点击显示 收起
1 From the Department of Surgery, The University of Texas Medical Branch, Galveston; the Shriners' Burns Hospital, Galveston, TX; and the Department of Internal Medicine, Washington University School of Medicine, St Louis.
2 Supported by National Institutes of Health Grant DK-51969, American Heart Association Texas Affiliate grant 96G-1618, a Fellowship from the European Society of Parenteral and Enteral Nutrition, General Clinical Research Center of the University of Texas grant 00073, and Shriners Hospital Grant 15849. 3 Address reprint requests to B Mittendorfer, Washington University School of Medicine, Department of Internal Medicine–Gastroenterology, 660 S Euclid, Box 8031, St Louis, MO 63110-1093. E-mail: mittendb{at}medicine.wustl.edu.
ABSTRACT
Background: High-carbohydrate (HC) diets are recommended for lowering the risk of coronary heart disease because they decrease plasma LDL-cholesterol concentrations. However, an unfavorable effect of HC diets is an increase in plasma triacylglycerol concentrations. The underlying mechanisms of this effect are still unclear.
Objective: We examined the effect of diet composition on VLDL-triacylglycerol metabolism using in vivo isotopically labeled VLDL-triacylglycerol tracers.
Design: Six healthy subjects were studied on 2 occasions: after 2 wk of an HC diet (75% carbohydrates, 10% fat, and 15% protein) and after 2 wk of an isoenergetic high-fat (HF) diet (30% carbohydrates, 55% fat, and 15% protein).
Results: The plasma VLDL-triacylglycerol concentration was higher after the HC diet than after the HF diet (690 ± 186 compared with 287 ± 104 µmol/L; P < 0.05) because of higher rates of VLDL-triacylglycerol production (0.76 ± 0.12 compared with 0.45 ± 0.15 µmol• kg-1•min-1; P < 0.05) rather than diminished VLDL-triacylglycerol clearance (1.5 ± 0.5 compared with 1.7 ± 0.5 mL•kg-1•min-1 after the HC diet than after the HF diet, respectively). The increase in VLDL-triacylglycerol production was probably mediated by a decrease in hepatic fatty acid oxidation after the HC diet (0.13 ± 0.02 compared with 0.69 ± 0.24 µmol•kg-1•min-1; P < 0.05), which presumably increased hepatic fatty acid availability for triacylglycerol synthesis.
Conclusions: The increase in fasting plasma triacylglycerol concentrations in response to short-term HC diets is due to accelerated VLDL-triacylglycerol secretion. Increased hepatic fatty acid availability, resulting from reduced hepatic fatty acid oxidation, is most likely responsible for the observed increase in VLDL-triacylglycerol secretion.
Key Words: Stable isotopes • fatty acid • liver • triacylglycerol • substrate oxidation • high-carbohydrate diet • high-fat diet
INTRODUCTION
High-carbohydrate (HC) diets are recommended for lowering the risk of coronary heart disease (CHD) because they decrease plasma LDL-cholesterol concentrations (1–4); however, they also increase fasting plasma triacylglycerol concentrations (1, 2, 4–7). On the basis of epidemiologic studies, elevated plasma triacylglycerol concentrations are associated with an increased risk of the development of (8–11) and mortality from (12) CHD. The significance of this association, however, is debatable; in some studies, plasma triacylglycerol concentration is a cholesterol-independent risk factor (8, 11), whereas in other studies it is not (12–14). Furthermore, elevated plasma triacylglycerol concentrations are associated with alterations of the coagulation system (15) and, in severe cases, are thought to play a role in the development of pancreatitis (16). Thus, understanding why HC diets elevate plasma triacylglycerol concentrations may aid in designing dietary guidelines that will promote the maintenance of good health.
