Another similar in vitro starch hydrolysis method was also shown to have a good correlation with the in vivo GI assay (Goni et al., 1996). Different cereal, legume and vegetable foods were studied, and the HI values were calculated. Starch hydrolysis at 90 min correlated even slightly better (r = 0.909, P < 0.05) with in vivo glycaemic responses than the HI values (r = 0.894), and was suggested as a simpler way of predicting the GI.
Analytical methods have been developed to classify starch to rapidly available glucose (RAG) and slowly available glucose (SAG) (Englyst et al., 1999, 2003).The enzymatic procedure included incubation with both pepsin and and a mixture of amylolytic enzymes, and pH, temperature, viscosity and mechanical mixing were adjusted to mimic the gastrointestinal conditions (Englyst et al., 1999). It was concluded that these parameters can explain the GI and II values of the cereal foods studied (Englyst et al., 2003). SAG and fat content together accounted for 73.1 per cent of the variance in GI, and RAG and protein content together 45.0 per cent of the variance in II.
The chewing/dialysis method does not mimic the events of gastric emptying. A method has been developed that would better take into account the particles size of the chewed food, and the gastric emptying (Liukkunen et al. unpublished). The samples are chewed in the mouth (four persons). Then the masticated residues will be treated with pepsin, mimicking the stomach conditions (mixing, acid, pepsin, +37 ëC). After dilution, the particles are photographed by camera and the particle size distribution is determined from all the masticated material by image analysis. This method gave good correlation with II (Liukkonen et al. unpublished). The particles originating from pasta were clearly largest, and those of normal wheat bread smallest.
The metabolic syndrome as a concurrence of disturbed glucose and insulin metabolism, overweight and abdominal fat distribution, dyslipidaemia and hypertension constitutes a major threat to public health because of its association with increased risk of type 2 diabetes mellitus and cardiovascular disease. The hallmark of the syndrome is resistance to biological effects of insulin in target tissues, and is also known as the insulin resistance syndrome (Reaven, 1988). The ultimate pathogenesis of this syndrome is unknown, but obesity and sedentary lifestyle in association with a Western type of diet and genetic factors clearly interact to produce it. With increasing occurrence of overweight and obesity, the problem of the metabolic syndrome as a threat to public health will increase.
Despite the abundant research that has been published on the metabolic syndrome, definitions of the metabolic syndrome have been numerous. The World Health Organization (WHO, 1999) consultation for the classification of diabetes and its complications, and the National Cholesterol Education Program Expert Panel (2001), have recently published definitions of the metabolic syndrome. The manifestations of cardiovascular risk factors such as dyslipidaemia, hypertension, endothelial dysfunction, inflammation, hypercoagulability and impaired fibrinolysis, obesity and abnormal insulin and glucose metabolism predispose persons with the metabolic syndrome to another major consequence of the metabolic syndrome, cardiovascular disease. In the Finnish population-based prospective study applying defined criteria of metabolic syndrome the overall mortality was also nearly two times higher in men with the metabolic syndrome (Lakka et al., 2002). Adjustment for conventional and non-conventional risk factors that were not part of the definition of the metabolic syndrome did not attenuate the risks.
The pathogenesis of the metabolic syndrome is poorly understood. Skeletal muscle is a major determinant of whole-body glucose disposal and the defects in insulin signalling in muscle tissue contribute to lowered insulin-stimulated glucose uptake (Kelley and Mandarino, 2000). Further, an abdominal distribution of fat is particularly deleterious. Abdominal fat can also be divided into subcutaneous and visceral compartments. Abdominal obesity has been hypothesized to mediate its deleterious effects on carbohydrate and lipid metabolism through the increased lipolytic activity of especially omental fat, which drains directly into the portal-venous system (BjoÈrntorp, 1991). This in turn results in higher non-esterified fatty acid concentrations, with consequent insulin resistance in the liver and skeletal muscle. In addition to abdominal subcutaneous and visceral fat, the lipid accumulation in skeletal muscle and liver has also been shown to be powerful determinants of insulin sensitivity (Ravussin and Smith, 2002).
