![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
Food Science and Human Nutrition Department, Iowa State University, Ames, IA 50011
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
ACKNOWLEDGMENTS
FOOTNOTES
LITERATURE CITED
ABSTRACT 
In liquid enteral formulations, high molecular weight soluble starches may be able to replace glucose and low molecular weight glucose polymers that have high glycemic indices. Male rats were fed either commercial cornstarch, dextrose, modified soluble potato (70-75% amylopectin) starch, or modified soluble amylomaize-7 (70% amylose) starch for 4 wk. Body weights did not differ among the groups. Food consumption was significantly higher in the two modified starch-fed groups than in the two control groups. Commercial cornstarch, dextrose, modified potato starch and modified amylomaize-7 starch were 100 ± 0, 100 ± 0, 69.0 ± 1.0 and 91.5 ± 0.8% digestible, respectively (n = 9, mean ± SEM). The modified potato starch-fed group deposited the least fat, protein and energy. In both modified starch-fed groups, liver weights were significantly greater than in the two control groups. In food-deprived rats, serum free fatty acid concentrations in the modified potato starch-fed group were significantly higher than in the two control groups, and serum glucose concentrations were significantly higher in the two modified starch-fed groups than in the controls. The insulin to glucagon ratios were significantly lower in the modified potato starch-fed and amylomaize-7 starch-fed groups than in the dextrose-fed control group. Serum protein concentrations, measured after food deprivation, were significantly lower in the modified potato starch-fed group than in the other three groups. Gluconeogenesis from fermentation products might account for the high serum glucose concentrations in the two experimental groups. These data indicate that only the modified amylomaize-7 starch may be useful in the development of food products for liquid nutritional supplements because of the high digestibility and the low resultant insulin levels.
KEY WORDS: rats · amylopectin starch · amylose starch · digestibility · blood glucoseINTRODUCTION
Carbohydrates constitute 40-80% of total energy intake in human diets. In the 1985 American diet, about 47% of the total energy intake was from carbohydrates of which less than half was from complex carbohydrate sources (U.S. Department of Agriculture 1985).
Most starches can be classified as amylose or amylopectin. Amylose is an essentially linear molecule in which the D-glucose units are linked by 
-(1,4) glucosidic links. Amylopectin contains both -(1,4)-linked glucose and -(1,6) linkages, resulting in a branched structure. Both amylose and amylopectin are insoluble in cold water. In most starches, amylopectin comprises 70-80% of total starch, and amylose comprises the remaining 20-30% (Cummings and Englyst 1995
). However, among cereals such as barley, corn and rice genotypes, varying amylose to amylopectin ratios are available. The composition of starch can vary from virtually pure amylopectin, such as in waxy barleys and rice, to high levels (>70%) of amylose, such as in the wrinkled pea (Pisum sativum) and in amylomaize (Annison and Topping 1994). In general, amylose is thought to be less digestible than amylopectin.
Because of high digestibility and water solubility, carbohydrates with low molecular weights such as dextrose, fructose, sucrose, lactose and glucose polymers are widely used in liquid nutritional supplements for adults with compromised digestive function, infant formulas, liquid carbohydrate supplements for exercise, soft drinks and in solid food products. These all pose one major problem. The low molecular weight carbohydrates exert a degree of osmolarity that in many situations is undesirable because of the subsequent gastrointestinal effects such as diarrhea or distention, and they frequently have a high glycemic index. In addition, the osmotic pressure of these carbohydrates in the stomach and intestine can limit the availability of water for hydration purposes. Furthermore, the high insulin responses to dextrose and glucose polymers can lead to acutely elevated postprandial insulin secretion, lipogenesis, reactive hypoglycemia, and to elevated blood lipid levels and increased risk for obesity and cardiovascular disease.
