Effects of Different Lipid Substrates on Glucose Metabolism Postabsorptive Humans C. Broussolle,
M. Beylot,
D. Chassard,
in Normal
and B. Beaufrere
We investigated the effects on glucose metabolism of the infusion of either long-chain triglycerides (LCT), a mixture of long-chain and medium-chain triglycerides (MCT/LCT), o-@hydroxybutyrate (D-f3-OHB), or saline in normal postabsorptive subjects. Plasma insulin, C-peptide, and glucagon concentrations were unchanged in all groups. LCT and MCT/LCT infusions increased levels of plasma free fatty acids (FFA) compared with those of the saline group, whereas D-6-OHB decreased them. Plasma ketone body concentrations were higher during the o-(3-OHB and triglyceride infusions than during the saline test. Glucose concentrations and appearance (R.) and disappearance (Rd) rates were not modified during saline infusion. Glucose levels decreased only in the o-6-OHB and MCT/LCT groups (P < .05), whereas they were unchanged during LCT infusion. Glucose R, decreased slightly by 15% to 17% in LCT, MCT/LCT, and o-6-OHB groups (P < .05 v saline). Glucose Rd decreased by 14% to 16% in each lipid-infusion group (P < .05 v saline). Glucose clearance rates decreased by 14% only in the LCT group (P < .OOl). Glucose oxidation rates did not change significantly during the lipid substrate infusions compared with saline infusion. In conclusion, (1) the effects of fatty acids on glucose metabolism appear to depend on the fatty acid chain length, since only LCT infusion significantly impaired glucose utilization; and (2) in subjects with normal endocrine pancreas function, we found no adverse effects of a short-term increase in lipid substrate availability on glucose production rate and concentration. Copyright 0 1992 by W. B. Saunders Company
E
VER SINCE THE INITIAL reports of Randle et al,‘.* numerous in-vitro and in-vivo studies have demonstrated that fatty acids compete with glucose for oxidative pathways, and that excessive fatty acid availability impairs insulin-stimulated glucose utilization and/or carbohydrate oxidation.3-r1 These studies and studies showing abnormalities of fatty acid metabolism in patients with non-insulindependent diabetes mellitus (NIDDM)r2-l5 led to the proposition that excessive utilization of lipid substrates could play an important role in the pathogenesis of NIDDM.i.16 However, with the exception of the recent report of Boden and Jadali,” previous in-vivo studies investigated the effects of free fatty acids (FFA) on glucose metabolism in the presence of elevated insulin levels. To play a significant part in the pathogenesis of NIDDM, FFA should also be able to restrain glucose utilization in basal, postabsorptive conditions. Moreover, since postabsorptive hyperglycemia in NIDDM patients appears to be related to the development of increased endogenous glucose production (EGP), FFA should also have a significant effect on the other determinant of glucose level, ie, hepatic glucose production. This possibility is supported by animal studies showing that hepatic fatty acid oxidation stimulates gluconeogenesis,1R-20 and that fatty acids can inhibit in-vitro liver glycolysis2* and activate hepatic glycogen phosphorylase.22 However, the physiological relevance of these studies remains uncertain, since in adult humans, the impact of elevated FFA levels on glucose production and its response to insulin has been
From the INSERM U.197, Fact& de Medecine Alexis Carrel, Lyon; and the Service d’Endoctinologie, Hopital Edouard-Hem’ot, Lyon, France. Supported in part by the Reseau de Recherche Chnique No. 488004 (INSERM and Hospices Civils de Lyon), and by a gift from B. Braun Laboratories (Melsungen, Germany) and Solvay (Bruxelles, Belgium). Address reprint requests to C. Broussolle, MD, Pavilion SC, Centre Hospitalier Lyon-Sud, 69310 Pierre Benite, France. Copyright 0 1992 by U?B. Saunders Company 0026-0495/92/4112-0002$03.0010 1276
found to be mild7 or moderate,7~8~10*11 and it depends on the hormonal environment of the liver.7,8,11 Moreover, the metabolic effects of fatty acids might depend on their structure, particularly their carbon chain length. For example, long-chain and medium-chain fatty acids have distinct effects on leucine metabolism23+24and, in dogs, pure medium-chain triglyceride (MCT) infusion results in hypoglycemia. 25 Last, elevated plasma FFA levels result in increased ketone body levels in vivo, which by themselves have effects on glucose metabolism,26-29 and could thus interfere with the direct effect of fatty acids during in-vivo studies. Therefore, to gain further insight into the relationship between fatty acids and glucose metabolism, we investigated the effects of plasma FFA levels increased by the intravenous infusion of either long-chain triglycerides (LCT) or a .50%:50% mixture of MCT and LCT (MCT/LCT) emulsion, and the effects of a D-B-hydroxybutyrate (D-P-OHB) infusion on glucose turnover rates and carbohydrate oxidation. These studies were conducted in normal subjects under basal, postabsorptive conditions. SUBJECTS AND METHODS Subjects
Twenty-two healthy male adult volunteers were studied; their characteristics are summarized in Table 1. No subject had a family history of diabetes mellitus, and none was taking any medication. At the time of the study, they were consuming a weightmaintaining diet containing at least 2.50g carbohydrates/d. Before their participation in the study began, the nature, purpose, and potential risks of the study were explained to all subjects; a written voluntary consent was obtained. The experimental protocol was approved by the Ethical Committee of University Claude Bernard of Lyon. Materials
o-[6,6-*Hz]glucose (99% atom % excess) was obtained from the Commissariat a I’Energie Atomique (Gif-sur-Yvette, France). It was tested to be sterile and pyrogen-free, dissolved in sterile normal saline solution, and passed through 0.22~km membrane Metabolism,
Vol41, No
12 (December), 1992: pp 1276-1283
LIPID SUBSTRATES AND GLUCOSE METABOLISM
Table 1. Characteristics
of Subjects in Each Group
Plasma
Age (VI
Weight (kg1
Height (ml
BMI (kg/m?
Saline (n = 4)
24.0 f 2.3
67.5 ? 4.0
1.76 2 3.92
21.6 t 1.1
LCT (n = 6)
24.0 + 1.2
69.9 + 1.7
1.80 t- 3.37
21.5?
MCT/LCT (n = 6)
26.0 +- 1.9
72.7 2 2.9
1.78 + 1.43
22.4 + 0.7
D+-OH8
25.0 r 0.4
73.2 2 1.9
1.82 2 3.06
22.1 rt 0.3
(n = 6)
0.5
NOTE. Results are means + SEM. Abbreviation: BMI, body mass index
filters (Millipore, Bedford, MA) before administration. Twenty percent (wtivol) LCT (Endolipide) and 20% MCTiLCT (Medialipide) were kmdly provided by Bruneau (Boulogne-Billancourt, France). The composition of these two emulsions is identical in terms of egg lecithin (1.2 gi 100 mL) and glycerol (2.5 g/l00 mL), and differs only with regard to triglycerides (LCT, soy oil; MCTi LCT, 50% soy oil and 50% coconut oil). D-P-OHB was obtained from Solvay (Bruxelles, Belgium). A mixture of sodium D-p-OHB (33 g/L) and D-P-OHbutyric acid (0.33 g/L) was prepared in sterile water to obtain a final isotonic solution with pH 6.8. The solution was passed through a 0.22~km filler and tested fo be pyrogen-free before infusion to the subjects.
FFA levels were determined
Analytical Procedures Blood for metabolite determinations was collected in ice-cold perchloric acid (6% vol/vol), and lactate, glycerol, acetoacetate, and D-P-OHB were assayed by enzymatic methods, as previously described.‘” Plasma glucose concentrations were measured using an autoanalyzer (Beckman Instruments France. Gagny, France).
Plasma medi-
SfafisficalAnalysis Results are expressed as means 2 SEM. All statistical calculations were performed using a one-way or two-way ANOVA, where appropriate, and then Dunnett’s t test to locate the differences.
