 SUBJECTS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Background: The glycemic index (GI) characterizes foods by using the incremental area under the glycemic response curve relative to a similar amount of oral glucose. Its ability to differentiate between curves of different shapes, the peak response, and other aspects of the glycemic response is debatable.

Objective: The objective was to explore the association between a food's GI and the shape of the curve in healthy individuals.

Design: A large database of 1126 foods tested by standardized GI methodology in 8–12 healthy subjects was analyzed systematically. Each food's absolute and incremental blood glucose concentrations were compared at individual time points with the GI. The average curve was generated for low-GI (≤55), medium-GI (56–69), and high-GI (≥70) foods within major food categories.

Results: The GI of individual foods was found to correlate strongly with the incremental and actual peak (Spearman's correlations of r = 0.76 and r = 0.73, respectively), incremental and actual glucose concentration at 60 min (r = 0.70 and r = 0.66, respectively), and maximum amplitude of glucose excursion (r = 0.68) (all P < 0.001). In contrast, there was only a weak correlation between the food's GI and the 120-min glucose concentration (incremental r = 0.20, P < 0.001; absolute r = 0.16, P < 0.001). Within food groups, the mean GI, 30- and 60-min glucose concentrations, and maximum amplitude of glucose excursion varied significantly for foods classified as having a low, medium, or high GI (P < 0.001).

Conclusions: The GI provides a good summary of postprandial glycemia. It predicts the peak (or near peak) response, the maximum glucose fluctuation, and other attributes of the response curve.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
SUBJECTS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
In medical nutrition therapy, the American Diabetes Association emphasizes the amount of carbohydrate required to optimize glucose control and slow the development of complications (1). However, a large body of evidence indicates that the carbohydrate source is also an important consideration (2, 3). Changes in glycemia after the ingestion of equivalent carbohydrate exchanges vary widely, depending on food factors such as the final physical and chemical state of the starch (4), the viscosity and structural integrity of the fiber (5), and the nature of the monosaccharides and disaccharides present (6).

The glycemic index (GI) of foods was developed in an attempt to systematically classify the carbohydrates in different foods according to time-integrated effects on postprandial glycemia. Foods with high GI values have been shown to be more rapidly digested and absorbed, causing greater fluctuations in blood glucose per unit of carbohydrate than foods with lower GI values (7). Developments in food processing over the past 50 y, such as high-temperature, high-pressure extrusion technology, increase the degree of starch gelatinization, which results in easier enzymic accessibility and faster digestion (8, 9). As a consequence, the GI of many modern starchy foods falls within the high range. Even among whole-grain breads and breakfast cereals, GI values can vary widely (10).

The GI is defined methodologically as the incremental area under the curve (AUC) for the blood glucose response after consumption of a food relative to that produced by a reference food given in an equivalent carbohydrate amount (50 or 25 g) (11). Some investigators have expressed concern that use of the AUC, which integrates the entire blood glucose response over time, may not be appropriate because it fails to take into account the shape of the curve (12, 13). Theoretically, it is possible that the carbohydrates in some foods elicit a sharp glucose "spike" that disappears quickly and undershoots the baseline value and yet have the same AUC as the carbohydrates that provoke a more gradual rise and fall in blood glucose. The distinction is important because emerging data suggest that the postprandial glycemic spike and degree of blood glucose fluctuation per se may be more clinically adverse than sustained glycemia (14, 15).

For these reasons, we explored the association between a food's GI and the shape of the curve in healthy individuals. Our working hypothesis was that the GI would reliably predict multiple attributes of the glycemic response. We considered the absolute and incremental peak glucose value (the "spike"); the 60-, 90-, and 120-min values; and the maximum excursion within 120 min (ie, maximum amplitude of the glucose excursion, or MAGE). We analyzed the absolute blood glucose responses recorded in a large database of foods that had been tested by our laboratory using the standard GI protocol. Average glucose curves were then generated for large numbers of foods classified as having a low, medium, or high GI within major food groups.


