| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Other Original Articles |
Department of Clinical and Experimental Medicine (A.A., L.G., S.d.K., A.T.), University of Padova Medical School, 35128 Padova, Italy; Department of Preventive Medicine (R.M.W.), Keck School of Medicine, University of Southern California, Los Angeles, California 90089; Institute of Systems Science and Biomedical Engineering (G.P.), National Research Council, 35127 Padova, Italy
Address all correspondence and requests for reprints to: Angelo Avogaro, M.D., Dipartimento Medicina Sperimentale, Policlinico, Via Giustiniani 2, 35128 Padova, Italy. E-mail: . angelo.avogaro{at}unipd.it
Abstract
Moderate alcohol (ETOH) intake has been associated with a significant reduction in risk for infarction among general populations. In this study, we assessed the effects of low-dose ETOH (40 g over 3-h period as vodka) on the interaction between glucose (G), insulin, and lactate (L) during the insulin-modified frequently sampled iv glucose tolerance test (FSIGT) (0.3 U/kg body weight between 20–25 min) in eight normal volunteers. In the control (C) study, water was administered. An insulin-independent two-compartment model was used to describe G and L kinetics. Insulin sensitivity (SI) was significantly higher in the ETOH study than in the C study (2.49 ± 0.52 vs. 0.92 ± 0.20 10-4 min-1µU-1ml, C vs. ETOH; P = 0.0391). No significant differences were observed in G effectiveness (0.029 ± 0.004 vs. 0.033 ± 0.004 min-1). Blood L levels were higher during FSIGT when ETOH was administered [area under the curve (AUC), 201 ± 16 vs. 123 ± 23 mmol/liter in 240 min; P = 0.0001]. The dynamic analysis of blood L concentrations showed that ETOH also significantly decreased L clearance (0.0016 ± 0.0011 vs. 0.0029 ± 0.0002 min-1; P = 0.0156), whereas no difference was observed for the fractional conversion of the rate of G disappearance to L (0.0033 ± 0.0012 vs. 0.0031 ± 0.0005 min-1). ETOH decreased baseline plasma FFA concentration; AUC of FFA was markedly reduced with ETOH (65 ± 14 vs. 109 ± 17 mmol/liter in 240 min; P = 0.0063) and inversely correlated with SI (r = 0.693; P = 0.0029). The amount of C-peptide in 240 min as well as the amounts before and after insulin administration were not different between the two tests. We concluded that G and L kinetics derived from FSIGT shows that moderate ETOH intake: 1) improves insulin action; 2) decreases L clearance; and 3) does not affect ß cell function. Because ETOH at moderate doses has a marked antilipolytic action, it might improve insulin action by improving substrate competition. The present findings suggest that moderate alcohol consumption in the diet should not be discouraged.
ALCOHOL (ETOH) ACCOUNTS for 4–6% of the average energy intake in most Western countries and is affected by both environmental and inherited biological mechanisms (1, 2). As a nutrient, it influences the metabolism of most tissues in the body, with marked effects on glucose (G) homeostasis (3). The effects of ETOH on G metabolism in vivo are complex and sometimes contradictory because response to alcohol intake varies, depending on the nutritional state of an individual (4). There are more than 60 prospective studies that suggest an inverse relation between moderate alcoholic beverage consumption and coronary artery disease (5, 6, 7, 8). Approximately half of the beneficial effects of alcohol on cardiovascular disease have been ascribed to an increase in high-density lipoprotein. The same dose-dependent effect of ETOH holds for insulin action in that high doses decrease whereas light-to-moderate ETOH consumption enhances hormone activity in term of G disposal (9, 10, 11). However, ETOH not only has direct effects on G metabolism but also interferes with intermediary metabolism, namely with lactate (L) metabolism, further complicating the metabolic scenario (12, 13). Much of the information gained on the effect of alcohol on G metabolism comes from tracer and clamp studies that cannot dissect the simultaneous effects of ETOH on G and L metabolism and insulin secretion. Thus, the aim of the present study was to investigate the acute effects of a moderate amount of ETOH by analyzing frequently sampled iv glucose tolerance test (FSIGT) data with a method that provides a measurement not only of insulin sensitivity and G effectiveness, but also parameters of L kinetics and the interaction of G and L (14).