In rats, there is strong evidence that hepatic VLDL-triacylglycerol production is greater in HC than in high-fat (HF) diets (17–22). In humans, however, there is still debate as to whether the increase in plasma triacylglycerol concentrations in response to HC intake is linked to increased hepatic VLDL-triacylglycerol secretion (23–28), diminished VLDL-triacylglycerol utilization (3, 29), or both (27). The discrepancy among these findings is most likely due to the differences in 1) the individuals studied [ie, healthy subject (3, 23, 29) or subjects with abnormalities in lipid metabolism (26, 27), atherosclerotic heart disease (25), or type 2 diabetes (3, 24, 27, 28)], 2) diet composition (from fat-free liquid to low-fat solid food diets with varying amounts of simple and complex carbohydrates), and 3) the duration of the studies (from several days to weeks). Furthermore, in some studies, VLDL-triacylglycerol kinetics were determined in the postabsorptive state, whereas in other studies they were measured during the ingestion of a dietary formula. Finally, the diversity of methods used to measure VLDL-triacylglycerol kinetics may have contributed to the different results.
Using a newly developed tracer technique, we showed that short-term hyperglycemia or hyperinsulinemia, with concomitant intralipid and heparin infusion to maintain a constant plasma fatty acid concentration, increases VLDL-triacylglycerol production but has no effect on VLDL-triacylglycerol clearance (30). The increase in VLDL-triacylglycerol secretion in this study was likely due to decreased splanchnic fatty acid oxidation (30), and HC diets may affect triacylglycerol metabolism in a similar way. However, the short duration of the study and the differences in the amount and mode of carbohydrate administration between the previously mentioned study (30) and HC dietary interventions preclude the extrapolation of these conclusions to such circumstances.
In the present study we measured VLDL-triacylglycerol secretion rates with the use of our newly developed VLDL-triacylglycerol tracer technique (30), while simultaneously assessing splanchnic fatty acid oxidation in subjects who had consumed an HC or HF diet for 2 wk.
SUBJECTS AND METHODS
Subjects
Six subjects [5 men and 1 woman; age: 35 ± 4 y; body weight: 74 ± 2 kg; body mass index (in kg/m2) 22 ± 1] participated in the study. All subjects were healthy, as indicated by a comprehensive medical history and results of a physical examination and standard blood and urine tests. All subjects had normal oral glucose tolerance, were sedentary, and did not exercise during the dietary intervention studies. The study protocol was approved by the Institutional Review Board and the General Clinical Research Center (GCRC) of the University of Texas Medical Branch at Galveston. Informed consent was obtained from all subjects before enrollment in the study.
Experimental design
Each subject was admitted to the GCRC at the University of Texas Medical Branch on 2 different occasions. On one occasion, the subjects consumed an HC diet and on the second occasion they consumed an isoenergetic-HF diet; both diets were consumed for 14 d (Table 1). Although it is not known whether 2 wk of dietary modification are sufficient to achieve a new, steady, physiologic state, it was shown previously that diet-induced changes in enzyme concentrations and activities (31), substrate oxidation (32), and plasma triacylglycerol concentrations (2, 26) occur and appear to stabilize after 2 wk. Thus, 2 wk of dietary modification should be sufficient to investigate its effects on substrate metabolism. The 2 diets were randomized in order and were consumed within 5–6 wk of each other. For each diet, 2 different menus, which consisted exclusively of regular food items (eg, pasta, meat, salads, fruit, and dairy products but no dietary formulas), were prepared by the metabolic kitchen at the GCRC and provided to subjects alternately every other day. The total daily energy supply was estimated to meet each subject's resting energy expenditure in addition to an extra 20% of supplied energy to account for their daily activities. Breakfast accounted for 20%, lunch for 35%, and dinner for 30% of total daily energy intake; 15% of total daily energy was consumed in the form of small snacks (eg, cereal bars and cookies). When at the GCRC, subjects were not allowed to leave the facility for >15 min at one time so that their food intake could be monitored closely. All subjects maintained their weight during the trials and during the time between trials.
