Hostname: page-component-7c8c6479df-hgkh8 Total loading time: 0 Render date: 2024-03-28T18:11:39.528Z Has data issue: false hasContentIssue false

Glycaemic index and glycaemic load values of cereal products and weight-management meals available in the UK

Published online by Cambridge University Press:  01 July 2007

C. Jeya K. Henry*
Affiliation:
Nutrition and Food Science Group, School of Life Sciences, Oxford Brookes University, Gipsy Lane Campus, Headington, Oxford, OX3 0BP, UK
Helen J. Lightowler
Affiliation:
Nutrition and Food Science Group, School of Life Sciences, Oxford Brookes University, Gipsy Lane Campus, Headington, Oxford, OX3 0BP, UK
Lis M. Dodwell
Affiliation:
Nutrition and Food Science Group, School of Life Sciences, Oxford Brookes University, Gipsy Lane Campus, Headington, Oxford, OX3 0BP, UK
Jacqueline M. Wynne
Affiliation:
Nutrition and Food Science Group, School of Life Sciences, Oxford Brookes University, Gipsy Lane Campus, Headington, Oxford, OX3 0BP, UK
*
*Corresponding author: Professor C. J. K. Henry, fax +44 (0) 1865 483618, email jhenry@brookes.ac.uk
Rights & Permissions [Opens in a new window]

Abstract

There is currently an increased global interest in the published glycaemic index (GI) values of foods. The aim of the present work was to supplement a previous study on the glycaemic response of 140 foods available in the UK by studying a further forty-four foods. One hundred and twenty-two healthy subjects, with a mean age of 32·4 (sd 11·4) years and a mean BMI of 23·6 (sd 3·6) kg/m2, were recruited to the study. Subjects were served equivalent available carbohydrate amounts (50 or 30 g) of test foods (cereal products and weight-management meals) and a standard food (glucose) on separate occasions. Capillary blood glucose was measured from finger-prick samples in fasted subjects (0 min) and at 15, 30, 45, 60, 90 and 120 min after starting to eat each test food. For each test food, the GI value was determined, and the glycaemic load was calculated as the product of the GI and the amount of available carbohydrate in a reference serving size. The GI values of the foods tested ranged from 23 to 83. Of the forty-four foods tested, thirty-three were classified as low-GI, eight as medium-GI and three as high-GI foods. Most GI values of the foods tested compared well with previously published values for similar foods. In summary, this study provides reliable GI and glycaemic load values for a range of foods, further advancing our understanding of the glycaemic response of different foods. The data reported here make an important addition to published GI values.

Type
Full Papers
Copyright
Copyright © The Authors 2007

The glycaemic index (GI), first introduced by Jenkins and colleagues (Reference Jenkins, Wolever, Taylor, Barker, Fielden, Baldwin, Bowling, Newman, Jenkins and Goff1981), is a classification of the blood glucose-raising potential of carbohydrate foods. It is defined as the incremental area under the blood glucose curve of a 50 g carbohydrate portion of a test food expressed as a percentage of the response to 50 g carbohydrate of a reference food taken by the same subject, on a different day (FAO/WHO, 1998). The principle is that the slower the rate of carbohydrate absorption, the lower the rise in blood glucose level and the lower the GI value (Augustin et al. Reference Augustin, Franceschi, Jenkins, Kendall and La Vecchia2002). Indeed, high-GI foods are characterised by fast-release carbohydrate and higher blood glucose levels. A GI value of 70 or more is considered high, one of 56–69 is medium and one of 55 or less is low (where glucose = 100; Brand-Miller et al. Reference Brand-Miller, Foster-Powell and Colagiuri2003).

Since the concept of GI was first introduced, many studies have investigated the potential health benefits of low-GI foods. Recent data support the preventive potential of a low-GI diet against the development of type 2 diabetes and cardiovascular disease (Salmeron et al. Reference Salmeron, Ascherio, Rimm, Colditz, Spiegelman, Jenkins, Stampfer, Wing and Willet1997a, Reference Salmeron, Manson, Stampfer, Colditz, Wing and Willetb; Frost et al. Reference Frost, Leeds, Dore, Madeiros, Brading and Dornhorst1999). There is also an interest in the potential of low-GI diets for body-weight management. Several studies have shown that low-GI foods, or lowering the GI of a food, reduces hunger and results in a lower energy intake (Ludwig, Reference Ludwig2000; Warren et al. Reference Warren, Henry and Simonite2003).