Protocols All subjects were studied in the postabsorptive state after a 12-hour overnight fast in the Clinical Study Unit of the Endocrinology and Nutrition Department (HBpital Edouard Herriot, Lyon, France). A retrograde catheter was inserted into a dorsal hand vein and kept patent by infusion of normal saline; this hand was kept in a hot blanket (60°C) to collect arterialized venous blood. Another forearm vein was catheterized in the contralateral arm to infuse tracer and either saline or substrate. A priming dose of D-[6,6-‘Hllglucose (13.2 pmolikg) was administered, and then o-[6,6-“H2]glucose (9.9 pmol/kg h) wascontinuously infused for 8 hours at a constant rate. Normal saline was infused during the first 3 hours of the study (basal period). It was followed during the last 5 hours by an infusion of either LCT (n = 6 subjects, 0.15 g/kg- h, LCT group), MCTiLCT (n = 6 subjects, O.I:i g/kg. h, MCT/LCT group), D-P-OHB (n = 6 subjects, 540 kmol/kg h, D -P-OHB group), or normal saline (n = 4. control group) using a volumetric infusion pump. Serial blood samples for substrate and hormone measurements and isotopic analysis were obtained before tracer infusion, during the last 40 minutes of the basal period, and then every 30 minutes during the lipid infusion and every 20 minutes during the last hour of the study. Oxygen consumption and carbon dioxide production were measured during the last 60 minutes of the basal period and during the last 80 minutes of the tests by open-circuit, indirect calorimetry using a Deltatrac Metabolic Monitor (Datex Instrumentarium, Hebsinki, Finland). Each measurement involved a period of time during which a stabilization of readings was achieved; data were then collected continuously. The subjects had been previously acclimated to the hood of the Deltatrac Metabolic Monitor. Urine samples were collected for the determination of urinary nitrogen excretion.
enzymatically.J’
um-chain fatty acid (octanoate and decanoate) concentrations were determined by gas chromatography on methylene chloride extracts of plasma, using nonanoate as the internal standard.32 Plasma insulin,33 glucagon,34 and C-peptide35 levels were measured by radioimmunoassay. Plasma o-[6,6-2HJglucose enrichments were measured using the acetyl-bisbutaneboronyl derivatives of glucose36 by selected ion-monitoring gas chromatography-mass spectrometry (Hewlett Packard 5971A-MSD, Hewlett Packard, Palo Alto, CA). The glucose turnover rate during the last 40 minutes of the basal period and during the last hour of the lipid infusion was calculated by the use of steady-state equations. Between these two periods, the rates of appearance (R,) and disappearance (Rd) of glucose were calculated using the non-steady-state approximation of Steele et al.)’ The metabolic clearance rate of glucose was calculated during steady-state periods by dividing glucose Rd by the plasma glucose concentration. Indirect calorimetry data and urinary nitrogen excretion were used to calculate carbohydrate oxidation according to the method of Frayn.3R
RESULTS
Substratesand Hormones No significant hormonal (insulin, C-peptide, and glucagon) changes occurred in any of the four groups during the study period, as shown in Table 2. During the D-P-OHB infusion, blood pH and bicarbonate levels increased from 7.41 t 0.01 and 19.8 ? 0.2 mmol/L to 7.47 ? 0.01 and 23.0 2 0.7 mmol/L, respectively (P < .Ol). Plasma total FFA concentrations (Fig 1) increased similarly during the LCT (317 t 52 to 999 + 68 kmol/L; P < ,001 1: basal and P < .Ol v saline) and MCT/LCT infusions (342 -C 39 to 1,020 ? 97 p,mol/L, P < .OOLv basal and saline). In the MCT/LCT group, the increase in FFA was mainly due to the octanoate and decanoate, which increased from < 10 to 271 2 26 and 180 ? 14 kmol/L, respectively, and represented two thirds of the total FFA increase at the end of the infusion. During the D-P-OHB infusion, FFA concentrations decreased from 452 f 135 to 122 * 42 pmol/L at 90 minutes (P < .OOl v basal and saline); they then increased progressively, but remained lower than basal values and saline values throughout the study period. Plasma glycerol concentrations (data not shown) followed the same pattern as FFA concentrations during the D-P-OHB infusion; they increased slightly during saline infusion. LCT and MCT/LCT infusions induced a sharp increase in glycerol concentrations to nearly 200 Fmol/L since both preparations contain large amounts of glycerol. The increase in total ketone body concentrations (Fig 1) was slightly but not significantly higher during LCT (82 t 13 to 514 -t 83 p,mol/L) than during saline infusion (134 2 31 to 447 c 104 p,mol/L); it was more pronounced in the MCT/LCT group (84 f 12 to 639 * 124 pmol/L, P < .0.5 v saline). As expected, total ketone body levels increased from 180 + 60 to 1,647 + 275 p,mol/L in the D-P-OHB group (P < ,001 v other groups). The P-OHB/ acetoacetate ratio remained unchanged during MCTiLCT
BROUSSOLLE
278
o 1200 2
Saline LCT
+
MCTiLCT
+
D-O-OH8 .
.
1000
g 2 s 0x
600
z.. gj
400
: I;
*
ET AL
800
200 0
-6 5
-60
0
+
60
120
180
240
300
Time(min)
Fig 1. Evolution of plasma FFA and blood total ketone body concentrations in the four tests. Results are means f SEM. tP c .05, lP c .Ol: different from the corresponding value of the saline infusion.
(2.24 k 0.42 to 2.06 k 0.5 kmol/L) and saline infusions (1.59 ? 0.15 to 1.81 2 0.26 l_t.mol/L), while it increased slightly during LCT (1.06 2 0.24 to 1.94 2 0.15 pmol/L; P < .OOl) and D-P-OHB infusions (1.68 +- 0.20 to 2.87 ‘_ 0.7 p,mol/L; P < .OOl). Plasma glucose concentrations were not significantly modified during the study period in saline (4.9 -C 0.1 to 4.6 c 0.1 mmol/L) and LCT groups (4.9 ? 0.1 to 4.7 +- 0.1 mmol/L), despite a trend for a progressive decrease. They decreased significantly during the D-P-OHB infusion (5.0 ? 0.1 to 4.5 2 0.1 mmol/L; P < .OOl v basal). In the MCT/LCT group, the decrease in plasma glucose concentrations was more pronounced (4.9 r 0.2 to 4.1 r 0.2 mmol/L, P < .OOl v basal). Figure 2 shows that in comparison with the saline group the decrease in plasma glucose concentrations was significant by 30 minutes in the D-pOHB or MCT/LCT groups, and that it remained significant throughout the study period. Plasma lactate concentrations were not modified during LCT and MCT/LCT infusions; they increased in the D-P-OHB group (565 ? 38 to 685 * 33 kmol/L; P < .02 v basal and saline; data not shown). Glucose Metabolism Glucose R, was not significantly modified compared with the baseline value of 2.21 ? 0.10 mg/kg . min in the saline group. It decreased significantly by 30 minutes during the MCT/LCT infusion from a basal value of 2.35 A 0.07 mg/ kg. min to a minimal value of 1.87 2 0.11 mg/kg . min at 280 minutes (P < .Ol). In the D-P-OHB group, glucose R, decreased by 30 minutes from a basal value of 2.35 ? 0.11 to a minimal value of 1.86 2 0.11 mgikg . min at 60 minutes
1279
LIPID SUBSTRATES AND GLUCOSE METABOLISM
a
Saline
+
MCTlLCT
+
LCT
*
II-R-OHB
-0,6
-0,8
._
a
-1,0
I0
60
120
180
240
300
Time(min) Fig 2. Variations in plasma glucose concentrations expressed as the difference in plasma glucose concentration at each time sample during the infusion and the mean glucose concentration during the basal period. Results are means -+ SEM. lP < .Ol, different from the corresponding value of the saline infusion.