   SUBJECTS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 SUBJECTS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Since 1995, the University of Sydney has conducted contract GI testing for the food industry alongside research on the GI and its relevance to health. Between 2001 and June 2007, raw data were routinely entered into a specially developed database that permitted systematic analysis of all aspects of GI testing. During that period, 1126 individual food items were tested in healthy subjects according to the standard GI testing protocol; the responses to the test foods were compared with the responses to a reference food (16). In most cases, the reference food was 50 g glucose dissolved in 250 mL water and tested on 2 or 3 separate occasions within a 3-mo period for each subject. In a small minority of cases, 25-g carbohydrate portions were used to accommodate more dilute sources of carbohydrate. The protocol was approved by the University of Sydney Human Research Ethics Committee, and the subjects gave written informed consent.

For each test, 8–12 healthy subjects were recruited and presented to the metabolic kitchen in the morning after a 10–12-h fast. They consumed the food with 250 mL water within 10–15 min, and finger-prick capillary blood was collected at 0 (start of the meal), 15, 30, 45, 60, 90, and 120 min. Glucose was determined in duplicate either with HemoCue (HemoCue AB, Ängelholm, Sweden) (17) in whole blood or in plasma after 0.8 mL blood was collected into an Eppendorf tube containing 10 U heparin. In the latter case, plasma was separated after centrifugation and cooled on ice until same-day assay with the glucose hexokinase assay on a centrifugal analyzer. For each test, the incremental AUC was calculated according to the trapezoidal method and compared with the average AUC (n = 2 or 3) for the reference food in that particular subject (16). Any area under the baseline (fasting value) was ignored. The GI was determined as the mean ± SE in 8–12 different subjects and rounded to the nearest integer. In instances in which the value for any one individual was unusually large or small (>2 SD from the mean), the value was removed and the mean ± SE was recalculated. Thus, the GI of any one food was based on ≥480 separate glucose determinations (10 subjects, 8 time points, duplicate assays, 2 reference foods, and 1 test food). In the database, 51% of the foods were compared with 3 reference tests (glucose standard) and the remainder against 2. The within-subject CV for the 2–3 tests of the reference food was 17.4 ± 5.4%.

The database
A relational database was created by using FileMaker Pro 6 (FileMaker Inc, Santa Clara, CA) to link information regarding the subjects (eg, sex, ethnicity, height, and relevant medical information) and foods (eg, nutrient composition and food group classification) with test data (eg, age and weight of subjects and test conditions) and all assay results (ie, the glucose result at each time point). The database automatically calculated all of the variables investigated in this study (AUC, GI, CV, peaks and nadirs, MAGE, and all other relative and absolute glucose values for each test).

Analysis of the database
In the first analysis, all foods that had been GI-tested according to the standard protocol were considered (n = 1126). To determine the ability of the GI to predict postprandial glycemia at any one time point, the GI was correlated against both the absolute concentration and the incremental change in blood glucose at 30, 60, and 120 min. The degree of fluctuation, also called MAGE, was determined as the difference between the highest and lowest concentration over 120 min.

For the second analysis, all foods that could be categorized into discrete food groups were included (n = 780 foods). Twelve types of foods were identified: potatoes, breads (white and whole-wheat), breakfast cereals, rice, pasta, snack bars, potatoes, dairy products, fruit, fruit juices, and vegetables (other than potato). Within each category, foods were classified as high GI (≥70), medium GI (56–69), or low GI (≤55) according to recommendations of Standards Australia (16). The average curve was generated from the mean data at each time point. The peak responses and blood glucose concentrations at 60, 90, and 120 min for the high-, medium-, and low-GI foods within each category were then compared.