View this table:
TABLE 1.. Diet composition1
Isotope infusion
On the evening of day 14 at 1900, the subjects ingested their last meal and then they fasted until the completion of the tracer infusion (15 h later) the next day. At 2030 polytetrafluoroethylene catheters were placed percutaneously into an antecubital vein for isotope infusion and into a contralateral dorsal hand vein for blood sampling. The hand vein was heated during sampling to obtain arterialized blood samples. At 2100 a bolus of NaH13CO3 (25 µmol/kg) was administered and a continuous infusion (0.035 µmol•kg-1•min-1) of a [U-13C]fatty acid mixture (86% enriched) bound to human albumin was initiated and maintained until the completion of the study the next day. At 0700 the next day, the subjects were transferred to the Angiography and Interventional Radiology Center of the University of Texas Medical Branch, where catheters were inserted into the femoral artery and a hepatic vein (see below). After catheter placement, the subjects were transferred back to the GCRC for completion of the study. The [U-13C]fatty acid infusion continued uninterrupted during catheter placement by use of a battery-operated pump. In 2 of the 6 subjects, the tracer infusion protocol was performed without the placement of the hepatic venous catheter.
Sixty to 90 min after catheter placement, arterial and hepatic venous blood samples were drawn and a primed, constant infusion of labeled VLDL-triacylglycerol was initiated and maintained for 120 min. The VLDL-triacylglycerol infusion rate was different for each subject, depending on the amount of labeled triacylglycerol produced in vivo (see below). Sixty percent of the available tracer was administered as a bolus at the beginning of the study to prime the VLDL-triacylglycerol pool; the remaining 40% was given at a constant rate (27 mL/h, ie, 2.5 µmol labeled VLDL triacylglycerol/h) over the remaining 2 h of the study.
Calibrated syringe pumps were used for all tracer infusions. For each infusion study, aliquots of the isotope infusates were analyzed for the exact enrichment and concentration of tracers. All tracer infusions were stopped after the last blood sample was obtained and the catheters were removed immediately; pressure was applied to the catheter insertion site and the subjects were monitored in the GCRC for several hours before discharge.
Blood and breath sampling
To determine background blood carbon dioxide and fatty acid enrichments, blood samples were drawn on the evening of day 14, before the tracer infusion began. Blood samples (10 mL) from the femoral artery and the hepatic vein were obtained both immediately before the VLDL-triacylglycerol tracer infusion began and every 10 min during the last 30 min of the tracer infusion to determine plasma substrate and hormone concentrations, plasma fatty acid enrichment, VLDL-triacylglycerol glycerol enrichment, and blood carbon dioxide carbon enrichment and concentration. To determine substrate and hormone concentrations and fatty acid and VLDL-triacylglycerol glycerol enrichment, blood samples were collected in prechilled tubes that contained EGTA, which was shown to prevent ex vivo lipolysis of triacylglycerol in the plasma of subjects given heparin as an anticoagulant (33). Samples were iced until the end of the study, at which time plasma was separated by centrifugation. One-third of the plasma was stored at -20°C until analyzed; the remaining two-thirds were stored at 4°C until the separation of VLDL-triacylglycerol by ultracentrifugation shortly after completion of the study. To determine blood carbon dioxide concentration, blood samples were collected into prechilled tubes containing sodium heparin and were analyzed immediately with the use of a 965 Ciba Corning CO2 Analyzer (Corning Inc, Eastham, MA). To determine blood 13CO2 enrichment, 1 mL blood was injected into a sealed Vacutainer (Becton Dickinson Co, Franklin Lakes, NJ) containing 5–10 µL hyperphosphoric acid (85%). Whole-body oxygen consumption (
Catheter placement
The right groin was prepared and draped in a sterile fashion and a lead glove was placed over the genitalia before the placement of femoral arterial and hepatic venous catheters. Next, a short, straight 4 Fr catheter connected to a pressurized flush setup was placed retrograde into the right common femoral artery. Thereafter, the right common femoral vein was punctured and a 6 Fr sheath was inserted. Through this sheath, a straight 5 Fr Simmon's catheter (Becton Dickinson Co) with several side holes near its tip was manipulated into the right or middle hepatic vein. After the catheter was positioned into the hepatic vein, a digital venogram was performed to verify its placement. The position of the catheter was confirmed again by a plain view abdominal X-ray at the end of the study. After suturing both catheters and sheath in place, a sterile transparent dressing was used to cover the vascular entry sites. All catheters were kept patent by a slow (15–20 mL/h) controlled infusion of heparin-treated (1000 U/L) normal saline.
