点击显示 收起
1 From the US Department of Agriculture/Agricultural Research Service Childrens Nutrition Research Center, Department of Pediatrics, Baylor College of Medicine and Texas Childrens Hospital, Houston.
2 This work is a publication of the US Department of Agriculture/Agricultural Research Service Childrens Nutrition Research Center, Department of Pediatrics, Baylor College of Medicine and Texas Childrens Hospital, Houston. The contents of this publication do not necessarily reflect the views or policies of the US Department of Agriculture, and mention of trade names, commercial products, or organizations does not imply endorsement by the US government. 3 Supported in part by the Mars Corporation and by the USDA/ARS under Cooperative Agreement no. 58-6250-6-001. 4 Reprints not available. Address correspondence to NF Butte, Childrens Nutrition Research Center, Baylor College of Medicine, 1100 Bates Street, Houston, TX 77030. E-mail: nbutte{at}bcm.tmc.edu.
ABSTRACT
Background: Difficulty adapting to high-fat (HF) and highcarbohydrate (HC) diets may predispose children to obesity and diabetes.
Objective: We tested the hypothesis that children have metabolic flexibility to adapt to HF and HC diets.
Design: In protocol 1, 12 children aged 69 y and 12 adolescents aged 1316 y were randomly assigned in a crossover design to consume low-fat (LF), HC (25% and 60% of energy, respectively) or HF, low-carbohydrate (LC) (55% and 30% of energy, respectively) diets. In protocol 2, 12 adolescents aged 1316 y were randomly assigned in a crossover design to consume an LF-HC diet with 11% or 40% of carbohydrate as fructose. Total energy expenditure, nonprotein respiratory quotients (NPRQs), and substrate utilization were measured by using 24-h calorimetry. Effects of sex, puberty, body fat (dual-energy X-ray absorptiometry), intraabdominal fat (magnetic resonance imaging), and fitness on substrate utilization were tested.
Results: Substrate utilization was not affected by puberty, body fat, intraabdominal fat, or fitness. Total energy expenditure was not affected by diet. In protocol 1, NPRQs and carbohydrate and fat utilization were significantly affected by diet (P = 0.001) and sex (P = 0.005). NPRQs and carbohydrate utilization increased with the LF-HC diet. NPRQs decreased and fat utilization increased with the HF-LC diet; changes in substrate utilization were less pronounced in females than in males. In protocol 2, 24-h NPRQs and 24-h substrate utilization were not significantly affected by fructose, although net carbohydrate and fat utilization were significantly lower and higher, respectively, with the high-fructose diet during fasting (P = 0.01) and in the subsequent feeding period (P = 0.05).
Conclusion: Healthy, nonobese children and adolescents adapt appropriately to HF and HC diets.
Key Words: Substrate utilization energy expenditure fat oxidation carbohydrate utilization fructose children adolescents
INTRODUCTION
Difficulty adapting to high-fat (HF) and high-carbohydrate (HC) diets may predispose children toward the development of obesity and diabetes, conditions that are increasing at alarming rates in the United States. Because of their palatability and energy density, HF diets may lead to increased energy intake and cumulative positive fat balances. HF diets may also decrease energy expenditure because of the lower thermic effect of feeding (1). There is no apparent autoregulation between fat intake and fat oxidation; fat oxidation is suppressed by the intake of other macronutrients (2). In contrast, carbohydrate oxidation shows tight autoregulation with carbohydrate intake because the capacity for glycogen storage is limited and de novo lipogenesis is likely to be quantitatively unimportant. The thermic effect of feeding associated with HC diets theoretically would be higher than that associated with HF diets because of glycogen storage and possible glucose stimulation of the sympathetic nervous system (3).
The type of carbohydrate also may influence thermogenesis and nutrient utilization. Because of the unique way in which fructose is metabolized, high intakes of it may alter carbohydrate and lipid utilization. This has become a potential concern because the 19941996 US Department of Agriculture food consumption survey found that 16% of total daily energy intake was from added sweeteners (4). Fructose is rapidly taken up by the liver and is therefore less available for muscle or adipose tissues; fructose enters into glycolysis, glycogenesis, lipogenesis, or gluconeogenesis at the triose phosphate level, bypassing the phosphofructokinase regulatory step (5). Several studies in adults (69) showed higher energy expenditure and carbohydrate utilization after a fructose load than after a glucose, sucrose, or starch load. The higher thermogenesis of fructose is attributed to the higher cost of glycogen storage (10).