GI values represent the glycaemic response of equivalent available carbohydrate amounts of foods and are therefore not always representative of the glycaemic effect of a typical serving of that food. To quantify the overall glycaemic effect of a standard portion of food, the concept of glycaemic load (GL) was introduced (Salmeron et al. Reference Salmeron, Ascherio, Rimm, Colditz, Spiegelman, Jenkins, Stampfer, Wing and Willet1997a, Reference Salmeron, Manson, Stampfer, Colditz, Wing and Willetb). This is the product of the amount of available carbohydrate in that serving and the GI of the food divided by 100. It is often necessary to consider the GL alongside GI values, especially when the carbohydrate content of the food is relatively small. A GL value of 10 or less is considered low, a GL value of 11–19 is medium and one of 20 or more is high (Brand-Miller et al. Reference Brand-Miller, Foster-Powell and Colagiuri2003).

Carbohydrate foods consumed in equivalent available carbohydrate amounts produce different glycaemic responses depending on many factors, such as particle size, cooking and food processing, other food components (e.g. fat, protein, dietary fibre), the proportion and type of sugars and starch, and the starch structure (Björck et al. Reference Björck, Granfeldt, Liljeberg, Tovar and Asp1994). Consequently, there is often considerable variation in the GI of the same food produced in different countries or by different manufacturers. The publication of reliably measured GI and GL values is needed to facilitate consumer application and to reduce unnecessary regional duplication. Until recently, the vast majority of published GI values have been Australasian or Canadian (Foster-Powell et al. Reference Foster-Powell, Holt and Brand-Miller2002). Henry et al. (Reference Henry, Lightowler, Strik, Renton and Hails2005) published a paper detailing GI and GL values for 140 foods commonly consumed in the UK. Thus, the aim of the current work was to provide additional GI and GL values for a wider range of foods available in the UK.

Methods

Subjects

One hundred and twenty-two healthy subjects, with a mean age of 32·4 (sd 11·4) years and a mean BMI of 23·6 (sd 3·6) kg/m2, were recruited via posters distributed throughout Oxford Brookes University, in addition to announcements in lectures and through personal networks. Exclusion criteria were as follows: age less than 18 or over 60 years, a BMI of 27 kg/m2 or more, and a fasting blood glucose value of over 6·1 mmol/l. Ethical approval for the study was obtained from the University's Research Ethics Committee. Subjects were given full details of the study protocol and had the opportunity to ask questions. All subjects gave written informed consent prior to participation.

Study protocol

The protocol used was adapted from that described by Wolever et al. (Reference Wolever, Jenkins, Jenkins and Josse1991) and is in line with procedures recommended by the FAO/WHO (1998). The FAO/WHO state that, to determine the GI of a food, tests should be repeated on six or more subjects; thus in the current study, each product was tested on a minimum of ten subjects. Overall, subjects tested between one and twelve different foods during the study. On the day prior to a test, subjects were asked to restrict their intake of alcohol and caffeine-containing drinks and to restrict their participation in intense physical activity. Subjects were also told not to eat or drink after 21.00 hours on the night before a test, although water was allowed in moderation.

Test foods

Forty-four different foods were tested, including breads, breakfast cereals, mixed meals (breakfast cereals with milk), snack bars and weight-management meals, representing a diverse range of foods commonly consumed in the UK. All foods were tested in equivalent available carbohydrate amounts (50 or 30 g) and compared with a reference food (glucose). In the case of foods with a low to moderate carbohydrate density, it is justified to reduce the carbohydrate load to avoid an unrealistically large meal size; such reductions are shown to produce similar GI values (Brouns et al. Reference Brouns, Bjorck, Frayn, Gibbs, Lang, Slama and Wolever2005). For each food product, the experimental portion was determined using data for available carbohydrate provided by the relevant food manufacturer. In the case of the mixed meals, both the breakfast cereals and milk contributed to the 50 g available carbohydrate (Table 1).

