Effect of difference in glucose infusion rate on quantification of hepatic glucose production.

0021-972X/90/7005-1354$02.00/0 Journal of Clinical Endocrinology and Metabolism Copyright © 1990 by The Endocrine Society

Vol. 70, No. 5 Printed in U.S.A.

Effect of Difference in Glucose Infusion Rate on Quantification of Hepatic Glucose Production* WAYNE H.-H. SHEUf, CLARIE B. HOLLENBECK, MIN-SHUNG WU, JOHNATHAN B. JASPAN, Y.-D. IDA CHEN, AND GERALD M. REAVEN Department of Medicine, Stanford University School of Medicine, and Geriatric Research, Education, and Clinical Center, Veterans Administration Medical Center, Palo Alto, California 94304

ABSTRACT. This study was carried out to determine whether hepatic glucose prodution (HGP) could be suppressed in normal subjects by infusing different amounts of glucose, in the absence of significant changes in steady state plasma glucose (SSPG) or insulin (SSPI) concentrations. Conseqently, subjects were infused with somatostatin (215 nmol/h), insulin (28.7 pmol/m2min), and amounts of glucose varying from 0-200 ^mol/m 2 min in the absence or presence of glucagon (5.2 pmol/m2-min). SSPI concentrations were constant (60-70 pmol/L) during these studies, and values for the total glucose appearance rate (glucose

A

LTHOUGH resistance to insulin-stimulated glucose uptake has been well recognized (1-4) as a characteristic defect in patients with noninsulin-dependent diabetes mellitus (NIDDM) for some time, a consensus as to the ability of insulin to suppress hepatic glucose (HGP) has been slower to develop. However, it now appears that a defect in insulin suppression of HGP can also be discerned during glucose clamp studies (5-8), particularly when they are performed at somewhat lower insulin concentrations. Indeed, it has been more recently suggested that hepatic and peripheral tissues are equally resistant to insulin in patients with NIDDM (8). Conclusions as to the effect of NIDDM on hepatic insulin resistance depend upon accurate measurements of the total glucose turnover rate (Ra) under basal conditions and after an increase in plasma insulin concentrations during glucose clamp studies. Basal HGP is assumed to equal the basal glucose turnover rate, while HGP during the glucose clamp study is determined by subtracting the glucose infusion rate needed to maintain a constant plasma glucose concentration during the study from the Received September 7, 1989. Address all correspondence and requests for reprints to: G. M. Reaven, M.D., Geriatric Research, Education, and Clinical Center 182B, Veterans Administration Medical Center, 3801 Miranda Avenue, Palo Alto, California 94304. * This work was supported by research grants from the NIH (RR00070 and DK-30732) and the Nora Eccles Treadwell Foundation. t Postdoctoral Scholar, Department of Medicine, Tri-Service General Hospital, Taipei, Taiwan, Republic of China.

infusion rate plus HGP) and SSPG did not vary significantly as

a function of the rate of exogenous glucose infusion. However, values for HGP fell in response to increases in glucose infusion rate and could be suppressed to approximately 50% of the original value despite the fact that SSPG, SSPI, and glucose appearance rate did not change significantly. These data indicate that HGP can be regulated by varying the rate of exogenous glucose infusion during glucose clamp studies. (J Clin Endocrinol Metab 70: 1354-1360, 1990)

total glucose turnover rate. Ra is determined during glucose clamp studies by infusing [3H] 3-glucose and calculating the turnover rate by conventional isotope dilution methods, using the nonsteady state equations of Steele (9). We have previously questioned the ability of this approach to generate accurate estimates of basal HGP during the period of time conventionally used to make these measurements (10, 11). The current study was initiated to challenge a basic assumption underlying the measurement of HGP during glucose clamp studies, namely that differences in the rate at which the variable glucose infusion is given do not affect quantification of HGP as long as plasma glucose and insulin concentrations are kept relatively constant. Glucose must be infused at a considerably greater rate in normal subjects compared to insulin-resistant patients with NIDDM in order to maintain a constant plasma glucose concentration during the clamp study. This difference in the rate of glucose infused is ignored when HGP is calculated as long as plasma glucose and insulin concentrations during the clamp study are comparable. We have examined this premise in normal individuals, and the results presented below strongly suggest that HGP can be suppressed by simply raising the rate of glucose infusion at a time when plasma glucose and insulin concentrations did not vary significantly.

