1 Introduction

There is an ongoing scientific debate about the prevention of chronic diseases particularly about replacing saturated fats with unsaturated fats or carbohydrates (1;2) . However, excessive consumption of certain carbohydrates can also be a risk factor for
cardiovascular disease, and certain carbohydrate-rich foods are implicated in the rapidly increasing incidence of obesity and type 2 diabetes. It is therefore important to establish the "metabolic quality" of carbohydrates so that consumption of the most beneficial types can be encouraged.

Consumption of simple sugars or rapidly digestible starch can lead to a rapid increase in serum glucose levels (high glycemic index foods), an increase in serum insulin levels and increased synthesis of triglycerides. This metabolic response is thought to be detrimental to longer-term optimal health. One parameter in considering the "metabolic quality" of carbohydrates is the nature of the serum glucose response or the glycemic index (3) . This response is dependent on: gastric emptying, intestinal transit, intestinal a-amylase activity, inhibition of endogenous glucose production, glycogen synthesis, stimulation of glucose utilisation and oxidation. When attempting to modulate the glycemic index of food products the many steps involved in this metabolic process need to be monitored. This can be achieved by labelling food products; the best way to do this is by using stable isotopes. In contrast to radioactive isotopes, stable isotopes are non-radioactive, not harmful for humans and are already naturally present in biological substances, for example, the 13 C atom is present naturally at about 1.1%. By increasing the amount of naturally enriched 13 C in plant material and consuming these products, their subsequent digestion and metabolism can be traced by measuring the 13 C-glucose in blood, 13 CO 2 in breath and 13 C-metabolites in faeces.

Carbohydrates that are not digested in the small intestine pass into the colon and are
partly fermented to form a variety of metabolites; previously this was thought to be waste processing. However we now know that the formation of metabolites during fermentation and their subsequent metabolism in the colonocytes are linked to the occurrence of diarrhoea, constipation, energy salvage, colon cancer, and immune status. More recently evidence has emerged that colonic fermentation not only affects gut metabolism but can also influence metabolic processes in other tissues and organs. For example, short and longer term consumption of starch that is resistant to small intestinal digestion has been shown to beneficially increase insulin sensitivity in healthy subjects (4;5) . The underlying mechanism of this phenomenon is not known, but short-chain fatty acids (SCFA's), like acetate and propionate, which are products of starch fermentation, could be involved. These SCFA's are metabolised to some extent by the colonic epithelial cells but also enter the portal circulation. Effects of SCFA's on liver metabolism have been reported (6-8) and recently SCFA's have been identified to be ligands for the orphan G protein-coupled receptor GPR41 - primarily expressed in adipose tissue (9) . Adipose tissue is known to secrete various signalling peptides - the adipokines - influencing for example, insulin sensitivity, feeding behaviour and inflammation (10) and could constitute a link between colonic fermentation and peripheral metabolic effects. However, the scientific evidence of the interrelationship between fermentation products and systemic effects is limited, partly due to the inaccessibility of the colon.

Thus, the aim of EUROSTARCH was to increase the knowledge of various stages of the processes in carbohydrate digestion and fermentation, to define food products most suitable for the prevention of chronic disease.

The aims of this project were

  1. To investigate the metabolic effects of starchy foods related to the rate and extent of digestion of starch: acute, semi-acute and long term effects.
  2. To investigate the biological effects of cognitive function and satiety of starchy foods
  3. To investigate fermentation of resistant starch in the colon.

Aim 1 To investigate the metabolic effects of starchy foods related to the rate and extent of digestion of starch: acute, semi-acute and long term effects.

Five sets of studies were conducted to investigate the acute, semi-acute and long term metabolic effects of starchy foods.

Study 1: Short-term metabolic effect of a low-glycemic (LGI) or a high-glycemic index (HGI) breakfast in overweight non-diabetic subjects
The aim of this study was to compare the effects of a low-glycemic index breakfast on postprandial glucose kinetics with that of a high-glycemic index breakfast in overweight subjects.

Study 2: Digestive characteristics of whole wheat bread (WB) determined in vivo
In many diets in different societies bread is the main source of starch. However, bread made from refined flour and even wholemeal bread, if made from finely ground whole grains, has features of rapidly digestible starch which is thought to be detrimental to health. This is based on in vitro measurements and glycemic index. However, no information is available about the digestive starch characteristics in vivo . The objective of this study was to characterise starch digestion of wholemeal wheat bread and postprandial glucose kinetics in healthy volunteers.

Study 3: Systemic rate of appearance of starch derived glucose and the secretion of incretin hormones
Glucagon-like peptide-1 (GLP-1) and glucose-dependent insulinotropic polypeptide (GIP) play a role in the control of glucose homeostasis: GIP is implicated in the regulation of energy storage. The capacity of carbohydrates to induce secretion of these incretin hormones could be one of the factors determining the metabolic quality of different types of carbohydrates . The objective was to analyze the correlation between the rate of intestinal absorption of (starch derived) glucose and plasma concentrations of GLP-1 and GIP after ingestion of glucose and starchy foods with different contents of rapidly and slowly available glucose.

