DETERMINATION OF THE GLYCAEMIC INDEX OF SELECTED FOODS (WHITE BREAD AND CEREAL BARS) IN HEALTHY PERSONS

UNLABELLED
The glycaemic index (GI) is a measure of the food power to raise blood glucose (B-glucose) concentration after a meal. For healthy eating, foods with low GI are recommended. However, for many foods in the European Union the GI has not been defined yet. The aims of this prospective open-label study were: (1) to determine the GI of white bread and juicy cereal bars FIT (Usovsko, Czech Republic) by means of the glucometer Optium (Abbott/Medisense); (2) to compare the GI of tested foods determined in the morning and in the evening hours; (3) to compare the GI of tested foods in men and women and (4) to assess the variability of the GI.


METHODS
To determine the GI, measured portions of food containing 50 g of carbohydrates were eaten by 11 healthy volunteers. B-glucose curves were constructed from B-glucose values at time 0, 15, 30, 45, 60, 60, 120 min after the meal. The GI was calculated by dividing the incremental area under the curve (IAUC) for the tested food by that for the standard food (IAUCS). In each volunteer each food was tested 5 times so that 5 GI's was obtained and the average was calculated. The GI for each tested food was calculated as the mean from the respective average GI's of the 11 volunteers. MS Excel and the statistical program SPSS v. 10.1 were used to analyze the data.


RESULTS
(1) The mean values of the GI for white bread was 70.3 % and for juicy cereal bars was 101.0 %, as determined in a total of 139 tests in the whole group of 11 volunteers. There was a difference when comparing white bread vs. glucose (p = 0.012) and white bread vs. cereal bars (p = 0.026) but no difference between glucose and cereal bars. (2) There was no significant difference between the GI determined in the morning and in the evening hours either for the total of 139 tests or for the individual tested foods. (3) No significant difference could be seen between the GI in men and women when comparing glucose, cereal bars and white bread. (4) There was a wide variability of GI in all tested foods: the standard deviation of GI for white bread was 30.7 %, for juicy cereal bars 38.0 %.


CONCLUSIONS
The GI's for white bread and juicy cereal bars were determined. There was no difference either between the GI values determined in the morning vs. the evening hours or between the values in men vs. women. The results show wide variability. An accurate standard method for the determination of GI needs to be defined, carefully used and re-evaluated to enable a comparison of the results with various methods of other working groups.


INTRODUCTION
The glycaemic index (GI) is an important parameter of food quality which compares the hyperglycaemic eff ect of a tested meal with pure glucose (or of another defi ned standard food).The GI is a measure of the food power to raise B-glucose concentration after a meal.The GI is defi ned as relation of the incremental area under the B-glucose response curve (IAUC) of a tested meal containing 50 g of digestible carbohydrates and the incremental area under the B-glucose response curve of the standard food, i.e.50 g pure glucose (IAUCS).Carbohydrates that breakdown quickly during digestion have a high GI because their B-glucose response is fast and high.Carbohydrates that breakdown slowly have a low GI 11 .
For healthy eating, particularly in persons with diabetes, obesity and insulin resistance, foods with low GI are recommended as they may help keep the euglycaemia and the normal spectrum of lipoproteins 3,4,8,12,22 .These eff ects result in decreased cardiovascular danger and probably also in reduced risk for colon and breast cancer 1 .
On the other hand, the GI values have a large interand intraindividual variability 18,19,24 .For many foods in the European Union the glycaemic index has not been defi ned yet.Even the methods for defi ning the GI are not standardized 13 .Therefore, any eff ort to enable the determination and practical use of GI may support establishing optimum dietary recommendations and good eating habits.

AIMS
Aims of this prospective open-label study were: 1. to determine the GI of white bread and cereal bars by means the glucometer system Optium; 2. to compare the GI of tested foods determined in the morning and in the evening hours; 3. to compare the GI of tested foods in men and women; 4. to assess the variability of the GI.