Materials and Methods The study population consisted of 13 volunteers, with normal glucose tolerance, as defined by the National Diabetes Data

1354

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 November 2015. at 07:03 For personal use only. No other uses without permission. . All rights reserved.

HGP AND GLUCOSE INFUSION RATE Group (12). Their age ranged from 35-65 yr, and their body mass index from 21.5-27.9 kg/m2. All subjects were in good health, and none was taking any medications known to affect carbohydrate metabolism. This study was approved by the Stanford University Human Subjects Committee, and each individual signed a consent form upon admission to the Stanford General Clinical Research Center. Two series of experiments were performed. In the first series, infusion studies were performed on two occasions in seven individuals. The basic protocol consisted of infusing insulin (28.7 pmol/m2-min) plus somatostatin (SRIF; 215 nmol/h) and either zero glucose or 111 /anol/m2• min glucose for 4 h. In four of the subjects a third infusion consisting of insulin, SRIF, and glucose at a rate of 200 ftmol/m2 • min was performed. A second series of experiments was carried out in six individuals, in which a constant infusion of glucagon (5.2 pmol/m2-min) was given. In these studies glucose was infused at a rate of either 0 or 183 jimol/m2 • min. Blood samples were obtained every 30 min for the first 210 min and every 10 min during the last 30 min from an indwelling catheter in a hand vein, kept in a radiant warmer at 70 C to provide arterialized samples. Plasma was separated immediately, frozen, and subsequently analyzed for measurement of plasma glucose (13) and insulin (14) concentrations. To quantify total glucose turnover, 30 fiCi [3H]3-glucose were injected as an iv bolus dose at time zero, followed by a constant infusion of 0.38 /iCi/min for the remainder of the study. Aliquots of plasma were obtained as described above and precipitated with Ba(0H) 2 and ZnSO4, and the plasma glucose concentration and radioactivity were measured in the protein-free supernatant. The specific activity of radioactive glucose was then calculated, and Ra was determined, using the equation of Steele to correct for nonsteady state conditions (9). This equation describes a single compartment model, with a total volume of distribution for glucose of 134 mL (67% of total extracellular water space, 200 mL). The model is valid in normal subjects only after 2-h tracer equilibrium to reach a quasisteady state condition (11). Statistical analysis Data are expressed as the mean ± SEM and were analyzed by the Statistical Analysis System (SAS) program (SAS, Inc., Cary, NC), using the general linear models procedure. Comparison of insulin, glucagon, glucose, and Ra values in response to the different infusions was achieved by three-way analysis of variance, with the factors being patient, time, and infusion rate.

Results Mean (±SEM) plasma glucose concentrations throughout the 4-h infusion studies performed in the absence (left panel) or presence (right panel) of glucagon are shown in Fig. 1. These data show that the plasma glucose concentration fell below the basal value during both studies carried out in the absence of glucagon replacement. However, it can be seen that increasing the glucose infusion rate from 0 to 111 /umol/m2• min had a relatively small effect on the resultant plasma glucose concentra-