(P < .Ol). In the LCT group, glucose R, decreased by 150 minutes (2.64 t 0.16 to 2.08 2 0.20 mgikg . min at 150 minutes; P < 0.01 1’ basal). Variations in glucose R,, expressed as the difference between glucose R, at each time sample during the lipid infusion and the mean glucose R, during the basal period, were significantly higher in the three lipid-substrate infusion groups than in the saline group (P < .05; Table 3). In the LCT group, glucose R, decreased by 0.56 2 0.15 mg/kg . min at time 150 minutes (P < ,001 1’ saline). In the MCT/LCT group, variations in glucose R, reached -0.51 +- 0.05 mg/kg . min at time 120 minutes (P < .Ol v saline). The D-P-OHB infusion decreased glucose R, by 0.50 t 0.16 mg/kg . min after 60 minutes (P < ,001 1, saline). Glucose Rd was not significantly modified in the saline group. In the LCT group, glucose Rd decreased by 90 minutes, reaching 2.06 ? 0.20 mgikg min after 150 minutes (P < .05 v basal). Glucose Rd decreased significantly by 120 minutes during the MCT/LCT infusion, reaching a minimal value of 1.85 ? 0.10 mgikg min (P < .Ol v basal). In the D-P-OHB group, glucose Rd decreased by 30 minutes and reached a minimal value of 1.86 ? 0.00 mgikg min at 180 minutes (P < .Ol v basal). Decreases in glucose Rd were significantly higher in the three lipid-infusion groups than in the saline group (Table 4). Glucose Rd followed roughly the same patterns as R, in the four groups of subjects. However, during the non-steadystate periods, glucose R, and Rd were not equal in all four groups. In the saline and LCT groups, the means of glucose R, (2.06 f 0.05 and 2.25 2 0.04 mg/kg . min, respectively) were not different from the means of glucose Rd (2.07 ? 0.05 and 2.26 + 0.04 mgikg . min, respectively). On the contrary, in the MCT/LCT and the D-P-OHB groups, the means of glucose R, during the infusion period (1.96 t 0.03 and 1.94 2 0.02 mgikg . min, respectively) were slightly
280
BROUSSOLLE
ET AL
lower than the means of glucose Rd (2.00 ? 0.04 and 1.97 t 0.03 mg/kg . min, respectively). Glucose metabolic clearance rate was not modified in saline and MCT/LCT groups (Fig 3). It decreased slightly but insignificantly in the D+OHB group, from 2.56 * 0.15 mL/kg . min during the basal period to 2.31 + 0.06 mL/kg - min during the last hour of the study. The decrease was more pronounced in the LCT group, from 2.96 t 0.19 during the basal period to 2.55 2 0.22 mL/kg . min during the last hour of the study (P < ,001). Glucose oxidation rates calculated by indirect calorimetry decreased similarly in LCT (1.69 ? 0.13 to 1.09 * 0.17 mg/kg . min; P < .Ol), MCTiLCT (2.09 5 0.17 to 1.35 f 0.14; P < .Ol), and D-l3-OHB (1.48 ? 0.13 to 0.92 -+ 0.17; P < .Ol) groups. These decreases were not different from that observed in the saline group (1.76 * 0.18 to 1.32 ? 0.13; P < .05) (Fig 4). DISCUSSION
In the present report, we investigated the effects of an infusion of three lipid-related substrates on glucose metabolism in healthy young volunteers in the postabsorptive state and in the presence of basal insulin levels. Metabolic substrates were modified as expected during the infusions. During D-P-OHB infusion, FFA and glycerol levels decreased and lactate concentrations increased, as previously shown.27~28~39 The infusion of MCT/LCT or LCT induced an acute elevation in plasma FFA, glycerol, and ketone body concentrations. The more pronounced increase in FFA and ketone body concentrations with MCT/LCT than with LCT infusion is consistent with the more rapid hydrolysis and oxidation of MCT.40 We failed to observe any modification in plasma pancreatic hormone concentrations throughout the studies; in particular, insulin levels remained at constant basal fasting levels. The absence of stimulation of insulin secretion during LCT infusion is in agreement with previously published results in man,5,41-43except for the recent study of Boden and Jadali,r7 who found that a 20% LCT emulsion induced a slight increase in C-peptide concentrations. However, this stimulatory effect was obtained for a higher FFA concentration than that used in our study. A stimulatory effect of ketone bodies on insulin secretion has been shown both in vitro44 and in vivo in animals,z9 and in some2’ but not a1128.45 human studies. Few
Saline
LCT
MCTILCT
D-0-OHB
Fig 3. Glucose metabolic clearance rates during the last 40 minutes of the basal period and during the last hour of the saline or lipid-related substrate infusion. Results are means f. SEM. lP < .OOl, different from basal value.
1281
LIPID SUBSTRATES AND GLUCOSE METABOLISM
”
Saline
LCT
MCTiLCT
D-R-OH6
Fig 4. Glucose oxidation during the last 40 minutes of the basal period and during the last hour of the saline or lipid-related substrate infusion. Results are means + SEM. lP c .05, l*P c .Ol: different from basal value.
are available concerning the effects of MCT on insulin secretion. A study in dogs25 did not show any modification in insulin levels following MCT infusion. The infusion of LCT, MCTILCT, or D-P-OHB induced a moderate decrease of EGP when compared with saline infusion. The results obtained with D-P-OHB are consistent with previous studies using either the nonphysiological racemic mixture of D,L-P-OHB or acetoacetate.27.28 As previously shown, alkalemia cannot be responsible for the modification of glucose production.27.28 The effects of MCTI LCT infusion also agree with results obtained previously in dogs.Z5 This decrease of EGP during D-P-OHB and MCT/ LCT infusions cannot be explained by a stimulation of insulin secretion, since we observed no increase of insulin or C-peptide levels. A possible explanation could be a decreased hepatic uptake of some gluconeogenic precursors such as lactate and pyruvate, since ketone bodies inhibit lactate removal-and glucose production-by isolated hepatocytes, probably by competing with lactate for its plasma membrane transporter.4h Our results with LCT may be consistent with studies in dogs showing an overall inhibitory effect of fatty acids on hepatic glucose production.47.3KThey disagree with the results of Boden and Jadali.” who found no decrease in glucose production during LCT-heparin infusion and a moderate increase when a pancreatic clamp was associated with LCT-heparin infusion. There is no clear explanation for this discrepancy, bul some differences in the experimental protocol should be noted. Boden and Jadali infused heparin and obtained higher FFA and glycerol concentrations, which might have resulted in different effects on glucose production. Moreover, LCT emulsions have a high proportion of unsaturated fatty acids; it is conceivable that the protocol used by Boden and Jadali resulted in a more important increase of plasma unsaturated fatty acid levels than did ours. Since unsaturdata
ated but not saturated fatty acids, activate hepatic glycogen phosphorylase, 22the effects on hepatic glucose metabolism might have been different. Conflicting results concerning the effects of increased FFA levels on glucose production in response to insulin have also been reported.7,R,10.11Taken together, these results show that the effects of fatty acids on hepatic glucose metabolism are more complicated than suggested by initial studies. In addition to the ambient insulin and glucagon concentrations,7,s these effects could also depend on the total concentration and structure of the fatty acids used. Glucose Rd decreased during LCT. MCTILCT, and D-P-OHB infusions when compared with the saline test. However, during the MCT/LCT and D-P-OHB tests, the decreased utilization rates appeared to be explained more by the moderate decrease in glucose levels than by a direct glucose-sparing effect of the infused substrates on glucoseusing tissues, since the glucose clearance rate was not significantly modified. On the contrary, the decreases in both the Rd and clearance rate during LCT infusion are in favor of a direct inhibitory effect of long-chain fatty acids on glucose utilization in the presence of basal insulin levels. These results are in agreement with the report of Boden and Jadali” and with previous studies performed during hyperinsulinemia.5-1’ This effect of LCT does not appear to be mediated through the increase in ketone body concentrations, since it was not reproduced by D-P-OHB infusion. In the postabsorptive state, muscle contributes to only approximately 20% of systemic glucose metabolism.4y In these conditions, any effect on the Randle cycle can be only minor. In conclusion, we found that LCT, MCT/LCT, and D-P-OHB infusions have different effects on basal glucose metabolism. All three substrates induced a slight decrease of EGP, whereas only LCT decreased the rate of glucose clearance. As a result, the glucose level decreased during MCT/LCT and D-P-OHB infusions and remained at control values during LCT infusion. A11 of the modifications observed were quantitatively minor and suggest that in subjects with normal endocrine pancreas secretion. a shortterm increase in lipid substrate availability has no major adverse effects on basal carbohydrate metabolism. However, the consequences can be different in subjects with abnormal p-cell function,17,5” and/or during prolonged exposure to excessive fat availability.5’) ACKNOWLEDGMENT We wish to thank M. Bruguier and A. Barb&x for their help in the preparation of D-[6.6-2H2]glucose; R. Cohen for performing the radioimmunoassays; and C. Urbain, D. Bomel, 1. Mercier, and M. OdCon for their technical assistance.