Table 1 Amount of available carbohydrate from mixed meals

In accordance with FAO/WHO recommendations, subjects tested each test food once and the reference food three times randomly on separate days, with a gap of at least 1 d between measurements to minimise carry-over effects (FAO/WHO, 1998). Subjects were studied in the morning after a 12 h overnight fast. Subjects consumed the reference food/test product within 15 min. The test products and the reference food were served with 200 ml water, and a further 200 ml water were given during the subsequent 2 h. Subjects remained sedentary during each session.

Blood glucose measurements

A fasting blood sample was taken at 0 min, and the reference food/test product was consumed immediately after this. Further blood samples were taken at 15, 30, 45, 60, 90 and 120 min after starting to eat. Blood was obtained by finger-prick using the Unistik 2 single-use lancing device (Owen Mumford). Prior to a finger-prick, subjects were encouraged to warm their hand to increase blood flow. Fingers were not squeezed to extract blood from the fingertip, in order to minimise plasma dilution. Blood glucose was measured using Ascensia Contour automatic blood glucose meters (Bayer HealthCare). The blood glucose meters were calibrated daily using control solutions from the manufacturer, and were also regularly calibrated against a clinical dry chemistry analyser (Reflotron Plus; Roche) and the HemoCue Glucose 201+ analyser (HemoCue Ltd).

Figure 1 shows the Pearson regression and Bland–Altman analyses for a random selection of 1400 blood samples simultaneously measured using the Ascensia Contour and the HemoCue Glucose 201+ analyser. There was a very strong correlation (r = 0·960, P < 0·001) and good agreement (mean difference 0·10 mmol; 95 % CI 0·07, 0·12; limits of agreement 0·88 and 1·08) between blood glucose measurements using the automatic analyser and the HemoCue analyser.

Fig. 1 Pearson regression and Bland–Altman analyses of 1400 random blood glucose measurements between the Ascensia Contour (ASC) and the HemoCue 201+ analyser (HEM).

Calculation of glycaemic index and glycaemic load

The incremental area under the blood glucose response curve (IAUC), ignoring the area beneath the baseline, was calculated geometrically for each food (FAO/WHO, 1998). The IAUC for each test product eaten by each subject was expressed as a percentage of the mean IAUC for the reference food eaten by the same subject:

The GI of each test product was taken as the mean for the whole group.

The GL of a specific serving of each food was calculated using the following equation:

The serving size of each food was taken from manufacturers' information or, when this was not available, from standard food portion sizes (Food Standards Agency, 2004).

Statistical analysis

Statistical analysis was performed using the Statistical Product and Service Solutions software (SPSS version 11.0.1; SPSS Inc., Chicago, IL, USA). To examine the correlation and agreement between the automatic analyser and the HemoCue Glucose 201+ analyser, Pearson's correlation coefficient and the method of Bland & Altman (Reference Bland and Altman1986) were used. Pearson's correlation coefficient and Spearman's correlation coefficient (rho) were used, where appropriate, to assess the relationship between the GI values and macronutrient content of the test foods. Statistical significance was set at P < 0·05.

Results

The GI and GL values for all forty-four tested foods are given in Table 2. Values are given as means with their standard errors. The GI values of the foods tested ranged from 23 (chocolate flavour drink) to 83 (hot oat cereal with water). Of the forty-four foods tested, thirty-three were classified as low-GI, eight as medium-GI and three as high-GI foods. The GL per serving ranged from 2·3 (choice grain crackers, rich tea biscuits) to 20·9 (chocolate soya drink).

Table 2 Glycaemic index (GI) and glycaemic load (GL) values for forty-four foods available in the UK

Serving size for breads, biscuits and crackers is one slice of bread or one cracker/biscuit.

* Semi-skimmed milk contained 3·4 g protein, 1·7 g fat and 4·7 g carbohydrate per 100 g.

Available carbohydrate (g/100 g) of test food and semi-skimmed milk.

Both the test food and the reference food contained 30 g available carbohydrate.