1355

tion. When the infusion were carried out in the presence of glucagon, plasma glucose concentrations were higher than the basal value. However, the increments in plasma concentration above basal were relatively similar when the glucose infusion rate was 0 or 183 jumol/m2 • min. Since steady state conditions were reached during the last 60 min of each study, seven measurements of plasma glucose concentration from 180-240 min were averaged to obtain the steady state plasma glucose (SSPG) concentration. These data appear in Fig. 2 and emphasize that the SSPG concentrations reached in response to the glucose infusion were comparable to those achieved when glucose was not infused. Steady state insulin concentrations were essentially identical in all four studies, ranging from 57 ± 7 to 65 ± 8 pM. Plasma glucagon rose to 32 ± 1 and 35 ± 1 pM during the two studies in which it was infused (Table 1). HGP values from 90-240 min during the infusion studies are illustrated in Fig. 3. These results demonstrate that HGP was lower throughout this period when glucose was infused in either the absence (left panel) or presence (right panel) of glucagon. Although absolute values were higher when glucagon was added to the infusate, the decrement in HGP seen when the glucose infusion rate was increased from 0 to 183 /xmol/m2 • min was even greater than when it was increased from 0 to 111 /imol/m2• min in the absence of glucagon. In four volunteers we were able to carry out three studies, infusing glucose at rates of 0, 100, and 200 /umol/ m2 • min. These measurements were made in the absence of glucagon, and the plasma glucose concentrations during the three infusions are shown in Fig. 4. As before, steady state conditions were reached by 180 min, and the SSPG and steady state plasma insulin (SSPI) concentrations during the last 60 min are shown in Fig. 5. Mean (±SEM) values for HGP during these three studies from 90-240 min are illustrated in Fig. 6. It is apparent that HGP fell progressively as the glucose infusion rate was increased from 0 to 100 to 200 Aimol/m2 • min. Thus, we were able to suppress HGP in these normal individuals to less than half the original value without any substantial change in SSPG or SSPI by simply increasing the rate at which exogenous glucose was infused.

Discussion The purpose of this study was to determine whether HGP could be suppressed, independent of significant changes in steady state plasma glucose or insulin concentrations, simply by raising the rate at which glucose was continuously infused into normal volunteers. The data presented in Figs. 1-6 illustrate that this goal was achieved, demonstrating that HGP decreased progressively as the glucose infusion rate was increased. Before


SHEU ET AL.

1356

JCE & M • 1990 Vol70«No5

10

10

+ Glucagon

- Glucagon

GInf

"Inf o Zero • 111 umol/m2/min

o Zero • 183|xmol/m2/min

•

0 U_i_

60

180

120

240

•

I

60

.

,

I

I

I

I

120

•

.

•

I

180

I

I

.

240

Time (min)

Time (min)

FIG. 1. Mean (±SEM) plasma glucose concentrations during 4-h glucose infusion studies at glucose infusion rates (Ginf) of 0, 111, or 183 /imol/m2min. Infusion studies were performed without {left panel) and with {right panel) glucagon. - Glucagon

FIG. 2. Mean (±SEM) SSPG concentrations in response to glucose infusion rates (GInf) of 0, 111, or 183 Mmol/m2min. Infusion studies were performed without {left panel) and with {right panel) glucagon replacement.

+ Glucagon

8 8

G•"Inf lr%l0

commenting upon the significance of these findings, it is necessary to discuss some of the practical aspects of the experiments presented. In the first place, the degree of variation in the SSPG (coefficient of variance, 5.1 ± 0.2%) and SSPI (coefficient of variance, 2.3 ± 0.3%) concentrations was similar during all studies, and values were within the limits that most investigators agree are acceptable during clamp studies. In addition, when glu-

G.,,,111 Inf

cagon was infused in order to achieve basal plasma concentrations during the infusions, the plasma glucagon concentrations were similar. Thus, there were no apparent differences in the plasma concentrations of the substrate and the two hormones thought to be the major regulators of HGP. Despite this, HGP decreased progressively as the glucose infusion rate was increased. If it is agreed that the infusion studies were technically


HGP AND GLUCOSE INFUSION RATE TABLE 1. Mean (±SEM) SSPI and SSPG concentrations during the last 60 min of the infusion studies Condition

Glucose infusion rate

Insulin

(/imol/m2 • min)

(pM)

—Glucagon

57 65 58 62

0 111 0 183

+Glucagon

Glucagon ( P M)