REFERENCES 1. Randle PJ, Garland PB, Hales CN, et al: The glucose fatty acid cycle: Its role in insulin sensitivity and the metabolic disturbances of diabetes mellitus. Lancet 1:785-789, 1963 2. Randle PJ, Newsholme EA, Garland PB: Regulation of glucose uptake by muscle. 8. Effects of fatty acids, ketone bodies and pyruvate, and of alloxan-diabetes and starvation, on the uptake
and the metabolic fate of glucose in rat heart and diaphragm muscles. Biochem J 93:652-665, 1964 3. Rennie MJ, Winder WW, Holloszy JO: A sparing effect of increased plasma fatty acids on muscle and liver glycogen in the exercising rat. Biochem J 156:647-655, 1976 4. Rennie MJ, Holloszy JO: Inhibition of glucose uptake and
1282
glycogenolysis by availability of oleate in well-oxygenated perfused skeletal muscle. Biochem J 168:161-170,1977 5. Thiebaud D, De Fronzo RA, Jacot E, et al: Effect of long chain triglyceride infusion on glucose metabolism in man. Metabolism 11:1128-1136,1982 6. Felley CP, Felley EM, Van Welle GD, et al: Impairment of glucose disposal by infusion of triglycerides in humans: Role of glycemia. Am J Physiol256:E747-E752,1989 7. Ferrannini E, Barrett EJ, Bevilacqua S, et al: Effect of fatty acids on glucose production and utilization in man. J Clin Invest 7211737-1747, 1983 8. Chambrier C, Picard S, Vidal H, et al: Interactions of glucagon and free fatty acids with insulin in control of glucose metabolism. Metabolism 39:976-984, 1990 9. Balasse 0, Neef MA: Operation of the “glucose-fatty acid” cycle during experimental elevations of plasma free fatty acids in man. Eur J Clin Invest 4:247-252,1974 10. Lee KU, Lee KH, Koh CS, et al: Artificial induction of intravascular lipolysis by lipid-heparin infusion leads to insulin resistance in man. Diabetologia 31:285-290,198s 11. Baron AD, Brechtel G, Edelman SV: Effects of free fatty acids and ketone bodies on in vivo non-insulin-mediated glucose utilization and production in humans. Metabolism 38:1056-1061, 1989 12. Taskinen MR, Bogardus C, Kennedy A, et al: Multiple disturbances of free fatty acid metabolism in non insulin dependent diabetes. Effect of oral hypoglycemic therapy. J Clin Invest 761637-644, 1985 13. Reaven GM, Hollenbeck C, Jeng CY, et al: Measurement of plasma glucose, free fatty acids, lactate and insulin for 24 hours in patients with NIDDM. Diabetes 37:1020-1025.1988 14. Chen YDI, Golay D, Swislocki ALM. et al: Resistance to insulin suppression of plasma free fatty acids concentrations and insulin stimulation of glucose uptake in non insulin dependent diabetes mellitus. J Clin Endocrinol Metab 64:17-21,1987 15. Groop LC, Bonnadonna RC, Del Prato S, et al: Glucose and free fatty acid metabolism in non insulin dependent diabetes. Evidence for multiple sites of insulin resistance. J Clin Invest 84:205-219,1989 16. Reaven GM: Banting Lecture 1988: Role of insulin resistance in human disease. Diabetes 37:1595-1667,198s 17. Boden G. Jadali F: Effects of lipid on basal carbohydrate metabolism in normal man. Diabetes 40:686-692, 1991 18. Ferre P, Sabatin P, El Manoubi L, et al: Relationship between ketogenesis and gluconeogenesis in isolated hepatocytes from newborn rats. Biochem J 200:429-433,198l 19. Ferre P, Pegorier JP, Williamson DH, et al: Interactions in vivo between oxidation of non esterified fatty acids and gluconeogenesis in the newborn rat. Biochem J 182593-5981979 20. Pegorier JP, Leturque A, Ferre P, et al: Effect of mediumchain triglyceride feeding on glucose homeostasis in the newborn rat. Am J Physiol244:E329-E334, 1983 21. Hue L, Malsin L, Rider MH: Palmitate inhibits liver glycolysis. Involvement of fructose 2,6 biphosphate in the glucose/fatty acid cycle. Biochem J 251:1595-1667, 1988 22. Gomez-Munoz A, Hales P, Brindley DN: Unsaturated fatty acids activate glycogen phosphorylase in cultured rat hepatocytes. Biochem J 276:209-215, 1991 23. Tessari P, Nissen SL, Miles JM, et al: Inverse relationship of leucine flux and oxidation in free fatty acid availability in vivo. J Clin Invest 77:575-581, 1986 24. Rodriguez NR, Schwenk WF, Beaufrere B, et al: Trioctanoin infusion increases in vivo leucine oxidation: A lesson in isotope modelling. Am J Physiol251:E343-E348, 1986
EROUSSOLLE
ET AL
25. Beaufrere B, Miles JM: Inhibition of hepatic glucose production by medium-chain fatty acids. Diabetes 33:47A, 1984 (abstr) 26. Robinson AM, Williamson DH: Physiological role of ketone bodies as substrates and signals in mammalian tissues. Physiol Rev 60:143-187,198O 27. Miles JM, Haymond MW, Gerich JE: Suppression of glucose production and stimulation of insulin secretion by physiological concentrations of ketone bodies in man. J Clin Endocrinol Metab 52:34-37,198l 28. Beylot M, Khalfallah Y, Riou JP, et al: Effects of ketone bodies on basal and insulin-stimulated glucose utilization in man. J Clin Endocrinol Metab 63:9-15,1986 29. Muller MJ, Paschen U, Seitz HJ: Effect of ketone bodies on glucose production and utilization in the miniature pig. J Clin Invest 74:249-256, 1984 30. Beylot M, Riou JP, Bienvenu F, et al: Increased ketonemia in hyperthyroidism. Evidence for a beta-adrenergic mechanism. Diabetologia 19:505-510, 1980 31. Okabe M, Usi Y, Nagashima K, et al: Enzyme determination of free fatty acids in serum. Clin Chem 13:476-480, 1973 32. Beaufrere B, Tessari P, Cattalini M, et al: Apparent decreased oxidation and turnover of leucine during infusion of medium chain triglycerides. Am J Physiol249:E175-E182, 1985 33. Hales CM, Randle PJ: Immunoassay of insulin with insulin antibody precipitate. Biochem J 88:137-148,1963 34. Harris J, Faloona GR, Unger RH: Glucagon, in Joffe BM, Behrman HR (eds): Methods of Hormone Radioimmunoassay. San Diego, CA, Academic, 1979. pp 643-671 35. Horwitz D, Starr J, Mako M, et al: Proinsulin, insulin, and C peptide concentrations in human portal and peripheral blood. J Clin Invest 55:1278-1283. 1975 36. Bier DM, Arnold KJ, Sherwan WR, et al: In vivo measurements of glucose and alanine metabolism with stable isotope tracers. Diabetes 26:22-27, 1977 37. Steele R, Wall JS, De Bodo C, et al: Measurement of size and turnover rate of body glucose pool by the isotope dilution method. Am J Physiol 187:15-24, 1956 38. Frayn KN: Calculation of substrate oxidation rates in vivo from gaseous exchange. J Appl Physiol55:628-634.1983 39. Shaw JHF, Wolfe RR: Influence of p-hydroxybutyrate infusion on glucose and free fatty acid metabolism in dogs. Am J Physiol247:E756-E764.1984 40. Johnson RC, Young SK, Cotter R, et al: Medium chain triglyceride lipid emulsion: Metabolism and tissue distribution. Am J Clin Nutr 52:502-508, 1990 41. Felber JP. Vannotti A: Effects of fat infusion on glucose tolerance and insulin plasma levels. Med Exp 10~153-156,1964 42. Gomez F, Jequier E, Chabot V, et al: Carbohydrate and lipid oxidation in normal human subjects: Its influence on glucose tolerance and insulin response to glucose. Metabolism 21:381-390, 1972 43. Carol1 KF, Nestel PJ: Effect of long-chain triglyceride on human insulin secretion. Diabetes 21:923-929,1972 44. Governa R, Tamarit J. Osorio J, et al: Action of betahydroxybutyrate, acetoacetate and palmitate on insulin release from the perfused isolated rat pancreas. Horm Metab Res 6:256261, 1976 45. Nair KS, Welle SL, Halliday D, et al: Effect of P-hydroxybutyrate on whole-body leucine kinetics and fractional mixed skeletal muscle protein synthesis in humans. J Clin Invest 82:198-205, 1988 46. Metcalfe HK, Monson JP, Welch SG, et al: Inhibition of lactate removal by ketone bodies in rat liver. Evidence for a quantitatively important role of the plasma membrane transporter in lactate metabolism. J Clin Invest 78743.747, 1986
LIPID SUBSTRATES
AND GLUCOSE METABOLISM
47. Seyffert WA, Madison LL: Physiologic effects of metabolic fuels on carbohydrate metabolism. I. Acute effect of elevation of plasma free fatty acids on hepatic glucose output, peripheral glucose utilization, serum insulin and plasma glucagon levels. Diabetes 16:765-776, 1967 48. Wolfe RR, Shaw JHF: Inhibitory effect of plasma free fatty acids on glucose production in the conscious dog. Am J Physiol ‘4h:E181-Elgh, 1984
