| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
ORIGINAL RESEARCH COMMUNICATION |
1 From the Department of Nutritional Sciences, Faculty of Medicine, University of Toronto, Toronto, ON, Canada
2 Supported by an unrestricted grant from the International Life Sciences Institute North America, the Bristol Myers Squibb/ Mead Johnson Freedom to Discover Award (to GHA), and a postgraduate scholarship from the Natural Sciences and Engineering Research Council of Canada (to TA). McCain Foods Ltd donated the pizzas for the experimental meals. 3 Reprints not available. Address correspondence to GH Anderson, Department of Nutritional Sciences, Faculty of Medicine, University of Toronto, 150 College Street, M5S 3E2, Toronto, ON, Canada. E-mail: harvey.anderson{at}utoronto.ca.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Objective: Two experiments were conducted to determine the effect of solutions containing sucrose, HFCS, or various ratios of glucose to fructose (G:F) on food intake (FI), average appetite (AA), blood glucose (BG), plasma insulin, ghrelin, and uric acid (UA) in men.
Design: Sugar solutions (300 kcal/300 mL) were (in %) G20:F80, HFCS 55 (G45:F55), sucrose, and G80:F20 (experiment 1, n = 12) and G20:F80, G35:F65, G50:F50, sucrose, and G80:F20 (experiment 2, n = 19). The controls were a sweet energy-free control (experiment 1) and water (both experiments). Solutions were provided in a repeated-measures design. AA, BG, and FI were measured in all subjects. Hormonal responses and UA were measured in 7 subjects in experiment 2. Measurements were taken from baseline to 75 min. FI was measured at 80 min.
Results: Sucrose and HFCS (experiment 1) and sucrose and G50:F50 (experiment 2) had similar effects on all dependent measures. All sugar solutions similarly reduced the AA area under the curve (AUC). FI and plasma UA concentrations were significantly (P < 0.05) lower after high-glucose solutions than after low-glucose solutions. The lower FI was associated with a greater BG AUC (P < 0.05) and smaller AA and ghrelin AUCs (P < 0.01). Insulin and BG AUCs were positively associated (P < 0.001).
Conclusion: Sucrose, HFCS, and G50:F50 solutions do not differ significantly in their short-term effects on subjective and physiologic measures of satiety, UA, and FI at a subsequent meal.
Key Words: Fructose • glucose • sucrose • high-fructose corn syrup • blood glucose • insulin • ghrelin • uric acid • appetite • food intake
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
However, the role of increased availability of sugars—and specifically of HFCS—in the national food supply, as a significant independent contributor to the current epidemic of obesity, is uncertain for several reasons. First, the availability of sugars has not increased disproportionately to the increased availability of total fat, protein, and energy per capita (3). Second, sugars suppress short-term food intake (FI) in children (4, 5) and adults (6-9), and the magnitude of this effect is inversely related to the glycemic response that those sugars elicit (10, 11). Third, HFCS is a nutritive sweetener containing an unbound form of the same monosaccharides as sucrose (sugar). Sucrose is composed of 50% fructose and 50% glucose linked together by 
-1–4 glycosidic bonds. The most common forms of HFCS are HFCS 55% and 42%. HFCS 55%, used primarily in beverages, is composed of 45% glucose and 55% fructose, and HFCS 42%, used primarily in foods, is composed of 58% glucose and 42% fructose (2).
Nevertheless, it is biologically plausible that the ratio of glucose to fructose (G:F) in solutions is a determinant of FI. Fructose does not increase the satiety signals of blood glucose (BG) and insulin to the same extent as does sucrose or glucose do (12-14). Short-term FI is inversely related to the glycemic (10, 11) and insulin (15) responses to sugars, and it has been proposed that fructose does not suppress ghrelin, a gastric appetite hormone (2). Therefore, we hypothesized that high G:F consumed in solutions would lead to a greater response in satiety hormones, subjective satiety, and FI at a later meal than would low G:F, but that there would be no differences among equicaloric solutions containing sucrose, its monosaccaride components, and HFCS.
|
SUBJECTS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
7.0 mmol/L) or liver or kidney disease, who had undergone a major medical or surgical event within the past 6 mo, or who were breakfast skippers, dieters, or under medication were excluded from all sessions. Restrained eaters were also excluded on their identification by a score of 11 on an eating habits questionnaire (16). Written informed consent was obtained from all subjects. The study protocol was approved by the Human Subjects Review Committee, Ethics Review Office, University of Toronto.