As alternatives, two starches were modified to have high molecular weights and high water solubility to possibly address some of these concerns. High amylopectin-containing (70-75% amylopectin) modified potato starch and high amylose-containing (70% amylose) amylomaize-7 starch were derived from potato and amylomaize-7 starches, respectively, by hydrolysis with hydrochloric acid in various alcohols (Fox and Robyt 1992, Ma and Robyt 1987). These starches have molecular weights ranging from 4750 to 126,670 depending on the process used in the preparation. The two modified starches studied in this report have molecular weights of 40,500 (modified potato starch) and 39,200 (modified amylomaize-7 starch). All of the modified starches are readily soluble in hot water, yielding crystal-clear solutions up to a concentration of at least 200 g/L. These starches should produce little osmotic pressure in comparison with low molecular weight glucose polymers and simple sugars. In addition, variation of the amount of amylose in these starches may influence the rate of digestion and the metabolic responses. To our knowledge, this is the first report about digestibility, endocrine responses and metabolic responses in rats fed modified water-soluble starches. The findings may be useful in the development of liquid nutritional supplements.
MATERIALS AND METHODS
, Ma and Robyt 1987). The two modified soluble starches were supplied by Dr. John Robyt (Iowa State University, Ames, IA).
|
Table 1. Diet composition |
20°C until analysis for glucose, triglycerides, free fatty acids, protein, insulin and glucagon. Liver, heart and epididymal fat pad weights were recorded. The body weights after exsanguination were also recorded. The contents of the gastrointestinal tracts were removed as much as possible. The carcasses were weighed and stored at 20°C for future analysis. Each of the remaining four groups of rats were fed one of the four experimental diets for a 4-wk period with free access to the diets and water. Body weight and food consumption were recorded weekly. In the last week of feeding, the fecal marker, 2.5 g/kg Cr2O3 was mixed into each diet. Food consumption was recorded daily and feces were collected daily for 6 d. During this period, samples of diets that contained the fecal marker and all feces were weighed, lyophilized, ground and stored at 4°C before analysis. After 4 wk of feeding (9 wk of age), all the rats were weighed and killed between 0900 and 1200 h by gassing with CO2 . Blood was collected by cardiac puncture. Because all 36 rats could not be killed in 1 d, 16 rats (the first 4 rats in each group) were killed on 1 d; the remaining 20 rats were killed on the following day. Blood, tissues and carcasses were collected and treated as described above. The total food consumption over the 4-wk feeding period was calculated as the sum of the weekly food consumption. The total body weight gain was calculated as the difference between the final live body weight and the initial body weight.
Diet and fecal analyses.
Diet (250 mg) and 500 mg of lyophilized fecal samples were each ashed in a muffle furnace at 450°C overnight and the ash was weighed. Then 15 mL of 700 g/L nitric acid and 5 mL of 600 g/L perchloric acid were added to each ashed sample. All of the samples were digested at 204°C (Tecator Digester, Hoganas, Sweden) for ~2 h until the green color changed to orange. The samples were diluted with water to obtain a chromium concentration in the 1-5 mg/L range. The concentration of chromium was determined by atomic absorption spectrophotometer (Allied Analytical system, IL Video 12 aa/ae spectrophotometer, Waltham, MA) at 400 nm (Saha and Gilbreath 1991).
). The extracted starch samples were degraded by the amyloglycosidase procedure as follows: 177 kU/L amyloglycosidase, 0.05 mol/L sodium acetate, pH 4.5, in 0.2 mL plus 0.2 mL starch suspension were heated at 55-60°C for 30 min. The released glucose was determined enzymatically by the addition of 3.4 mL of indicator reagent that contained 1.07 mmol/L NAD+ , 1.03 mmol/L ATP, 4.73 kU/L hexokinase, and 0.18 kU/L bacterial glucose-6-phosphate dehydrogenase in 0.1 mol/L Tris, pH 7.5 at 25°C. The absorbance of NADH produced after 30 min was recorded at 340 nm (Beutler 1983). All of the enzymes, NAD+ and ATP were purchased from Sigma Chemical (St. Louis, MO). The fecal starch contents were calculated as glucose units. Fecal samples in which chromium excretion was at a constant level were used to calculate starch recovery. The digestibilities of experimental carbohydrates were calculated in the terms of glucose units as (intake feces)/intake. The results were calculated on a dry weight basis.