Study 4: Effects of glycemic index (GI) and content of indigestible carbohydrates from the evening meal on glucose tolerance at a subsequent standardised breakfast and other metabolic variables
Currently there is a lack of low-GI cereal foods on the market. In the design of such foods more knowledge is needed concerning the mechanisms for observed long-term metabolic benefits of low-GI diets, and concerning the possibilities for further enhancement of the metabolic benefits by formulating low-GI foods that are also rich in specific indigestible carbohydrates. The aim of these two experimental series (4A and 4B) was to evaluate the impact of several cereal based evening meals varying in GI and contents of indigestible carbohydrates on glucose tolerance, at a subsequent standardised breakfast.

Study 5: Effects of a five-week low-glycemic index regimen on weight control and cardiovascular risk factors in overweight, otherwise healthy subjects
The objective of this study was to evaluate the health benefit of a low glycemic index (LGI) diet, in comparison with a high glycemic index (HGI) diet, in overweight healthy men and women.

Aim 2 To investigate the biological effects of cognitive function and satiety of starchy foods

Study 1: Cognitive functions in relation to postprandial glycaemic response
One pilot study and one extended study were conducted to investigate cognitive functions in relation to postprandial glycaemic response. A primary objective was to find useful tools to study potential differences in cognitive function during the postprandial period, and to use these tools to evaluate functionality after test meals differing in post meal glycaemia. In addition, the aim was to study cognitive performance in relation to glucose tolerance status.

Study 2: Possible influence of botanical structure, fibre content and/or acetic acid on gastric emptying rate (GER) and satiety
The aim of the study was to evaluate to what extent certain food factors could modulate the GER and, as a result, the post meal satiety of cereal products in healthy subjects. The food factors studied were; the botanical integrity and/or fibre content of the product (flour based vs. coarse bread), and supplementation of bread with acetic acid (vinegar was used). The hypothesis was that starchy products which delay GER may favour a higher subjective rating of post-meal satiety.

Aim 3 To investigate fermentation of resistant starch in the colon

Techniques have been developed to measure colonic 13 C- short-chain-fatty acid production in blood and urine. For characterization of the colonic microbiota the Fluorescent in Situ Hybridisation (FISH) technique and the Denaturing Gradient Gel Electrophoresis (DGGE) have been optimized. To monitor the metabolic capacity of the colonic microbiota the 15 N-retention method, as well as proteomics techniques have been developed. Two studies were conducted in which one or more of these methods were applied.

Study 1: Influence of resistant starch on 15 N-incorporation
The objective was to investigate the influence of resistant starch on colonic ammonia metabolism by measuring 15 N-incorporation.

Study 2: Evaluation of the fermentation of resistant starch
The objective was to examine whether short-chain fatty acid (SCFA) production in vivo is different after the ingestion of 13 C barley as whole kernels, compared to 13 C barley porridge. The working hypothesis was that whole barley kernels contain a significantly greater proportion of resistant starch (RS), resulting in more RS reaching the colon, with greater SCFA production. Production of SCFA is thought to be involved in the reduced glycaemic response to a subsequent standardized breakfast observed after an evening meal of whole barley kernels, but not after barley porridge (c.f. results of study 4, Aim 1). The mechanism of reducing the glycaemic response is not fully understood but is likely to involve SCFA, either through transport to the portal circulation and alteration of hepatic glucose status or by signalling, via gut peptide hormones, such as PYY 3-36 , which influence gastro-intestinal motility.

2 Materials and methods

1 To investigate the metabolic effects of starchy foods related to the rate and extent of digestion of starch: acute, semi-acute and long term effects.

In studies 2, 3 and 4 healthy subjects took part, while studies 1 and 5 were conducted with overweight subjects.

Study design
The acute studies, 1, 2 and 3 used a cross-over with at least one week between studies. After an overnight fast the test meal were administered. Blood and breath baseline samples were collected and during 6 or 8 (study 1) hours, postprandial.

In the semi-acute studies, 4A and B, the test meals were taken in the evening, in a randomized manner. In the morning, the subjects received a standardized breakfast. Blood and breath samplers were collected before breakfast and at 3 hours postprandial.

Study 5 was a parallel, randomized study. The intervention period lasted 5 weeks in which either a low or high glycemic index diet was followed.

Test meals
13 C- enriched grain
Two batches of 13 C-enriched grain were obtained. The first consisted of 337g of barley grain enriched at 2.43 atom % excess 13 C and 142g of Durum wheat grain 3.89 atom % excess 13 C. The second one consisted of a total of 3850g of barley grain at an enrichment of 0.142 atom % excess 13 C and 506 g Durum wheat grain at an enrichment of 0.471 atom % excess 13 C.

Model breakfast foods
For studies 1 and 5, two differently processed model breakfast foods were developed
- plain biscuits and extruded cereals, respectively with high content of slowly available glucose (SAG = 26 g/100 g) and low content of SAG (1 g/100 g). Both cereal products had the same macronutrient composition (62% carbohydrate, 12% protein, 26% lipids) and were consumed with 180ml of half-fat skimmed milk.

These products were 13 C-labelled for study 1. In study 5 they were part of either the low or high glycemic index diet which was achieved with the use of two specific lists with either high GI or low GI commercialised starchy products.

In order to assure that the eventual differences observed between the low and high GI model foods were indeed due to differences in digestion, and were not due to, for example, differences in gastric emptying, the transit times, gastric emptying and orocaecal transit time of the model food were determined using dedicated breath tests, with both model foods containing 13 C-octanoic acid and lactose- 13 C-ureide.

