点击显示 收起
1 From the Departments of Endocrinology and Metabolism, Clinical Chemistry Laboratory of Endocrinology, and Biochemistry, Academic Medical Center, University of Amsterdam; the Center for Liver, Digestive and Metabolic Diseases, Academic Hospital Groningen, Groningen, Netherlands; and the Departments of Internal Medicine and Endocrinology, Leiden University Medical Center, Leiden, Netherlands.
2 Supported by grant 96.604 from the Dutch Diabetes Foundation. 3 Address reprint requests to PHLT Bisschop, Department of Endocrinology and Metabolism (F5), Academic Medical Center, University of Amsterdam, PO Box 22700, 1100 DE Amsterdam, Netherlands. E-mail: p.h.bisschop{at}amc.uva.nl.
ABSTRACT
Background: A high dietary fat intake is involved in the pathogenesis of insulin resistance.
Objective: The aim was to compare the effect of different amounts of dietary fat on hepatic and peripheral insulin sensitivity.
Design: Six healthy men were studied on 3 occasions after consuming for 11 d diets with identical energy and protein contents but different percentages of energy as fat and carbohydrate as follows: 0% and 85% [low-fat, high-carbohydrate (LFHC) diet], 41% and 44% [intermediate-fat, intermediate-carbohydrate (IFIC) diet], and 83% and 2% [high-fat, low-carbohydrate (HFLC) diet]. Insulin sensitivity was quantified by using a hyperinsulinemic euglycemic clamp (plasma insulin concentration: 190 pmol/L).
Results: During hyperinsulinemia, endogenous glucose production was higher after the HFLC diet (2.5 ± 0.3 µmol•kg-1•min-1; P < 0.05) than after the IFIC and LFHC diets (1.7 ± 0.3 and 1.2 ± 0.4 µmol•kg-1•min-1, respectively). The ratio of dietary fat to carbohydrate had no unequivocal effects on insulin-stimulated glucose uptake. In contrast, insulin-stimulated, nonoxidative glucose disposal tended to increase in relation to an increase in the ratio of fat to carbohydrate, from 14.8 ± 5.1 to 20.6 ± 1.9 to 26.2 ± 2.9 µmol•kg-1•min-1 (P < 0.074 between the 3 diets). Insulin-stimulated glucose oxidation was significantly lower after the HFLC diet than after the IFIC and LFHC diets: 1.7 ± 0.8 compared with 13.4 ± 2.1 and 19.0 ± 2.1 µmol•kg-1•min-1, respectively (P < 0.05). During the clamp study, plasma fatty acid concentrations were higher after the HFLC diet than after the IFIC and LFHC diets: 0.22 ± 0.02 compared with 0.07 ± 0.01 and 0.05 ± 0.01 mmol/L, respectively (P < 0.05).
Conclusion: A high-fat, low-carbohydrate intake reduces the ability of insulin to suppress endogenous glucose production and alters the relation between oxidative and nonoxidative glucose disposal in a way that favors storage of glucose.
Key Words: Glucose metabolism • insulin • dietary fat • dietary carbohydrate • euglycemic clamp • glucose turnover rate • men
INTRODUCTION
In addition to genetic predisposition, dietary factors have been linked to the pathogenesis of insulin resistance, especially a high intake of dietary fats. In rats, high fat feeding induces insulin resistance at both the peripheral and hepatic levels (1). However, the effects of fat intake on insulin action and glucose metabolism in humans are less clear. Yost et al (2) found that glucose disposal at plasma insulin concentrations of 400 pmol/L was not significantly different after 16 d of diets containing either 25% or 50% fat. At comparable insulin concentrations, Cutler et al (3) showed that even intakes of diets containing 75% of energy as fat did not alter peripheral or hepatic insulin sensitivity. However, insulin concentrations of 400 pmol/L result in complete suppression of endogenous glucose production (4) and are therefore less suitable for quantifying hepatic insulin sensitivity. Fukagawa et al (5) compared the effect of a habitual diet containing 40% of energy as fat with that of a diet containing 15% of energy as fat at lower insulin concentrations (200 pmol/L). They found a modest increase in peripheral insulin sensitivity after the low-fat diet, but no difference in hepatic insulin sensitivity. In humans, evidence of a possible dose-effect relation between dietary fat content and insulin sensitivity, on the basis of physiologically relevant insulin concentrations and the entire spectrum of isoenergetic dietary fat content, has not been documented.