Study design and treatments
Two experiments with a randomized repeated-measures design were conducted. Health young men were randomly assigned to receive 1 of 6 sugar solutions at weekly intervals. In experiment 1, the 4 sugar solutions were HFCS (G45:F55, 75% concentrate; Cargill Sweeteners Company, Wayzata, MN), sucrose (Redpath Sugar; Tate and Lyle North American Sugars, Toronto, Canada), G20:F80, and G80:F20. Two control solutions were used. One was a sweet energy-free control that contained water sweetened with sucralose (McNeil Specialty Products Company, New Brunswick, NJ), and one was water alone (Crystal Springs, Quebec City, Canada). Sweetness was equalized for all treatments except the water control by the addition of 54, 52, 27, and 480 g sucralose to the F20:G80, sucrose, HFCS, and sweetened control solutions, respectively. Sucralose was chosen as a noncaloric sweetener because it has no interaction with carbohydrate metabolism, BG, blood fructose, or insulin secretion and has no effect on the central nervous system (17, 18). To reduce sweetness and improve palatability, lemon juice (Equality; The Great Atlantic and Pacific Company of Canada Ltd, Toronto, Canada) was added. These test solutions were rated equally sweet and equally palatable by a test panel of 8 subjects. In experiment 2, subjects received solutions of G20:F80, G35:F65, G50:F50, sucrose, and G80:F20 as test treatments and water as a control. Treatments were not equalized for sweetness and palatability because no associations and interactions were found in experiment 1 between the sweetness or palatability of treatments and subjective appetite or FI.
Twelve subjects were used in experiment 1, which is consistent with the sample size used in many previous studies in which greater suppression of FI was found after a 75-g preload of sucrose or glucose (6, 10) or a preload of 50 g protein (19) than with the energy-free control. Because the sample size may have been too small to show differences among solutions in experiment 1, we calculated that a sample size of 18 ( error = 0.05; ß error = 0.20) was needed for identification of a 120-kcal difference in response between the sugar (G20:F80 and G80:F20) solutions. Therefore, in experiment 2, 19 subjects completed the study sessions.
Sugar solutions contained 300 kcal in 300 mL. In both experiments, treatment formulations used D-glucose monohydrate (Grain Process Enterprises Ltd, Scarborough, Canada) and pure fructose (Now Natural Foods, Bloomingdale, IL). Fructose was not added in proportions of >80% of the sugars in solutions because, when fructose is used alone, even small amounts (eg, 26 g) have resulted in symptoms of nausea and diarrhea in 50% of the population (20, 21). In a pilot study in our laboratory, male subjects reported gastrointestinal discomfort after they consumed 75 g (300 mL) of a G10:F90 solution.
The prepared solutions were stored in the refrigerator overnight and served chilled. An additional 100 mL of water was served after the solutions were consumed to reduce aftertaste.
Protocol
The protocol and procedures are similar to those reported in previous studies (6, 10, 19). Subjects chose a time between 1100 and 1400 at which to participate in the sessions, and they were asked to arrive at the same time and on the same day of the week for all sessions. They were required to fast for 10–12 h and then to consume a standard breakfast 4 h before arrival at the testing facilities of the Department of Nutritional Sciences, University of Toronto. The standard breakfast consisted of a single serving of a ready-to-eat cereal (Honey Nut Cheerios; General Mills, Mississauga, Canada), a 250-mL box of 2%-fat milk (Sealtest Skim Milk; Sealtest, Markham, Canada), a 250-mL box of orange juice (Tropicana Products Inc, Bradenton, FL), and tea or coffee without sugar or sweetener. The subjects were asked not to eat or drink anything between their breakfast and the study session except water, which was allowed up to 1 h before the session. They were also instructed to refrain from alcohol consumption and any unusual exercise and activity the night before a study session.