In vivo VLDL-triacylglycerol tracer production
Because a VLDL-triacylglycerol tracer is not commercially available, it was produced in vivo 3–4 d before the infusion studies began. Subjects ingested 3 g [13C3]glycerol dissolved in 60 mL water and plasmapheresis was performed 4 h later. VLDLs that contained [13C3]glycerol were isolated from plasma (400 mL) under sterile conditions by overlaying 30 mL plasma with a solution with a density = 1.006 kg/L (26) and spinning it in an SW 28 rotor (model L7-55; Beckman Instruments, Palo Alto, CA) for 16 h at 78000 x g. The newly labeled VLDL solution was tested for sterility and then stored at 4°C until used as a tracer 2–3 d later. This procedure does not deplete subjects of any red blood cells and produces enough tracer for 2–3 h of infusion, plus the priming dose. Because of the in vivo production procedure, only autologous infusions were used.
Blood flow
Splanchnic blood flow was determined by using a constant infusion of indocyanine green (ICG) dissolved in normal saline. The dye was infused through the femoral artery catheter at a rate of 0.5 mg/min for 55 min, beginning 75 min before the end of the tracer infusion. Blood samples were drawn simultaneously from the hepatic vein and a hand vein at 25, 35, 45, and 55 min after the ICG infusion began. The ICG dye concentration in the infusate and in serum samples was determined using a spectrophotometer set at = 805 nm.
Materials
All isotopes were purchased from Cambridge Isotope Laboratories (Andover, MA). The composition of the [U-13C]fatty acid mixture was 51.7% palmitic acid (16:0), 9.3% palmitoleic acid (16:1), 5.5% stearic acid (18:0), 7.9% oleic acid (18:1), 14.1% linoleic acid (18:2) and 1.5% linolenic acid (18:3). Human albumin (5%) was purchased from Baxter Healthcare Corporation (Glendale, CA) and ICG from AKORN, Inc (Buffalo Grove, IL).
Analytic procedures
The ratio of 13CO2 to 12CO2 ratio in blood samples was determined by isotope ratio mass spectrometry (SIRA VG Isotech, Cheshire, United Kingdom) as previously described (30, 34). To determine plasma fatty acid enrichments and concentrations, fatty acids were extracted from plasma and isolated by thin-layer chromatography as described previously (35). After extraction from the thin-layer chromatography silica gel, fatty acids were combusted and the ratio of the 13CO2 to 12CO2 in the headspace was determined by isotope ratio mass spectrometry. Plasma glucose and lactate concentrations were measured with the use of a 2300 STAT analyzer (Yellow Springs Instruments Co, Yellow Springs, OH). Plasma ketone concentrations were determined enzymatically and plasma insulin concentrations were determined by radioimmunoassay.
Plasma VLDLs were isolated by overlaying 2 mL plasma with a solution with a density = 1.006 kg/L, followed by ultracentrifugation in a 70.1 Ti rotor for 16 h at 110000 x g and 15°C. Next, the fraction containing VLDL was carefully removed by slicing the top of the tube and was stored at 4°C until further processing.
The triacylglycerol concentration in the VLDL fraction was determined enzymatically. To determine VLDL-triacylglycerol glycerol enrichment, triacylglycerols in the VLDL fraction were isolated by thin-layer chromatography (35), hydrolyzed, and the fraction containing glycerol was collected and dried in a Speed-Vac concentrator (Savant Instruments, Farmingdale, NY). One hundred milliliters heptafluorobutyryl anhydride:ethylacetate (1:1, by vol) were added to the dried residue and the samples were heated at 70°C for 10 min. The samples were transferred to autosampler vials and analyzed by gas chromatography–mass spectrometry.