As the metabolic syndrome becomes more severe, interplay between genetic susceptibility, insulin resistance and dietary patterns may lead in susceptible individuals to progressive 3-cell failure and impaired insulin secretion capacity. As 3-cell function declines, impaired glucose tolerance (IGT) develops (Kahn, 2003). Roughly 5±10 per cent of persons with IGT convert to type 2 diabetes yearly (Edelstein et al., 1997). With time, hyperglycaemia leads to further loss of pancreatic 3-cell function. It has not been fully resolved whether this loss of pancreatic function results primarily from excessive secretion of insulin (i.e., 3- cell exhaustion) or toxicity to 3-cells because of hyperglycaemia. However, from this mechanism one may anticipate that a diet that produces higher plasma glucose concentrations and greater demand for insulin would increase the risk of type 2 diabetes. By definition, high-GI forms of carbohydrate produce high concentrations of plasma glucose and increased insulin demand and may therefore contribute to an increased risk of type 2 diabetes. The individual response to a given carbohydrate load is influenced by the degree of underlying insulin resistance, which is, in turn, determined primarily by degree and type of adiposity, physical activity, genetics and other aspects of diet. Thus, it might be expected that the adverse metabolic effects of high-GI foods would be pronounced in sedentary, overweight or genetically susceptible persons and be quite modest in healthy graduate students participating frequently in metabolic studies.
One of the problems with the GI concept is that the insulin response is not proportional to the glucose response. Holt et al. (1997) compared the effect of isoenergetic amounts of foods on the insulin secretory response and found that the postprandial insulin responses were not closely related to the carbohydrate content or to the glycaemic effects of the foods. Whereas the glycaemic response was a significant predictor of the insulin response, it accounted for only 23 per cent of the variability in the insulinaemia. This is in line with our experience in which in healthy subjects with normal glucose tolerance there was no difference in the glycaemic responses between wheat or different rye breads, whereas insulin responses of rye breads were markedly different so that the same amount of carbohydrate required less insulin. Therefore it seems that in healthy persons the use of GI is hampered by many difficulties, and insulin responses could be used instead. However, even if the total glucose responses did not differ between the different types of breads (Juntunen et al., 2002), the shapes of the glucose curves showed interesting patterns ± after 3 hours from ingestion of the wheat bread plasma glucose levels had declined below the fasting level whereas after rye bread ingestion glucose levels were still above fasting level. The rapid decrease of blood glucose may increase the feeling of hunger and may cause more frequent eating.
Recent nutritional guidelines usually recommend high intakes of carbohydrate. The main reason was that high-fat diets enhanced the development of cardiovascular diseases. Recently, it has been acknowledged that the type of fat is at least as important as the quantity of fat. It seems that we are encountering the same phenomenon as regards to type of carbohydrates in the diet; the quality is at least as important as quantity. Epidemiological studies addressing the association of dietary factors with the metabolic syndrome using the WHO or NCEP definitions have not yet been published and there is an urgent need for controlled randomized studies in these group of subjects. However, the large cohort studies assessing the association of incidence of diabetes mellitus, or cardiovascular disease nonetheless suggest that a diet that has a low glycaemic index, high in fibre or high in fruit and vegetable intake may decrease the risk for obesity, type 2 diabetes or cardiovascular disease and its risk factors. Herein we review the effects of low-GI diets and their effects on the treatment of overt diabetes, prevention of type 2 diabetes, obesity and serum lipoproteins.
Low-GI carbohydrates may have acute effects on the clinical management of people with diabetes, e.g. patients treated with short-acting prandially administered insulin analogues may need to inject their insulin after eating. However, the main goal of treatment of diabetes is to prevent long-term complications by improving metabolic control, which is commonly assessed by glycated proteins, such as fructosamine or haemoglobin A1c. In studies conducted mostly in patients with type 2 diabetes the low-GI diets have shown an improvement of 10 per cent as compared to high-GI foods (FAO/WHO, 2003). This effect is clinically significant and comparable to the effects of monounsaturated fat compared with high-carbohydrate diets.
Recently, a meta-analysis identified 14 studies, comprising 356 participants, that met strict inclusion criteria and that showed that low-GI diets reduced HbA1c by 0.43 per cent points above that produced by high-GI diets (Brand-Miller et al., 2003). Taking both HbA1c and fructosamine data together and adjusting for baseline differences, glycated proteins were reduced 7.4 per cent more on the low-GI diet than on the high-GI diet. Therefore, choosing low-GI foods in place of conventional or high-GI foods has a small but clinically useful effect on medium-term glycaemic control in patients with diabetes and the benefit is similar to that offered by pharmacological agents that also target postprandial hyperglycaemia.
In a carefully conducted Swedish study (JaÈrvi et al., 1999) in patients with type 2 diabetes the baseline diet for both study groups was in line with the current recommendations but differed in GI. Although both diets improved insulin sensitivity and serum lipids, food modification by lowering GI about 30 per cent resulted in lower LDL-cholesterol, glucose and insulin responses. Low-GI has been shown to have beneficial effects on LDL-cholesterol levels in patients with type 2 diabetes during weight-losing diets (Heilbronn et al., 2002). In subjects with type 1 diabetes EURODIAB study showed that low-GI was associated with better metabolic control, more favourable lipoprotein pattern and smaller waist circumference (Buyken et al., 2001).