Wheat bread
For study 2, 11.00g of 13 C-enriched wheat and 355g of unlabelled wheat of the same variety were milled in a electric household grain mill (SAMAP Elsasser grain mill F 100) on the finest grade. The wholemeal flour, 320g lukewarm water, 6.4g dry yeast and 6.4g salt were used to bake bread in a Panasonic home bread maker.

Low glycemic index evening meals
For study 4A, the following evening meals, with comparable glycemic indices but differing content of resistant starch (RS) and dietary fiber (DF) were administered and compared to white wheat bread:

Wheat kernels (high DF and RS), spaghetti (low DF and low RS), spaghetti with added purified wheat bran (high DF, low RS), intact barley kernels (high DF and high RS).

For the follow-up study 4B, spaghetti or white bread + barley dietary fibre (DF) corresponding to the content of DF in intact barley kernels, spaghetti + double amounts DF, spaghetti + oat DF, barley flour porridge, and white bread were consumed as six separate evening meals.

2 To investigate biological effects (cognitive function, satiety) of starchy foods

Cognitive function

Subjects and study design
Forty healthy subjects between 50 - 70 years old participated. Each was tested on two occasions, approximately one week apart. Cognitive tests of working memory were performed 4 times after consuming the test breakfast, at 35, 90, 120 and 150 minutes. A test of selective attention was included in the late postprandial period (170 min). Capillary blood glucose tests were taken before breakfast and during the 3 h postprandial period.

Test meals
To simulate two distinct glycaemic excursions, one with high GI and one with low GI glycaemic features - 50 g glucose dissolved in water was consumed, either as a single dose or portioned out into six smaller doses, half an hour apart.

Cognitive tests:

Working-memory (WM)
This test deals with a memory system that stores and processes information at the same time. The capacity of working memory is limited and in this test the span of the system was determined. The tests were oral and each test comprised of12 sets of four sentences that could be either correct or incorrect. Immediately after each sentence, orally presented, the subject had to indicate if the sentence was correct or not. After each set of four sentences the subject was told to recall the first word in each sentence. The task was scored for correct recalled words.

Selective attention (SA)
This computerized test consisted of 78 different pictures shown at two second intervals, one by one, on a computer screen. The pictures consisted of a square, divided in four smaller squares. One of the four smaller squares was red and one was green. Each time a new picture emerge on the screen either the green, the red or neither of the coloured squares was in the same position as in the previous picture. The subjects had to mark which of the three alternatives occurred when a new picture emerge. The task was scored for correct responses, and the reaction time.


Fourteen healthy subjects took part in a crossover trial. The meals were served in a random order, after an over-night fast.

Test meals
The test meals were vinegar supplemented (28g vinegar) bread products; white wheat bread, wholemeal bread (80% wholemeal wheat flour) and coarse bread (80% pre-boiled wheat kernels) respectively. White wheat bread without vinegar was used as the reference.

Gastric emptying rate (GER):
GER was assessed by using real-time ultrasonography. Subjects were examined with an EBU 400 ultrasonograph. Gastric emptying was monitored indirectly by determining the longitudinal and anteroposterior cross-sectional diameters of a single section of the gastric antrum. The intragastric volume was assumed to be directly proportional to the cross-sectional area of the gastric antrum. At each observation, 3 measurements were taken and a mean value calculated. Measurements were taken at 25 and 100 minutes after the start of the meal, when the area of the antrum was at maximum and minimum size, respectively.

Satiety rating:

The feeling of satiety/hunger was estimated before and at 15, 30, 45, 60, 90 and 120 min after the start of the meal, using a visual analogue scale. The volunteers were asked to put a cross at the position on the scale that best corresponded to the question "How do you feel right now, concerning hunger-satiety?" The endpoints were respectively labelled with the description extreme hunger and extreme satiety.

3. To investigate fermentation of resistant starch in the colon.

Technique development

Colonic 13 C-short-chain-fatty acid (SCFA) production determination from blood
GC/MS is a powerful analytical tool for the stable isotope studies, but it requires high isotopic enrichments of 13 C-labelled SCFA ( ³ 1 Mole Percent Excess; MPE) in the plasma sample. To measure low isotopic enrichments of 13 C-SCFA (or natural abundance < 1 MPE), a method in GC-C-IRMS using a Solid Phase Micro-Extraction (SPME) was developed. After removal of plasma proteins, the SPME fiber was plunged in the liquid sample for 40 min at 40°C. Then, acetate was directly desorbed into GC-C-IRMS. The accuracy of the isotopic enrichment measurement was determined using plasma spiked with 13 C-acetate and 13 C-butyrate solution from 0 to 1 MPE.

13 C-volatile fatty acid (VFA) flux determined in urine
In this approach metaprobes (tracer materials produced by chemical synthesis) were used orally, in combination with urine sampling. This permitted non-invasive measurement of acetate production, over extended (~24 hours) periods, typical of non-digestible carbohydrate fermentation. When combined with improved analytical methods that were developed in parallel, the approach can be applied to measure low enrichments of 13 C metaprobes with accuracy. Initial validation studies in two healthy adult males were undertaken using plasma sampling over 8 hours and urine sampling over 24 hours. This was to determine i), that urine could be used where multiple blood samples had previously been used, and ii), that an oral metaprobe bolus could be used where previously a primed constant i.v. infusion had been used. To facilitate this, 1- 13 C-acetate and 2 H 3 -acetate were used simultaneously.