Calculations
Fatty acid kinetics
The whole-body rate of fatty acid appearance in plasma was determined by dividing the tracer infusion rate by the average enrichment of fatty acids in arterial plasma during the last 30 min of the tracer infusion. Splanchnic fatty acid kinetics were calculated by use of the average femoral arterial and hepatic venous fatty acid and carbon dioxide enrichments and concentrations in blood and plasma samples obtained in the last 30 min of the tracer infusion and splanchnic blood and plasma flow measurements during the last 30 min of the ICG infusion. Fatty acid net balance (NB) across the splanchnic region was calculated with the use of the following equation:
RESULTS
Tracer enrichment
When measurements were made for the calculation of fatty acid and VLDL-triacylglycerol kinetics, an isotopic plateau was reached for VLDL triacylglycerol, plasma fatty acid, and blood carbon dioxide carbon enrichment (Figure 1).
FIGURE 1. . Mean (±SEM) plasma VLDL triacylglycerol (TG), plasma fatty acid (FA), and blood carbon dioxide carbon enrichments in the femoral artery during the last 30 min of tracer infusion after the high-fat (HF; •) and high-carbohydrate (HC; ) diets. n = 6.
Substrate concentrations and blood flow
During the last 30 min of the tracer infusion, plasma glucose, fatty acid, total cholesterol, and insulin concentrations were similar after the HC and HF diets (Table 2). However, the plasma lactate concentration was higher and the plasma ß-hydroxybutyrate concentration was lower after the HC diet than after the HF diet (Table 2). Both total triacylglycerol and VLDL-triacylglycerol concentrations were higher (P < 0.01) after the HC diet (Figure 2). Splanchnic blood flow was 943 ± 102 mL/min after the HC diet and 1041 ± 148 mL/min after the HF diet (NS).
View this table:
TABLE 2.. Average substrate, insulin, and cholesterol concentrations in the femoral artery after the high-fat (HF) and high-carbohydrate (HC) diets1
FIGURE 2. . Mean (±SEM) total and VLDL-triacylglycerol (TG) concentrations in arterial plasma after the high-fat (HF) and high-carbohydrate (HC) diets. *Significantly different from HF diet, P < 0.05. n = 6.
VLDL-triacylglycerol kinetics
The rate of appearance of VLDL triacylglycerol was significantly higher (P < 0.04) after the HC diet (0.76 ± 0.12 µmol•kg-1•min-1) than after the HF diet (0.45 ± 0.15 µmol•kg-1•min-1; Figure 3), whereas there was no difference in VLDL-triacylglycerol clearance between the HC and HF diet groups (1.5 ± 0.5 compared with 1.7 ± 0.5 mL•kg-1•min-1, respectively).
FIGURE 3. . Mean (±SEM) rate of appearance (Ra) and clearance of VLDL triacylglycerol (TG) after the high-fat (HF) and high-carbohydrate (HC) diets. *Significantly different from HF diet, P < 0.05. n = 6.
Whole-body and splanchnic fatty acid kinetics and oxidation
The rate of appearance of plasma fatty acids was 30% lower after the HC diet (7.3 ± 0.8 µmol•kg-1•min-1) than after the HF diet (10.8 ± 1.9 µmol•kg-1•min-1; P < 0.05). Whole-body fat oxidation (indirect calorimetry) was significantly lower after the HC diet (2.4 ± 0.6 µmol•kg-1•min-1) than after the HF diet (4.5 ± 0.4 µmol•kg-1•min-1), resulting in a higher respiratory quotient after the HC than after the HF diet (0.89 ± 0.02 compared with 0.8 ± 0.01). Similarly, oxidation of plasma-derived fatty acids in the splanchnic region was lower after the HC than after the HF diet (0.13 ± 0.02 compared with 0.69 ± 0.24 µmol•kg-1•min-1, respectively). The decrease in splanchnic fatty acid oxidation was not related to a decline in fatty acid availability to the splanchnic region because the splanchnic fatty acid net balance (36 ± 16 compared with 67 ± 15 µmol/L, respectively) was similar after the HC and HF diets. However, the percentage of fatty acid that were used and oxidized to carbon dioxide in the splanchnic region was lower after the HC (23 ± 4%) than after the HF (49 ± 7%) diet.
