Digestion, absorption and energy value of carbohydrates

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, GI³ 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, GI³ 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

*GI = glycemic index (white bread = 100), mean ± SEM of mean values from various studies.
**Number of studies.

^a Bagel, stuffing mix, hamburger bun, rolls, melba toast.

^b Bran buds, Bran chex, Cheerios, Corn bran, Corn chex. Cream of wheat, Crispix, Golden Grahams, Grapenuts, Grapenuts flakes, Life, Pro stars, Sustain, Team, Total (GI range, 83-127)

^c Bran buds with psyllium, Red River, Special K (Australia), Sultana bran (Australia) (GI range 67-77)

^d Sweetened with sugar

^e Artificially sweetened

^f Cherries, fruit cocktail, grapefruit, grapefruit juice, grapes, plum, pineapple juice

^g Pineapple, raisins, rockmelon, sultanas, watermelon