Characterization of the colonic microbiota: Fluorescent in situ hybridisation (FISH)
FISH is a new technique for identifying bacterial strains using 16S rRNA-targeted oligonucleotide probes. Each cell contains 10,000-60,000 ribosomes, thus the binding of fluorescently labelled probes to the 16S rRNA molecules of the ribosomes will make the cell visible under a fluorescent microscope. The probes are directed at different phylogenetic levels (Domain, Family, Genus, Species) of the bacterial Kingdom. With FISH, it is possible to identify quantitatively and qualitatively a large number of different bacterial species of the gut ecosystem. Several probes have been developed to analyse the bacterial composition as completely as possible, and to validate these in human studies.

Characterization of the colonic microbiota: Denaturing Gradient Gel electrophoresis (DGGE)
With DGGE, DNA fragments of the same length but with different base-pair sequences can be separated based on differences in the melting behaviour. The electrophoresis is carried out on gels with an increasing gradient of denaturant perpendicular, or parallel to the direction of electrophoresis.
With DGGE changes in the composition of the colonic microbiota can be monitored.

Characterization metabolic capacity of the colonic microbiota: 15 N-retention
Upon oral administration of lactose[ 15 N,  15 N]ureide, the bond between the sugar and ureide moiety of this molecule resists enzymatic degradation during passage through the small intestine, whereas it is hydrolysed upon arrival in the colon by the activity of specific bacteria. The [ 15 N,  15 N]urea generated in this way is consequently rapidly hydrolysed to [ 15 N]NH 3 . Ammonia can be taken up by the bacteria for their own metabolism, followed by faecal excretion, or it can be absorbed through the colonic mucosa and, after conversion in the liver to urea, excreted in urine. Hence, the ratio between faecal and urinary excretion of 15 N reflects the fate of ammonia in the colon.

Characterization of the metabolic capacity of the colonic microbiota: Proteomics
Proteomics techniques were set up for analyzing bacterial protein profiles. The proteomic technique chosen was the SELDI-TOF Protein Chip technology, which has a high-throughput capacity and is less complicated to perform than the classical 2D-gel electrophoresis followed by MALDI-TOF MS. For identification of proteins, LS-MS/MS was used. Pilot experiments with monocultures of bacteria ( Bifidobacterium breve, Bifidobacterium animalis, Bifidobacterium longum ) incubated with different media (with glucose, lactose or without sugar) have been conducted to examine whether with these techniques bring about changes in enzyme activity/protein profiles which can be detected.

Studies 1 and 2

Subjects and study design

In study 1, ten healthy volunteers participated in a randomized placebo-controlled cross-over study in which they received either RS (2 x 15g/d) or placebo (maltodextrins) for 2 weeks, followed by a 2-week wash-out, and a second period of 2 weeks during which they received the opposite treatment. Ten other volunteers performed a similar study in which they received either RS (2 x 15 g/d) or RS combined with wheat bran (2 x 6 g/d). At the start and at the end of each intervention period, the volunteers consumed a pancake test meal labelled with 75mg lactose [ 15 N, 15 N']-ureide after which they performed a fractionated 48-h urine collection.

6 healthy subjects took part in study 2 which was performed in a cross-over manner with at least one week apart. After an overnight fast subjects received an intravenous infusion of 2 H-labelled short-chain fatty acids as sodium salts ( 2 H 3 -acetate, 2 H 7 -butyrate, 2 H 5 -propionate) over 14 hours. Before and after ingestion of the test meals blood, breath and urine samples were collected in regular intervals. After 6, 9 and 12 hours, subjects received a standard light meal which was not 13 C enriched and did not contain resistant carbohydrates.

Test meals

In study 1, "C Actistar", a commercially available RS 3-type starch (Cerestar, Vilvoorde , Belgium ) originating from Tapioca starch ( Manihot esculenta ) was consumed. Maltodextrin was used as the placebo. Coarse wheat bran (AVEVE, Belgium ) was administered as the source of bran.

In study 2, 13 C-enriched barley, either as whole kernels cooked in water or as barley porridge (barley flour cooked in water) was administered. 14 C-labelled inuline carboxylic acid was also used for the test meals for measurement of the orocecal transit time.

3 Results

1 To investigate the metabolic effects of starchy foods related to the rate and extent of digestion of starch: acute, semi-acute and long term effects.

Study 1: Short-term metabolic effect of a low-glycemic (LGI) or a high-glycemic index (HGI) breakfast in overweight non-diabetic subjects
The two model breakfast did not differ in gastric emptying rate and oro-cecal transit time and therefore differences in glycemic or insulinemic index are due to difference in the digestion rate and can not be attributed to differences in transit time. The area under the curve of glycemia, as well as the rate of appearance of exogenous glucose (RaE) was lower after the LGI breakfast, as compared to that after the HGI breakfast. Insulinemia after both breakfasts did not differ. Endogenous glucose production was less inhibited after the LGI breakfast than that after HGI breakfast, despite a similar level of insulinemia. 68.7% of the glucose derived from the LGI breakfast had appeared in plasma, 270 minutes postprandially, as compared to 96.3% from the HGI breakfast. There was a trend towards a lower total carbohydrate oxidation after the LGI breakfast, due to lower glycemia.