On arrival, subjects completed questionnaires assessing their sleep habits and stress factors and their compliance with fasting and their pattern of activity on the preceding day. If they reported significant deviations from their usual patterns, they were asked to reschedule. Before subjects consumed the test solutions, they completed visual analogue scale (VAS) questionnaires measuring motivation to eat (7, 22) and physical comfort (7), and a blood sample was obtained. If BG was >6 mmol/L, which suggested that the subject had eaten recently or may be insulin resistant, he or she was rescheduled. Participants were moved to another room (feeding room) where they received one of the test solutions. They were instructed to consume the treatment within 3 min and to return to the experiment room to complete questionnaires assessing the sweetness and palatability of the treatments (6). At 15, 30, 45, 60, and 75 min after consumption of the drinks, the VAS scales were completed and blood samples collected. Subjects remained seated throughout the experimental session and were allowed to read or listen to music.
Pizza (McCain Foods Ltd, Florenceville, NB) and water (Crystal Springs) intakes were measured at an ad libitum lunch 80 min after the subjects consumed the preload solutions. In addition, in experiment 2, to measure thirst before eating, subjects received a bottle of water (500 mL) at 75 min. The bottle was removed and replaced by another before the subjects received the pizza tray.
Three varieties of pizza (Deluxe, Pepperoni, and Three Cheese Deep 'N Delicious pizza; McCain Foods Ltd) were purchased from local retailers. All 3 varieties were similar in contents—averaging 10.0 g protein, 7.6 g fat, 26.6 g carbohydrate, and 226 kcal/100 g energy—and in size (5-in diameter). Because of the lack of a thick outer crust, the pizzas have a uniform energy content that eliminates the possibility that participants would eat the energy-dense filling and leave the outside crust of the pizza. Participants ranked the pizzas according to their preference at screening, and their same choices were provided at each of the 6 sessions. Each pizza (cooked for 8 min at 430 °F and cut in 4 slices) was weighed before serving. Subjects were provided 3 varieties of pizzas to reduce the effect of sensory-specific satiety on test meal intake. The pizzas were served to the subjects on trays at 10-min intervals; in each case, the previous tray was removed and the remaining pizza weighed, until the subjects declined further trays. Each tray contained 2 pizzas of their first choice and 1 pizza each of their second and third choices. Subjects were instructed to eat until they were comfortably full.
The energy intake from the pizza was calculated from the weight consumed and the compositional information provided by the manufacturer. Water intake was measured by weight. Cumulative energy intakes were calculated by adding the energy consumed from the sugar solution to the energy consumed at the test meal. Caloric compensation at the test meal for that consumed in the preloads was calculated by the formula [(kcal consumed at the test meal after the water control –kcal consumed at the test meal after the sugar solution)/300 kcal (in sugar solution)] x 100. A 100% caloric compensation indicates that, at sessions when he was given the 300-kcal treatment, a subject had a lunch intake 300 kcal lower than that at sessions when he was given the control preload. Caloric compensation of <100% indicated that the subject had low compensation for the preload energy at the test meal, whereas scores > 100% indicated overcompensation for preload energy at the test meal.
In each experiment, BG was measured in 12 subjects by finger-prick with the use of a glucometer (Accu-Chek Compact; Roche Diagnostics Canada, Laval, Canada). The same glucometer was used for the same subject for all 6 sessions. Subjects cleaned their fingers before and after each finger-prick with an alcohol swab (Ingram and Bell Medical, Don Mills, Canada). The first drop of blood was wiped off, and the next drop was placed on glucometer strip for measurement of the BG.