Adaptation to HF and HC diets has been studied in adults by using 24-h respiration calorimetry to measure total energy expenditure (TEE) and substrate utilization (1118); but similar studies in children have not been conducted. Taking 24-h measurements of substrate utilization over the fed and fasted states has the advantage of capturing the total effect of diet composition, because the effects may continue beyond dietary absorption and transport. Changes may occur in the pool sizes of different nutrient stores such as glycogen and in the circulating concentrations of free fatty acids; moreover, there may be changes in tissue sensitivity to key hormones such as insulin. Therefore, we designed random crossover studies to test the metabolic flexibility of healthy, nonobese children and adolescents to adapt to HF, HC, and high-fructose diets through changes in substrate utilization. We also tested the effects of sex, puberty, body composition, and fitness on substrate utilization.
SUBJECTS AND METHODS
Study design
In protocol 1, a random crossover design was used to investigate the adaptation of children and adolescents to isoenergetic low-fat (LF), HC (25% and 60% of energy, respectively) and HF, low-carbohydrate (LC) (55% and 30% of energy, respectively) diets. In protocol 2, we investigated the adaptation of adolescents to the LF-HC diet with 11% or 40% of carbohydrate as fructose. After a 7-d diet equilibration at home, the children and adolescents were admitted to the Metabolic Research Unit of the Childrens Nutrition Research Center in Houston for 2 d on 2 occasions 28 wk apart. The study was approved by the Baylor Affiliates Review Board for Human Subject Research. Written informed consent was obtained for all studies.
Subjects
The children and adolescents were required to be healthy, as determined by a physical examination and standard blood chemistry panel; to be prepubertal (Tanner stage 1) or adolescent (Tanner stage 4 or 5); to have a normal weight-for-height (below the 85th percentile for National Center for Health Statistics data); and to have normal body composition, defined as < 28% body fat as measured by dual-energy X-ray absorptiometry. Children and adolescents were excluded if they had an obese parent [body mass index (in kg/m2) > 28] or a first-degree relative with diabetes.
Body size and composition
Before each study, body weight and height were measured by using an electronic balance (Healthometer, Bridgeview, IL) and stadiometer (Holtain Limited, Croswell, Crymych, United Kingdom), respectively. Fat-free mass (FFM) and fat mass (FM) were measured by dual-energy X-ray absorptiometry (QDR2000, software version 5.56; Hologic Inc, Madison, WI). Dual-energy X-ray absorptiometry allows for measurement of 3 compartments: lean tissue mass, FM, and bone mineral content. For the total body, FFM was defined as the sum of lean tissue mass and bone mineral content. Intraabdominal fat (IAF) and subcutaneous abdominal fat were measured by using 1.5-T magnetic resonance imaging scanners [General Electric Sigma 5.4 (General Electric, Milwaukee) and Philips Gyro 6.1 and Intera 7.1 (Philips, Amsterdam)]. A 1-cm, single-slice image was made at the level of the umbilicus (L4L5) by using T-1 weighting to emphasize fat, giving it a bright signal intensity. Using image-analysis software (General Electric or Phillips EASY VISION workstation), each image was carefully inspected and drawn by hand for intraabdominal adipose areas. A fat tissuehighlighting technique was used to determine the IAF and subcutaneous abdominal fat. Magnetic resonance imaging data are expressed as cross-sectional area (cm2).
