Introduction

Much has been learned about carbohydrate digestion and absorption over the last 20 years, and this new knowledge has, in many ways, changed completely the way we think about dietary carbohydrates. We now know that starches are not completely digested, and, indeed, some are quite poorly digested. We have learned that the undigestible carbohydrates are not just neutral bulking agents, but have important physiologic effects, and even contribute energy to the diet. "Sugar" is not bad for health, and starches are not all equal in their effects on blood glucose and lipids. However, knowledge in all these areas is far from complete. In addition, there is unresolved controversy about how to define and how to measure dietary fibre and starch, and different methods are in use in different parts of the world. This presents a major challenge for those who have the responsibility of formulating policies and recommendations about dietary carbohydrates and how the energy value and carbohydrate composition of foods is determined.

Energy value of carbohydrates

Many different methods have been used to determine how much of the energy in foods is available for human metabolism, termed metabolizable energy (ME). The total amount of energy in a food (TE) can be determined by calorimetry, but ME is less than TE because not all the energy in food is absorbed and some is absorbed, but lost in the urine. Most of the energy not absorbed ends in the feces, but some is lost in the gases and heat produced during colonic fermentation.

The most common approach for determining the energy content of foods is the factorial method (68) in which the amount of energy contained in each of the various components of the food (ie. fat, protein, carbohydrate, alcohol) is calculated, and the sum of the resulting figures is taken as the amount of energy in the food. Determining the energy value of carbohydrate presents a conceptual challenge because carbohydrates vary in their gross energy content per gram, the degree to which they are digested and absorbed, and the fact that the undigestible carbohydrates provide an amount of energy which depends upon the degree to which they are fermented in the colon. This may vary from 0 to 100%. Alternative empirical models have been proposed based on regression equations developed from experiments where gross energy intake and energy excretion in urine and stool were measured on a variety of diets. Here, metabolizable energy in the diet is equal to gross energy intake minus energy losses, the latter being estimated from nitrogen and unavailable carbohydrate intakes. It has been argued that empirical models for determining the energy content of the diet are more accurate than the factorial approach because they have fewer and smaller sources of error (68). Nevertheless, it seems unlikely that the factorial approach will be replaced, at least in the near future, because it is ingrained in food labelling regulations and food tables.

Digestion and absorption of carbohydrates

Polysaccharides and oligosaccharides must be hydrolyzed to their component monosaccharides before being absorbed. The digestion of starch begins with salivary amylase, but this activity is much less important than that of pancreatic amylase in the small intestine. Amylase hydrolyzes starch, with the primary end products being maltose, maltotriose, and a -dextrins, although some glucose is also produced. The products of a -amylase digestion are hydrolyzed into their component monosaccharides by enzymes expressed on the brash border of the small intestinal cells, the most important of which are maltase, sucrase, isomaltase and lactase (69). With typical refined Western diets, carbohydrate digestion is rapid and carbohydrate absorption occurs primarily in the upper small intestine. This is reflected by the presence of finger-like villi in the mucosa of the upper small intestine, with wider and shorter villi in the lower half of the small intestine. However, carbohydrate digestion and absorption can occur along the entire length of the small intestine, and is shifted toward the ileum when the diet contains less readily digested carbohydrates, or when intestinal glucosidase inhibitors which may be used to treat diabetes are present. In this situation, the upper small intestine exhibits wide villous structures with leaf-like arrays, while in the ileum the villi become longer and more finger-like.

Monosaccharides

Only D-glucose and D-galactose are actively absorbed in the human small intestine. D-fructose is not actively absorbed, but has a rate of diffusion greater than would be expected by passive diffusion. The sodium dependent glucose transporter, SGLT1, is responsible for the active transport of glucose or galactose with an equimolar amount of sodium against a concentration gradient into the cytoplasm of the enterocyte. Fructose is taken up by facilitated transport by the glucose transporter 5 (GLUT5). Glucose is pumped out of the enterocyte into the intracellular space by the glucose transporter 2 (GLUT2) (70). The complete mechanism of fructose absorption in the human intestine is not understood. When fructose is given alone in solution, 40-80% of subjects have malabsorption, and some subjects can absorb less than 15g fructose. Flatulence and diarrhoea are common if doses of fructose over 50g are given by mouth. However, if fructose is given in combination with glucose or starch, fructose is completely absorbed, even in subjects who malabsorb fructose alone (71). Since fructose rarely occurs in the diet in the absence of other carbohydrates, fructose malabsorption is really only a problem for studies involving oral fructose loads.

Disaccharides

Intestinal brush border glucosidases tend to be inducible. For example, there is evidence that a high sucrose intake increases the postprandial insulin and the gastric inhibitory polypeptide responses to large loads of oral sucrose (72), which probably reflects an increased rate of absorption due to induction of intestinal sucrase activity. Lack of brush border glucosidases results in an inability to absorb specific carbohydrates. This occurs rarely, except for lactase deficiency which is common in non-Caucasian populations. The latter may be complete or partial and results in a reduced ability to digest and absorb lactose.