1283
49. De Fronzo RA. Jacot E, Jequier E. et al: The etfect of insulin on the disposal of intravenous glucose: Results from indirect calorimetry and hepatic and femoral venous catheterization. Diabetes 30:1000-1007. 1981 50. Pascoe WS, Storlien LH: Inducement hy fat feeding of basal hyperglycemia in rats with abnormal p cell function: Model for study of etiology and pathogenesis of NIDDM. Diabetes 39:226233.1990
M. Beylot,
D. Chassard,
in Normal
and B. Beaufrere
We investigated the effects on glucose metabolism of the infusion of either long-chain triglycerides (LCT), a mixture of long-chain and medium-chain triglycerides (MCT/LCT), o-@hydroxybutyrate (D-f3-OHB), or saline in normal postabsorptive subjects. Plasma insulin, C-peptide, and glucagon concentrations were unchanged in all groups. LCT and MCT/LCT infusions increased levels of plasma free fatty acids (FFA) compared with those of the saline group, whereas D-6-OHB decreased them. Plasma ketone body concentrations were higher during the o-(3-OHB and triglyceride infusions than during the saline test. Glucose concentrations and appearance (R.) and disappearance (Rd) rates were not modified during saline infusion. Glucose levels decreased only in the o-6-OHB and MCT/LCT groups (P < .05), whereas they were unchanged during LCT infusion. Glucose R, decreased slightly by 15% to 17% in LCT, MCT/LCT, and o-6-OHB groups (P < .05 v saline). Glucose Rd decreased by 14% to 16% in each lipid-infusion group (P < .05 v saline). Glucose clearance rates decreased by 14% only in the LCT group (P < .OOl). Glucose oxidation rates did not change significantly during the lipid substrate infusions compared with saline infusion. In conclusion, (1) the effects of fatty acids on glucose metabolism appear to depend on the fatty acid chain length, since only LCT infusion significantly impaired glucose utilization; and (2) in subjects with normal endocrine pancreas function, we found no adverse effects of a short-term increase in lipid substrate availability on glucose production rate and concentration. Copyright 0 1992 by W. B. Saunders Company
E
VER SINCE THE INITIAL reports of Randle et al,‘.* numerous in-vitro and in-vivo studies have demonstrated that fatty acids compete with glucose for oxidative pathways, and that excessive fatty acid availability impairs insulin-stimulated glucose utilization and/or carbohydrate oxidation.3-r1 These studies and studies showing abnormalities of fatty acid metabolism in patients with non-insulindependent diabetes mellitus (NIDDM)r2-l5 led to the proposition that excessive utilization of lipid substrates could play an important role in the pathogenesis of NIDDM.i.16 However, with the exception of the recent report of Boden and Jadali,” previous in-vivo studies investigated the effects of free fatty acids (FFA) on glucose metabolism in the presence of elevated insulin levels. To play a significant part in the pathogenesis of NIDDM, FFA should also be able to restrain glucose utilization in basal, postabsorptive conditions. Moreover, since postabsorptive hyperglycemia in NIDDM patients appears to be related to the development of increased endogenous glucose production (EGP), FFA should also have a significant effect on the other determinant of glucose level, ie, hepatic glucose production. This possibility is supported by animal studies showing that hepatic fatty acid oxidation stimulates gluconeogenesis,1R-20 and that fatty acids can inhibit in-vitro liver glycolysis2* and activate hepatic glycogen phosphorylase.22 However, the physiological relevance of these studies remains uncertain, since in adult humans, the impact of elevated FFA levels on glucose production and its response to insulin has been
From the INSERM U.197, Fact& de Medecine Alexis Carrel, Lyon; and the Service d’Endoctinologie, Hopital Edouard-Hem’ot, Lyon, France. Supported in part by the Reseau de Recherche Chnique No. 488004 (INSERM and Hospices Civils de Lyon), and by a gift from B. Braun Laboratories (Melsungen, Germany) and Solvay (Bruxelles, Belgium). Address reprint requests to C. Broussolle, MD, Pavilion SC, Centre Hospitalier Lyon-Sud, 69310 Pierre Benite, France. Copyright 0 1992 by U?B. Saunders Company 0026-0495/92/4112-0002$03.0010 1276
found to be mild7 or moderate,7~8~10*11 and it depends on the hormonal environment of the liver.7,8,11 Moreover, the metabolic effects of fatty acids might depend on their structure, particularly their carbon chain length. For example, long-chain and medium-chain fatty acids have distinct effects on leucine metabolism23+24and, in dogs, pure medium-chain triglyceride (MCT) infusion results in hypoglycemia. 25 Last, elevated plasma FFA levels result in increased ketone body levels in vivo, which by themselves have effects on glucose metabolism,26-29 and could thus interfere with the direct effect of fatty acids during in-vivo studies. Therefore, to gain further insight into the relationship between fatty acids and glucose metabolism, we investigated the effects of plasma FFA levels increased by the intravenous infusion of either long-chain triglycerides (LCT) or a .50%:50% mixture of MCT and LCT (MCT/LCT) emulsion, and the effects of a D-B-hydroxybutyrate (D-P-OHB) infusion on glucose turnover rates and carbohydrate oxidation. These studies were conducted in normal subjects under basal, postabsorptive conditions. SUBJECTS AND METHODS Subjects
Twenty-two healthy male adult volunteers were studied; their characteristics are summarized in Table 1. No subject had a family history of diabetes mellitus, and none was taking any medication. At the time of the study, they were consuming a weightmaintaining diet containing at least 2.50g carbohydrates/d. Before their participation in the study began, the nature, purpose, and potential risks of the study were explained to all subjects; a written voluntary consent was obtained. The experimental protocol was approved by the Ethical Committee of University Claude Bernard of Lyon. Materials
o-[6,6-*Hz]glucose (99% atom % excess) was obtained from the Commissariat a I’Energie Atomique (Gif-sur-Yvette, France). It was tested to be sterile and pyrogen-free, dissolved in sterile normal saline solution, and passed through 0.22~km membrane Metabolism,
Vol41, No
12 (December), 1992: pp 1276-1283
LIPID SUBSTRATES AND GLUCOSE METABOLISM
Table 1. Characteristics
of Subjects in Each Group
Plasma
Age (VI
Weight (kg1
Height (ml
BMI (kg/m?
Saline (n = 4)
24.0 f 2.3
67.5 ? 4.0
1.76 2 3.92
21.6 t 1.1
LCT (n = 6)
24.0 + 1.2
69.9 + 1.7
1.80 t- 3.37
21.5?
MCT/LCT (n = 6)
26.0 +- 1.9
72.7 2 2.9
1.78 + 1.43
22.4 + 0.7
D+-OH8
25.0 r 0.4
73.2 2 1.9
1.82 2 3.06
22.1 rt 0.3
(n = 6)
0.5
NOTE. Results are means + SEM. Abbreviation: BMI, body mass index
filters (Millipore, Bedford, MA) before administration. Twenty percent (wtivol) LCT (Endolipide) and 20% MCTiLCT (Medialipide) were kmdly provided by Bruneau (Boulogne-Billancourt, France). The composition of these two emulsions is identical in terms of egg lecithin (1.2 gi 100 mL) and glycerol (2.5 g/l00 mL), and differs only with regard to triglycerides (LCT, soy oil; MCTi LCT, 50% soy oil and 50% coconut oil). D-P-OHB was obtained from Solvay (Bruxelles, Belgium). A mixture of sodium D-p-OHB (33 g/L) and D-P-OHbutyric acid (0.33 g/L) was prepared in sterile water to obtain a final isotonic solution with pH 6.8. The solution was passed through a 0.22~km filler and tested fo be pyrogen-free before infusion to the subjects.
FFA levels were determined
Analytical Procedures Blood for metabolite determinations was collected in ice-cold perchloric acid (6% vol/vol), and lactate, glycerol, acetoacetate, and D-P-OHB were assayed by enzymatic methods, as previously described.‘” Plasma glucose concentrations were measured using an autoanalyzer (Beckman Instruments France. Gagny, France).
Plasma medi-
SfafisficalAnalysis Results are expressed as means 2 SEM. All statistical calculations were performed using a one-way or two-way ANOVA, where appropriate, and then Dunnett’s t test to locate the differences.