Study 2: Digestive characteristics of wholemeal wheat bread (WB) determined in vivo
Neither the peak glucose concentrations nor the area under the curve (AUC) of the WB differed significantly from that of glucose. WB resulted in a lower insulin response as compared to glucose, as seen by a lower peak concentration (P = 0.01) as well as a lower 0 - 120 min AUC (P = 0.037). In the early postprandial phase (0 - 120 min) the rate of appearance of exogenous glucose (RaE) after WB was the same as that after glucose (p = 0.834), however after 120 min the RaE, after WB, declined significantly slower than after glucose. The differences in RaE kinetics resulted in a 160% higher 120 - 240 AUC after WB, as compared to glucose (p = 0.005). After ingestion of WB, the endogenous glucose production (EGP) was significantly more suppressed, from 60 min onwards, than after ingestion of glucose. Endogenous glucose production returned to near baseline values at 180 min after glucose, whereas it was still suppressed 6 hours after ingestion of WB.

Study 3: Systemic rate of appearance of starch derived glucose and the secretion of incretin hormones
GLP-1 concentrations were significantly increased 180 to 300 min after ingestion of uncooked cornstarch (starch product with a high content of slowly available glucose). A high GIP response in the early postprandial phase (15 - 90 min) was observed after consumption of glucose. There was a strong positive within-subject correlation between the rate of appearance of exogenous glucose and GIP concentrations (r = 0.73, P < 0.01) across the test meals.

Study 4: Effects of GI and content of indigestible carbohydrates at the evening meal on glucose tolerance at a subsequent standardised breakfast and other metabolic variables
Study 4A: The blood glucose response (0-120 min) to the standardised breakfast was significantly lower after consuming barley kernels in the evening, compared with evening meals with white wheat bread (WWB) (P=0.019) or spaghetti+wheat bran (P=0.046). There were no significant differences in insulin concentrations at breakfast. Breath hydrogen excretion at breakfast was significantly higher after an evening meal with barley kernels compared with WWB, wheat kernels or spaghetti (P=0.026, 0.026 and 0.015, respectively), and the concentration of plasma propionate at breakfast was significantly higher following an evening meal with barley kernels compared with an evening meal with WWB (P=0.041). In parallel, free fatty acid concentrations were significantly lower after barley kernels compared with WWB (P=0.042) or spaghetti evening meals (P=0.019).
Study 4B: The blood glucose response to the standardised breakfast was significantly lower after consuming spaghetti + double amounts of barley DF in the evening compared with oat porridge (P=0.012). This spaghetti meal also resulted in the highest breath H 2 excretion (P<0.02). Serum butyrate concentrations in the morning showed a significant negative correlation to the incremental glucose areas (r = -0.98, P = 0.001). Thus only high levels of barley DF added to the spaghetti evening meal improved glucose tolerance in an overnight perspective. In contrast with previous studies which showed effects of intact barley kernels, barley flour porridge showed no effects.

Study 5: Effects of a five-week low-glycemic index regimen on weight control and cardiovascular risk factors in overweight, otherwise healthy subjects
Self-reported data from subjects' dietary record books showed that they were compliant, and tolerance of the diet was also very good. There were no significant differences between groups at baseline in GI (LGI: 64±7 vs. HGI: 69±7). After the 5 week nutritional intervention, dietary survey results showed that all subjects in the LGI group reached a low glycemic index (46±3). No subjects from the HGI group reached the defined high glycemic index target (i.e. >70) but remained at a high level with no significant variation of their mean GI after 5 weeks of the diet (69±7 at baseline vs. 67±6 at the end).

The Low GI diet was associated with:
- a significant weight loss of -1.1 ± 0.3 kg (p= 0.004)
- no changes in fat mass (-0.7 ± 0.6 %, p= 0.15)
- a significant 10% decrease in total cholesterol (both HDL and LDL), (p<0.001)
- a significant 7% decrease in C peptide (mostly in women) with no change in insulinemia (p<0.05)

The High GI diet was associated with:
- a significant 2.4% decrease in basal glycemia mostly in women (p=0.002)

2 To investigate biological effects (cognitive function, satiety) of starchy foods

Study 1: Cognitive functions in relation to postprandial glycaemic response
The cognitive tests used (WM and SA) allowed discrimination of cognitive performance, as related to differences in postprandial glycaemia. A breakfast that had the capacity to maintain a higher net increment in blood glucose, in the later postprandial period, i.e. a low-GI breakfast, showed an advantage in cognitive performance, compared with a high-GI breakfast. After adjusting for glucose tolerance, the subjects performed significantly better in the later postprandial period during low-GI condition in both the WM-test and the test of SA, compared with the high-GI condition. But, even though a low-GI breakfast seemed to be preferable for cognitive function over all, a high GI-breakfast was shown to enhance the learning capacity. Learning capacity at repeated WM testing (35 - 150 min) was enhanced when subjects were challenged with a high-GI breakfast at the first test occasion. Also, the results showed that individual glucose tolerance has an important influence on cognitive performance. Subjects with higher glucose tolerance performed better in the cognitive tests, therefore suggesting that cognitive performance can be affected by glucose regulation, even if the glucose tolerance is within the normal range.