Fitness
Peak oxygen consumption (
The LF-HC diet consisted of 15% of energy as protein, 25% as fat, and 60% as carbohydrate, of which 21% was from fructose, 13% from sucrose, 5% from glucose, and 4% from lactose. The HF-LC diet consisted of 15% of energy as protein, 55% as fat, and 30% as carbohydrate, of which 20% was from fructose, 12% from sucrose, 5% from glucose, and 5% from lactose. In protocol 2 the fructose content of the LF-HC diet was modified to equal 11% and 40% of energy from carbohydrates in the low- and high-fructose diets, respectively. The energy and macronutrient contents of the diets were analyzed by using the Minnesota Nutrition Data System (version 2.8; Nutrition Coordinating Center, Minneapolis). Subjects were fed an amount of the LF-HC diet or the HF-LC diet to achieve energy balance. For the home diet, total energy intake was determined from the subjects calculated basal metabolic rate according to Schofield et al (19) and multiplied by 1.72.0, depending on the subjects level of physical activity as assessed by interview (20). Three meals/d and 2 snacks/d were weighed, prepacked by the research kitchen staff, and delivered to the subjects home. Subjects were instructed to eat to satiate their appetite and to choose equally from all the food items provided. Any food not consumed was returned to the kitchen and weighed; the actual energy and macronutrient intakes were calculated on the basis of the difference between the weight of the leftover food and that of the delivered food. A dietitian was in frequent telephone contact with the family to evaluate any problems and the subjects adherence to the diet. The dietitian made adjustments to the diet if the amounts were judged to be insufficient or excessive by the family or if large amounts of food were returned to the kitchen on day 3. To compensate for the sedentary level of physical activity in the calorimeter, energy intake was reduced to achieve energy balance. Energy intake for the maintenance of energy balance was estimated from the subjects calculated basal metabolic rate (19), which was multiplied by 1.41.5. The meals were consumed as 4 meals daily: dinner at 1730 and a snack at 1830 on the first day and breakfast and lunch at 0830 and 1200, respectively, on the next day.
Room respiration calorimetry
Blood chemistries were determined from a sample taken at 0600 after an overnight fast. Plasma glucose concentrations were measured by using a Yellow Springs Instruments analyzer (Yellow Springs Instruments, Yellow Springs, OH), and triacylglycerol concentrations were measured enzymatically with the use of lipase on the Vitros automatic analyzer (Johnson & Johnson Clinical Diagnostics, Inc, Rochester, NY). Nonesterified fatty acids (NEFAs) were detected colorimetrically after treatment with acyl-CoA synthetase (EC 6.2.1.3) and acyl-CoA oxidase (EC 1.3.3.6) (Wako Chemicals USA, Inc, Richmond, VA). Insulin and C-peptide concentrations were measured by using a double-antibody radioimmunoassay (Linco Research, Inc, St Charles, MO). The detection limits were 1.4 pmol/L for insulin and 0.1 µg/L for C-peptide. In protocol 2 only, lactate was analyzed enzymatically by using a Yellow Springs Instruments analyzer.
Statistical analysis
Data are summarized as means ± SDs. Descriptive statistics, Pearson correlations, one-sample t test, and multiple regression analysis were performed by using MINITAB (release 13; Minitab, Inc, State College, PA). Repeated-measures analysis of variance (Procedures 2V and 5V; BMDP Statistical Software Inc, Los Angeles) was used to test the effects of diet on energy expenditure and substrate utilization. The basic model included the repeated factor (LF-HC or HF-LC), grouping factors (sex and puberty status), covariates (weight, or FFM and FM, and energy balance), and two-way and three-way interaction terms between diet, sex, and puberty status. Significant interactions were further examined by reanalyzing the diet effect within a group with repeated-measures analysis of variance and by making comparisons between the males and females by using one-way analysis of variance.
RESULTS
Subject description
In protocol 1, the sample consisted of 11 white, 8 African American, and 5 Hispanic children and adolescents (Table 1). The prepubertal children were between 6 and 9 y of age, and the adolescents were between 13 and 16 y of age. As expected, the adolescent males were significantly heavier than the adolescent females, and the adolescents were significantly heavier than the children (sex x puberty interaction, P = 0.002). FFM in the adolescent males significantly exceeded that in the adolescent females, and FFM was significantly higher in the adolescents than in the children (sex x puberty interaction, P = 0.001). FM differed significantly (P = 0.001) by puberty status (adolescent > prepubertal), and percentage of FM (%FM) was significantly higher in the adolescent females than in the adolescent males (sex x pubertal interaction, P = 0.02). IAF did not differ significantly by sex or puberty status. In protocol 2, 9 white, 1 African American, and 2 Hispanic adolescents aged 1316 y were studied. The males were significantly heavier and taller than the girls (P < 0.05). The males also had significantly higher FFM but significantly lower FM and %FM than did the girls (P < 0.002). Subcutaneous abdominal fat and IAF were not significantly different between the males and females.