To obtain sufficient blood for the measurement of insulin, ghrelin, and uric acid (UA) concentrations in experiment 2, an indwelling intravenous catheter was inserted by a registered nurse into the antecubital fossa vein in the arm of 7 different subjects upon their arrival. The blood samples were drawn into chilled heparinized tubes (Vacutainer; Becton Dickinson, Rutherford, NJ). Two lavender-capped Vacutainer tubes, coated with potassium oxalate–sodium fluoride anticoagulant (EDTA; at 1 mg/mL blood), were used at baseline and 15, 30, 45, 60, and 75 min. For analysis of BG, insulin, and UA, a blood sample was collected in 1 tube (5 mL). For analysis of ghrelin, a separate sample was obtained in a 4-mL Vacutainer tube, and 400 µL aprotinin was added to the tube immediately (<30 s) after blood collection. The tubes were centrifuged at 4 °C for 10–15 min at 2000 x g, and the plasma was stored at –80 °C for analysis. Plasma glucose was measured by using a glucose analyzer (Cobas Integra 800; Roche Diagnostics GmbH, Mannheim, Germany), insulin was measured by using an electrochemiluminescence immunoassay (Roche Diagnostics GmbH), total ghrelin was measured by using a radioimmunoassay (GHRT-89HK; Linco Research Inc, St Charles, MO), and UA was measured by using an enzymatic colorimetric test (11875426 216; Roche Automated Modular Systems, Basel, Switzerland).
Statistical analysis
We used SAS software (version 8.2; SAS Institute Inc, Cary, NC) to conduct the statistical analyses. To test for the effect of the treatments on the outcome variables, one-factor repeated-measures analysis of variance (ANOVA) using the general linear model (PROC GLM procedure) was performed on data for food and water intakes at 80 min, perceived sweetness, palatability of treatments and pizza, physical comfort, and net areas under the curve (AUCs) for average appetite (AA), BG, UA, insulin, and ghrelin. The net AUCs were calculated by applying the trapezoid rule (23), and they included areas over and under the baseline values.
In both experiments, 2-factor repeated-measures ANOVA (GLM) was applied to test for the effects of treatment and time and for treatment x time interaction for AA scores, for the individual VAS used in calculating AA scores, and for BG, insulin, ghrelin, and UA concentrations over 75 min. When an interaction was statistically significant, a one-factor ANOVA using a GLM procedure was followed by Tukey's post hoc test to identify mean differences among treatments at each time of measurement.
In experiment 2, a 2-factor ANOVA (PROC GLM) was performed to determine the effect of the route of blood sampling (finger-prick, n = 12; intravenous catheter, n = 7) and of treatment on BG, AA, and FI. When venous blood was collected through the indwelling catheter, subjects had significantly lower BG AUC, AA AUC, and FI (P < 0.01 for all) than did subjects from whom capillary blood was collected by finger-prick. However, because no significant interaction (P > 0.05) was found between treatment and the method of blood sampling, the effect of treatment was consistent across both sampling methods. Therefore, the data for the dependent measures are reported for the pooled sample.
A composite score of the 4 motivation-to-eat VASs was calculated, as described previously by us (6, 10) and others (22), to obtain the AA score. The AA score was reflective of the individual scores on the motivation-to-eat questions and was used here as a summary measure of subjective appetite for analyses.
Correlation analyses of dependent measures were made by using Pearson's correlation coefficients. Significance was set at P < 0.05. Data are presented as means ± SEMs.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Food intake
In both experiments, treatments affected FI. In experiment 1, all sugar solutions except G20:F80 suppressed FI at the test meal significantly (P = 0.0001) more than did the water control, but only the G80:F20 and sucrose solutions suppressed FI significantly (P = 0.0001) more than did the sweet control (Table 1
). There were no significant differences in FI among HFCS and the other sugar solutions.
|
). Subjects had significantly (P = 0.0001) lower FIs after consuming the G80:F20 and sucrose solutions than after consuming the G20:F80 and G35:F65 solutions. However, there were no significant differences in FI between the G50:F50 solution and sucrose or between the G20:F80 and G35:F65 solutions.
|
). The HFCS and sucrose solutions did not differ significantly from each other or the water control, but they resulted in significantly (P = 0.0001) higher cumulative energy intakes than did the noncaloric sweet control (sucralose). Cumulative intakes did not differ significantly after water or the G80:F20, HFCS, and sucrose solutions.