Protocols All subjects were studied in the postabsorptive state after a 12-hour overnight fast in the Clinical Study Unit of the Endocrinology and Nutrition Department (HBpital Edouard Herriot, Lyon, France). A retrograde catheter was inserted into a dorsal hand vein and kept patent by infusion of normal saline; this hand was kept in a hot blanket (60°C) to collect arterialized venous blood. Another forearm vein was catheterized in the contralateral arm to infuse tracer and either saline or substrate. A priming dose of D-[6,6-‘Hllglucose (13.2 pmolikg) was administered, and then o-[6,6-“H2]glucose (9.9 pmol/kg h) wascontinuously infused for 8 hours at a constant rate. Normal saline was infused during the first 3 hours of the study (basal period). It was followed during the last 5 hours by an infusion of either LCT (n = 6 subjects, 0.15 g/kg- h, LCT group), MCTiLCT (n = 6 subjects, O.I:i g/kg. h, MCT/LCT group), D-P-OHB (n = 6 subjects, 540 kmol/kg h, D -P-OHB group), or normal saline (n = 4. control group) using a volumetric infusion pump. Serial blood samples for substrate and hormone measurements and isotopic analysis were obtained before tracer infusion, during the last 40 minutes of the basal period, and then every 30 minutes during the lipid infusion and every 20 minutes during the last hour of the study. Oxygen consumption and carbon dioxide production were measured during the last 60 minutes of the basal period and during the last 80 minutes of the tests by open-circuit, indirect calorimetry using a Deltatrac Metabolic Monitor (Datex Instrumentarium, Hebsinki, Finland). Each measurement involved a period of time during which a stabilization of readings was achieved; data were then collected continuously. The subjects had been previously acclimated to the hood of the Deltatrac Metabolic Monitor. Urine samples were collected for the determination of urinary nitrogen excretion.
enzymatically.J’
um-chain fatty acid (octanoate and decanoate) concentrations were determined by gas chromatography on methylene chloride extracts of plasma, using nonanoate as the internal standard.32 Plasma insulin,33 glucagon,34 and C-peptide35 levels were measured by radioimmunoassay. Plasma o-[6,6-2HJglucose enrichments were measured using the acetyl-bisbutaneboronyl derivatives of glucose36 by selected ion-monitoring gas chromatography-mass spectrometry (Hewlett Packard 5971A-MSD, Hewlett Packard, Palo Alto, CA). The glucose turnover rate during the last 40 minutes of the basal period and during the last hour of the lipid infusion was calculated by the use of steady-state equations. Between these two periods, the rates of appearance (R,) and disappearance (Rd) of glucose were calculated using the non-steady-state approximation of Steele et al.)’ The metabolic clearance rate of glucose was calculated during steady-state periods by dividing glucose Rd by the plasma glucose concentration. Indirect calorimetry data and urinary nitrogen excretion were used to calculate carbohydrate oxidation according to the method of Frayn.3R
RESULTS
Substratesand Hormones No significant hormonal (insulin, C-peptide, and glucagon) changes occurred in any of the four groups during the study period, as shown in Table 2. During the D-P-OHB infusion, blood pH and bicarbonate levels increased from 7.41 t 0.01 and 19.8 ? 0.2 mmol/L to 7.47 ? 0.01 and 23.0 2 0.7 mmol/L, respectively (P < .Ol). Plasma total FFA concentrations (Fig 1) increased similarly during the LCT (317 t 52 to 999 + 68 kmol/L; P < ,001 1: basal and P < .Ol v saline) and MCT/LCT infusions (342 -C 39 to 1,020 ? 97 p,mol/L, P < .OOLv basal and saline). In the MCT/LCT group, the increase in FFA was mainly due to the octanoate and decanoate, which increased from < 10 to 271 2 26 and 180 ? 14 kmol/L, respectively, and represented two thirds of the total FFA increase at the end of the infusion. During the D-P-OHB infusion, FFA concentrations decreased from 452 f 135 to 122 * 42 pmol/L at 90 minutes (P < .OOl v basal and saline); they then increased progressively, but remained lower than basal values and saline values throughout the study period. Plasma glycerol concentrations (data not shown) followed the same pattern as FFA concentrations during the D-P-OHB infusion; they increased slightly during saline infusion. LCT and MCT/LCT infusions induced a sharp increase in glycerol concentrations to nearly 200 Fmol/L since both preparations contain large amounts of glycerol. The increase in total ketone body concentrations (Fig 1) was slightly but not significantly higher during LCT (82 t 13 to 514 -t 83 p,mol/L) than during saline infusion (134 2 31 to 447 c 104 p,mol/L); it was more pronounced in the MCT/LCT group (84 f 12 to 639 * 124 pmol/L, P < .0.5 v saline). As expected, total ketone body levels increased from 180 + 60 to 1,647 + 275 p,mol/L in the D-P-OHB group (P < ,001 v other groups). The P-OHB/ acetoacetate ratio remained unchanged during MCTiLCT
BROUSSOLLE
278
o 1200 2
Saline LCT
+
MCTiLCT
+
D-O-OH8 .
.
1000
g 2 s 0x
600
z.. gj
400
: I;
*
ET AL
800
200 0
-6 5
-60
0
+
60
120
180
240
300
Time(min)
Fig 1. Evolution of plasma FFA and blood total ketone body concentrations in the four tests. Results are means f SEM. tP c .05, lP c .Ol: different from the corresponding value of the saline infusion.
(2.24 k 0.42 to 2.06 k 0.5 kmol/L) and saline infusions (1.59 ? 0.15 to 1.81 2 0.26 l_t.mol/L), while it increased slightly during LCT (1.06 2 0.24 to 1.94 2 0.15 pmol/L; P < .OOl) and D-P-OHB infusions (1.68 +- 0.20 to 2.87 ‘_ 0.7 p,mol/L; P < .OOl). Plasma glucose concentrations were not significantly modified during the study period in saline (4.9 -C 0.1 to 4.6 c 0.1 mmol/L) and LCT groups (4.9 ? 0.1 to 4.7 +- 0.1 mmol/L), despite a trend for a progressive decrease. They decreased significantly during the D-P-OHB infusion (5.0 ? 0.1 to 4.5 2 0.1 mmol/L; P < .OOl v basal). In the MCT/LCT group, the decrease in plasma glucose concentrations was more pronounced (4.9 r 0.2 to 4.1 r 0.2 mmol/L, P < .OOl v basal). Figure 2 shows that in comparison with the saline group the decrease in plasma glucose concentrations was significant by 30 minutes in the D-pOHB or MCT/LCT groups, and that it remained significant throughout the study period. Plasma lactate concentrations were not modified during LCT and MCT/LCT infusions; they increased in the D-P-OHB group (565 ? 38 to 685 * 33 kmol/L; P < .02 v basal and saline; data not shown). Glucose Metabolism Glucose R, was not significantly modified compared with the baseline value of 2.21 ? 0.10 mg/kg . min in the saline group. It decreased significantly by 30 minutes during the MCT/LCT infusion from a basal value of 2.35 A 0.07 mg/ kg. min to a minimal value of 1.87 2 0.11 mg/kg . min at 280 minutes (P < .Ol). In the D-P-OHB group, glucose R, decreased by 30 minutes from a basal value of 2.35 ? 0.11 to a minimal value of 1.86 2 0.11 mgikg . min at 60 minutes
1279
LIPID SUBSTRATES AND GLUCOSE METABOLISM
a
Saline
+
MCTlLCT
+
LCT
*
II-R-OHB
-0,6
-0,8
._
a
-1,0
I0
60
120
180
240
300
Time(min) Fig 2. Variations in plasma glucose concentrations expressed as the difference in plasma glucose concentration at each time sample during the infusion and the mean glucose concentration during the basal period. Results are means -+ SEM. lP < .Ol, different from the corresponding value of the saline infusion.