Therefore, we studied the effects of 3 euenergetic diets representing a wide range of fat contents (from 0% to 83% of energy) on insulin sensitivity in 6 healthy men. We used a hyperinsulinemic, euglycemic clamp to produce plasma insulin concentrations of 200 pmol/L to determine both hepatic and peripheral insulin sensitivity.
SUBJECTS AND METHODS
Subjects
The subjects were 6 healthy men aged 29–55 y with a body mass index (in kg/m2) of 21–26. The subjects were in good health, had no family history of diabetes, and used no medications. All subjects were recruited from among hospital employees and participated because of their special interest in this field of research. All subjects gave written, informed consent and the study was approved by the Medical Ethical Committee of the Academic Medical Center.
Diets
The subjects were studied on 3 occasions, each time after they had consumed a different diet for 11 d. The experiments in each subject were separated by an interval of 8–10 wk, during which time the subjects resumed their habitual diets. The sequence of the 3 studies was determined by balanced assignment. The 3 diets consisted of liquid formulas containing identical amounts of protein (15% of energy) and an identical protein composition. The diets were custom-made (Nutricia, Zoetermeer, Netherlands). The low-fat, high-carbohydrate (LFHC) diet provided 0% of energy as fat and 85% of energy as carbohydrate; the intermediate-fat, intermediate-carbohydrate (IFIC) diet provided 41% of energy as lipids and 44% of energy as carbohydrate; and the high-fat, low-carbohydrate (HFLC) diet provided 83% of energy as lipids and 2% of energy as carbohydrate. The ratio of saturated to monounsaturated to polyunsaturated lipids was 2:2:1 for the 2 diets that contained fat and all 3 diets contained 15 g fiber.
The energy requirements of each subject were assessed by a dietitian by means of a 3-d dietary journal. Liquid meals with predetermined amounts of energy were consumed at 6 fixed time points each day between 0800 and 2130 for 11 d. Compliance with the diets was assessed by measuring the respiratory quotient, which reflects the ratio of carbohydrate to fat intake (6). Respiratory quotients were measured after 10 and 11 d of the experimental diet, after an overnight fast. Subjects refrained from alcohol consumption during the experimental diets and physical activity was limited to usual daily activities. In addition to the diets, the subjects were allowed to drink only water ad libitum.
Protocol
The subjects were admitted to the Clinical Research Center and studied in the supine position. At 0700, after the subjects had fasted overnight (for 14 h), a catheter was inserted into an antecubital vein of each arm. One catheter was used to sample arterialized blood with use of a heated hand box (60°C). The other catheter was used to infuse [6,6-2H2]glucose, a 20%-glucose solution, and insulin. At 0900, after a blood sample was taken to measure the background enrichment of plasma glucose, a primed, continuous infusion of [6,6-2H2]glucose (>99% enriched; Cambridge Isotope Laboratories, Cambridge, MA) was started at a rate of 0.33 µmol•kg-1•min-1 (prime: 26.7 µmol/kg). After 150, 165, and 180 min, blood samples were drawn for measurement of basal endogenous glucose production and fatty acid concentrations. At 1200, a primed, continuous infusion of insulin (100 kU Actrapid/L; Novo Nordisk Farma BV, Zoeterwoude, Netherlands) was started at a rate of 20 mU•m body surface area-2•min-1. Plasma glucose concentrations were measured every 5 min with a Glucose Analyzer 2 (Beckman, Palo Alto, CA) and the 20%-glucose solution was infused at a variable rate to maintain euglycemia at 5.0 mmol/L. [6,6-2H2]Glucose was added to the infusate containing the 20%-glucose solution to achieve glucose enrichments of 2%. This was done to minimize changes in isotopic enrichment resulting from changes in the infusion rate of exogenous glucose and thus to allow for accurate quantification of endogenous glucose production (7, 8). Blood samples were taken for isotopic enrichment of plasma glucose and insulin and fatty acid concentrations 180, 195, and 210 min after the insulin infusion began. During the study, subjects were allowed to drink only water.