Subjects and Methods
Subjects
FSIGTs were performed twice on eight normal healthy volunteers (six males and two females) who participated in the study after giving written informed consent. Their mean age and body mass index were 24 ± 3 yr and 23.5 ± 1.5 kg/m2, respectively, and all were customary wine drinkers [<500 ml of wine (10–12% wt/vol) per day]. The subjects were in good health. For at least 3 d before each test, they consumed a controlled diet containing more than 250 g carbohydrate and maintained a regular life style with no strong exercising or smoking. The experimental protocol was approved by the Ethical Committee of the School of Medicine of the University of Padova.
Experimental procedures
Each subject participated in two tests, an ETOH study and a control study, in random order on separate days. At least 1 wk elapsed between two consecutive tests. Subjects were admitted to the University Hospital of Padova at 0700 h, after an overnight fast, to perform the study. A 20-gauge butterfly needle was inserted into a dorsal hand vein at 0730 h, and a blood sample was taken. The hand was then placed in a box heated at 60 C to arterialize venous blood. The patency of the needle was maintained with a controlled saline infusion throughout the study. Then, an 18-gauge cannula was placed into the contralateral antecubital vein for injection of the G load that occurred at 0830 h and for the administration of insulin (see below).
For the ETOH study, 40 g of ETOH (administered as vodka 40% wt/vol) were sipped from time -60 min until the end of the FSIGT (240 min). Blood samples for hormones and substrates were obtained 30 min, 15 min, and immediately before the G injection. At time zero, the G load (0.3 g/kg of body weight) was administered over 1 min. Frequent samples for plasma G, L, insulin, C-peptide, and FFA concentrations were obtained at 2, 3, 4, 5, 6, 8, 10, 12, 15, 20, 25, 30, 35, 40, 60, 80, 100, 120, 140, 180, 210, and 240 min. At 20 min, plasma insulin concentration was enhanced by an exogenous square-wave iv infusion of regular insulin (0.03 U/kg) lasting 5 min. During the control study, an amount of tap water similar to the vodka of the other protocol was sipped throughout the test.
Analytical methods
Plasma G was measured with the G oxidase method on a Beckman Glucose Analyser. Blood L was measured with a fluorometric technique that had an overall coefficient of variation (CV) of 5.2 ± 3.8% (15). The total FFA concentration was determined with a microenzymatic technique (16) (CV, 6.9 ± 2.3%). Plasma insulin and C-peptide were measured by conventional RIA (17) (CV, 6 ± 4% and 5.3 ± 3.2%, respectively). Blood ethanol was determined with an enzymatic technique (18) (CV, 12 ± 8%).
Data analysis
The G and L kinetics from the FSIGT was analyzed using the compartmental model developed by Watanabe et al. (14). This model links the kinetics of L to the minimal model of G kinetics (14, 19), providing quantitative estimates of both G and L kinetics, as well as their interaction. The model can be described by the following set of differential equations.
 | (1) |
 | (2) |
 | (3) |
 | (4) |
Equations 1
and 2 describe the classic minimal model of G kinetics that has been described previously in detail (19, 20, 21). Briefly, the model segregates the disappearance of G from plasma into two separate processes. The first is the effect of G to enhance its own disappearance at basal insulin. This is accounted for by parameter SG and is termed G effectiveness. The second assumes that insulin in a compartment remote from plasma, presumably interstitial fluid (21, 22), augments the disappearance of G. The remote insulin effect X(t) is described by parameters p2 and p3, which characterize the transport into and degradation of insulin in the remote compartment. Insulin sensitivity is derived as the ratio of these two parameters.
The kinetics of L is linked to the G model via Eq 3. This equation describes a single pool of metabolic intermediates (M), presumably glycolytic intermediates (e.g. triose phosphates, pyruvate, etc.), through which G is converted to L. The fraction of the rate of total G disappearance destined for L is accounted for by the term 2f (SG + X(t)) G(t). The multiplicative factor of two accounts for the 2:1 molar ratio between L and G. G metabolism via alternative pathways including glycogen synthesis is accounted for by (1 - f) (SG+X(t)) G(t). Thus, this model assumes that L production during the FSIGT is due to G alone. Parameter K01 accounts for loss of metabolic intermediates to pathways that branch off from glycolysis and subsequently do not contribute to L formation. Equation 4 characterizes the kinetics of L in blood. The fractional rate of conversion of metabolic intermediates to L and subsequent release into blood is described by parameter K21, and conversion of L back to intermediate is represented by K12. Fractional disappearance of L from the blood compartment is represented by K02. VG, VM, and VL describe the distribution spaces for G, metabolic intermediates, and L, respectively.