(P < .Ol). In the LCT group, glucose R, decreased by 150 minutes (2.64 t 0.16 to 2.08 2 0.20 mgikg . min at 150 minutes; P < 0.01 1’ basal). Variations in glucose R,, expressed as the difference between glucose R, at each time sample during the lipid infusion and the mean glucose R, during the basal period, were significantly higher in the three lipid-substrate infusion groups than in the saline group (P < .05; Table 3). In the LCT group, glucose R, decreased by 0.56 2 0.15 mg/kg . min at time 150 minutes (P < ,001 1’ saline). In the MCT/LCT group, variations in glucose R, reached -0.51 +- 0.05 mg/kg . min at time 120 minutes (P < .Ol v saline). The D-P-OHB infusion decreased glucose R, by 0.50 t 0.16 mg/kg . min after 60 minutes (P < ,001 1, saline). Glucose Rd was not significantly modified in the saline group. In the LCT group, glucose Rd decreased by 90 minutes, reaching 2.06 ? 0.20 mgikg min after 150 minutes (P < .05 v basal). Glucose Rd decreased significantly by 120 minutes during the MCT/LCT infusion, reaching a minimal value of 1.85 ? 0.10 mgikg min (P < .Ol v basal). In the D-P-OHB group, glucose Rd decreased by 30 minutes and reached a minimal value of 1.86 ? 0.00 mgikg min at 180 minutes (P < .Ol v basal). Decreases in glucose Rd were significantly higher in the three lipid-infusion groups than in the saline group (Table 4). Glucose Rd followed roughly the same patterns as R, in the four groups of subjects. However, during the non-steadystate periods, glucose R, and Rd were not equal in all four groups. In the saline and LCT groups, the means of glucose R, (2.06 f 0.05 and 2.25 2 0.04 mg/kg . min, respectively) were not different from the means of glucose Rd (2.07 ? 0.05 and 2.26 + 0.04 mgikg . min, respectively). On the contrary, in the MCT/LCT and the D-P-OHB groups, the means of glucose R, during the infusion period (1.96 t 0.03 and 1.94 2 0.02 mgikg . min, respectively) were slightly
280
BROUSSOLLE
ET AL
lower than the means of glucose Rd (2.00 ? 0.04 and 1.97 t 0.03 mg/kg . min, respectively). Glucose metabolic clearance rate was not modified in saline and MCT/LCT groups (Fig 3). It decreased slightly but insignificantly in the D+OHB group, from 2.56 * 0.15 mL/kg . min during the basal period to 2.31 + 0.06 mL/kg - min during the last hour of the study. The decrease was more pronounced in the LCT group, from 2.96 t 0.19 during the basal period to 2.55 2 0.22 mL/kg . min during the last hour of the study (P < ,001). Glucose oxidation rates calculated by indirect calorimetry decreased similarly in LCT (1.69 ? 0.13 to 1.09 * 0.17 mg/kg . min; P < .Ol), MCTiLCT (2.09 5 0.17 to 1.35 f 0.14; P < .Ol), and D-l3-OHB (1.48 ? 0.13 to 0.92 -+ 0.17; P < .Ol) groups. These decreases were not different from that observed in the saline group (1.76 * 0.18 to 1.32 ? 0.13; P < .05) (Fig 4). DISCUSSION
In the present report, we investigated the effects of an infusion of three lipid-related substrates on glucose metabolism in healthy young volunteers in the postabsorptive state and in the presence of basal insulin levels. Metabolic substrates were modified as expected during the infusions. During D-P-OHB infusion, FFA and glycerol levels decreased and lactate concentrations increased, as previously shown.27~28~39 The infusion of MCT/LCT or LCT induced an acute elevation in plasma FFA, glycerol, and ketone body concentrations. The more pronounced increase in FFA and ketone body concentrations with MCT/LCT than with LCT infusion is consistent with the more rapid hydrolysis and oxidation of MCT.40 We failed to observe any modification in plasma pancreatic hormone concentrations throughout the studies; in particular, insulin levels remained at constant basal fasting levels. The absence of stimulation of insulin secretion during LCT infusion is in agreement with previously published results in man,5,41-43except for the recent study of Boden and Jadali,r7 who found that a 20% LCT emulsion induced a slight increase in C-peptide concentrations. However, this stimulatory effect was obtained for a higher FFA concentration than that used in our study. A stimulatory effect of ketone bodies on insulin secretion has been shown both in vitro44 and in vivo in animals,z9 and in some2’ but not a1128.45 human studies. Few
Saline
LCT
MCTILCT
D-0-OHB
Fig 3. Glucose metabolic clearance rates during the last 40 minutes of the basal period and during the last hour of the saline or lipid-related substrate infusion. Results are means f. SEM. lP < .OOl, different from basal value.
1281
LIPID SUBSTRATES AND GLUCOSE METABOLISM
”
Saline
LCT
MCTiLCT
D-R-OH6
Fig 4. Glucose oxidation during the last 40 minutes of the basal period and during the last hour of the saline or lipid-related substrate infusion. Results are means + SEM. lP c .05, l*P c .Ol: different from basal value.
are available concerning the effects of MCT on insulin secretion. A study in dogs25 did not show any modification in insulin levels following MCT infusion. The infusion of LCT, MCTILCT, or D-P-OHB induced a moderate decrease of EGP when compared with saline infusion. The results obtained with D-P-OHB are consistent with previous studies using either the nonphysiological racemic mixture of D,L-P-OHB or acetoacetate.27.28 As previously shown, alkalemia cannot be responsible for the modification of glucose production.27.28 The effects of MCTI LCT infusion also agree with results obtained previously in dogs.Z5 This decrease of EGP during D-P-OHB and MCT/ LCT infusions cannot be explained by a stimulation of insulin secretion, since we observed no increase of insulin or C-peptide levels. A possible explanation could be a decreased hepatic uptake of some gluconeogenic precursors such as lactate and pyruvate, since ketone bodies inhibit lactate removal-and glucose production-by isolated hepatocytes, probably by competing with lactate for its plasma membrane transporter.4h Our results with LCT may be consistent with studies in dogs showing an overall inhibitory effect of fatty acids on hepatic glucose production.47.3KThey disagree with the results of Boden and Jadali.” who found no decrease in glucose production during LCT-heparin infusion and a moderate increase when a pancreatic clamp was associated with LCT-heparin infusion. There is no clear explanation for this discrepancy, bul some differences in the experimental protocol should be noted. Boden and Jadali infused heparin and obtained higher FFA and glycerol concentrations, which might have resulted in different effects on glucose production. Moreover, LCT emulsions have a high proportion of unsaturated fatty acids; it is conceivable that the protocol used by Boden and Jadali resulted in a more important increase of plasma unsaturated fatty acid levels than did ours. Since unsaturdata
ated but not saturated fatty acids, activate hepatic glycogen phosphorylase, 22the effects on hepatic glucose metabolism might have been different. Conflicting results concerning the effects of increased FFA levels on glucose production in response to insulin have also been reported.7,R,10.11Taken together, these results show that the effects of fatty acids on hepatic glucose metabolism are more complicated than suggested by initial studies. In addition to the ambient insulin and glucagon concentrations,7,s these effects could also depend on the total concentration and structure of the fatty acids used. Glucose Rd decreased during LCT. MCTILCT, and D-P-OHB infusions when compared with the saline test. However, during the MCT/LCT and D-P-OHB tests, the decreased utilization rates appeared to be explained more by the moderate decrease in glucose levels than by a direct glucose-sparing effect of the infused substrates on glucoseusing tissues, since the glucose clearance rate was not significantly modified. On the contrary, the decreases in both the Rd and clearance rate during LCT infusion are in favor of a direct inhibitory effect of long-chain fatty acids on glucose utilization in the presence of basal insulin levels. These results are in agreement with the report of Boden and Jadali” and with previous studies performed during hyperinsulinemia.5-1’ This effect of LCT does not appear to be mediated through the increase in ketone body concentrations, since it was not reproduced by D-P-OHB infusion. In the postabsorptive state, muscle contributes to only approximately 20% of systemic glucose metabolism.4y In these conditions, any effect on the Randle cycle can be only minor. In conclusion, we found that LCT, MCT/LCT, and D-P-OHB infusions have different effects on basal glucose metabolism. All three substrates induced a slight decrease of EGP, whereas only LCT decreased the rate of glucose clearance. As a result, the glucose level decreased during MCT/LCT and D-P-OHB infusions and remained at control values during LCT infusion. A11 of the modifications observed were quantitatively minor and suggest that in subjects with normal endocrine pancreas secretion. a shortterm increase in lipid substrate availability has no major adverse effects on basal carbohydrate metabolism. However, the consequences can be different in subjects with abnormal p-cell function,17,5” and/or during prolonged exposure to excessive fat availability.5’) ACKNOWLEDGMENT We wish to thank M. Bruguier and A. Barb&x for their help in the preparation of D-[6.6-2H2]glucose; R. Cohen for performing the radioimmunoassays; and C. Urbain, D. Bomel, 1. Mercier, and M. OdCon for their technical assistance.