In experiment 2, the G20:F80 and G35:F65 solutions resulted in significantly (P = 0.0001) higher cumulative energy intakes than did the sucrose and G80:F20 solutions, but the energy intakes did not differ significantly from those with the water control or the G50:F50 solution (Table 2). No statistically significant differences in cumulative energy intakes were observed among sucrose, G50:F50, and the water control.
Caloric compensation
In experiment 1, caloric compensation of 92% and 89% for the G80:F20 and sucrose solutions, respectively, was significantly (P < 0.05) greater than that of 45% for G20:F80. At 63%, HFCS did not differ significantly from the other sugar solutions (Table 1). In experiment 2, G80:F20 resulted in a caloric compensation (155%) that was significantly (P = 0.0001) greater than that of 46%, 41%, and 56% seen for G20:F80, G35:F65 and, G50:F50, respectively (Table 2). Sucrose at 118% did not differ significantly from the other sugar solutions.
Water intake
In experiment 1, subjects had the highest and lowest water intakes at the test meal after HFCS and the sweet control (sucralose), respectively (Table 1). Water intakes after all other solutions were intermediate and not significantly different from either the HFCS or sucralose solution. In experiment 2, neither cumulative water intakes nor water consumed before or within the test meal differed significantly among the treatments (Table 2).
Average appetite score
All sugar solutions except G20:F80 and HFCS (experiment 1) and G50:F50 (experiment 2) lowered subjective AA AUCs significantly (P < 0.01) more than did the water control. However, AA AUCs did not differ significantly among the sugar solutions in either experiment (Table 3).
|
). The interaction is explained by the significantly (P < 0.05) greater and earlier increase in AA score with time after the controls than after the sugar solutions. The decrease from baseline in AA score after the sugar solutions was the greatest at 15 and 30 min; the AA score then returned to baseline or rose above it at 75 min. In both experiments, AA scores at each time did not differ significantly among the sugar solutions over the 75-min span.
|
Blood glucose
In both experiments, all sugar solutions resulted in significantly (P < 0.0001) higher BG AUC than did the controls (Table 3). In experiment 1, of the sugar solutions, G80:F20 and G20:F80 resulted in the highest and lowest BG AUCs, respectively; sucrose and HFCS were intermediate and did not differ significantly from each other (Table 3). In experiment 2, the highest and lowest BG AUCs were seen after the G80:F20 and G20:F80 solutions, respectively. The BG AUCs for sucrose, G50:F50, and G35:F65 solutions did not differ significantly (Table 3).
In both experiments, BG was significantly (P < 0.0001) affected by treatment, time, and treatment x time interaction. In experiment 1, the increase in BG was significantly (P < 0.0001) higher after all sugar solutions than after the water and sucralose controls at 15, 30, and 45 min; G80:F20 and G20:F80 resulted in the highest and lowest BG responses, respectively (Table 5). At 60 and 75 min, BG remained significantly (P < 0.0001) higher than the water control only after the G80:F20 solution. Sucrose and HFCS resulted in an intermediate and identical significant (P < 0.0001) increase in BG. In experiment 2, BG was significantly (P < 0.0001) higher after all of the sugar solutions—except the G20:F80 (at 45 and 60 min) and G35:F65 (60 min) solutions—than after the water control at 60 min (Table 5). However, at 75 min, BG was significantly (P < 0.0001) higher than the water control only after the G80:F20 solution. There was no significant difference between G50:F50 or sucrose at any time.
|
). Of the sugar solutions, G80:F20 and G20:F80 had the lowest and highest UA AUCs, respectively. UA AUCs did not differ significantly after the G35:F65, G50:G50, and sucrose solutions.
|
). At 75 min, UA concentrations were highest after G20:F80. The sucrose and F50:G50 solutions each resulted in significantly (P < 0.05) lower UA concentrations than did the G20:F80 solution, but they did not differ significantly from any other solutions.
|
). The G50:F50 and sucrose solutions did not differ significantly from each other or from any other sugar solution.