Individual estimates of the nine kinetic parameters and two distribution volumes of the model as described cannot be determined from the G and L time courses. To facilitate identification of the parameters, Eqs 3 and 4 are reparameterized to derive a new set of equations resulting in an identifiable set of parameters. This identifiable system consists of the following set of differential equations:
 | (5) |
|
| (6) |
 | (7) |
The equations describing the minimal model remain unaltered (Eqs 5 and 6), as do the physiologic interpretation of the parameters. Equations 3 and 4 are combined to yield Eq 7 with an identifiable set of three parameters defined as:
 | (8) |
 | (9) |
 | (10) |
Parameters A1 and A2 broadly characterize the mixing and clearance kinetics of L in blood, respectively. However, because these are now lumped parameters, it is not possible to distinctly assign a single physiologic process to either. Parameter A3 describes the fractional rate of conversion of the rate of G disappearance to L. This parameter is analogous, but not equivalent, to calculating the fraction of G that is converted to L.
Numerical methods
Parameter identifications were performed using MLAB software for the personal computer (Civilized Software, Inc., Bethesda, MD), which uses a Marquardt-Levenburg iterative weighted least-squares algorithm. The weighting was determined using an internal weighting function provided in MLAB. This weighting function performs a five-point moving average to estimate the SD in the data and computes the inverse variance for each time point.
The insulin sensitivity index was also calculated by using the classic minimal model of G disappearance (20) to compare the values with those obtained with the G/L model.
The accuracy of the parameter estimates was assessed using the CV. The SD was obtained by the square root of the main diagonal of the covariance matrix calculated during the least-squares estimation procedure (23).
Calculations and statistical analysis
Parameter KG, the G tolerance index, was calculated as the slope of the natural logarithm of G concentration vs. time from 10–20 min for the period before insulin infusion, and also from 20–40 min. The areas under the insulin and C-peptide concentration curves were computed by integration using the trapezoid rule. The acute insulin response to G (AIRG) was calculated as the mean suprabasal insulin concentration from 3–10 min after the G injection. Insulin clearance was assessed by dividing the insulin dose by the area under the suprabasal insulin concentration curve from 20–240 min (24). Statistical differences in mean values were assessed by the Wilcoxon matched pairs test. Differences in sampling times during the FSIGT between groups were assessed by ANOVA, taking into account the subject and time effects. A P value less than 0.05 was considered statistically significant. Values are reported as mean ± SE, except when otherwise designated.
Results
FSIGT
Blood ETOH levels were constant during the ETOH study (3 ± 2, 4 ± 1, and 2 ± 1 mg/dl at time 0, 120, and 240, respectively). The intrasubject CV for ETOH concentration was 10 ± 4%.
Basal values (Table 1
) showed no difference in fasting conditions before both studies. The average time courses of G concentrations during FSIGT are shown in Fig. 1. In both cases, the insulin administration at 20 min yielded a marked acceleration of G disappearance as shown by the KG values (Table 1). In the control studies, G and insulin concentrations during the FSIGT reattained fasting levels by the end of the test (P > 0.75), whereas C-peptide was significantly lower (0.97 ± 0.12 ng/ml; P = 0.006). In the ETOH study, G, insulin, and C-peptide concentrations at 240 min were all lower than fasting values [85 ± 4 mg/dl (P = 0.038), 5.7 ± 0.4 µU/ml (P = 0.007), and 0.81 ± 0.11 ng/ml (P = 0.002), respectively].
|
|
, top), although no difference was detected in the basal levels at the time of G injection (0.555 ± 0.075 vs. 0.489 ± 0.039 mmol/liter; P = 0.411). ETOH decreased baseline plasma FFA concentration (Fig. 2, bottom) from 459 ± 57 to 159 ± 39 µmol/liter (P < 0.01), and the mean plasma FFA levels during the FSIGT were also significantly lower when ETOH was consumed (218 ± 120 vs. 412 ± 200 µmol/liter; P < 0.01). The AUC of FFA was markedly reduced with ETOH (65 ± 14 vs. 109 ± 17 mmol/liter in 240 min; P = 0.0063) and inversely correlated with SI (r = 0.693; P = 0.0029) (Fig. 3).
|
|
insulin [0–15]), was not different between the two protocols (Table 1), suggesting that differences observed in G disappearance and tolerance are not due to a different amount of acting insulin. Total insulin during the test, calculated as the area of insulin concentration up to 240 min, was not different in the two tests as well as KG calculated after exogenous insulin (Table 1).