±6 ±8 ±3 ±4

22 22 32 35

±2 ±3 ±1 ±1

adequate, it is necessary to explain how an increase in the rate of glucose infusion could bring about the inhibition of HGP. The most obvious possibility is that the rate of utilization of glucose by the liver increased as the glucose infusion rate was raised, secondary to a decrease in hepatic phosphorylase activity and/or an increase in glycogen synthase activity. There is certainly abundant evidence that an increase in hepatic glucose utilization can lead to the enzyme changes postulated (15, 16), and it is not surprising that HGP would be suppressed if these events transpired. However, it is not immediately apparent how hepatic glucose utilization could increase in response to an increase in the glucose infusion rate, with little or no change in the SSPG concentration. The only answer to this question is that SSPG must tend to increase as the glucose infusion rate is increased, but this leads to such an increase in the hepatic glucose utilization rate that substantial changes in SSPG cannot be discerned. In other words, a considerable increment in

hepatic glucose utilization can occur in response to small increases in the plasma glucose concentration. However, the changes in plasma glucose concentration are relatively small in magnitude and difficult to demonstrate during conventional glucose clamp studies. At any given insulin concentration a point will be reached when the ability of the individual to respond to an increase in the glucose infusion rate by simultaneously increasing glucose uptake and decreasing HGP to the degree needed to prevent a rise in SSPG will be exceeded. When this happens, a substantial increase in SSPG concentration will be observed. However, this point was never reached within the range of glucose infusion rates used in this study, and HGP was progressively suppressed with little or no change in SSPG as the rate of glucose infusion was increased. Although we believe that the formulation outlined above offers a credible explanation for why HGP fell progressively as the glucose infusion rate was increased, the implication of our observations is not dependent upon acceptance of the hypothesis presented. These studies were initiated to examine the assumption that estimates of HGP would be constant, regardless of an increase in the glucose infusion rate, as long as plasma glucose and insulin concentrations were relatively comparable. The results presented demonstrate that this was not the case, no matter what explanation is used to account for the phenomena described in Figs. 1-6. Our 600

600

+ Glucagon

- Glucagon

500

500

400

400

I 300

300

200

f200

.£

1357

E

100

o Zero • 111 u.mol/m2/min

90

120

150

180

Time (min)

210

" T"

i

100

240

i r oZero • 183 u,mol/m2/min

90

120

150

180

210

240

Time (min)

FIG. 3. Mean (±SEM) HGP in response to glucose infusion rates (Ginf) of 0, 111, or 183 jmiol/m2-min. Infusion studies were performed without (left panel) and with (right panel) glucagon replacement.


SHEU ET AL.

1358

JCE & M • 1990 Vol70«No5

10

o Zero 0 m 111 (imol/m^/min • 200 |imol/rrr/min

CD O O

.2

4

I

120

60

180

I

I

I

240

Time (min) FIG. 4. Mean (±SEM) plasma glucose concentrations during 4-h glucose infusion studies at glucose infusion rates (Ginf) of 0,100, or min in the absence of glucagon replacement.

SSPI

SSPG

FIG. 5. Mean (±SEM) SSPG and SSPI concentrations at glucose infusion rates (Ginf) of 0,100, and 200 Mmol/m2 • min in the absence of glucagon replacement.

O 2

G,n,°

observation that HGP decreased in response to progressive increases in the glucose infusion rate, despite comparable values for SSPG and SSPI concentrations, is not totally unexpected given the results of similar studies recently carried out in humans (17, 18) and swine (19). The protocols and experimental results in both of these studies were comparable to ours, and it was concluded that substantial decrements in HGP can occur with little

G...111 "Inf

G.,.,200 Inr

or no increase in SSPG. Based upon their results, both Wolfe et al. (17,18) and Miiller et al. (19) concluded that glucose itself has the ability to suppress HGP. It is obvious that the current results as well as those of Wolfe and associates (17, 18) and Muller and colleagues (19) raise serious questions as to the significance of measurements of HGP made during conventional glucose clamp studies. Whether an increase in the glucose


HGP AND GLUCOSE INFUSION RATE

1359

400 r~

300

200 CL

O 100

o Zero • 111 [imol/rrr/min • 200 umol/rrr/min I

90

120

180

150

,

•

I 210

240

Time (min) FIG. 6. Mean (±SEM) HGP during 4-h glucose infusion studies at glucose infusion rates (Ginr) of 0, 100, or 200 fimol/m2• min in the absence of glucagon replacement.