Study 2: Possible influence of botanical structure, fibre content and/or acetic acid on gastric emptying rate (GER) and satiety
GER was delayed and satiety increased by consuming coarse bread with added vinegar. Thus, the early (25 min) and late (100 min) median cross-sectional area of gastric antrum were larger after ingestion of the coarse bread with vinegar, than after ingestion of white wheat bread. Coarse bread with vinegar also gave significantly higher satiety scores at all time points (15-120 min), compared with white wheat bread. Furthermore a significant correlation was seen between early median cross sectional area (25 min) and early (0-15 min) and total satiety area (0-120 min). Thus, the early intragastric volume, as well as satiety area, increased in the following order, white wheat bread, white wheat bread with vinegar, wholemeal bread with vinegar and coarse bread with vinegar.

3 To investigate fermentation of resistant starch in the colon.

Technique development

Colonic 13 C- short-chain-fatty acid (SCFA) production determined in blood
Good linearity and repeatability (RSD < 5%) were obtained for acetate and butyrate. Plasma acetate, propionate and butyrate concentrations were also determined relative to 3-Methylvalerate (Internal Standard). Good linearity and repeatability were observed from 0 to 400 µM for acetate, from 0 to 20 µM for propionate and from 0 to 10 µM for butyrate.

13 C- volatile fatty acid (VFA) flux determined in urine
A robust GC/MS method to measure the concentration and 2 H-enrichment of VFA in urine and plasma was developed. It ensured that the sample preparation procedure yielded sufficient sample, in the correct form, for both GCMS and GC-combustion-Isotope Ratio MS (GC-C-IRMS) analysis. GC-C-IRMS is a method for compound-specific analysis at low 13 C-enrichment. This is a novel development, and it has been demonstrated that this procedure can be extended to propionate analysis.

The principle of this approach is to compare total acetate production (Area Under Curve; AUC) following two test meals, given under similar conditions, on different days. One meal contained resistant starch or another non-digestible carbohydrate (NDC) added to the usual meal, and the other comprised the same (unamended) meal. The difference in AUC is a measure of VFA production from the NDC. The AUC method is particularly suited to non-invasive measurements in urine.

Initial experiments confirmed the equivalence of 2 H 3 -acetate and 1- 13 C-acetate. Furthermore, urine sampling showed that the area under the curve of urinary tracer excretion yields essentially the same result as plasma plateau-derived acetate production. This also showed a bolus tracer dose to be equivalent to a 'pseudo' constant infusion (multiple oral doses).

A very important development has been the extension of this method to measure total acetate production to allow simultaneous measurement of total propionate production.

Characterization of the colonic microbiota: FISH
An extensive set of probes (15 probes) was developed to analyze the composition of the colonic microbiota. The technique has been automated in order to obtain a reasonable throughput and application in several studies. It was shown that there is a large inter-individual difference of microbiota composition.

Characterization of the colonic microbiota: DGGE
The DGGE method has been proven suitable to monitor changes in composition of the colonic microbiota.

Characterization metabolic capacity of the colonic microbiota: 15 N-retention
It has been shown that lactose-[ 15 N]-ureide is an adequate tool to monitor changes in the fate of ammonium in the colon.

Characterization of the metabolic capacity of the colonic microbiota: Proteomics
The protein profiles obtained were reproducible. Significant differences in protein profiles of each Bifidobacterium strain in media with glucose, lactose or without sugar could be detected with SELDI-TOF-MS, as well as significant differences in protein profiles of the 3 bifidobacterial strains in the same medium.

Study 1: Influence of resistant starch on 15 N-incorporation
Administration of RS for 2 weeks resulted in a significant decrease in 15 N-excretion over 48h (from 50.93 ± 14.41% at baseline, to 38.65 ± 12.04 after RS intake). The effect on the cumulative 15 N-excretion was mainly due to a decrease in the 6 - 24h fraction. The effect of RS administration was significantly different from that of placebo intake, whereas no difference could be found between RS and RS+WB. A tendency towards a more pronounced effect with RS was observed.

Study 2: Evaluation of the fermentation of resistant starch
It was shown that a high level of 13 C-enrichment of barley meals was necessary for in vivo determination of short-chain fatty acids. Based on the hydrogen data both test meals were shown to be incompletely digested, and that part of the test meals reached the colon where it fermented. The oro-cecal transit time of the test meals was the same. The 13 C-short-chain fatty acid measurements in urine are complete yet the results need to be worked out and interpreted. The 13 C-short-chain fatty acid measurement in plasma measurements also needs to be completed.

5 Conclusions 

1 To investigate the metabolic effects of starchy foods related to the rate and extent of digestion of starch: acute, semi-acute and long term effects

13 C-enriched substrates, grains and model foods have been successfully produced and applied.

Industrial food processing has been shown to dramatically modify the glucose response. Critical food processing parameters have been identified to achieve low gelatinised starch products which result in a low glycemic response.

With stable isotope techniques more insight has been gained in glucose kinetics after the ingestion of starchy foods, and the relationship with hormonal responses. The influx rate of exogenous glucose (RaE) is not the only determinant of the insulinemic response after consumption of starchy foods.