. In the presence of ATP and CoA, acyl-CoA synthetase (Sigma) catalyzed activation of free fatty acids to form acyl-CoA and AMP. The AMP formed was monitored by sequential reactions catalyzed by myokinase (Fisher Chemical, Springfield, NJ), pyruvate kinase, and L-lactate dehydrogenase (Sigma). The absorbance of NADH was followed at 340 nm. Serum total protein was assayed by using biuret reagent (White et al. 1976). Serum insulin and glucagon concentrations were assayed by using RIA procedures as described in the insulin and glucagon RIA kits (ICN Biomedicals, Costa Mesa, CA).
Liver composition.
Determination of the composition of liver was not originally planned, but was analyzed after differences in liver weight among groups were noticed. Each liver was cut into small pieces and placed into a homogenizing tube followed by addition of 15 mL water. The liver was homogenized with the homogenizing tube in an ice bath. The entire homogenate was transferred into a tared tube, frozen at 20°C and subsequently lyophilized. Lyophilized samples were weighed and stored in desiccators at 4°C prior to assay. The water content of liver was calculated as fresh liver weight minus lyophilized liver weight. The liver protein was assayed by using biuret reagent (White et al. 1976). The liver lipid was extracted from weighed lyophilized samples by chloroform/methanol (2:1, v/v) and determined gravimetrically (Griminger and Gamarsh 1972). Because differences in liver weights were unexpected, the liver samples were not quickly frozen. Therefore, the glycogen contents of the livers were not assayed. The calculations of total liver composition were based on fresh liver weight.
Body composition.
Each frozen carcass, minus liver, was chopped into small pieces and homogenized in a 3.78-L blender with addition of sufficient weighed amounts of water so that the blender could operate freely. An aliquot of homogenized carcass was transferred into a tared Erlenmeyer flask, frozen at 20°C and subsequently lyophilized. Lyophilized samples were weighed and stored in desiccators at 4°C for future analysis. Water content of each carcass was calculated as the fresh carcass weight minus the lyophilized carcass weight. Total body water was calculated as water content of carcass plus water content of liver. Carcass lipid was extracted from weighed lyophilized samples by chloroform/methanol (2:1 v/v), and determined gravimetrically (Griminger and Gamarsh 1972). Total body lipid was calculated as carcass lipid plus liver lipid. For the carcass protein determination, weighed lyophilized samples were digested in 18 mol/L H2SO4 , 3 g/L SeO2 at 420°C (Tecator Digester) for 1 h. Then the digested samples were diluted and the resultant nitrogen was determined as (NH4)2SO4 colorimetrically at 625 nm (Chaney and Marbach 1962). The protein factor of 6.25 was used in calculation for carcass protein content. The total body protein was calculated as carcass protein plus liver protein. The weight of the rat without its gastrointestinal contents was the weight used in the calculations of total body composition.
Energy determination.
Samples of the lyophilized diets and fecal samples were pressed into pellets that weighed between 1.0 and 1.3 g. The energy content of the diets and each fecal sample was determined by using the Parr 1242 adiabatic oxygen bomb calorimeter (Parr Instrument, Moline, IL). The digestible energy of each diet was calculated as (intake feces)/intake. Total energy intake for the 4-wk feeding period was calculated as (energy content/g diet) × (total food intake for 4-wk feeding period). For body energy content determination, the same procedure was employed. Body energy gain was calculated as the difference between the final body energy content of each experimental rat and the average carcass energy content of the group killed at the start of the feeding period. Energy efficiency was calculated as the percentage of body energy gained divided by energy consumption during the feeding period.
Statistics.