View this table:
TABLE 1 . Subject characteristics1
Protocol 1
The fasting plasma glucose, NEFA, and triacylglycerol concentrations after consumption of the LF-HC diet were not significantly different from those after consumption of the HF-LC diet (Table 2). Fasting plasma insulin concentrations differed significantly by diet (LF-HC diet > HF-LC diet), sex (M < F), and pubertal status (Tanner stage 1 < Tanner stage 4 or 5) (P = 0.001). C-peptide concentrations were significantly higher after consumption of the LF-HC diet than after consumption of the HF-LC diet (P = 0.02) and were significantly higher in the adolescents than in the prepubertal children (P = 0.001). Nevertheless, both the plasma insulin and C-peptide concentrations in all the children and adolescents after consumption of both diets were within the normal postabsorptive range.
View this table:
TABLE 2 . Fasting plasma concentrations: protocol 11
The total energy and macronutrient intakes of the subjects while they followed the LF-HC and HF-LC diets at home and in the calorimeter are summarized in Table 3. As planned, there was a significant difference (P = 0.001) in food quotient between the 2 diets.
View this table:
TABLE 3 . Dietary intakes at home and in the calorimeter: protocol 11
Twenty-four-hour TEE, energy balance, and substrate utilization are summarized in Table 4. Energy balance differed significantly by sex (P = 0.02) and Tanner stage (P = 0.03) but not by diet. Adjusted for FFM, FM, activity counts, and energy balance, TEE and NPEE were not significantly different by diet, sex, or Tanner stage. The energy expenditure under basal conditions and during sleep and exercise after consumption of the LF-HC diet did not differ significantly from that after consumption of the HF-LC diet (data not shown). As might have been anticipated from the differences in dietary macronutrient composition, the RQ during sleep and exercise, but not during basal conditions, was significantly higher after consumption of the LF-HC diet than after consumption of the HF-LC diet (P = 0.001) and was significantly higher in the females than in the males (sleep, P = 0.002; exercise, P = 0.03) (data not shown).
View this table:
TABLE 4 . Total energy expenditure (TEE) and substrate utilization: protocol 11
Adjusted for energy balance, 24-h RQ and nonprotein respiratory quotient (NPRQ) changed appropriately in response to dietary composition (P = 0.001), and there was a sex difference (P = 0.02) but no pubertal effect. Adjusted for energy balance, protein oxidation [percentage of TEE (%TEE)] did not differ significantly by diet. Adjusted for energy balance, net carbohydrate utilization [(%TEE and percentage of NPEE (%NPEE)] was significantly higher after consumption of the LF-HC diet than after consumption of the HF-LC diet (P = 0.001); net carbohydrate utilization was also significantly higher in the females than in the males (P = .02). Adjusted for energy balance, net fat utilization (%TEE and %NPEE) was significantly higher (P = 0.001) after consumption of the HF-LC diet than after consumption of the LF-HC diet and was significantly higher (P = 0.02) in the males than in the females. Although net fat and carbohydrate utilization (expressed as %TEE or %NPEE) were significantly correlated with weight, body mass index, FFM, and
FIGURE 1. . Net substrate balances of the prepubertal boys (n = 6) and girls (n = 6) and of the adolescent males (n = 6) and females (n = 6) after consumption of the low-fat, high-carbohydrate (LF-HC) and high-fat, low-carbohydrate (HF-LC) diets. Adjusted for energy balance, net protein balance () did not differ significantly by diet. Adjusted for energy balance, net carbohydrate () and fat () balances differed significantly (P = 0.001) by diet, and net carbohydrate balance differed significantly (P = 0.007) by sex.