Findings from epidemiological studies suggest that total carbohydrate intake is not associated with increased risk of type 2 diabetes (Feskens et al., 1995; Salmeron et al., 1997a, b; Meyer et al., 2000) whereas the intakes of whole grain cereal products, such as ryebread, have been shown to reduce the risk (Meyer et al., 2000; Liu et al., 2000b; Montonen et al., 2003), and highly refined grain products such as white wheat bread and white rice to increase the risk (Salmeron et al., 1997a, b). The results are conflicting concerning the risk increase due to refined grain products (Liu, 2003).
Multifactorial prevention of type 2 diabetes including increasing fibre intake as a component has been shown to prevent the development of type 2 diabetes. In the Finnish Diabetes Prevention Study (DPS) 523 obese persons with IGT were randomized in five centres into an intervention and control group. The study was stopped prematurely after consultation with an expert panel because of highly significant difference in the incidence of diabetes between the intervention and control groups (Tuomilehto et al., 2001). During the trial the risk reduction of diabetes was 58 per cent. When the results were analysed according to the success score, none developed diabetes in either group if they achieved four or five intervention goals (weight loss at least 5 per cent, physical activity at least 4 h/week, fibre intake > 15 g/1000 kcal, intake of total fat less than 30 per cent of energy and that of saturated fat less than 10 per cent of energy).
What are the theoretical mechanisms by which low-GI foods prevent development of type 2 diabetes? The fundamental defects are insulin resistance and impaired -cell function. In a study by Frost et al. (1996) women were randomly assigned to consume high- or low-GI diets for 3 weeks. Insulin resistance measured in vivo and in cultured adipocytes was greater in women consuming the high-GI diet. The adverse effects of the high-GI diet appeared to be due to an increased production of free fatty acids in the late postprandial state. In a study by Kiens and Richter (1996) an increase in insulin resistance was not found in seven healthy, lean young men after the subjects had consumed a high-GI diet. It should be noted that subjects were quite healthy and had normal insulin sensitivity, emphasizing the concept that metabolic effects of GI are difficult to observe in healthy population. A recent crossover study involving 11 overweight subjects showed that insulin sensitivity measured by the euglycaemic hyperinsulinaemic clamp improved after subjects consumed a whole-grain diet compared with a refined-grain diet for 6 weeks, independent of body weight (Pereira et al., 2002). Thus there is suggestive evidence of relationship of GI with insulin resistance but further studies are required.
-Cell failure may be induced by high-GI foods by repeated postprandial hyperinsulinaemia leading to its overstimulation and exhaustion (Ludwig, 2002b; Grill and Bjorklund, 2001). In a crossover study of six healthy adults, Jenkins et al. (1987) found that a low-GI diet containing mainly intact whole grains reduced C-peptide concentrations (a 32 per cent reduction) compared with a high-GI diet containing primarily refined grain products. Recently, we showed that long-term ingestion of rye bread in elderly women enhanced acute insulin response to intravenous glucose stimulus (Juntunen et al., 2003a), which could be a novel mechanism for the effects of low GI on glucose and insulin.
In theory, low-GI foods may benefit weight control by promoting satiety and fat oxidation at the expense of carbohydrate oxidation. Most studies have suggested that low-GI meals increase fullness to a greater extent than do comparable high-GI meals (Ludwig, 2002a). This may be due to `ups and downs’ or hyperglycaemic and hypoglycaemic effects of high-GI foods that could partly explain the lower satiety observed in the postprandial period. Further, prolonged transit time in the gastrointestinal tract may have a prolonged stimulation of receptors signalling to central satiety centres. Slabber et al. (1994) studied 30 obese females with two hypocaloric diets (high-GI versus low-GI) in a 12-week study that was followed by a 12-wk crossover study. Both diets produced weight loss during the first 12 weeks (9.4 compared with 7.4 kg, p = NS). During the follow-up crossover study, the low-GI diet produced greater weight loss (7.4 compared with 4.5 kg; P = 0.04) than did the high-GI diet. The further satiety mechanism may be due to the second-meal effect; studies with variable GI and load have suggested that the lower the glycemic index and load of the first meal, the less food is consumed in the subsequent meal (Wolever et al., 1988; Liljeberg and BjoÈrck, 2000).