Insulin clearance increased with ETOH, whereas no differences were observed for insulin secretion as characterized by C-peptide measurements (Table 1). The amount of C-peptide in 240 min as well as the amounts before and after insulin administration were not different between the two tests, suggesting that insulin secretion was not affected by ETOH intake.
Metabolic parameters of G homeostasis
Insulin sensitivity was significantly higher in the ETOH study than in the control study (Table 2). These values of insulin sensitivity were compared with those obtained with the classic minimal model (P = 0.179 and 0.909, respectively, when compared with the respective SI obtained with the G/L model). The two measurements of insulin sensitivity significantly correlated (r = 0.603; P = 0.013). No significant differences were observed in G effectiveness.
|
), whereas no difference was observed for the fractional conversion of the rate of G disappearance to L or the mixing characteristics. Discussion
The aim of the present work was to test the effect of in vivo moderate ETOH administration on the interaction between G, insulin, and L. This approach, which is based on a two-compartment model that describes G and L kinetics, shows that moderate ETOH intake improves insulin-mediated G uptake, decreases L clearance, and does not affect ß-cell function. Literature findings on the effect of ETOH on G metabolism are contradictory. In vivo experimental studies using the euglycemic-hyperinsulinemic clamp and tracer methodology showed that ETOH decreases insulin- mediated G disposal (25). This effect appears to be dose-dependent because an inhibition of the G use by insulin is usually observed at ethanol concentrations above 40 mg/dl. In contrast, several population-based studies have shown that low-to-moderate doses of ETOH improve insulin action (10, 26). Our finding of an increased insulin sensitivity endorses the results of the Insulin Resistance and Atherosclerosis Study (27), which showed that there is an inverse U-shaped relationship between whole-body insulin sensitivity and ETOH intake, and agrees with those of Facchini et al. (26) who showed that light-to-moderate drinkers (10–30 g/d) had lower integrated plasma G and insulin responses to the G challenge.
It was not possible in this whole-body study to directly assess the cellular or organic mechanisms responsible for this ETOH-induced increased G uptake. It has been recently shown that ETOH may induce the recruitment of specific G transporters at least in vitro (28). ETOH might positively act on G storage, although findings on this issue are lacking. During insulin-stimulated conditions in humans, ETOH did not appreciably affect glycogen synthase activity. Thus, the effects on nonoxidative G disposal should deserve further investigation (4). We would like to emphasize that all of the subjects included in this study were habitual ETOH consumers; therefore we cannot extrapolate a similar effect of ETOH on insulin sensitivity in nondrinkers at similar achieved concentration. Because it has been recently shown that insulin sensitivity correlated positively with ETOH consumption (29), it will be of interest to assess this possible relationship in future studies.
One plausible explanation for the ETOH-induced amelioration in insulin action comes from our FFA findings. In this study, we confirm that moderate ETOH doses can decrease plasma FFA concentration (30). A small amount of ETOH can exert an antilipolytic action. We therefore speculate that although ETOH may negatively affect pyruvate dehydrogenase, it can stimulate G oxidation by improving the competition between lipid and G at peripheral tissue level. This hypothesis is graphically depicted in Fig. 3, which relates the changes in SI with plasma FFA AUC in both the presence and absence of ETOH. Thus, it appears that one possible mechanism by which ETOH might improve insulin action is indeed an amelioration of the so-called Randle cycle. Future studies are needed to confirm this hypothesis.
We found that ETOH did not modify noninsulin-mediated G uptake (G effectiveness). The central nervous system accounts for almost 70% of noninsulin-mediated G uptake, and the nonnervous tissues account for the remaining 30%, with 13% being due to muscle uptake (31). Thus, moderate acute ETOH intake does not seem to affect brain G uptake, although it has been shown in experimental animals that the chronic consumption of ethanol in a free-choice situation may impair neuronal G uptake and glycolytic flux (32, 33, 34).