Indirect calorimetry
Oxygen consumption (
Plasma samples for glucose enrichment of [6,6-2H2]glucose were deproteinized with methanol (9). The aldonitril pentaacetate derivative of glucose (10) was injected into a gas chromatograph– mass spectrometer system (HP 6890 series II gas chromatograph equipped with a split-splitless injector and an HP 5973 model mass selective detector; Hewlett-Packard, Palo Alto, CA). Separation was achieved on a DB17 column (30 m x 0.25 mm, film thickness of 0.25 µm; J&W Scientific, Folsom, CA). Glucose was monitored at mass-to-charge ratios of 187, 188, and 189. Within each series, 3 control samples with known enrichments were measured for quality control. Glucose enrichments were calculated by dividing the area of the mass-to-charge 189 peak by total peak area. Xylose was used as an internal standard to measure glucose concentrations.
Analytic procedures
The plasma insulin concentration was determined by radioimmunoassay (Insulin RIA 100; Pharmacia Diagnostic AB, Uppsala, Sweden) with an intraassay CV of 3–5%, an interassay CV of 6–9%, and a detection limit of 15 pmol/L. Serum fatty acids were measured with an enzymatic method (NEFAC; Wako Chemicals GmbH, Neuss, Germany) with an intraassay CV of 2–4%, an interassay CV of 3–6%, and a detection limit of 0.02 mmol/L.
Calculations and statistics
When endogenous glucose production (Ra) and glucose disposal (Rd) are calculated, the added source of labeled glucose entering the system and the exogenous glucose infusate should be taken into account. Thus, Ra and Rd were calculated with a modified form of the Steele equations as described by Finegood et al (11):
RESULTS
Dietary compliance was assessed by measuring the postabsorptive respiratory quotient after 10 and 11 d of the experimental diets. The respiratory quotient decreased as the ratio of dietary fat to dietary carbohydrate increased from 0.86 ± 0.02 to 0.81 ± 0.01 to 0.73 ± 0.01 (P < 0.01).
Plasma insulin concentrations and glucose kinetics
Basal plasma insulin concentrations were 38 ± 3, 37 ± 3, and 25 ± 4 pmol/L after the LFHC, IFIC, and HFLC diets, respectively (LFHC and IFIC diets compared with the HFLC diet: P < 0.05 ). During the hyperinsulinemic clamp study, insulin concentrations were 193 ± 12, 189 ± 12, and 174 ± 8 pmol/L, respectively (NS).
Basal plasma glucose concentrations were 5.17 ± 0.17, 5.11 ± 0.11, and 4.65 ± 0.21 mmol/L, respectively (LFHC and IFIC diets compared with the HFLC diet: P < 0.05). During the hyperinsulinemic clamp, glucose concentrations were 4.9 ± 0.04, 4.9 ± 0.04, and 4.9 ± 0.07 mmol/L, respectively (NS). The rates of endogenous glucose production are presented in Table 1. Basal endogenous glucose production was inversely related to the dietary fat content. Insulin decreased endogenous glucose production in all diet groups, but was less effective after the HFLC diet (P = 0.002).
View this table:
TABLE 1. Endogenous glucose production at basal insulin concentrations and during euglycemic hyperinsulinemic clamp–induced hyperinsulinemia and the relative suppression of endogenous glucose production compared with basal values by insulin in healthy men after consumption of low-fat, high-carbohydrate; intermediate-fat, intermediate-carbohydrate; and high-fat, low-carbohydrate diets for 11 d1
Peripheral glucose metabolism
Dietary fat and carbohydrate contents had no unequivocal effect on total insulin-stimulated glucose disposal (Figure 1), but oxidative glucose disposal was significantly lower after the HFLC diet than after the other 2 diets, both at basal insulin concentrations and during hyperinsulinemia. Nonoxidative glucose disposal during the hyperinsulinemic clamp tended to increase as the dietary fat content increased (P = 0.074 between the diets).
FIGURE 1. . Mean (±SE) total, oxidative, and nonoxidative glucose disposal at basal insulin concentrations () and during hyperinsulinemia () in 6 healthy men after consumption of low-fat, high-carbohydrate; intermediate-fat, intermediate-carbohydrate; and high-fat, low-carbohydrate diets for 11 d. Means with different superscript letters are significantly different, P < 0.05. For each diet, nonoxidative glucose disposal was significantly greater during hyperinsulinemia than at basal insulin concentrations, P < 0.05.