Statistical analysis
Spearman's correlation coefficients were calculated between the GI and MAGE and the glucose concentration at the 30-, 60-, 90-, and 120-min time points (absolute and incremental). Where relations appeared to be only approximately linear, loess curves were fitted to indicate the actual shape of the relation. Other relations were also explored, including quadratic and cubic equations. One-way analyses of variance were used to compare differences between foods classified as low GI, medium GI, and high GI with pairwise comparisons subject to Bonferroni corrections for multiple tests. All statistical analyses were carried out by using the SPSS statistical package (version 14.0; SPSS Inc, Chicago, IL), and tests used a significance level of 0.05.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 SUBJECTS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
A total of 546 different volunteers (58% female) participated in GI testing over the time frame of this study. Their mean ± SD age was 24.0 ± 4.8 y and body mass index (in kg/m2) was 22.2 ± 2.3. The mean (± SD) fasting glucose concentration was 5.11 ± 0.36 mmol/L in plasma (n = 8056) and 4.57 ± 0.47 mmol/L in whole blood (n = 3841). Most subjects were of European white origin (74%), and the largest minority group was Chinese (18%).

Within the total database (n = 1126 foods), the GI value was found to correlate strongly with the food's incremental peak (Spearman's r = 0.76), actual peak (r = 0.73), incremental glucose concentration at 60 min (r = 0.70), actual 60-min glucose concentration (r = 0.66), incremental 90-min glucose concentration (r = 0.43), actual 90-min glucose concentration (r = 0.34), and MAGE (r = 0.68); P < 0.001 for all (Figure 1). Actual peak concentration and MAGE were highly correlated (r = 0.83, P < 0.001). In contrast, only a weak correlation between the food's GI and the 120-min glucose concentration was found (incremental r = 0.20, P < 0.001; absolute r = 0.16, P < 0.001; Figure 2A); although weak, these correlations were significant because of the large amount of data.


Figure 1
View larger version (14K):
[in this window]
[in a new window]

  		FIGURE 1. Scattergrams with fitted Loess curves that show the relation between (A) absolute peak glucose concentration compared with the glycemic index (GI) of each food (n = 1126; Spearman's r = 0.73, P < 0.001, (B) absolute 60-min glucose concentration compared with the GI of each food (n = 1126; r = 0.66, P < 0.001), and (C) the maximum amplitude of the glucose excursion (MAGE) compared with the GI of each food (n = 1126; r = 0.68, P < 0.001. Each food was tested in differing groups of 8–12 normal subjects.

 

Figure 2
View larger version (34K):
[in this window]
[in a new window]

  		FIGURE 2. A: Scattergram with fitted Loess curve that shows absolute 120-min glucose concentration compared with the glycemic index (GI) of each food (n = 1126; Spearman's r = 0.14, P < 0.001. B: The rank order of 120-min incremental blood glucose concentration across 34 different food groups. Starchy foods generally gave higher values than foods rich in simple sugars, whether these were naturally occurring or added.

 
We also explored nonlinear alternatives, but found little improvement. For example, for the 60-min time point versus GI, r2 was 0.40 for linear, 0.40 for quadratic, and 0.43 for cubic fits. Nonparametric loess curves were fitted to indicate the actual nature of the relation (Figures 1 and 2) and suggested a steeper relation between absolute peak versus GI when the GI was less than {approx}55 (Figure 1A). For absolute glucose concentration at 60 min, the relation was steepest in the intermediate range (Figure 1B). For MAGE (Figure 1C), the relation was less steep for GI values above {approx}40. No consistent pattern was apparent at the 120-min time point.

With one exception, starchy foods, irrespective of their GI, tended to remain above the baseline level at 120 min, whereas foods high in sugars (both naturally occurring and added) tended to undershoot the baseline (Figure 2B). The foods with the highest 120-min values were the high-GI rices (mean difference from baseline for 22 products = 0.4 mmol/L) and high-GI whole-grain breads (mean incremental level in 8 products = 0.4 mmol/L). The foods with the lowest 120-min values were soft drinks (mean difference from baseline for 7 products = –0.5 mmol/L) and fruit juices (mean difference from baseline for 29 products = –0.5 mmol/L). Potato products were the one exception and behaved like sugary foods, undershooting the baseline by 120 min (mean of 9 products = –0.4 mmol/L).