The Glycemic Index

The blood glucose responses of carbohydrate foods can be classified by the glycemic index (GI). The GI is considered to be a valid index of the biological value of dietary carbohydrates. It is defined as the glycemic response elicited by a 50g carbohydrate portion of a food expressed as a percent of that elicited by a 50g carbohydrate portion of a standard food (73). The glycemic response is defined as the incremental area under the blood glucose response curve, ignoring the area beneath the fasting concentration (i.e. the area beneath the curve) (74-76). The standard food has been glucose or white bread. If glucose is the standard, (ie. GI of glucose = 100) the GI values of foods are lower than if white bread is the standard by a factor of 1.38 because the glycemic response of glucose is 1.38 times that of white bread. GI values for several hundred foods have been published (77,78) (see Table 8).

The Glycemic Index and Mixed Meals

The validity of the GI has been the subject of much controversy, mostly because of supposed lack of application to mixed meals. Much of the controversy has been because of application of inappropriate methods to estimate the expected glycemic responses for mixed meals. When properly applied, the GI predicts, with reasonable accuracy, the relative blood glucose responses of mixed meals of the same composition but consisting of different carbohydrate foods (79).

Implications of the Glycemic Index

There are a number of long-term implications of altering the rate of absorption, or GI, of dietary carbohydrate. There is good evidence that reducing diet GI improves overall blood glucose control in subjects with diabetes (80) and reduces serum triglycerides in subjects with hypertriglyceridemia (81).

There is also some evidence that the glycemic index is relevant to sports nutrition and appetite regulation. Low GI foods eaten before prolonged strenuous exercise increased endurance time and provided higher concentrations of plasma fuels toward the end of exercise (82). However, high GI foods led to faster replenishment of muscle glycogen after exercise (83).

TABLE 8 Glycemic index of selected foods (continues on next page)



GI* n**

GI* n** BAKED GOODS GRAINS Cakes 87±5 9 Pearled barley 36±3 4 Cookies 90±3 14 Cracked barley 72 1 Crackers, wheat 99±4 8 Buckwheat 78±6 3 Muffins 88±9 8 Bulgur 68±3 4 Rice cakes 123±6 2 Couscous 93±9 2 Cornmeal 98±1 3 BREADS Barley kernel 49±5 3 Sweet corn 78±2 7 Barley flour 95±2 2 Millet 101 1 Rye kernel 71±3 6 Rice, white 81±3 13 Rye flour 92±3 10 Rice, low amylose 126±4 3 Rye crispbread 93±2 5 Rice, high amylose 83±5 3 White bread 101±0 5 Rice, brown 79±6 3 Whole-meal flour 99±3 12 Rice, instant 128±4 2 Other productsa 100±4 5 Rice, parboiled 68±4 13 Specialty rices 78±2 5 BREAKFAST CEREALS Rye kernels 48±4 3 All bran 60±7 4 Tapioca 115±9 1 Cornflakes 119±5 4 Wheat keenelsa 59±4 4 Muesli 80±14 4 Oat bran 78±8 2 DAIRY PRODUCTS Porridge oats 87±2 8 Ice cream 84±9 6 Puffed rice 123±11 3 Milk, whole 39±9 4 Puffed wheat 105±3 2 Milk, skim 46 1 Shredded wheat 99±9 3 Yogurt d 48±1 2 Other, GI ³ 80b 103±3 15 Yogurt e 27±7 2 Other, GI<80c 72±2 4 FRUIT LEGUMES Apple 52±3 4 Baked beans 69±12 2 Apple juice 58±1 2 Black-eyed peas 59±12 2 Apricots, dried 44±2 2 Butter beans 44±3 3 Apricots, canned 91 1 Chickpeas 47±2 3 Banana 83±6 5 Canned chickpeas 59±1 2 Banana, underripe 51±8 2 Haricot beans 54±8 5 Banana, overripe 82±8 2 Kidney beans 42±6 7 Kiwifruit 75±8 2 Kidney beans, canned 74 1 Mango 80±7 2 Lentils 38±3 6 Orange 62±6 4 Lentils, green 42±6 3 Orange juice 74±4 3 Lentils, green canned 74 1 Paw paw 83±3 2 Lima beans 46 1 Peach, canned 67±12 3 Peas, dried green 56±12 2 Pear 54±4 4 Peas, green 68±7 3 Other, GI<80f 54±7 7 Pinto beans 61±3 3 Other, GI ³ 80g 92±4 5 Soya beans 23±3 3 Split peas, yellow 45 1 PASTA SNACKS Linguine 71±4 6 Jelly beans 114 1 Macaroni 64 1 Lifesavers 100 1 Macaroni, boxed 92 1 Chocolate (various) 84±14 2 Spaghetti, white 59±4 10 Popcorn 79 1 Spaghetti, durum 78±7 3 Corn chips 105±2 2 Spaghetti, brown 53±7 2 Potato chips 77±4 2 Other 59±3 8 Peanuts 21±12 3 POTATOES SOUPS Instant 118±2 5 Bean soups (various) 84±7 4 Baked 121±16 4 Tomato 54 1 New 81±8 3 SUGARS White, boiled 80±2 3 Honey 104±21 2 White, mashed 100±2 3 Fructose 32±2 4 French fries 107 1 Glucose 138±4 8 Sweet potato 77±11 2 Sucrose 87±2 5 Yam 73 1 Lactose 65±4 2