Plasma fatty acid concentrations and fat oxidation rates
Postabsorptive fatty acid concentrations were 0.36 ± 0.05, 0.36 ± 0.04, and 0.78 ± 0.12 mmol/L with the LFHC, IFIC, and HFLC diets, respectively (LFHC and IFIC diets compared with the HFLC diet: P < 0.05). During the clamp studies, plasma fatty acid concentrations were 0.05 ± 0.01, 0.07 ± 0.01, and 0.22 ± 0.02 mmol/L, respectively (LFHC and IFIC diets compared with the HFLC diet: P < 0.05). Fat oxidation was significantly higher after the HFLC diet than after the other 2 diets at both basal insulin concentrations and during hyperinsulinemia (Figure 2). Fat oxidation was significantly lower during hyperinsulinemia than at basal insulin concentrations during the LFHC and IFIF diets but there was no significant difference between the 2 conditions during the HFLC diet.
FIGURE 2. . Mean (±SE) fat oxidation at basal insulin concentrations () and during hyperinsulinemia () in 6 healthy men after consumption of low-fat, high-carbohydrate; intermediate-fat, intermediate-carbohydrate; and high-fat, low-carbohydrate diets for 11 d. Means with different superscript letters are significantly different, P < 0.05.
DISCUSSION
The results of the present study indicate that the HFLC diet modulated insulin action on postabsorptive glucose metabolism. The HFLC diet attenuated the suppressive action of insulin on endogenous glucose production. Although fat intake had no unambiguous effect on peripheral insulin-stimulated glucose disposal, the HFLC diet suppressed the stimulatory effects of insulin on glucose oxidation almost totally, whereas it increased the effects of insulin on nonoxidative glucose disposal.
The diets used in the present study provided a wide range of fat intakes, from 0% to 83% of total energy. Because the diets were euenergetic, higher fat intakes inevitably led to lower carbohydrate intakes. This approach was used to avoid the influence of overfeeding or underfeeding on endogenous glucose production and insulin-mediated peripheral glucose metabolism. In a typical ad libitum high-fat diet, energy intake usually increases, which should be borne in mind when comparisons are made with the HFLC diet used in the present study. In addition, liquid diets contain a higher percentage of simple sugars than do solid-food diets. It was shown previously that carbohydrate composition affects fatty acid synthesis (13). Because the effects of carbohydrate composition on insulin action remain to be elucidated, caution should be exercised when making generalizations about diets consumed by the general population.
The present study showed that insulin was less effective in suppressing endogenous glucose production after the HFLC diet than after the other 2 diets. Insulin suppresses endogenous glucose production via direct effects on the liver, but also indirectly via a reduction in fatty acid concentrations (14–16). In the present study, this indirect effect of insulin was inhibited by the HFLC diet because fatty acid concentrations were suppressed less effectively during the hyperinsulinemic clamp compared with the other 2 diets. Because, in general, there is a positive relation between fatty acid concentrations and the rate of appearance of fatty acids, which reflects the rate of lipolysis, these observations indicate insulin resistance with respect to the effects of insulin on lipolysis. It is possible that the impaired action of insulin on endogenous glucose production after the HFLC diet was related to the higher fatty acid concentrations during the hyperinsulinemic clamp (17). In the postabsorptive state, both plasma insulin concentrations and endogenous glucose production were lower after the HFLC diet than after the other 2 diets. This finding suggests that hepatic insulin sensitivity increased. However, this may not merely be a reflection of altered insulin sensitivity, but rather a different mechanism—hepatic glycogen depletion (18, 19).
The effect of the dietary fat content on insulin sensitivity with respect to the effects of insulin on glucose disposal was not conclusive because insulin sensitivity was not significantly different between the high- and low-fat diets even though we established a maximum euenergetic difference in fat intake. Thus, there appears to be no dose-response relation between the dietary (euenergetic) fat content and peripheral insulin sensitivity with respect to the effects of insulin on glucose disposal. Our findings support the notion expressed in the literature that dietary fat does not directly cause peripheral insulin resistance with respect to glucose uptake (2, 3, 20).