The 24-h period of calorimetry was partitioned into three 8-h periods to examine the effects of fed (16002400 and 08001600) and fasted (24000800) states on substrate utilization (Figure 2). RQ was also examined at 1-h intervals to depict the 24-h trends in RQ that are not discernable from the 24-h means (Figure 3). As anticipated, RQ and net carbohydrate utilization (%TEE and %NPEE) were significantly higher (P = 0.001) after consumption of the LF-HC diet than after consumption of the HF-LC diet during the fed and fasted states, although the effect was significantly less in the males than in the females (P = 0.02) (data not shown). The converse was true for net fat utilization (data not shown).
FIGURE 2. . Mean respiratory quotients (RQs) of the prepubertal and adolescent males (; n = 12) and females (; n = 12) during fed (16002400 and 08001600) and fasted (24000800) periods of following a low-fat, high-carbohydrate (LF-HC) or high-fat, low-carbohydrate (HF-LC) diet. Adjusted for energy balance, RQs during the fed and fasted states differed significantly by diet (P = 0.001) and sex (P = 0.02), and RQs at the basal metabolic rate (BMR) differed significantly by sex (P = 0.05) but not by diet.
FIGURE 3. . Hourly mean respiratory quotients (RQs) of the prepubertal and adolescent males (; n = 12) and females (- - - -; n = 12) throughout the 24-h calorimetry study after consumption of the low-fat, highcarbohydrate (LF-HC) and high-fat, low-carbohydrate (HF-LC) diets.
Protocol 2
The fasting plasma concentrations of glucose (P = 0.10), insulin (P = 0.17), C-peptide (P = 0.94), lactate (P = 0.82), NEFAs (P = 0.66), and triacylglycerol (P = 0.07) after consumption of the low-fructose diet were not significantly different from those after consumption of the high-fructose diet (Table 5). Although all values were within the normal range, fasting plasma glucose and insulin concentrations were significantly higher in the females than in the males (P = 0.02), and there was no consistent effect of increasing dietary fructose on fasting plasma triacylglycerol concentrations, which increased in 7 adolescents (mean increase: 0.36 mmol/L; range: 0.090.62 mmol/L) and remained unchanged or decreased in 5 adolescents (mean decrease: 0.11 mmol/L; range: -0.16 to 0.02 mmol/L).
View this table:
TABLE 5 . Fasting plasma concentrations: protocol 21
The total energy and macronutrient intakes of the subjects while they followed the low- and high-fructose diets at home and in the calorimeter are summarized in Table 6. As per the design, the low-fructose diet provided 20.5 ± 1.0 g fructose/d whereas the high-fructose diet provided 128.5 ± 3.8 g fructose/d (P = 0.001).
View this table:
TABLE 6 . Dietary intakes at home and in the calorimeter: protocol 21
Twenty-four-hour TEE, energy balance, and substrate utilization are summarized in Table 7. Energy balance differed significantly by sex (P = 0.003) but not by diet. Adjusted for FFM, FM, activity counts, and energy balance, TEE did not differ significantly by diet or sex. The energy expenditure during basal conditions, sleep, and exercise (data not shown) and over 24 h after consumption of the low-fructose diet did not differ significantly from that after consumption of the high-fructose diet. Adjusted for energy balance, the 24-h RQ (P = 0.06), NPRQ (P = 0.06), and net carbohydrate and fat utilization (P = 0.08) were not significantly different after consumption of the high- and low-fructose diets. During sleep but not during basal conditions or exercise, RQ was significantly lower after consumption of the high-fructose diet than after consumption of the low-fructose diet (P = 0.04) (data not shown).
View this table:
TABLE 7 . Total energy expenditure (TEE) and substrate utilization: protocol 21
Factors affecting substrate utilization were explored. Substrate utilization did not differ significantly between the males and the females. For both the males and the females, weight and FFM were negatively correlated with NPRQ and carbohydrate utilization (%NPEE) (r = -0.58 to -0.64, P < 0.01), and %FM was positively correlated with NPRQ and carbohydrate utilization (%NPEE) [r = 0.40 (P = 0.05) for NPRQ and 0.42 (P < 0.05) for carbohydrate utilization]. However, there were no sex-independent effects of body size and composition on substrate utilization. No correlations were observed between IAF or fitness and TEE or substrate utilization. Fasting plasma insulin concentrations were positively correlated with RQ, NPRQ, and net carbohydrate utilization after consumption of the high-fructose diet.