DISCUSSION
In the present study we examined 1) the effect of an HC diet consumed for 2 wk on VLDL-triacylglycerol metabolism in healthy subjects by determining the absolute rate of VLDL-triacylglycerol secretion with the use of in vivo-labeled VLDL-triacylglycerol tracers, and 2) fatty acid kinetics and oxidation across the splanchnic region. We found that the increase in plasma VLDL-triacylglycerol concentration after consumption of the HC diet was accompanied by, and most likely a result of, increased VLDL-triacylglycerol secretion. This increase in VLDL-triacylglycerol secretion may be linked to decreased oxidation of plasma-derived fatty acid by splanchnic tissues, which is expected to increase hepatic fatty acid availability for triacylglycerol synthesis. VLDL-triacylglycerol clearance, on the other hand, was not different in response to the 2 diets. The absolute rate of VLDL-triacylglycerol uptake per given time, and thus the utilization of VLDL-triacylglycerol, was greater after the HC diet than after the HF diet. This is because clearance is not a measure of the absolute rate of utilization but an estimate of the efficiency of VLDL-triacylglycerol uptake. Accelerated VLDL-triacylglycerol utilization after the HC diet, however, was not sufficient to prevent an increase in plasma VLDL-triacylglycerol concentration.
Total plasma cholesterol concentrations were not affected by diet composition. Likewise, we observed no effect of diet composition on fasting plasma glucose, fatty acid, and insulin concentrations. These findings are similar to previous reports (29, 38) on the effect of HC diets on plasma substrate, cholesterol, and insulin concentrations of healthy subjects.
The observed increase in the rate of fasting hepatic VLDL-triacylglycerol secretion agrees with the results from most (23–28) but not all (3, 29) studies that examined the effect of diet composition on VLDL-triacylglycerol metabolism in humans. However, only a few of these studies investigated the effect of HC diets in healthy persons (3, 23, 29); others included only 2 or 3 healthy persons (26, 27), which makes interpretation of the results difficult. Furthermore, several of the studies that included healthy people were performed with subjects who were in the fed state (3, 23, 26, 27), in which case the acute meal effect may have obscured the effect of long-term HC intake. In 1 (3) of the 2 studies (3, 29) in which VLDL-triacylglycerol secretion was not altered by the HC diet, plasma triacylglycerol concentrations also did not change. This contradicts most reports in the literature (1, 2, 4–7, 23–29) and may explain the lack of an effect of an HC diet on VLDL-triacylglycerol metabolism. In the other study (29), the increase in plasma triacylglycerol concentration, which was similar to the concentration in our study, was accompanied by a transport rate that was 20% higher. However, the difference was not significant, which was possibly due to insufficient statistical power to detect a difference of this magnitude that, in turn, could have resulted from the small difference in carbohydrate content between the HF (50% of total energy intake) and HC (68%) diets compared with this and other studies. The proportional contribution of simple and complex carbohydrates, however, was similar to our study. In contrast with other studies, the HC diet used by Parks et al (29) was also high in fiber. However, we believe that the fiber content was not responsible for the observed differences in VLDL-triacylglycerol metabolism because high amounts of fiber in the diet are generally associated with low plasma total and LDL-cholesterol concentrations (39–41), whereas triacylglycerol concentrations are not affected (39, 40). Parks et al (29), however, found no difference in plasma cholesterol concentrations, whereas plasma triacylglycerol concentrations increased significantly.