infusion rate can suppress HGP without any change in SSPG, or whether a very small increase in SSPG is required to accomplish this task does not seem to be the crucial issue. In either case, all of these data demonstrate that the glucose infusion rate per se is an important regulator of HGP and raise serious questions concerning our ability to interpret the results of glucose clamp studies aimed at estimating suppression of HGP. When a conventional euglycemic hyperinsulinemic clamp is performed in individuals with widely different degrees of insulin sensitivity, the greater the degree of insulin resistance, the lower the glucose infusion rate. If HGP were suppressed to a variable degree in these individuals, the conventional conclusion would be that this was due to differences in hepatic sensitivity to insulin. However, the data presented in this paper demonstrate that this need not be the interpretation, and the variations in HGP among these individuals could just as easily be a reflection of the differences in the glucose infusion rate. In conclusion, the results presented here have demonstrated that variations in the glucose infusion rate during clamp studies can lead to differences in HGP, despite the fact that SSPI and SSPG concentrations were comparable. As a result of these findings it is necessary to question the premise that comparisons of HGP between individuals is possible as long as the SSPG and SSPI concentrations are maintained comparable from patient to patient. Perhaps these results can be best summarized by Wolfe and colleagues (18), who

commented upon "the precision with which hepatic glucose output was suppressed to an amount exactly equal to the infusion rate," and emphasized "the sensitivity of the liver to small changes in the concentration of glucose in regulation of glucose output." As a consequence, previous estimates of the suppressibility of HGP, based upon the assumption that differences in glucose infusion rate during the study are irrelevant, are very likely confounded by ignoring this variable. It seems obvious that new approaches to measurement of HGP, taking into account the effect of differences in glucose infusion rate, be developed. Our laboratory is now planning to address this issue, and we hope that the results presented in this communication encourage other groups to join in this important task. References 1. Shen S, Reaven G, Farquhar J. Comparison of impedance to insulin-mediated glucose uptake in normal subjects and subjects with latent diabetes. J Clin Invest. 1980;49:2151-60. 2. Ginsberg H, Kimmerling G, Olefsky JM, Reaven GM. Demonstration of insulin resistance in untreated adult onset diabetic subjects with hyperglycemia. J Clin Invest. 1975;55:454-61. 3. Reaven GM. Insulin resistance in noninsulin-dependent diabetes mellitus: does it exist and can it be measured? Am J Med. 1983;74:3-17. 4. Reaven GM, Chen Y-DI, Donner CC, Fraze E, Hollenbeck CB. How insulin resistant are patients with noninsulin-dependent diabetes mellitus? J Clin Endocrinol Metab. 1985;61:32-6. 5. Kolterman O, Gray R, Griffin J, et al. Receptor and postreceptor defects contribute to the insulin resistance in noninsulin-dependent diabetes mellitus. J Clin Invest. 1981;68:957-69.


1360

SHEU ET AL.

6. Bogardus C, Lillioja S, Howard B, et al. Relationships between insulin secretion, insulin action and fasting plasma glucose concentration in nondiabetic and noninsulin-dependent diabetic subjects. J Clin Invest. 1984;74:1238-46. 7. Revers R, Fink R, Griffin J, et al. Influence of hyperglycemia on insulin's in vivo effects in type II diabetes. J Clin Invest. 1984;73:664-72. 8. Campbell PJ, Mandarino LJ, Gerich JE. Quantification of the relative impairment in actions of insulin on hepatic glucose production and peripheral glucose uptake in non-insulin-dependent diabetes mellitus. Metabolism. 1988;37:15-21. 9. Steele R. Influences of glucose loading and of injected insulin on hepatic glucose output. Ann NY Acad Sci. 1959;82:420-30. 10. Chen Y-DI, Jeng C-Y, Hollenbeck CB, Wu MS, Reaven GM. Relationship between plasma glucose and insulin concentration, glucose production, and glucose disposal in normal subjects and patients with non-insulin dependent diabetes. J Clin Invest. 1988;82:21-5. 11. Chen Y-DI, Swislocki ALM, Jeng C-Y, Juang JH, Reaven GM. Effect of time on measurement of hepatic glucose production. J Clin Endocrinol Metab. 1988;67:1084-8. 12. National Diabetes Data Group. Classification and diagnosis of

13. 14. 15. 16.