Insulin concentrations were significantly (P < 0.05) affected by treatment, time, and treatment x time interaction. All sugar solutions increased insulin concentrations at 15, 30, and 45 min significantly (P < 0.05) more than did the water control. The G20:F80 solution resulted in significantly (P < 0.05) lower insulin concentrations than did the sucrose and G80:F20 solutions at 45 min (Figure 1). At 60 min, insulin concentrations after the G20:F80 solution did not different from those after the control. At 75 min, insulin concentrations remained significantly (P < 0.05) higher than those after the control only after the G80:F20 and sucrose solutions, but there were no significant differences among G20:F80, G35:F65, sucrose, and G50:F50.
Ghrelin
All sugar solutions resulted in significantly (P < 0.05) lower ghrelin concentrations than did the water control. Thus, the ghrelin AUC was negative and significantly (P < 0.05) larger after all sugar solutions than after the water control (Table 6).
Ghrelin concentrations were significantly (P < 0.05) affected by treatment, time, and treatment x time interaction (Figure 1). The interaction appears to be explained by a significantly (P < 0.05) greater and earlier suppression in ghrelin by sucrose and G50:F50 than by G80:F20 and G20:F80, but no sugar solutions resulted in a significant difference in ghrelin concentrations at 30 min. All sugar solutions except G80:F20 (at 15 and 45 min) and G20:F80 (at 15 min) resulted in significantly (P < 0.05) lower ghrelin concentrations than did the water control.
Relations among dependent measures
FIs were positively correlated with AA score at 75 min (experiment 1: r = 0.42, P < 0.001; experiment 2: r = 0.27, P < 0.005), and inversely correlated with BG (experiment 2: r = –0.41, P < 0.01) and insulin (experiment 2: r = –0.39, P < 0.05) concentrations at 45 min. FIs were inversely correlated with BG AUC (experiment 1: r = –0.30, P < 0.01; experiment 2: r = –0.38, P < 0.0001) (Figure 2). The lower AA as reported by AUC was correlated with lower FIs (experiment 1: r = 0.41, P < 0.001; experiment 2: r = 0.18, P = 0.06) and higher BG AUCs (experiment 1: r = –0.29, P < 0.05; experiment 2: r = –0.31, P < 0.001) (Figure 2).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Sucrose, HFCS, and G50:F50 solutions induced similar BG and hormonal responses and decreases in subjective satiety and FIs at the test meal. Sucrose and HFCS were similar to each other in their effects on BG in experiment 1, and it would be expected that the insulin responses were similar because it was previously shown that sucrose and HFCS have similar effects on postprandial BG and insulin concentrations (24). Furthermore, when sucrose or HFCS beverages were served at meals to make up 
17% of the energy, no differences were found in BG, insulin, leptin, and ghrelin concentrations measured at frequent intervals over 24 h (25). In experiment 2, a solution of G50:F50 was compared with sucrose because, when sucrose is added to acidic solutions, as in soft drinks or fruit-flavored drinks, that sucrose is primarily hydrolyzed before consumption (26). Again, the disaccharide and the monosaccharide mixtures led to the same results in all dependent measures.