The comparison and interpretation of the available data from the literature is further complicated by the variety of techniques that have been used to measure VLDL-triacylglycerol kinetics; although they all have been used for a long time, no one technique has been generally accepted. This is because all such techniques are tainted by several assumptions (42) and there is evidence that substantially different results can be obtained depending on the mathematical model (27) or applied method (43). In most of the studies mentioned previously, the VLDL-triacylglycerol secretion rate was estimated based on the fractional catabolic (3, 25–28) or synthetic (23, 29) rate of the VLDL-triacylglycerol pool. In the present study we used a newly developed methodology that involved in vivo VLDL-triacylglycerol labeling and reinfusion as a tracer on a different occasion. This technique is based on the simple tracer dilution principle and, therefore, involves fewer assumptions with regard to, for example, the appropriate precursor pool or mathematical model than do other techniques. In addition, it allows for the direct measurement of absolute rates of VLDL-triacylglycerol secretion.
The underlying mechanisms mediating the observed increase in VLDL-triacylglycerol secretion in response to an HC diet are not entirely clear. From studies in vitro (44) and in healthy persons (45, 46) it is known that fatty acids directly stimulate triacylglycerol production; insulin, on the other hand, has an inhibitory effect (45, 47). Plasma insulin and fatty acid concentrations were similar after 2 wk of the HC and HF diet. However, hepatic availability of plasma fatty acid was higher after the HC than after HF diet. This is because hepatic fatty acid oxidation decreased in response to the HC diet, whereas net fatty acid balance across the splanchnic region was similar after the 2 diets. As a result, more of the fatty acids that entered the splanchnic region may have been channeled toward pathways other than oxidation, presumably triacylglycerol synthesis, after the HC diet. Increased fatty acid availability to the splanchnic region may, therefore, have caused a higher rate of VLDL-triacylglycerol secretion after the HC diet than after the HF diet in our study.
The mechanisms responsible for the suppression of splanchnic fatty acid oxidation after the HC diet cannot be deduced from the present study. However, it is likely that the ingestion of high amounts of carbohydrate for a prolonged period of time diminishes fatty acid oxidation by decreasing the rate of fatty acid entry in the mitochondria. We presented evidence previously suggesting that increased systemic glucose concentrations inhibit whole-body and splanchnic fatty acid oxidation (48, 49) in such a manner (49). In the present study, systemic plasma glucose concentrations and glucose net balance across the splanchnic region were similar after the HC and HF diets. It is known, however, that basal glycogenolysis and glycolytic flux are high after an HC diet, as indicated by higher rates of carbohydrate oxidation (29, 32) and lactate concentrations (in present study). Thus, the observed inhibition of splanchnic fatty acid oxidation after the HC diet may have been mediated by an increase in hepatic glycolytic flux, which resulted in the inhibition of fatty acid entry into mitochondria, similar to what we observed during hyperglycemia (48, 49).
Our finding that HC diets elevate plasma triacylglycerol concentrations by increasing the rate of VLDL-triacylglycerol secretion, presumably by increased hepatic fatty acid availability, suggests that measures should be taken to lower hepatic fatty acid availability to prevent the increase in plasma VLDL-triacylglycerol concentration. Low to moderate intensity exercise, for instance, stimulates plasma-derived fatty acid oxidation (50, 51), which may decrease fatty acid availability to the liver and thereby counteract the effect of HC diets on plasma VLDL-triacylglycerol concentrations. Furthermore, there is evidence that exercise stimulates muscle lipoprotein lipase concentrations and activities (52), which may lower plasma VLDL-triacylglycerol concentrations by increasing VLDL-triacylglycerol clearance. An alternative measure would be to replace carbohydrates in the diet with n–3 fatty acids, which were found to decrease plasma triacylglycerol concentration (53, 54).
In summary, we showed that the HC diet–induced increase in fasting plasma triacylglycerol concentration is accompanied by, and probably results from, an accelerated VLDL-triacylglycerol secretion rather than from diminished VLDL-triacylglycerol use. The increase in triacylglycerol secretion may be linked to an increase in hepatic fatty acid availability resulting from lower splanchnic fatty acid oxidation.
ACKNOWLEDGMENTS
We gratefully appreciate the help of the nursing and dietary staff of the General Clinical Research Center at the University of Texas Medical Branch in Galveston and thank O Ozkan for the placement of the catheters and Y Zheng and G Zhao for their technical assistance.
REFERENCES