The 24-h net substrate balances were 24.2 ± 11.1 g carbohydrate/d, 7.0 ± 9.1 g fat/d, and 0.2 ± 2.4 g protein/d after consumption of the low-fructose diet and 58.7 ± 8.3 g carbohydrate/d, -5.4 ± 7.9 g fat/d, and 4.0 ± 1.9 g protein/d after consumption of the high-fructose diet. Net carbohydrate balance differed significantly by diet (P = 0.01); the positive net carbohydrate balance was significantly different from zero after consumption of the high-fructose diet (P < 0.001). In contrast, fat balance was neither significantly different by diet nor significantly different from zero. The 24-h calorimetry was partitioned into three 8-h periods to examine the effects of fed (16002400 and 08001600) and fasted (24000800) states on substrate utilization (data not shown). Net carbohydrate and fat utilization (%NPEE) did not differ significantly by diet during the initial fed period (16002400). However, during the fasting period (24000800), net carbohydrate utilization was significantly lower (47% compared with 56%) and net fat utilization was significantly higher (53% compared with 44%) after consumption of the high-fructose diet than after consumption of the low-fructose diet (P = 0.01). During the next 8-h fed period (08001600), net carbohydrate utilization remained significantly lower (65% compared with 70%) and net fat utilization (35% compared with 30%) remained significantly higher after consumption of the high-fructose diet than after consumption of the low-fructose diet (P = 0.05). RQ values at 1-h intervals throughout the 24 h are shown for all the adolescents combined after consumption of the low- and high-fructose diets (Figure 4).
FIGURE 4. . Hourly mean respiratory quotients (RQs) of the adolescent males (n = 6) and females (n = 6) throughout the 24-h calorimetry study after consumption of the low-fructose (- - - -) and high-fructose () diets.
DISCUSSION
In this random crossover challenge of HF-LC and LF-HC diets in children and adolescents, rates of energy expenditure were not significantly affected by diet composition. The children and adolescents adapted appropriately to HC and HF diets, but the adaptive response was not totally complete. Interestingly, the males adapted more readily than did the females to the changes in diet composition. A high-fructose diet did not alter thermogenesis, but slightly lowered net carbohydrate utilization and increased net fat utilization during fasting and in the subsequent feeding period.
Adaptation to LF-HC and HF-LC diets was studied by using room calorimeters to measure TEE and substrate utilization in adults (1118) but not in children. In agreement with our results, most adult studies (1218) did not find an effect of diet composition on TEE if energy intake was fixed.
As anticipated in our study, 24-h net carbohydrate utilization was significantly higher after consumption of the LF-HC diet than after consumption of the HF-LC diet. Twenty-four-hour net fat utilization was significantly higher after consumption of the HF-LC diet than after consumption of the LF-HC diet. The same pattern of substrate utilization was seen in the fed and fasted states. The significant differences in fuel utilization in both the 8-h fed and fasted states indicate at least transient changes in body nutrient stores. The 24-h net carbohydrate balance was positive and the fat balance was negative after consumption of the LF-HC diet in the males and the females, whereas the 24-h net fat balance was positive and the carbohydrate balance was negative after consumption of the HF-LC diet in the females only.
After a 7-d diet equilibration in the present study, adjustment of substrate utilization was apparently incomplete in the males and the females after consumption of the LF-HC diet and in the females after consumption of the HF-LC diet, which is in agreement with some (12, 13), but not all, other investigations (14). The experimental diets may have been such a departure from the subjects habitual diet that a longer period for complete adaptation may have been required. We also must recognize several potential sources of experimental error: diet composition based on the Nutrition Data System database; subject dietary compliance at home; variation in physical activity at home, which may have affected glycogen stores; urinary nitrogen measurement; and assumptions underlying calculation of substrate utilization from respiratory calorimetry (23).