Low-GI foods may also have relevance with fuel oxidation. Postprandial rises in glucose and insulin concentrations increase carbohydrate oxidation and decrease fatty acid oxidation but during later postprandial phases increased release of counter-regulatory hormones restores euglycaemia and elevates free fatty acid (FFA) levels. Increased availability and oxidation of fatty acids may in turn decrease carbohydrate oxidation (Jenkins et al., 2002b). Whether high-GI diets, which induce chronic hyperglycaemia and hyperinsulinaemia, can reduce the body’s capacity to oxidize fat and significantly increase body fat storage remains unresolved. There is an urgent need for randomized, controlled, multicentre intervention studies comparing the effects of conventional and low-GI diets on body weight, composition and fuel metabolism to confirm or cancel the hypothesis that the faster digestion and absorption and higher insulin responses after high-GI meals affect in different ways satiety and energy partitioning that over the long term favour expansion of the fat stores.
High carbohydrate consumption is associated with increased serum triglyceride and low HDL-cholesterol levels, both of which are hallmarks of the metabolic syndrome and increase the risk of cardiovascular diseases. Serum triglycerides change more rapidly in metabolic studies, whereas epidemiological studies have shown an association between low HDL-cholesterol and GI.
A retrospective analysis of 7-day weighted records of 1420 British adults performed in 1986±1987 and showed a significant negative correlation between serum HDL cholesterol and GI (P < 0.0001 for women, for men P = 0.02) (Frost et al., 1999). The difference in the HDL-cholesterol between the highest and lowest GI quintiles was 0.25 mmol/L, which is a clinically significant difference. Data from United States of 14000 subjects showed consistent findings in this regard: dietary GI and plasma HDL-cholesterol concentrations showed an inverse association and the differences in HDL-cholesterol concentrations between the lowest and highest GI quintiles was 0.10 mmol/L (Ford and Liu, 2001).
There has been an enormous amount of work concerning the impact of longer-term consumption of carbohydrate-rich diets on blood lipids, which has been reviewed recently (Parks and Hellerstein, 2000). When the content of carbohydrates is increased above 50 per cent of the daily intake of energy at the expense of fat, and especially monounsaturated fat, it has been repeatedly shown to induce hypertriglyceridaemia coupled with lowered HDL-cholesterol. Even if the mechanisms for hypertriglyceridaemia are unresolved, it has been suggested to result from the overproduction of both VLDL triglycerides and VLDL particles together with impairment of VLDL clearance.
However, if high-carbohydrate diets, which are composed of natural foods, have been rich in starchy unrefined and whole products, they have not shown elevations in triglyceride concentrations as compared with high-fat, low-carbohydrate diets. In a recent meta-analysis of 11 studies triglycerides and total cholesterol were reduced on average by 9 and 6 per cent, respectively, coupled with improvements in glycaemic control and insulin sensitivity when low- vs. high-GI diets were compared. While changing only around half the carbohydrates from conventional foods to low-GI ones, which is around 25±30 per cent from dietary calories, measurable gains have been observed (Miller, 1994). It is therefore appealing to suggest that intake of low-GI foods overcomes the adverse changes in lipid metabolism induced by high-carbohydrate diets, and offers a practical means to treat and prevent diseases linked to metabolic syndrome.
The Nurses Health Study, consisting of 10 years follow-up on 75 000 women aged 38±63 years, showed that a high glycaemic load from dietary carbohydrates increases risk of coronary heart disease (Liu et al., 2000a). The glycaemic load was obtained by using a food-frequency questionnaire. During the follow-up, 763 myocardial infarctions occurred. Defined glycaemic load was associated with risk of myocardial infarction after adjustment for other risk factors, like age, smoking, total energy intake, hypertension and dyslipidaemia. The relative risk of myocardial infarction and glycaemic load was most pronounced in overweight women. However, this large study has been criticized for the food-frequency data used (Pi-Sunyer, 2002).
Low-grade inflammation is a rather novel characteristic of obesity, metabolic syndrome and atherosclerosis (Ridker et al., 2003). The GI may have a role in the liberation of proinflammatory cytokines and acute phase proteins and a high-GI diet may induce the generation of reactive oxygen species and lowering of serum antioxidants, thus contributing to oxidative damage (Jenkins et al., 2002b). The relation of GI to cancer has been also been suggested (Terry et al., 2003). Mental performance should also be a subject of further studies and recently it has been shown in elderly people with type 2 diabetes that acute ingestion of high-GI carbohydrate contributes to the memory impairment (Greenwood et al., 2003).