REFERENCES 1. Randle PJ, Garland PB, Hales CN, et al: The glucose fatty acid cycle: Its role in insulin sensitivity and the metabolic disturbances of diabetes mellitus. Lancet 1:785-789, 1963 2. Randle PJ, Newsholme EA, Garland PB: Regulation of glucose uptake by muscle. 8. Effects of fatty acids, ketone bodies and pyruvate, and of alloxan-diabetes and starvation, on the uptake
and the metabolic fate of glucose in rat heart and diaphragm muscles. Biochem J 93:652-665, 1964 3. Rennie MJ, Winder WW, Holloszy JO: A sparing effect of increased plasma fatty acids on muscle and liver glycogen in the exercising rat. Biochem J 156:647-655, 1976 4. Rennie MJ, Holloszy JO: Inhibition of glucose uptake and
1282
glycogenolysis by availability of oleate in well-oxygenated perfused skeletal muscle. Biochem J 168:161-170,1977 5. Thiebaud D, De Fronzo RA, Jacot E, et al: Effect of long chain triglyceride infusion on glucose metabolism in man. Metabolism 11:1128-1136,1982 6. Felley CP, Felley EM, Van Welle GD, et al: Impairment of glucose disposal by infusion of triglycerides in humans: Role of glycemia. Am J Physiol256:E747-E752,1989 7. Ferrannini E, Barrett EJ, Bevilacqua S, et al: Effect of fatty acids on glucose production and utilization in man. J Clin Invest 7211737-1747, 1983 8. Chambrier C, Picard S, Vidal H, et al: Interactions of glucagon and free fatty acids with insulin in control of glucose metabolism. Metabolism 39:976-984, 1990 9. Balasse 0, Neef MA: Operation of the “glucose-fatty acid” cycle during experimental elevations of plasma free fatty acids in man. Eur J Clin Invest 4:247-252,1974 10. Lee KU, Lee KH, Koh CS, et al: Artificial induction of intravascular lipolysis by lipid-heparin infusion leads to insulin resistance in man. Diabetologia 31:285-290,198s 11. Baron AD, Brechtel G, Edelman SV: Effects of free fatty acids and ketone bodies on in vivo non-insulin-mediated glucose utilization and production in humans. Metabolism 38:1056-1061, 1989 12. Taskinen MR, Bogardus C, Kennedy A, et al: Multiple disturbances of free fatty acid metabolism in non insulin dependent diabetes. Effect of oral hypoglycemic therapy. J Clin Invest 761637-644, 1985 13. Reaven GM, Hollenbeck C, Jeng CY, et al: Measurement of plasma glucose, free fatty acids, lactate and insulin for 24 hours in patients with NIDDM. Diabetes 37:1020-1025.1988 14. Chen YDI, Golay D, Swislocki ALM. et al: Resistance to insulin suppression of plasma free fatty acids concentrations and insulin stimulation of glucose uptake in non insulin dependent diabetes mellitus. J Clin Endocrinol Metab 64:17-21,1987 15. Groop LC, Bonnadonna RC, Del Prato S, et al: Glucose and free fatty acid metabolism in non insulin dependent diabetes. Evidence for multiple sites of insulin resistance. J Clin Invest 84:205-219,1989 16. Reaven GM: Banting Lecture 1988: Role of insulin resistance in human disease. Diabetes 37:1595-1667,198s 17. Boden G. Jadali F: Effects of lipid on basal carbohydrate metabolism in normal man. Diabetes 40:686-692, 1991 18. Ferre P, Sabatin P, El Manoubi L, et al: Relationship between ketogenesis and gluconeogenesis in isolated hepatocytes from newborn rats. Biochem J 200:429-433,198l 19. Ferre P, Pegorier JP, Williamson DH, et al: Interactions in vivo between oxidation of non esterified fatty acids and gluconeogenesis in the newborn rat. Biochem J 182593-5981979 20. Pegorier JP, Leturque A, Ferre P, et al: Effect of mediumchain triglyceride feeding on glucose homeostasis in the newborn rat. Am J Physiol244:E329-E334, 1983 21. Hue L, Malsin L, Rider MH: Palmitate inhibits liver glycolysis. Involvement of fructose 2,6 biphosphate in the glucose/fatty acid cycle. Biochem J 251:1595-1667, 1988 22. Gomez-Munoz A, Hales P, Brindley DN: Unsaturated fatty acids activate glycogen phosphorylase in cultured rat hepatocytes. Biochem J 276:209-215, 1991 23. Tessari P, Nissen SL, Miles JM, et al: Inverse relationship of leucine flux and oxidation in free fatty acid availability in vivo. J Clin Invest 77:575-581, 1986 24. Rodriguez NR, Schwenk WF, Beaufrere B, et al: Trioctanoin infusion increases in vivo leucine oxidation: A lesson in isotope modelling. Am J Physiol251:E343-E348, 1986
EROUSSOLLE
ET AL
25. Beaufrere B, Miles JM: Inhibition of hepatic glucose production by medium-chain fatty acids. Diabetes 33:47A, 1984 (abstr) 26. Robinson AM, Williamson DH: Physiological role of ketone bodies as substrates and signals in mammalian tissues. Physiol Rev 60:143-187,198O 27. Miles JM, Haymond MW, Gerich JE: Suppression of glucose production and stimulation of insulin secretion by physiological concentrations of ketone bodies in man. J Clin Endocrinol Metab 52:34-37,198l 28. Beylot M, Khalfallah Y, Riou JP, et al: Effects of ketone bodies on basal and insulin-stimulated glucose utilization in man. J Clin Endocrinol Metab 63:9-15,1986 29. Muller MJ, Paschen U, Seitz HJ: Effect of ketone bodies on glucose production and utilization in the miniature pig. J Clin Invest 74:249-256, 1984 30. Beylot M, Riou JP, Bienvenu F, et al: Increased ketonemia in hyperthyroidism. Evidence for a beta-adrenergic mechanism. Diabetologia 19:505-510, 1980 31. Okabe M, Usi Y, Nagashima K, et al: Enzyme determination of free fatty acids in serum. Clin Chem 13:476-480, 1973 32. Beaufrere B, Tessari P, Cattalini M, et al: Apparent decreased oxidation and turnover of leucine during infusion of medium chain triglycerides. Am J Physiol249:E175-E182, 1985 33. Hales CM, Randle PJ: Immunoassay of insulin with insulin antibody precipitate. Biochem J 88:137-148,1963 34. Harris J, Faloona GR, Unger RH: Glucagon, in Joffe BM, Behrman HR (eds): Methods of Hormone Radioimmunoassay. San Diego, CA, Academic, 1979. pp 643-671 35. Horwitz D, Starr J, Mako M, et al: Proinsulin, insulin, and C peptide concentrations in human portal and peripheral blood. J Clin Invest 55:1278-1283. 1975 36. Bier DM, Arnold KJ, Sherwan WR, et al: In vivo measurements of glucose and alanine metabolism with stable isotope tracers. Diabetes 26:22-27, 1977 37. Steele R, Wall JS, De Bodo C, et al: Measurement of size and turnover rate of body glucose pool by the isotope dilution method. Am J Physiol 187:15-24, 1956 38. Frayn KN: Calculation of substrate oxidation rates in vivo from gaseous exchange. J Appl Physiol55:628-634.1983 39. Shaw JHF, Wolfe RR: Influence of p-hydroxybutyrate infusion on glucose and free fatty acid metabolism in dogs. Am J Physiol247:E756-E764.1984 40. Johnson RC, Young SK, Cotter R, et al: Medium chain triglyceride lipid emulsion: Metabolism and tissue distribution. Am J Clin Nutr 52:502-508, 1990 41. Felber JP. Vannotti A: Effects of fat infusion on glucose tolerance and insulin plasma levels. Med Exp 10~153-156,1964 42. Gomez F, Jequier E, Chabot V, et al: Carbohydrate and lipid oxidation in normal human subjects: Its influence on glucose tolerance and insulin response to glucose. Metabolism 21:381-390, 1972 43. Carol1 KF, Nestel PJ: Effect of long-chain triglyceride on human insulin secretion. Diabetes 21:923-929,1972 44. Governa R, Tamarit J. Osorio J, et al: Action of betahydroxybutyrate, acetoacetate and palmitate on insulin release from the perfused isolated rat pancreas. Horm Metab Res 6:256261, 1976 45. Nair KS, Welle SL, Halliday D, et al: Effect of P-hydroxybutyrate on whole-body leucine kinetics and fractional mixed skeletal muscle protein synthesis in humans. J Clin Invest 82:198-205, 1988 46. Metcalfe HK, Monson JP, Welch SG, et al: Inhibition of lactate removal by ketone bodies in rat liver. Evidence for a quantitatively important role of the plasma membrane transporter in lactate metabolism. J Clin Invest 78743.747, 1986
LIPID SUBSTRATES
AND GLUCOSE METABOLISM
47. Seyffert WA, Madison LL: Physiologic effects of metabolic fuels on carbohydrate metabolism. I. Acute effect of elevation of plasma free fatty acids on hepatic glucose output, peripheral glucose utilization, serum insulin and plasma glucagon levels. Diabetes 16:765-776, 1967 48. Wolfe RR, Shaw JHF: Inhibitory effect of plasma free fatty acids on glucose production in the conscious dog. Am J Physiol ‘4h:E181-Elgh, 1984
1283
49. De Fronzo RA. Jacot E, Jequier E. et al: The etfect of insulin on the disposal of intravenous glucose: Results from indirect calorimetry and hepatic and femoral venous catheterization. Diabetes 30:1000-1007. 1981 50. Pascoe WS, Storlien LH: Inducement hy fat feeding of basal hyperglycemia in rats with abnormal p cell function: Model for study of etiology and pathogenesis of NIDDM. Diabetes 39:226233.1990