In adults adaptation to a HF diet was associated with fasting plasma insulin and
We and others (14) have seen sex differences in fuel utilization. If girls do not adjust to high-fat diets as readily as boys do, they may be more susceptible to cumulative positive fat balances because of day-to-day fluctuations in fat intake. Our observations of differences in diet adaptation between males and females could not be explained by body size, FFM, FM, %FM, IAF, or fitness level. Hormonal responses to the diets may have accounted for the sex effect. Our only evidence of this is that fasting plasma insulin concentrations were higher in the males than in the females; unfortunately, we did not take blood samples while the subjects were in the fed state. Other unexplored explanations for the sexual dimorphism include differences in sex hormones, FFM composition, glycogen stores, intracellular fatty acid availability in muscle, and lipolytic activity in adipose tissue.
In contrast with other researchers, we did not observe higher energy expenditure after consumption of the high-fructose diet. Previous investigations (69, 29, 30) involved adults and studied the thermogenic response after a single meal or bolus during a time frame that was shorter than that of our study. Blaak and Saris (9) reported that increases in energy expenditure occurred in lean, healthy males after they ingested 75 g fructose but that energy expenditure returned to baseline values after 210 min. Carbohydrate oxidation increased and fat oxidation decreased with fructose ingestion, and the greatest changes were observed at 60 and 30 min, respectively. Brundin and Wahren (30) found that a 75-g load of fructose increased oxygen uptake 9.5% from basal values and RQs to 0.97 after 2 h. Also, in contrast to other studies (79, 29), we did not observe an increase in 24-h net carbohydrate utilization with the high-fructose diet. The discrepancy is probably due to differences in study duration and the clear short-term and longer-term effects of fructose feeding. We observed decreased net carbohydrate utilization and increased net fat utilization during fasting and in the subsequent feeding period after consumption of the high-fructose diet, although the physiologic significance of these observations is unclear given the absence of a diet effect on substrate utilization in the first 8-h fed period. The lower NPRQ seen after consumption of the high-fructose diet may be attributed to alternative intermediate pathways for fructose metabolism. By bypassing the regulatory step in glycolysis catalyzed by phosphofructokinase, fructose is rapidly phosphorylated, providing increased substrate to metabolic pathways leading from triose phosphate (ie, glycolysis, glycogenesis, gluconeogenesis, and lipogenesis). The major products of fructose metabolism in the liver are glucose, glycogen, and lactate, whereas smaller amounts are oxidized to carbon dioxide or converted to ketone bodies and triacylglycerol (31). When subjects are in the fed state, fructose appears to be predominantly glycolyzed to lactate rather than converted to glycogen or glucose. Consequently, the postprandial insulin counterregulatory response would be lower after fructose consumption than after sucrose consumption. When subjects are in the fasted condition, gluconeogenesis and glucose production from fructose are active.
The longer-term effects of a high-fructose diet that were shown in rats include increased concentrations and utilization of nonesterified fatty acids (32, 33). Just as the utilization of glucose is decreased at the level of the whole body, it is also decreased at the level of individual tissues. In the liver, there is decreased utilization and oxidation of glucose (34, 35). Muscle, as well as adipose tissue, has a decreased ability to metabolize glucose and an increased ability to oxidize fatty acids after consumption of a high-fructose diet. Healthy persons consuming high-fructose diets may have normal fasting triacylglycerol concentrations; however, this is only an indicator of triacylglycerol clearance and does not give information on prandial and postprandial triacylglycerol concentrations, which may be elevated after consumption of a high-fructose diet. Plasma triacylglycerol has been shown to increase in humans consuming high-fructose diets (3639). Unfortunately, we did not take blood samples to examine the postprandial excursions in the concentrations of glucose, insulin, lactate, NEFAs, and triacylglycerol.
In conclusion, we showed that healthy, nonobese children and adolescents adapt appropriately to HF and HC diets, although the adaptive response in the subjects in our study was incomplete. The high-fructose diet did not significantly alter thermogenesis or 24-h net carbohydrate or fat utilization but did slightly lower net carbohydrate utilization and increase net fat utilization during fasting and in the subsequent feeding period; the physiologic significance of these observations remains unclear without more extensive study over a longer duration.
ACKNOWLEDGMENTS
We thank the children and adolescents who participated in this study and acknowledge the contributions of Sopar Seributra and Sandra Kattner for nursing and dietary support; Maurice Puyau, Firoz Vohra, Roman Shypailo, JoAnn Pratt, and Mary Thotathuchery for technical assistance; and Anne Adolph for data management.
REFERENCES