Even though the dietary fat content did not conclusively alter total glucose disposal, there were marked effects of dietary fat content on both oxidative and nonoxidative glucose disposal. Higher dietary fat contents resulted in increased insulin-stimulated nonoxidative glucose disposal and reduced carbohydrate oxidation, suggesting that insulin stimulates glycogen synthesis more effectively when dietary fat intakes increase and carbohydrate intakes decrease. This agrees with the increase in glycogen synthase activity by insulin observed after the consumption of high-fat diets (3). In contrast, the HFLC diet inhibited the stimulatory effects of insulin on glucose oxidation. Therefore, a high-fat, low-carbohydrate diet appears to result in a dissociation with respect to the effects of insulin on oxidative and nonoxidative glucose pathways.
A high-fat, low-carbohydrate diet should cause an increase in ketone production and oxidation. Ketones that are produced, but not oxidized, generate a respiratory quotient of 0, which tends to decrease the overall respiratory quotient and result in underestimated glucose oxidation rates. To quantify the potential effect induced by this metabolic process, we measured the amount of urinary 3-hydroxybutyrate excretion during the last 24 h of the HFLC diet (4.5 ± 1.4 mmol/24 h). Assuming that this amount represents the amount of 3-hydroxybutyrate produced but not oxidized,
During the consumption of high-fat, low-carbohydrate diets, fat is the major fuel source, as evidenced by a respiratory quotient of 0.7. Similar fuel selection occurs during starvation (23). Because of this similarity it was of interest to compare the effects of high-fat, low-carbohydrate diets with the known effects of starvation on hormonal and metabolic changes. Both high-fat, low-carbohydrate diets and starvation decrease insulin concentrations, basal glucose production, and basal glucose oxidation, whereas both conditions increase lipolysis. In addition, both conditions are known to decrease insulin-stimulated glucose oxidation (24). In contrast with these similarities, there are also distinct differences between the effects of high-fat, low-carbohydrate diets and starvation. For instance, high-fat, low-carbohydrate diets do not induce peripheral insulin resistance with respect to glucose uptake and stimulate nonoxidative glucose disposal, whereas starvation reduces insulin-mediated glucose uptake (24–26) and does not increase nonoxidative glucose disposal (24). Thus, although the effects of high-fat, low-carbohydrate diets and starvation on basal fuel selection are comparable, the effects are clearly different with respect to insulin-stimulated peripheral glucose metabolism.
High fat intakes are associated with insulin resistance and type 2 diabetes. In type 2 diabetes, the most apparent defect in peripheral glucose metabolism is diminished insulin-stimulated glucose uptake (27), resulting in decreased intracellular glucose availability. Consequently, both glycogen synthesis and glucose oxidation are impaired in patients with type 2 diabetes (28). When glucose transport was experimentally increased to normal concentrations (by hyperinsulinemia or hyperglycemia), only glucose oxidation remained impaired; glycogen synthesis was restored (29, 30). In contrast with the findings for type 2 diabetes, consumption of the HFLC diet in the present study did not conclusively suppress insulin-stimulated glucose transport. Although glucose transport, measured as glucose disposal, was not affected by a high fat intake, glucose oxidation was 90% lower after the HFLC diet. As mentioned above, glucose oxidation is also impaired in diabetes, but not to the same extent (26–28%) (30). Therefore, the alterations in insulin-mediated peripheral glucose metabolism induced by the HFLC diet in the present study differed both qualitatively and quantitatively from those characteristic of type 2 diabetes.
In conclusion, diets with a high-fat, low-carbohydrate content have differential effects on insulin action. High-fat, low-carbohydrate diets impair the action of insulin on endogenous glucose production, glucose oxidation, and probably lipolysis, whereas high-fat, low-carbohydrate diets do not unequivocally affect the action of insulin on total glucose disposal and tend to enhance the action of insulin on nonoxidative glucose disposal. Despite the large differences in the fat contents of the diets studied, we could not establish a dose-response relation between dietary fat content and all aspects of insulin sensitivity. Remarkably, in the context of diabetes risk, 2 aspects of glucose homeostasis actually improved after consumption of the HFLC diet: decreased basal endogenous glucose production and improved insulin-stimulated nonoxidative glucose disposal. This observation might prove critical in the design of future studies.
REFERENCES