The average curve shapes for the high-, medium-, and low-GI foods within 10 major food categories are illustrated in Figure 3 and Figure 4. Universally, the peak glucose values occurred at 30 min, and, almost always, the lowest reading was recorded at 120 min. When dairy products were tested in 25-g carbohydrate portions, the nadir occurred at 60 min. Potatoes as a group had the highest GI (mean of 9 products = 90), followed by the high-GI breakfast cereals (mean of 29 products = 82; Table 1). The foods with the lowest GI values included legumes (mean of 11 products = 38), protein bars (mean GI of 59 products = 28), and plain milk (mean of 13 products = 25). Notably, no soft drinks, fruit juices, or dairy products fell into the high-GI category. Whole-grain breads and breakfast cereals were available in all 3 GI categories. Foods classified as high GI, medium GI, or low GI within each category usually differed significantly with respect to mean GI and 30-, 60-, and 90-min glucose concentrations (absolute and incremental) (Table 1).


Figure 3
View larger version (33K):
[in this window]
[in a new window]

  		FIGURE 3. A–H: Incremental blood glucose profiles for high, medium, and low glycemic index (GI) foods within 8 different food categories. n = number of foods that were tested in each category in 8–12 subjects. All tests were 50-g portions of carbohydrate.

 

Figure 4
View larger version (20K):
[in this window]
[in a new window]

  		FIGURE 4. Incremental blood glucose profiles for high, medium, and low glycemic index (GI) of dairy products (A and B) and vegetables (C). n = number of foods that were tested in each category in 8–12 subjects. Some dairy product tests were given with both 25- and 50-g portions of carbohydrate, as indicated.

 

View this table:
[in this window]
[in a new window]

  		TABLE 1. Plasma glucose responses to consumption of foods with a low, medium, or high glycemic index (GI)1

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 SUBJECTS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
The current analysis was the first attempt to systematically characterize the relation between a food's GI and other attributes of the postprandial glycemic response. It was shown that the GI, a number derived from the incremental AUC, was correlated with the absolute peak response (the glycemic spike), the glucose concentration at 60 and 90 min, and MAGE (the difference between the peak and the nadir). Although these relations were approximately linear, loess curves indicated that the steepness of the relation varied somewhat over the full GI range. In univariate analysis, each variable showed a positive correlation with GI and accounted for 50–60% of the variation. Consequently, the 2 food groups with the highest GI values (potatoes and breakfast cereals) also had among the highest glucose peaks (absolute peak: 8.7 and 7.7 mmol/L, respectively). The foods with the lowest GI values (plain milk and protein bars) were also the 2 categories that had among the lowest peak responses (absolute peak: 5.5 and 6.1 mmol/L, respectively). MAGE ranged from a high of 4 mmol/L for potatoes to only 1.3 mmol/L for legumes and 1.1 mmol/L for protein bars.

This analysis was also the first to explore differences in the shape of the curve of within- and between-food categories. As is evident in Figures 3 and 4, the general curve shape is similar within each food category even though the AUCs differ. The notion that low-GI foods produce a sustained rise in blood glucose was not supported. Moreover, high-GI foods all differed significantly from similar foods with a low GI at several time points (Table 1). Although there was also a positive relation between the GI and the 2-h glucose concentration, the correlation coefficient was weak (r = 0.16), and considerable heterogeneity was observed between the different types of foods. In general, sugary foods, including soft drinks and fruit juices, irrespective of GI, were more likely to return to baseline sooner (between 60 and 75 min) and to show an undershoot at the 120-min time point (Figure 2). In contrast, with the exception of potatoes, starchy foods tended to remain above baseline at 120 min. These findings can be explained by the fact that simple sugars such as sucrose, lactose, and fructose—whatever their source—contribute fewer glucose moieties than the same weight of starch. Fifty grams of sucrose, for example, contains only 25 g of glucose equivalents versus 50 g of glucose equivalents in starch. Compared with 50 g glucose, 25 g glucose produces a curve that returns to baseline earlier (Figure 4). Although the difference was subtle, the "undershoot" seen for fruit juices, soft drinks, and potatoes may have physiologic ramifications for appetite and weight control (18).