17. 18. 19.

JCE & M • 1990 Vol 70 • No 5

diabetes mellitus and other categories of glucose intolerance. Diabetes. 1979;28:1039-57. Kadish AH, Litle RL, Sternberg JC. A new and rapid method for determination of glucose by measurement of rate of oxygen consumption. Clin Chem. 1968;14:116-31. Hales CN, Randle PJ. Immunoassay of insulin with insulin-antibody precipitate. Biochem J. 1963;88:137-46. Newman JD, Armstrong JM. On the activities of glycogen phosphorylase and glycogen synthase in the liver of the rat. Biochim Biophys Acta. 1978;544:225-33. Stalmans W, DeWulf H, Hue L, Hers H-G. The sequential inactivation of glycogen phosphorylase and activation of glycogen synthethase in liver after the administration of glucose to mice and rats. Eur J Biochem. 1974;41:127-34. Wolfe RR, Allsop SR, Burke JF. Glucose metabolism in man: responses to intravenous glucose infusion. Metabolism. 1979;28:210-20. Wolfe RR, Shaw JHF, Jahoor F, Herndon DN, Wolfe MH. Response to glucose infusion in humans: role of changes in insulin concentration. Am J Physiol. 1986;250:E306-ll. Miiller MJ, Moring J, Seitz HJ. Regulation of hepatic glucose output by glucose in vivo. Metabolism. 1988;37:55-60.


Metabolism: Mechanisms of hepatic glucose production revealed.

Insulin control of hepatic glucose production.

Autonomic regulation of hepatic glucose production.

Longitudinal changes in basal hepatic glucose production and suppression during insulin infusion in normal pregnant women.

Effect of cephalic glucose infusion on insulin secretion.

Incomplete suppression of hepatic glucose production in non-insulin dependent diabetes mellitus measured with [6,6-2H2]glucose enriched glucose infusion during hyperinsulinaemic euglycaemic clamps.

The Effect of Phloroglucinol, A Component of Ecklonia cava Extract, on Hepatic Glucose Production.

Glucose requirements following burn injury. Parameters of optimal glucose infusion and possible hepatic and respiratory abnormalities following excessive glucose intake.

Effect of starvation on hepatic glycogen metabolism and glucose homeostasis.

Intraportal glucose delivery enhances the effects of hepatic glucose load on net hepatic glucose uptake in vivo.

Persistent stimulatory effect of glucagon on glucose production despite downregulation.

Correction of glucose carbon recycling for the determination of 'true' hepatic glucose production rates by (1-13C1)glucose.

The effects of continuous glucose infusion on blood, plasma, and brain glucose in anesthetized rats.

The effect of glucose infusion on the plasma free amino acids in sheep.

Effect of glucagon infusion on plasma magnesium, glucose, and insulin in bull calves.

Effect of prolactin infusion on lactation, glucose kinetics, and pancreatic hormones in lactating goats.

The effect of 2-deoxy-D-glucose and D-glucose on the efferent discharge rate of sympathetic nerves.

Role of placental insufficiency and intrauterine growth restriction on the activation of fetal hepatic glucose production.

Tumor necrosis factor alpha abolished the suppressive effect of insulin on hepatic glucose production and glycogenolysis stimulated by cAMP.

Thyroid hormone treatment decreases hepatic glucose production and renal reabsorption of glucose in alloxan-induced diabetic Wistar rats.

Effect of somatostatin and a glucagon-specific analog on glucose homeostasis during arginine infusion.

Hepatic portal vein infusion of glucose and sodium solutions on the control of saline drinking in the rat.

Simulation of oral glucose tolerance tests and the corresponding isoglycemic intravenous glucose infusion studies for calculation of the incretin effect.

Effect of intact parathyroid hormone on hepatic glucose release in the dog.