Bread and crackers and breakfast cereals represented a wide range of GI values, most showing low (e.g. multiseed bread, choice grain crackers, high-fibre cereals), with some medium (e.g. rye crackers, cereal flakes with fruit) and two high (e.g. cereal biscuit, hot oat cereal I with water), values. All snack bars and sweet biscuits fell into the low-GI category. Most weight-management meal products fell into the low-GI category, with the exception of the chocolate soya drink (high), and vegetable and chicken, and mushroom soups (medium).

The addition of semi-skimmed milk to the breakfast cereals reduced the GI values from a high to a low classification. The GI values of cereal biscuits and hot oat cereal I were reduced from 72 to 47 (P = 0·112) and 83 to 47 (P = 0·011), respectively, when consumed with milk.

There was no relationship between the GI value and the amount of protein per 50 g available carbohydrate portion (Fig. 2; Pearson's r = − 0·288; P = 0·061). There was, however, a weak negative relationship between the GI value and amount of fat per experimental portion (Fig. 2; Spearman's rho = − 0·373; P = 0·014). When the weight-management meals (i.e. lower-fat products) were excluded, there was a strong negative relationship between GI value and amount of fat per experimental portion (Fig. 3; Spearman's rho = − 0·727; P < 0·001), but not between GI value and amount of protein per 50 g available carbohydrate portion (Fig. 3; Pearson's r = − 0·295; P = 0·095).

Fig. 2 Relationship between the glycaemic index (GI) value and the amount of protein (A) and fat (B) per 50 g available carbohydrate portion in all foods tested.

Fig. 3 Relationship between the glycaemic index (GI) value and the amount of protein (A) and fat (B) per 50 g available carbohydrate portion, excluding weight-management meals.

Discussion

This study provides the GI values of a number of foods not previously tested, further expanding our database of the glycaemic response of different foods available in the UK. The GI values of several foods and mixed meals reported in this study have not previously been published. Where, however, comparison with published values (Foster-Powell et al. Reference Foster-Powell, Holt and Brand-Miller2002) was possible, foods tested in the current study compared favourably. For example, in healthy subjects, the GI values for wheat biscuits (61–75), rye crisp bread (69), meal-replacement bars (30–45) and meal-replacement chocolate drink powder (26) reported in international GI tables (Foster-Powell et al. Reference Foster-Powell, Holt and Brand-Miller2002) are similar to those reported here.

Small differences of less than 10–15 units lie within the error associated with the measurement of GI (Wolever et al. Reference Wolever, Jenkins, Jenkins and Josse1991; Foster-Powell et al. Reference Foster-Powell, Holt and Brand-Miller2002), but there were a few values that were notably different from those previously reported. In particular, the GI value of white bread reported in the current study was lower than that reported for most white breads. This may be due to differences in processing conditions and the use of new food ingredients in the baking process. This therefore reconfirms the need to test food products in the country of consumption.

The weight-management meals were mostly low GI. There is considerable interest in the potential of low-GI foods for the management of obesity (Warren et al. Reference Warren, Henry and Simonite2003), and such products may play an important role in body-weight regulation. With the increasing consumption of weight-management meals in our society, the current GI table will enable consumers and researchers alike to select low-GI foods for their respective needs.

In the current study, ‘mixed meal’ testing of breakfast cereal with milk was conducted, in contrast to standard GI testing for breakfast cereals per se. Given the practical nature of GI, food companies, industry and individuals now want to know the GI values of food as eaten. The results from our study suggest that the addition of semi-skimmed milk to breakfast cereals may reduce the GI value. It is now well recognised that the low glycaemic response to milk does not solely depend on its lactose content. Milk protein has a strong insulinotropic effect (Nilsson et al. Reference Nilsson, Stenberg, Frid, Holst and Björck2004). Consequently, when testing foods with the addition of milk, the role that milk proteins play may also need to be considered. Casein and whey proteins are rich sources of leucine and phenylalanine. Leucine in particular has been shown to promote insulin secretion by simulating β-cell function (van Loon et al. Reference van Loon, Saris, Verhagen and Wagenmakers2000). The results obtained from our study may be interpreted on the basis of the above observations. The co-ingested milk protein stimulated insulin production, which in turn facilitated glucose utilisation, leading to a lower GI value. Further investigations on the association between GI and lactose and fructose would be of interest.