The strengths of this analysis include the size of the database, the large number of food categories examined, the use of standardized GI protocols with 2–3 reference foods tested for each test food, and the use of finger-prick capillary sampling. It is not yet widely recognized that changes in blood glucose immediately after a meal may be identified more readily at finger sites than at the forearm (19). The use of finger-prick blood sampling is also associated with less variability in resulting GI values (20). Our subjects also showed low intraindividual (within-subject) variability for repeat tests of the reference food ({approx}17%), which is substantially lower than that recorded by others; for example, Vega-Lopez et al (21) reported a variability of {approx}43%. The limitations, however, included the fact that the database was built on the testing of single foods in healthy individuals. This may make the findings less generalizable to mixed meals and to individuals with diabetes, although there is abundant evidence of support (22–25). The findings are also derived from the average of many foods, and the scattergrams clearly suggest wide prediction intervals. These data should therefore not be used to predict responses to individual foods or groups of foods.

A potential limitation of this analysis was the assumption that one large group of subjects has approximately the same degree of glucose tolerance as another large group. This assumption is not appropriate when small numbers of subjects (n {approx} 10) are involved; however, in the present context, the mean glycemic response within each food category (eg, high-GI white breads) was often the average of {approx}200 or more individuals. As a group, glucose tolerance in these {approx}200 subjects was no different from that of any other group of {approx}200 lean young healthy subjects (data not shown). Significant heterogeneity among subjects would work against the chance of detecting differences at various time points of the glycemic response.

Our study was unique from the standpoint of defining "normal" postprandial glycemia. The results indicate that the normal response depends on the choice of carbohydrate food. For example, MAGE varied over a 4-fold range and the peak varied >3-fold. The threshold for unacceptable postprandial glycemia defined by the American Diabetes Association and World Health Organization is 8.89 mmol/L (>160 mg/dL) at any time after the meal (25); however, in the present analysis, potatoes gave an average absolute peak of 8.67 mmol/L in healthy individuals consuming 50 g carbohydrate. Realistically, meals often contain larger quantities of carbohydrate. Postmeal glycemia. as elicited by modern carbohydrate foods, can therefore contribute to the dysglycemic "iceberg" or hidden glycemia that may be more important over the longer term than an increase in fasting glucose.

Mounting evidence suggests that the postprandial state contributes to the development of chronic disease, particularly atherosclerosis (26). In individuals with diabetes or prediabetes, the postprandial phase is characterized by a large and more prolonged increase in blood glucose concentrations. Observational studies have shown that postchallenge or postmeal glucose peaks are an independent risk factor for cardiovascular disease (14, 27, 28). Many cardiovascular disease risk factors, including coagulatory proteins, are modified in the postprandial phase in diabetic subjects and are directly affected by an acute increase in glycemia (29). Postmeal glycemic spikes within the normal range can also act directly to increase oxidative stress and inflammatory responses (30). Recently, we reported that increases in nuclear transcription factor-{kappa}B activation (an inflammatory marker) in mononuclear cells of healthy individuals were 3-fold higher after a high-GI carbohydrate meal or glucose challenge than after a low-GI carbohydrate meal (31).

In summary, this analysis supports the hypothesis that the GI reliably predicts multiple attributes of the glycemic response. We found that the overall shape of postprandial glycemia is similar for foods categorized as having a low, medium, or high GI according to Standards Australia criteria (16). The notion that a low-GI food has a uniquely long tail or extended glucose profile is not correct. Although a slowly digested starch or sugar may represent a slow-release form of energy, this does not imply that a low-GI food produces a sustained glucose response. The message to eat whole grains can result in the intake of carbohydrates with a GI no different from that in response to the consumption of white bread. Finally, if a reduction in postprandial glycemia is to be part of the strategy for preventing and managing diabetes and cardiovascular disease, the GI is as relevant as the amount of carbohydrate.