Determination of the glycaemic index a) Getting basic data
To determine the GI, measured portions of tested food containing 50 g of carbohydrates were eaten by each of the 11 healthy volunteers (Table 1) after an overnight fast; the same approach was used after an afternoon fast.Fingerprick blood samples were investigated at 15-30 minute intervals over the next two hours after the meal (at times 0, 15, 30, 45, 60, 90, 120 min; the beginning of the food intake was time 0).
Each volunteer measured his/her B-glucose concentrations by means of a glucometer Optium.At the end of the one-week test period the B-glucose values were transferred from the memory of the glucometer into a PC for further analysis.
b) Construction of B-glucose response curves.
The averages of the respective B-glucose concentrations after the meal were used to draw a B-glucose response curve for the two-hour period.The values at times 75 and 105 min were obtained by extrapolation.c) Exclusion of disturbed tests.
For the purpose of statistical evaluation, all tests that were not complete and all tests where the fi rst (i.e.fasting) B-glucose concentration was 7.0 mmol/l or higher were excluded.d) Calculations of individual GI values in every volunteer.
The incremental area under the curve (IAUC) was calculated for each meal in every volunteer separately (as the sum of the surface of triangles and trapezoids between the B-glucose curve and horizontal baseline going parallel to x-axis from the beginning of B-glucose curve at time 0 to the point at time 120 min) to refl ect the total rise in B-glucose concentration after eating the tested food.
The IAUCS for the standard reference food (i.e.50 g of pure glucose) was obtained similarly to the mean from the fi rst three independent IAUCS 1 , IAUCS 2 , IAUCS 3 in the same volunteer.
In the IAUC/IAUCS calculations, all B-glucose values in the course of the test lower than the fi rst value (at time 0) were equalized to the respective fi rst value.
In each volunteer, the GI (%) was calculated by dividing the IAUC for the tested food by the IAUCS for the standard food and multiplying by 100.The following formula was used: 1 IAUC -Incremental Area Under the blood glucose response Curve for the tested meal IAUCS -Incremental Area Under the blood glucose response Curve for the standard meal e) Working out the average of GI's for tested food in each volunteer.
In each volunteer each food item was tested 5 times so that 5 GI's was obtained and (after the exclusion of disturbed tests) the average was calculated.
f) Final calculation of the GI for each tested food.
The GI for each tested food was calculated as the mean from the respective average GI's of the 11 volunteers.
g) The variability of GI for each tested food was assessed according to standard deviation of the mean; histograms of GI values demontrated the frequency and range of results.

Healthy volunteers
The participants in this study were healthy persons recruited from the nursing staff , laboratory assistants and students.Thorough clinical and laboratory investigations were performed to establish that the volunteers were healthy.(Table 1).
The food was professionaly prepared in the expected quality and quantity; the portions were packed and marked with a set sign.Each serving contained 50 g of digestible carbohydrates.Glucose was dissolved in 300 ml of tea, coff ee or water before drinking.

Study design
1.Each volunteer received a glucometer Optium and 100 strips (Lot No 51322); everyone was trained in selfmonitoring and instructed how to keep to the principles of the study protocol: -to consume the tested and the standard food Determination of the glycaemic index of selected foods (white bread and cereal bars) in healthy persons

Blood glucose curves
A total of 164 tests (with each food 55 tests, one test with white bread was omitted) were performed.
The mean 2-hour B-glucose curves for pure glucose, white bread and cereal bars are drawn in Fig. 1.
The mean B-glucose curves in the morning (n = 77) and in the evening (n = 87) for all kinds of tested foods are shown in Fig. 2 and for the individual tested foods in Fig. 3-5.
The mean B-glucose curves for 5 men (n =74 and 6 women (n = 90) comparing all kinds of food together are demonstrated in Fig. 6 and curves comparing individual foods in Fig. 7-9.