Both experiments showed greater suppression of FI and increased caloric compensation associated with higher G:F ratios in solutions. All sugar solutions except the low-glucose-containing sugar solutions (G20:F80 and G35:F65) led to lower FIs at 80 min than did the water control. These results are consistent with previous reports that 75-g sucrose and glucose solutions, but not a G20:F80 solution, suppressed FIs 60 min later (6, 10) and that FI shortly after the consumption of sugar solutions is inversely related to the effect of sugar solutions on BG (11). The results contrast with reports that solutions containing fructose alone suppress FI at a later meal more than do glucose solutions (12). However, in these comparisons, this effect of fructose was due to its slow absorption when consumed in the absence of glucose, which resulted in gastrointestinal distress (19). For this reason, a solution of fructose alone was not included in the present studies
All sugar solutions increased BG and insulin and reduced ghrelin more than did the control. Although BG may contribute directly to satiety (7, 8), it is clearly not the only reason for the more favorable effects of the high-glucose solutions than of the high-fructose solutions on FI. High-glucose treatments and the higher BG responses derived from them also were associated with greater responses in the satiety hormone insulin and a decrease in the orexigenic hormone ghrelin. However, the role of BG and insulin in the suppression of ghrelin remains uncertain for several reasons. Although a previous study reports that glucose at 15% of the energy in meals (27) leads to lower postprandial ghrelin than does fructose at 15% of the energy in meals, this association does not define the mechanism controlling the suppression of ghrelin secretion. In the present study, insulin and ghrelin responses were not found to be inversely related, and this observation is consistent with more recent data showing that, whereas the presence of insulin is required, a postprandial insulin response is not required, and nutrient sensing by ghrelin-producing cells is an important regulator of ghrelin secretion (28). For example, fat ingestion leads to ghrelin suppression but does not increase postprandial insulin (29). Our data support the proposed role of ghrelin-producing cells in nutrient sensing, because all sugar solutions suppressed ghrelin similarly, even though BG and insulin were higher in the high-glucose than in the low-glucose solutions.
Plasma UA was measured because replacement of sucrose with HFCS has been suggested to play a causal role in the metabolic syndrome (30). Fructose consumption increases plasma UA by increasing purine biosynthesis and decreases renal clearance of UA by increasing plasma lactate concentrations (31). In the present study, all sugar solutions except the G80:F20 solution significantly increased UA above control values and the fructose content of the solutions associated positively with UA AUC (r = 0.69, P < 0.0001) and inversely with insulin AUC (r = –0.54, P < 0.001). However, the G50:F50 and sucrose solutions increased UA AUC equally, which suggests that a significant difference between HFCS and sucrose is unlikely. Therefore, whereas greater intakes of fructose may elevate plasma UA, as reported previously (31), the substitution of HFCS for sucrose is an unlikely contributor to the metabolic syndrome by this mechanism.
The differing G:F in solutions did not affect water consumption. The observation that subjects in experiment 1 drank more water at their meal after the HFCS than after the sweet control was pursued in experiment 2 because it was possible that fructose, because of its slower absorption compared with that of glucose (32), created a hyperosmolar environment in the small intestine and thus caused the retention of fluid, feelings of thirst, and gastrointestinal discomfort (19). Because all sugar solutions resulted in similar water intake before and during the test meal in experiment 2, the differences found between HFCS and the sucralose solution in experiment 1 were attributed to reduced thirst after the sucralose solution and not to increased thirst due to fructose in the sugar solutions. Although there is no mechanism to account for this effect of sucralose, it does not appear to be unique because, when consumed during a meal, aspartame-sweetened drinks reduced thirst more than did sucrose-sweetened drinks (33).
Sweetness was not equalized among solutions in experiment 2 because it was proposed that the greater sweetness of fructose would also be an independent factor affecting FI (2). However, no significant relation was found between the sweetness (r = –0.06) and palatability (r = 0.13) of treatments and FI at the test meal, as expected if the time interval between treatment and FI is >1 h (34). In experiment 1, whereas the sweetness of all solutions except the water control was equalized with the sweetest solution (G20:F80), sweetness was reduced by adding lemon juice. The subjects judged the palatability of the test solutions to be equal to the water control.