   ACKNOWLEDGMENTS
 
The authors' responsibilities were as follows—JCB-M and GD: conceived the study; GD and FA: created the database; and JCB-M, KS, and PP: analyzed the data. All authors contributed to the interpretation of the findings and the writing of the manuscript. JBM is a co-author of The New Glucose Revolution book series (Hodder and Stoughton, London, United Kingdom), director of a nonprofit GI-based food-endorsement program in Australia, and director of the University of Sydney GI testing service. FA and KS were employed by the University of Sydney for the purposes of commercial GI testing. PP and GD had no conflicts of interest to declare.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 SUBJECTS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. American Diabetes Association. Nutrition recommendations and interventions for diabetes: a position statement of the American Diabetes Association. Diabetes Care 2008;31:S61–78..[Free Full Text]
  2. Brand-Miller, J, Hayne, S, Petocz, P & Colagiuri, S. Low-glycemic index diets in the management of diabetes: a meta-analysis of randomised controlled trials. Diabetes Care 2003;26:2261–7..[Abstract/Free Full Text]
  3. Opperman, AM, Venter, C, Oosthuizen, W, Thompson, R & Voster, H. Meta-analysis of the health effects of using the glycaemic index in meal-planning. Br J Nutr 2004;92:367–81..[Medline]
  4. Gallant, DJ, Bouchet, B, Buleon, A & Perez, S. Physical characteristics of starch granules and susceptibility to enzymatic degradation. Eur J Clin Nutr 1992;46(Suppl):S3–16..
  5. Granfeldt, Y, Hagander, B & Bjorck, I. Metabolic responses to starch in oat and wheat products: on the importance of food structure, incomplete gelatinization or presence of viscous dietary fibre. Eur J Clin Nutr 1995;49:189–99..[Medline]
  6. Wahlqvist, ML, Wilmshurst, E & Richardson, E. The effect of chain length on glucose absorption and the related metabolic response. Am J Clin Nutr 1978;31:1998–2001..[Abstract/Free Full Text]
  7. Jenkins, DJA, Jenkins, AL, Wolever, TM, Collier, GR, Rao, AV & Thompson, LU. Starchy foods and fiber: reduced rate of digestion and improved carbohydrate metabolism. Scand J Gastroenterol Suppl 1987;129:132–41..[Medline]
  8. Brand, JC, Nicholson, P, Thorburn, A & Truswell, A. Food processing and the glycemic index. Am J Clin Nutr 1985;42:1192–6..[Abstract/Free Full Text]
  9. Ross, SW, Brand, J, Thorburn, A & Truswell, A. Glycemic index of processed wheat products. Am J Clin Nutr 1987;46:631–5..[Abstract/Free Full Text]
  10. Foster-Powell, K, Holt, SH & Brand-Miller, JC. International table of glycemic index and glycemic load values: 2002. Am J Clin Nutr 2002;76:5–56..[Abstract/Free Full Text]
  11. Jenkins, DJ, Wolever, TM, Taylor, RH, et al.. Glycemic index of foods: a physiological basis for carbohydrate exchange. Am J Clin Nutr 1981;34:362–6..[Abstract/Free Full Text]
  12. Franz, MJ. The glycemic index: not the most effective nutrition therapy intervention. Diabetes Care 2003;26:2466–8..[Free Full Text]
  13. Pi-Sunyer, FX. Glycemic index and disease. Am J Clin Nutr 2002;76:290S–8S..[Abstract/Free Full Text]
  14. Temelkova-Kurktschiev, TS, Koehler, C, Henkel, E, Leonhardt W, Fuecker, K & Hanefeld, M. Postchallenge plasma glucose and glycemic spikes are more strongly associated with atherosclerosis than fasting glucose or HbA1c level. Diabetes Care 2000;23:1830–4..[Abstract/Free Full Text]
  15. Monnier, L, Mas, E, Ginet, C, et al.. Activation of oxidative stress by acute glucose fluctuations compared with sustained chronic hyperglycemia in patients with type 2 diabetes. JAMA 2006;295:1681–7..[Abstract/Free Full Text]
  16. Standards Australia. The glycemic index of foods. Available from: http://www.saiglobal.com/shop/script/Details.asp?DocN=AS0733779662AT (cited 16 July 2007).