ETOH intake deeply influences not only G but also L metabolism. This is of great clinical importance because L itself is a major precursor of gluconeogenesis and can be metabolized within peripheral tissue. Thus, ETOH may directly affect not only G usage, but also its production by interfering with the Cori cycle. The combined analysis of G and L metabolism exploited in this study shows that, in the presence of ETOH, besides a substantial improvement of G uptake, there is a reduction in L clearance in the face of unchanged fractional conversion of G to L. These findings support the concept that, during ETOH intake, both reduced uptake and increased production play a role in determining higher blood L levels. Assuming no effect on G storage, a higher G uptake, in the presence of a right-shifted equilibrium of nicotinamide adenine dinucleotide/reduced nicotinamide adenine dinucleotide (NAD/NADH), should yield increased peripheral L production.
From our methodology, we cannot tell whether this is due to either a peripheral or hepatic reduction in L uptake. It is possible that the conversion of L to pyruvate is impaired, so that the carbons, which were otherwise destined for G synthesis, are not oxidized and hence exported from the liver as L into the bloodstream. Kreisberg (12) found that the main effect of ETOH in increasing L concentration is to decrease L conversion into G. Because we did not use G tracers, we could not gain information on hepatic G output (HGO); however, indirect evidence that ETOH may decrease HGO comes from the plasma G profiles. During the last part of the FSIGT (Fig. 1), plasma G concentration mostly depends on HGO. Thus, the lower G levels at the end of the test, compared with fasting, suggest an inhibitory action of ETOH on HGO.
In this study we have shown that moderate ETOH consumption induces several metabolic changes in G, L, and lipid metabolism. These changes all take place without a significant effect of ETOH on ß-cell secretion. In fact, we did not observe any changes in insulin secretory pattern estimated as the dynamic acute insulin response to the G challenge and with the area under the C-peptide concentration curve. This implies that ETOH does not cause significant modification in ß-cell function. However, the presence of reduced plasma insulin concentration during this experimental condition would support the hypothesis that ETOH might increase insulin clearance as we observed in the ETOH study. Nonetheless, we believe that further studies will be needed to specifically address this issue.
In conclusion, we have found that moderate ETOH consumption both improves insulin action without affecting noninsulin-mediated G uptake and decreases L clearance. Thus, the increase in blood L during ETOH intake is mainly due to an inhibition of its removal. Moreover, ETOH does not appear to affect ß-cell function to a significant extent. The present findings may explain the metabolic effects of moderate ETOH intake which, therefore, should not be discouraged.
Acknowledgments
Footnotes
This study was supported by a Veneto Region grant on Invecchiamento dell’anziano. R.M.W. was a NIH postdoctoral fellow (DK-09525) when these studies were performed and is currently supported by a Career Development Award from the American Diabetes Association. Preliminary results were presented June 2001 at the 61st annual American Diabetes Association Scientific Sessions in Philadelphia, Pennsylvania.
Abbreviations: AIRG, Acute insulin response to glucose; AUC, area under the curve; C, control; CV, coefficient of variation; ETOH, alcohol; FSIGT, frequently sampled iv glucose; G, glucose; HGO, hepatic glucose output; L, lactate.
Received August 10, 2001.
Accepted December 12, 2001.
References
This article has been cited by other articles:
 |
|
 |
|
  M. R. Rickels, A. Naji, and K. L. Teff Insulin Sensitivity, Glucose Effectiveness, and Free Fatty Acid Dynamics after Human Islet Transplantation for Type 1 Diabetes J. Clin. Endocrinol. Metab., June 1, 2006; 91(6): 2138 - 2144. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Brehm, K. Thomaseth, E. Bernroider, P. Nowotny, W. Waldhausl, G. Pacini, and M. Roden The Role of Endocrine Counterregulation for Estimating Insulin Sensitivity from Intravenous Glucose Tolerance Tests J. Clin. Endocrinol. Metab., June 1, 2006; 91(6): 2272 - 2278. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Avogaro, R. M. Watanabe, A. Dall'Arche, S. Vigili De Kreutzenberg, A. Tiengo, and G. Pacini Acute Alcohol Consumption Improves Insulin Action Without Affecting Insulin Secretion in Type 2 Diabetic Subjects Diabetes Care, June 1, 2004; 27(6): 1369 - 1374. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. R. Greenfield, K. Samaras, A. B. Jenkins, P. J. Kelly, T. D. Spector, and L. V. Campbell Moderate Alcohol Consumption, Estrogen Replacement Therapy, and Physical Activity Are Associated With Increased Insulin Sensitivity: Is abdominal adiposity the mediator? Diabetes Care, October 1, 2003; 26(10): 2734 - 2740. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