High FIs by young men given test meals after an overnight fast and a light breakfast have been reported in many studies, but this has not been found to compromise treatment effects (6-8, 10-11). The higher FIs in experiment 1 than in experiment 2 may be due to several reasons. First, experiment 1 was conducted in the summer, and experiment 2 was conducted in the winter. Higher ambient temperatures are associated with lower FIs than are low ambient temperatures (35). Second, on average, subjects came in 2 h later in experiment 2. Therefore, the test meal later in the day would also contribute to greater hunger and FI. Third, subjects in experiment 2 were on average 5.4 y younger (P < 0.004) and 5.7 kg heavier (P < 0.06) and had a BMI 1.2 greater (P < 0.06) than did subjects in experiment 1. Appetite and FI are reduced by age (36) and are increased in association with body weight and the duration of fasting (37). The overall treatment effects in the 2 experiments were similar, however, which suggests that these differences in characteristics and FIs between the 2 samples did not affect the outcomes.
In conclusion, solutions of HFCS, F50:G50, and sucrose were similar in their effects on subjective measures and physiologic signals of satiety, plasma UA concentrations, and FIs in young men. However high G:F in isocaloric sugar solutions result in higher BG and insulin concentrations and lower UA concentrations and FIs than did low G:F.
|
ACKNOWLEDGMENTS |
|---|
The authors' responsibilities were as follows—GHA: conceiving the hypothesis, designing the experiment, and writing the manuscript; and TA: conducting the experiments, collecting and analyzing the data, and writing the manuscript. GHA serves as the chair of the board of the International Life Sciences Institute and as a science advisor to the Canadian Sugar Institute and to Archer Daniels Midland (a producer of HFCS); he has no equity or other financial interests in either industry. GHA also has received unrestricted research grant funding from the US Sugar Association and the Canadian Sugar Institute. TA had no personal or financial conflict of interest.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. S. White Misconceptions about High-Fructose Corn Syrup: Is It Uniquely Responsible for Obesity, Reactive Dicarbonyl Compounds, and Advanced Glycation Endproducts? J. Nutr., June 1, 2009; 139(6): 1219S - 1227S. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. H. Moran Fructose and Satiety J. Nutr., June 1, 2009; 139(6): 1253S - 1256S. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. J. Angelopoulos, J. Lowndes, L. Zukley, K. J. Melanson, V. Nguyen, A. Huffman, and J. M. Rippe The Effect of High-Fructose Corn Syrup Consumption on Triglycerides and Uric Acid J. Nutr., June 1, 2009; 139(6): 1242S - 1245S. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Hamedani, T. Akhavan, R. A. Samra, and G H. Anderson Reduced energy intake at breakfast is not compensated for at lunch if a high-insoluble-fiber cereal replaces a low-fiber cereal Am. J. Clinical Nutrition, May 1, 2009; 89(5): 1343 - 1349. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. J Melanson, T. J Angelopoulos, V. Nguyen, L. Zukley, J. Lowndes, and J. M Rippe High-fructose corn syrup, energy intake, and appetite regulation Am. J. Clinical Nutrition, December 1, 2008; 88(6): 1738S - 1744S. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. R. Vasselli Fructose-induced leptin resistance: discovery of an unsuspected form of the phenomenon and its significance. Focus on "Fructose-induced leptin resistance exacerbates weight gain in response to subsequent high-fat feeding," by Shapiro et al. Am J Physiol Regulatory Integrative Comp Physiol, November 1, 2008; 295(5): R1365 - R1369. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
), G35:F65 (
), G50:F50 (
), sucrose (*), G80:F20 (
), and water (
). 
, change. Mean plasma uric acid, insulin, and total ghrelin at baseline were 295.9 ± 8.5 µmol/L, 45.3 ± 3.8 pmol/L, and 733.4 ± 40.7 pg/mL, respectively. Significant treatment, time, and treatment x time interaction effects (P < 0.05 for all) were found by 2-factor ANOVA (general linear model, GLM). One-factor ANOVA (GLM) followed by Tukey's post hoc was used to differentiate the effect of treatment at each measured time. Means at the same time with different superscript letters were significantly different, P < 0.05 (n = 7).