The presence of large amounts of fat and protein may also reduce the GI of a food (Wolever et al. Reference Wolever, Katzmanrelle, Jenkins, Vuksan, Josse and Jenkins1994). It is generally accepted that fat may lower the postprandial glucose response by delaying the rate of gastric emptying (Owen & Wolever, Reference Owen and Wolever2003). It has also been speculated that lipids may bind with the amylose fraction of starch, rendering it less susceptible to amylase (Siswoyo & Morita, Reference Siswoyo and Morita2001). Protein increases the amount of insulin secreted, causing blood glucose levels to be less affected, and may also form a protective network around the carbohydrate molecule, preventing the action of glycolytic enzymes (Bornet et al. Reference Bornet, Costagliola, Rizkalla, Blayo, Fontvieille, Haardt, Letanoux, Tchobroutsky and Slama1987). In the present study, although an effect of protein and fat was not observed across the entire group of foods, there was a strong negative association with fat content when the weight-management meals (i.e. lower-fat foods) were excluded. Therefore, this does not rule out the fact that the GI value of individual products may be determined by their protein and fat content.

Many of the products presented in the current paper are a result of a simple reformulation and alteration of processing conditions by food manufacturers to reduce the glycaemic response of the food; for example, a reformulated white bread in our study had a GI value of 59 compared with a GI value of 70 for standard white wheat flour bread (Foster-Powell et al. Reference Foster-Powell, Holt and Brand-Miller2002). This confirms the view that the demand for and interest in low-GI foods is an adequate stimulus for the food industry to develop such foods.

In conclusion, the present paper provides reliable GI and GL values for a range of different foods and mixed meals consumed in the UK, further advancing our understanding of the glycaemic response of different foods. The data reported here make an important addition to published GI values, enabling consumers to have a wider range and selection of low-GI foods to choose from.

Acknowledgements

We thank the following food companies for their contribution: British Bakels Ltd, Cambridge Manufacturing Company Limited, CSM Bakery Supplies Europe, Lyme Regis Fine Foods Ltd, United Biscuits and Warburtons Ltd.