  • The regulation of endogenous glucose production (EGP) is not predominantly insulin dependent. The mechanism of this phenomenon is unknown, possible factors involved could include gastrointestinal hormones or adipokines.
  • The metabolic response after ingestion of whole meal wheat bread is not only determined by the rate of starch digestion, but also by other food factors.
  • The incretin response to rapidly and slowly available carbohydrates is different. The rate of exogenous glucose appearance is strongly correlated with the plasma concentrations of GIP. Slowly available carbohydrates can induce a late and prolonged GLP-1 response.

It has been found that certain low GI-foods which are high in resistant starch and dietary fiber (barley kernels), not only acutely decrease glycemia, but also glycemia after subsequent high GI food (breakfast to lunch, evening meal to breakfast). The finding that barley flour porridge showed no effects indicates the importance of the preserved botanic structure in the food matrix of barley kernels and/or an additional effect of resistant starch to achieve the overnight second-meal effect on glucose tolerance. The results suggest that events of colonic fermentation of indigestible carbohydrates are involved in modulating glucose tolerance from the evening to breakfast.

Substituting a HGI diet with a LGI diet for 5 weeks has been shown to result in decreased body weight and improved lipid profile in overweight healthy subjects. The effects of the altered rate of starch digestion and that of the higher content of dietary fiber can not be separated.

2 To investigate biological effects (cognitive function, satiety) of starchy foods

The cognitive tools developed allowed detection of differences in cognitive performance, related to the postprandial blood glucose level; higher glycaemia in the late postprandial phase resulting in improved cognitive function. The results also indicate that even in healthy elderly subjects, there was a significant difference in cognitive performance related to the efficiency in blood glucose regulation; where subjects with less efficient blood glucose regulation displayed lower scores in the cognitive tests. Finally, learning ability was enhanced with high blood glucose levels.

It has been shown that it is possible to influence the intragastric volume and thereby indirectly, satiety, with food factors such as botanical integrity and/or fiber content, and acetic acid.

3 To investigate fermentation of resistant starch in the colon.

Material and tools are now available to study the effect of resistant starch/dietary fiber fermentation on the production of short chain fatty acids as well as on changes in colonic microbiota:

•  13 C-barley with high enrichment makes it possible to trace short-chain fatty acids in plasma and urine.

•  An analytical method based on the use of Solid-Phase-Microextraction coupled with the GC-C-IRMS has been developed to measure very low 13 C-acetate and 13 C-butyrate enrichments within physiological concentrations in human plasma after ingestion of 13 C -enriched resistant starch.

•  A non-invasive method using urine samples to measure total acetate production is now available, and has been successfully applied in the fermentation study. Simultaneous measurement of total propionate production is possible. Propionate is potentially a key mediator in moderating aspects of the metabolic syndrome, such as circulating triglyceride and possibly, cholesterol levels.

•  The 15 N-incorporation method is available, which is a non-invasive strategy based on stable isotope technology, and focussing on the metabolite ammonia to measure quantitatively the influence of non-digestible carbohydrates on the metabolic processes in the colon. Based on this methodology it has been established that the ingestion of resistant starch results in a higher level of removal of ammonia from the colonic lumen.

•  With the FISH technique and specific probes it is possible to identify quantitatively and qualitatively a large number of different bacterial species of the gut ecosystem. An extensive set of probes is available; however the large inter-individual variation in composition of colonic microbiota makes it difficult to detect effects of dietary intervention.

•  With the DGGE method changes in the general composition of the microbiota, as well as in individual strains, can be monitored.

•  SELDI-TOF Protein Chip technology is available to detect changes in protein profiles of monocultures of bacteria induced by various substrates.

These techniques now make it possible to study, in detail, the fermentation processes in the colon and relate these to potential health benefits of different diets. Recent studies indicated that the colon plays an important role in modulating whole body energy metabolism. This set of developed techniques can be used for systematically studying complex interactions. 

6 Exploitation and dissemination of results 

The following aspects of the EUROSTARCH project can subsequently be used in the following specific fields:

Scientific aspects
Scientific research of each of the participating groups in EUROSTARCH will be published in international peer reviewed journals. More than 20 publications will be produced and a large number of presentations will be given at various international conferences (information not collected).

Health care aspectsIncreased knowledge about digestion and fermentation of starch will be used in dietary advice for healthy people and various patient groups. Stimulation of the consumption of healthy starch products (containing slowly digestible starch and fiber) might lead to a reduction of chronic diseases like obesity, diabetes mellitus type 2, and cardiovascular diseases.

Consumer information
An important aspect is the communication of EUROSTARCH results to consumers. Information about healthy starch products could impact on their dietary choices and consumption which could not only lead to a reduction in the risk of chronic diseases, but could also lead to an improvement of cognitive functions in certain situations, which might be highly relevant.

Medical science
In the EUROSTARCH project the stable isotope technology was used extensively. This has led to new functional tests (gastric emptying, intestinal transit, colonic function) and analytic procedures which can be broadly applied in bio-medical practice.

Stable isotope technology was used and further developed in the EUROSTARCH project, which led to the following consequences:

a) Enforcement of the EC stable isotope network.

b) Availability of 13 C-substrates like 13 C-grains and 13 C-lactose.

c) Availability of several new methods using 13 C-technology.

The combination of the new substrates and new methods creates a new challenge - to enter fields which are currently difficult to access, for instance postprandial glucose kinetics and monitoring colonic fermentation product profiles.