Statistical evaluation of GI values for tested foods
Due to incomplete number of B-glucose estimations or due to high B-glucose concentration at start (> 7.0 mmol/l), 25 tests (15.1 %) had to be excluded from the statistical analysis.The mean values of the GI for glucose, cereal bars and white bread determined in a total of 139 tests in the whole group of 11 volunteers are shown in Table 3.There was a signifi cant diff erence when comparing white bread vs. glucose (p = 0.012) and white bread vs. cereal bars (p = 0.026) but no diff erence between glucose and cereal bars (p = 0.732) could be seen (Fig. 10).
There was no signifi cant diff erence between the GI determined in the morning and in the evening hours neither for the total of 139 tests nor for the individual tested foods.
No signifi cant diff erence could be seen between the GI in 5 men and the GI in 6 women when comparing glucose, cereal bars and white bread.However, in men, there was a signifi cant diff erence between the GI for white bread vs. glucose (p = 0.004) and cereal bars vs. glucose (p = 0.044) -see Fig. 11; in women, the only signifi cant diff erence was between white bread vs. cereal bars (p = 0.047) -see Fig. 12.

Variability of GI
The standard deviations of GI see Table 3.The histograms (Fig. 13-14) demonstrate the frequency of GI values in individual tested foods.

DISCUSSION
The concept of the glycaemic index of foods has been developed in the course of the last thirty years without having reached its fi nal version 22,25 .
Recent studies from Harvard School of Public Health indicate that the risks of diseases such as type 2 diabetes and coronary heart disease are strongly related to the GI of the overall diet.In 1999, the World Health Organisation (WHO) and Food and Agriculture Organisation (FAO) recommended that people in industrialised countries base their diets on low-GI foods in order to prevent the most common diseases, such as coronary heart disease, diabetes and obesity 5,7,9,10,14,15,17,23 .Some foods on the world market already show their GI rating on the nutrition information panel.Terms such as complex carbohydrates and sugars, which commonly appear on food labels, are now recognised as having little nutritional or physiological signifi cance.The WHO/FAO recommend that these terms be replaced with the total carbohydrate content of the food and its GI value.
In the Czech Republic, however, we were not able to fi nd any food product containing the nutritional lable with GI value.
International Tables of Glycaemic Index were published by the American Journal of Clinical Nutrition in 1995 and 2002 6 .Tables of the GI values contain about 600 diff erent foods.According to GI, foods may be divided into three groups: foods with low GI (GI = 55 % or less), foods with medium GI (GI = 56-69 %) and foods with high GI (GI = 70 % or more).
The GI values of foods must be measured using valid scientifi c methods 16,20,21 .The accuracy of the measurements of the GI is infl uenced particularly by the following factors 2 : 1. method for calculating IAUC; 2. method for measuring the B-glucose; 3. defi ning the amount of the tested food which contains 50 g of hyperglycaemic (i.e.absorbable, digestive) carbohydrates; 4. the usage of the standard food (defi ning the amount and the kind of the white bread that contains 50 g of digestive carbohydrates); 5. tested individuals; 6. the glycaemia variability from day to day; 7. time of the day when the test is carried out.
In this study, the glucometers Optium were used by 11 trained volunteers.Optium system enabled a reliable registration of all B-glucose values including exact times of measurement even though it was not possible to keep an eye on the performance of the test.Over 50 tests carried out with each tested food in order to investigate the GI exceeds the usual amount of tests used by other working bread and cereal bars

Table 1 .
Characteristics of healthy volunteers in the study.

Table 2 .
The time schedule for individual tests with food A (glucose), B (white bread) and C (juicy cereal bars) used by each volunteer.

Table 2 )
; no other food was allowed for breakfast and dinner; -to consume no food from dinner until breakfast and from lunch to dinner; drinking water, mineral water, tea and coff ee without sugar was allowed; -to keep to the same extent of physical exercise during the whole one-week test period; -to consume no alcohol and not to smoke.2. Each volunteer received 15 servings of standard and tested foods with an exact marking when to eat which serving. 3.Each volunteer kept a diary on food intake, exercise and results of B-glucose selfmonitoring.4. The PC Link was used to transfer the data from glucometer Optium to a PC.MS Excel and statistical program SPSS v. 10.1 were used to analyze the data.