References

Augustin, LS, Franceschi, S, Jenkins, DJA, Kendall, CWC & La Vecchia, C (2002) Glycemic index in chronic disease: a review. Eur J Clin Nutr 56, 10491071.CrossRefGoogle ScholarPubMed
Björck, I, Granfeldt, Y, Liljeberg, H, Tovar, J & Asp, NG (1994) Food properties affecting the digestion and absorption of carbohydrates. Am J Clin Nutr 59, S699S705.Google Scholar
Bland, JM & Altman, DG (1986) Statistical methods for assessing agreement between two methods of clinical measurement. Lancet I, 307310.CrossRefGoogle Scholar
Bornet, FRJ, Costagliola, D, Rizkalla, SW, Blayo, A, Fontvieille, A, Haardt, M, Letanoux, M, Tchobroutsky, G & Slama, G (1987) Insulinemic and glycemic indexes of six starch-rich foods taken alone and in a mixed meal by type 2 diabetics. Am J Clin Nutr 45, 588595.CrossRefGoogle Scholar
Brand-Miller, J, Foster-Powell, K & Colagiuri, S (2003) The New Glucose Revolution. New York: Marlowe.Google Scholar
Brouns, F, Bjorck, I, Frayn, KN, Gibbs, AL, Lang, V, Slama, G & Wolever, TMS (2005) Glycaemic index methodology. Nutr Res Rev 18, 145171.CrossRefGoogle ScholarPubMed
FAO/WHO (1998) Carbohydrates in Human Nutrition. Report of a Joint FAO/WHO Expert Consultation. Rome: FAO.Google Scholar
Food Standards Agency (2004) Food Portion Sixes, 3rd ed. London: TSO.Google Scholar
Foster-Powell, K, Holt, SHA & Brand-Miller, JC (2002) International table of glycemic index and glycaemic load values: 2002. Am J Clin Nutr 76, 556.CrossRefGoogle ScholarPubMed
Frost, G, Leeds, AA, Dore, CJ, Madeiros, S, Brading, S & Dornhorst, A (1999) Glycaemic index as a determinant of serum HDL-cholesterol concentration. Lancet 353, 10451048.Google Scholar
Henry, CJK, Lightowler, HJ, Strik, CM, Renton, H & Hails, S (2005) Glycaemic index and glycaemic load values of commercially available products in the United Kingdom. Br J Nutr 94, 922930.Google Scholar
Jenkins, DJA, Wolever, TMS, Taylor, RH, Barker, H, Fielden, H, Baldwin, JM, Bowling, AC, Newman, HC, Jenkins, AL & Goff, DV (1981) Glycemic index of foods: a physiological basis for carbohydrate exchange. Am J Clin Nutr 34, 362366.CrossRefGoogle Scholar
Ludwig, DS (2000) Dietary glycemic index and obesity. J Nutr 130, 280S283S.CrossRefGoogle ScholarPubMed
Nilsson, M, Stenberg, M, Frid, AH, Holst, JJ & Björck, IME (2004) Glycemia and insulinemia in healthy subjects after lactose-equivalent meals of milk and other food proteins: the role of plasma amino acids and incretins. Am J Clin Nutr 80, 12461253.CrossRefGoogle ScholarPubMed
Owen, B & Wolever, TMS (2003) Effect of fat on glycaemic response in normal subjects: a dose–response study. Nutr Res 23, 13411347.CrossRefGoogle Scholar
Salmeron, J, Ascherio, A, Rimm, EB, Colditz, GA, Spiegelman, D, Jenkins, DJ, Stampfer, MJ, Wing, AL & Willet, EC (1997 a) Dietary fiber, glycaemic load, and risk of NIDDM in men. Diabetes Care 20, 545550.Google Scholar
Salmeron, J, Manson, JE, Stampfer, MJ, Colditz, GA, Wing, AL & Willet, EC (1997 b) Dietary fiber, glycaemic load, and risk of non-insulin-dependent diabetes mellitus in women. JAMA 277, 472477.Google Scholar
Siswoyo, TA & Morita, N (2001) Influence of acyl chain lengths in mono- and diacyl-sn-glycerophosphatidylcholine on gelatinization and retrogradation of starch. J Agric Food Chem 49, 46884693.Google Scholar
van Loon, LJ, Saris, WH, Verhagen, H & Wagenmakers, AJ (2000) Plasma insulin responses after ingestion of different amino acid or protein mixtures with carbohydrate. Am J Clin Nutr 72, 96105.CrossRefGoogle ScholarPubMed
Warren, JM, Henry, CJK & Simonite, V (2003) Low glycaemic index breakfasts and reduced food intake in preadolescent children. Pediatrics 112, E414E419.CrossRefGoogle ScholarPubMed
Wolever, TMS, Jenkins, DJA, Jenkins, AL & Josse, RG (1991) The glycemic index: methodology and clinical implications. Am J Clin Nutr 54, 846854.Google Scholar
Wolever, TMS, Katzmanrelle, L, Jenkins, AL, Vuksan, V, Josse, RG & Jenkins, DJA (1994) Glycemic index of 102 complex carbohydrate foods in patients with diabetes. Nutr Res 14, 651669.Google Scholar
Figure 0

Table 1 Amount of available carbohydrate from mixed meals

Figure 1

Fig. 1 Pearson regression and Bland–Altman analyses of 1400 random blood glucose measurements between the Ascensia Contour (ASC) and the HemoCue 201+ analyser (HEM).

Figure 2

Table 2 Glycaemic index (GI) and glycaemic load (GL) values for forty-four foods available in the UK

Figure 3

Fig. 2 Relationship between the glycaemic index (GI) value and the amount of protein (A) and fat (B) per 50 g available carbohydrate portion in all foods tested.

Figure 4

Fig. 3 Relationship between the glycaemic index (GI) value and the amount of protein (A) and fat (B) per 50 g available carbohydrate portion, excluding weight-management meals.