Findings from different studies of diverse populations suggest that intake of whole-grain products can lower the risk of type 2 diabetes (Salmeron et al., 1997a, b; Meyer et al., 2000; Liu et al., 2000b; Fung et al., 2002; McKeown et al., 2002; Montonen et al., 2003). There is a need to develop greater variation of low-GL commercial cereal-based products to the market, but the greatest challenge is the management of good perceived texture. If a product contains fibre, the starch content will be lower and the blood glucose increasing potential lower than without fibre. Breads that have beneficial effects on glucose and insulin responses can be baked by either adding fibre or by lowering rate of starch digestibility.
Mutations that increase the ratio of amylose to amylopectin in starch-containing plants can be used for developing new cereal-based ingredients that induce low glycaemic responses. Maize varieties homozygous for the amylose extender or `ae’ gene, with amylose levels between 50 and 80 per cent, have been commercially available since the 1950s (Vineyard et al., 1958). High-amylose barley, maize and rice have been studied in relation to digestibility in vitro and in vivo.
A lower glycaemic and insulin index and higher content of resistant starch have been reported for a barley bread baked from a high-amylose barley genotype by using long-time/low-temperature baking conditions (AÊ kerberg et al., 1998). Recent studies have shown that breads baked from high-amylose wheat flour had significantly lower loaf volume than breads baked from normal flours (Morita et al., 2002). Concerning bread products the great challenge is to combine low starch digestibility with good sensory characteristics (Autio et al., unpublished).
In many studies rice products based on high-amylose varieties were shown to lower GI (Goddard et al., 1984; Juliano and Goddard, 1986; Miller et al., 1992). A study including 12 different rice products, found lower GI and II only for high-amylose rice products. With cooked rice, hardness increased and stickiness decreased with an increase in the amylose content (Kohyama et al., 1998). Hardness is also greatly dependent on the degree of cooking, and water content. High amylose rice is a potential new raw material, as amylose also improves the textural properties of cooked rice products, such as rice noodles (Yoenyongbuddhagal and Noomhorn, 2002).
High-amylose maize is available as a cereal grain and as a starch. Reduced postprandial responses of glucose and insulin in healthy subjects following ingestion of crackers made from high-amylose maize starch compared with a corresponding product made from low-amylose starch has been reported (Behall et al., 1988). Autoclaved high-amylose maize starch when incorporated in hot mixed lunches had beneficial effects on glucose and insulin responses (Van Amelsvoort and Weststrate, 1992).
The use of soluble fibres as ingredients for low-GI foods presumes that they are soluble and have high viscosity, which can be achieved by combining high-molecular weight gum with low concentration or low-molecular weight gum with high concentrations. Oat and guar gum are possible candidates for low GI foods. Brennan et al. (1996) have shown that the rate of starch hydrolysis was retarded significantly when the starch granules and surrounding bread matrix were coated with a layer of galactomannan. It is easier to solubilize gums with higher amounts of water and thus foods, such as drinks, soups and sauces could be more suitable than foods with lower water content. With an oat 0-glucan preparation of high-molecular weight, the maximum practicable concentration of 0-glucan in the soup was 0.5 per cent (Lyly et al., unpublished). With a lower molecular weight preparations of oat and barley, it was possible to add up to 2 per cent 0-glucan to the soup. The viscosities of 0.5 per cent high molecular weight oat, 2 per cent lower molecular weight oat and barley, at 50s—1 were 347, 1575 and 343 mPa s respectively. From the point of view of flavour release, barley 0-glucan was better than oat.
Bran supplementation or use of wholemeal flours that contain bran usually weakens the bread volume, loaf structure and mouthfeel (Rao and Rao, 1991; Zhang and Moore, 1999; Salmenkallio-Marttila et al., 2001). Prefermentation of wheat bran with yeast and with yeast and lactic acid bacteria improved the loaf volume, crumb structure and shelf-life. The bread had good flavour and homogeneous pore structure. A combination of commercial baking enzymes also had a positive effect on the volume, structure and flavour of bread containing wheat bran (Salmenkallio-Marttila et al., unpublished). The best result was obtained by combining the use of prefermentation and addition of baking enzymes. In another study, three different wholemeal breads (60 per cent wholemeal flour on the basis of flour weight) baked from oat, rye and wheat were fed to 15±20 healthy subjects, and their II and GL were determined (Autio et al., unpublished). The texture of wholemeal breads was improved by gluten addition in order to improve sensory characteristics of the breads. All wholemeal breads had an II about the same or slightly higher than normal wheat bread. All breads contained 5.9±7.0 per cent fibre, their GL were for oat 9.7, for rye 10 and for wheat 11 in comparison to 14.2 obtained for normal wheat bread.
Aucun commentaire:
Enregistrer un commentaire