Data are presented as means. All of the results were statistically evaluated by ANOVA (SAS, Version 6.06.01, SAS Institute, Cary, NC). Individual comparisons were made by the least significant difference (LSD), which uses the mean square of the error term from the ANOVA. The pooled SEM, derived from the mean square error term of the ANOVA, was used to describe variability of the each response criterion. Differences were considered significant at P < 0.05.
RESULTS
|
Table 2. Body weight, body weight gain, food consumption and dietary energy consumption in rats fed control or modified carbohydrates1 |
|
Table 3. Digestibility of experimental carbohydrates and dietary energy in rats fed control or modified carbohydrates1 |
|
Table 4. Body composition and energy in rats fed control or modified carbohydrates1 |
|
Table 5. Absolute and relative organ weights in rats fed control or modified carbohydrates1 |
|
Table 6. Liver composition in rats fed control or modified carbohydrates1 |
|
Table 7. Serum metabolites, insulin, and glucagon in food-deprived rats that had been fed control or modified carbohydrates for 4 wk1 |
DISCUSSION
). Different types of starches elicit significantly different postprandial blood glucose and insulin responses. The observed metabolic responses to various starchy foods depend on several factors, including the physical form of the starch (O'Dea et al. 1980), the rate of meal ingestion (Jenkins et al. 1982), the cooking method (Collings et al. 1981), the presence of fat (Holm et al. 1983) or protein (Tovar et al. 1990), and the ratio of amylose to amylopectin (Behall and Howe 1995, Behall et al. 1988 and 1989, Granfeldt et al. 1995, Juliano and Goddard 1986).
) and corn (Behall et al. 1988 and 1989, Granfeldt et al. 1995) were reported to have significantly lower postprandial serum glucose and insulin responses. Furthermore, long-term ingestion of a high amylose corn-based diet (70% amylose) improved fasting triglyceride and cholesterol concentrations in both healthy and hyperinsulinemic individuals, compared with a high amylopectin corn-based diet (Behall and Howe 1995, Behall et al. 1989). In contrast, others have reported different results. Rice varieties high in amylose showed higher (Rao 1971) or similar (Jiraratsatit et al. 1987) starch digestion rates and glycemic responses compared with rice varieties low in amylose. In another study, Panlasigui et al. (1991) reported that the digestibilities and glycemic responses were significantly different among rice varieties with similar amylose concentrations. These authors concluded that amylose content alone is not a good predictor of starch digestion rate. Rice varieties with similar high amylose contents can differ in physicochemical properties and this, in turn, may influence starch digestibility. Therefore, the amylose and amylopectin ratio is not the sole determining factor in rate of starch digestion and postprandial glycemic responses.
, Ma and Robyt 1987). Therefore, some of the physicochemical properties of the original starches were changed. These may account for the low digestibility of the modified potato starch. Furthermore, the differences in digestibility among the carbohydrates were reflected in the differences in digestible energy among the four diets (Table 3). The rats compensated for the lower digestibilities of the two experimental starches by increasing food and dietary energy consumption (Table 2).
).
, an oral propionate supplement had no detectable effect on carbohydrate or lipid metabolism in either nondiabetic or streptozocin-induced diabetic rats. However, Demigné et al. (1986) reported that short-chain fatty acids, particularly acetate and propionate, were absorbed in very large amounts in rats fed a high fiber diet. Gluconeogenesis was active in rats fed the high fiber diet. Propionate rectally infused into humans has been reported to raise the serum glucose and glucagon concentrations, consistent with the suggestion that colonic fermentation in nonruminants can generate propionate for gluconeogenesis (Wolever et al. 1991). In our experiment, the observation of colonic distention with gas suggests that colonic fermentation occurred in rats fed the two modified starches. Therefore, gluconeogenesis from products of fermentation may be responsible for the higher serum glucose concentrations in the two modified starch-fed groups after food deprivation.
ACKNOWLEDGMENTS
FOOTNOTES
Manuscript received 19 June 1996. Initial reviews completed 29 July 1996. Revision accepted 4 March 1997.
LITERATURE CITED
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