Economic activities
Research about the metabolic quality of starch has led to improved knowledge about digestion and fermentation of starch. This knowledge is the scientific basis for future development of starch based functional foods. In addition, this knowledge is a prerequisite for the basis of related health claims. Part of this project was also dedicated to the improvement of food processing, leading to healthier starchy products. Our industrial partner changed starch processing which led to products with slow-release starch characteristics. Given that slowly released starch is considered to reduce risk of chronic disease, this could also be the basis for selection of certain strains of grain diversity programmes or plant improvement programmes.

Environmental aspects
The stimulation of stable isotope technology can partly replace the use of radio-isotopes. This will contribute to the reduction of the environmental burden of radio-isotopes.

Societal aspects
It is well known that the efficacy of dietary advice to the healthy population is limited. Therefore, it is certainly worthwhile exploring how the food industry can contribute to improve food formulations by making healthier products, combined with efficient product labelling, and appropriate marketing of functional foods.

The EUROSTARCH project will be used as a model to study the process of development of functional foods and the role of various stakeholders in this process. Full results will be available after the completion of the project.

Dissemination of results

Results will be disseminated to stakeholders through various routes:

More than 20 papers will be published in international journals and presented in international meetings. During the EUROSTARCH project 5 workshop/conferences were organised around the scientific aspects of starch digestion and fermentation and stable isotope technology:

  1. "Application of Stable Isotopes in Gastroenterology and Nutrition"

1 st of September 2002, SIGN/EUROSTARCH workshop, Glasgow , UK

  1. "Application of Stable Isotopes in Gastroenterology and Nutrition"

1 st of October 2003, SIGN/EUROSTARCH/ISOMATS workshop, Rome , Italy

  1. "The glycemic index"

3 rd of October 2003, EUROSTARCH satellite symposium, FENS conference, Rome , Italy

  1. "International Workshop on Colon Fermentation and Metabolism"

14 th - 16 th of April 2004, EUROSTARCH workshop, Groningen , The Netherlands

  1. "Health aspects related to the digestion and fermentation of starch, results of the EUROSTARCH project in broader perspective"

27/28 th of April 2006, final EUROSTARCH conference, Groningen , The Netherlands

In presentation of the results to the public the role of starchy food in health has been explained on this website.

Food industry
Modification of food processing techniques for starch that leads to the development of products with improved health properties is an important issue for the food industry. The International Life Sciences Institute (ILSI) has proposed a workshop on 6 - 8 th of December 2006 in Nice. This will focus on the health aspects of GI-foods and it will certainly contribute to the further dissemination of knowledge about the health properties of low GI foods.

Policy makers
Information about improved dietary advice, based on the metabolic qualities of starch and better functional foods, along with reliable health claims, are important issues for preventive nutrition and health care. Moreover, as knowledge increases as a result of such scientific investigations, it can be applied by the food industry, health and medical professions and the wider public. These applications can be studied to give information about how best to manage and regulate future developments and their impact for society.

8 Literature cited

(1) Connor WE, Connor SL. Should a low-fat, high-carbohydrate diet be recommended for everyone? The case for a low-fat, high-carbohydrate diet. N Engl J Med 1997; 337(8):562-563.

(2) Katan MB, Grundy SM, Willett WC. Should a low-fat, high-carbohydrate diet be recommended for everyone? Beyond low-fat diets. N Engl J Med 1997; 337(8):563-566.

(3) Jenkins DJ, Wolever TM, Taylor RH, Barker H, Fielden H, Baldwin JM et al. Glycemic index of foods: a physiological basis for carbohydrate exchange. Am J Clin Nutr 1981; 34(3):362-366.

(4) Robertson MD, Currie JM, Morgan LM, Jewell DP, Frayn KN. Prior short-term consumption of resistant starch enhances postprandial insulin sensitivity in healthy subjects. Diabetologia 2003; 46(5):659-665.

(5) Robertson MD, Bickerton AS, Dennis AL, Vidal H, Frayn KN. Insulin-sensitizing effects of dietary resistant starch and effects on skeletal muscle and adipose tissue metabolism. Am J Clin Nutr 2005; 82(3):559-567.

(6) Anderson JW, Bridges SR. Short-chain fatty acid fermentation products of plant fiber affect glucose metabolism of isolated rat hepatocytes. Proc Soc Exp Biol Med 1984; 177(2):372-376.

(7) Kok N, Roberfroid M, Robert A, Delzenne N. Involvement of lipogenesis in the lower VLDL secretion induced by oligofructose in rats. Br J Nutr 1996; 76(6):881-890.

(8) Nishina PM, Freedland RA. Effects of propionate on lipid biosynthesis in isolated rat hepatocytes. J Nutr 1990; 120(7):668-673.

(9) Brown AJ, Goldsworthy SM, Barnes AA, Eilert MM, Tcheang L, Daniels D et al. The Orphan G protein-coupled receptors GPR41 and GPR43 are activated by propionate and other short chain carboxylic acids. J Biol Chem 2003; 278(13):11312-11319.

(10) Gimeno RE, Klaman LD. Adipose tissue as an active endocrine organ: recent advances. Curr Opin Pharmacol 2005; 5(2):122-128.



eurostarch logo

Fifth framework programme Quality of life and management of living resources


Members login