  17. Ashworth, L, Gibb, I & Alberti, K. HemoCue: evaluation of a portable photometric system for determining glucose in whole blood. Clin Chem 1992;38:1479–82..[Abstract/Free Full Text]
  18. Campfield, LA & Smith, F. Blood glucose dynamics and control of meal initiation: a pattern detection and recognition theory. Physiol Rev 2003;83:25–58..[Abstract/Free Full Text]
  19. Ellison, J, Stegmann, J, Colner, S, et al.. Rapid changes in blood glucose produce concentration differences at finger, forearm, and thigh sampling sites. Diabetes Care 2002;25:961–4..[Abstract/Free Full Text]
  20. Wolever, T, Brand-Miller, J, Abernethy, J, et al.. Measuring the glycemic index of foods: interlaboratory study. Am J Clin Nutr 2008;87:247S–57S..[Abstract/Free Full Text]
  21. Vega-Lopez, S, Ausman, L, Griffith, J & Lichtenstein, A. Interindividual variability and intraindividual reproducibility of glycemic index values for commercial white bread. Diabetes Care 2007;30:1412–7..[Abstract/Free Full Text]
  22. McMillan-Price, J, Petocz, P, Atkinson, F, et al.. Comparison of 4 diets of varying glycemic load on weight loss and cardiovascular risk reduction in overweight and obese young adults: a randomised controlled trial. Arch Intern Med 2006;166:1466–75..[Abstract/Free Full Text]
  23. Wolever, T, Yang, M, Zeng, X, Atkinson, F & Brand-Miller, J. Food glycemic index, as given in glycemic index tables, is a significant determinant of glycemic responses elicited by composite breakfast meals. Am J Clin Nutr 2006;83:1306–12..[Abstract/Free Full Text]
  24. Wolever, T, Jenkins, D, Josse, R, Wong, G & Lee, R. The glycemic index: similarity of values derived in insulin-dependent and non-insulin-dependent diabetic patients. J Am Coll Nutr 1987;6:295–305..[Abstract]
  25. Nansel, TR, Gellar, L & McGill, A. Effect of varying glycemic index meals on blood glucose control assessed with continuous glucose monitoring in youth with type 1 diabetes on basal-bolus insulin regimens. Diabetes Care 2008;31:695–7..[Abstract/Free Full Text]
  26. Heine, R, Balkau, B, Ceriello, A, Del Prato, S, Horton, E & Taskinen, M-R. What does postprandial hyperglycaemia mean? Diabet Med 2004;21:208–13..[Medline]
  27. Esposito, K, Ciotola, M, Carleo, D, et al.. Post-meal glucose peaks at home associate with carotid intima-media thickness in type 2 diabetes. J Clin Endocrinol Metab 2008;93:1345–50..[Abstract/Free Full Text]
  28. Hanefeld, M, Koehler, C, Schaper, F, Fuecker, K, Henkel, E & Temelkova-Kurktschiev, T. Postprandial plasma glucose is an independent risk factor for increased carotid intima-media thickness in non-diabetic individuals. Atherosclerosis 1999;144:229–35..[Medline]
  29. Ceriello, A. Postprandial glucose regulation and diabetic complications. Arch Intern Med 2004;164:2090–5..[Abstract/Free Full Text]
  30. Brownlee, M. Biochemistry and molecular biology of diabetes complications. Nature 2001;414:813–20..[Medline]
  31. Dickinson, S, Hancock, D, Petocz, P, Ceriello, A & Brand-Miller, J. High glycemic index carbohydrate increases NF-kB activation in mononuclear cells of young lean healthy subjects. Am J Clin Nutr 2008;87:1177–83..
Received for publication May 1, 2008. Accepted for publication September 18, 2008.





This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
89/1/97    most recent
ajcn.2008.26354v1
Right arrow Purchase Article
Right arrow View Shopping Cart
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Brand-Miller, J. C
Right arrow Articles by Denyer, G.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Brand-Miller, J. C
Right arrow Articles by Denyer, G.
Agricola
Right arrow Articles by Brand-Miller, J. C
Right arrow Articles by Denyer, G.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS