The Journal of Nutrition Vol. 128 No. 6 June 1998, pp. 1023-1030

Insulin-Dependent Glucose Utilization in Intensively Milk-Fed Veal Calves Is Modulated by Supplemental Lactose in an Age-Dependent Manner^1,2

Doris Hugi, Luc Tappy^*, Helga Sauerwein, Rupert M. Bruckmaier, and Jürg W. Blum³

Division of Nutrition Pathology, Institute of Animal Breeding, University of Berne, 3012 Berne, Switzerland; ^* Institute of Physiology, University of Lausanne, 1005 Lausanne, Switzerland; and  Institute of Physiology, Technical University of Munich, 85350 Freising-Weihenstephan, Germany

	ABSTRACT

Abstract Introduction Methods Results Discussion References

Postprandial insulin resistance with excessive hyperinsulinemia, hyperglycemia and glucosuria develops with increasing age in veal calves intensively fed milk replacers. We tested the age dependency of insulin resistance, modulated by high lactose intake, glucose oxidation and insulin receptor number and affinity after an overnight period without food. Male calves were fed a milk replacer containing 290 or 423 g lactose and 310 and 541 g total sugar/kg from 69-195 kg body weight. At mean body weights of 95 and 170 kg, insulin-dependent glucose metabolism was studied in euglycemic-hyperinsulinemic glucose clamps (EGC), and glucose-dependent insulin responses were tested in hyperglycemic clamps (HGC). EGC were combined with infusions of [¹³C₆]glucose to measure glucose kinetics and glucose oxidation by determination of ¹³CO₂ exhalation. During EGC and HGC, insulin concentrations were similar in both groups, indicating comparable insulin secretion and metabolic clearance rates. On the basis of glucose infusion rates required to maintain eu- or hyperglycemia in EGC and HGC, respectively, insulin-dependent glucose utilization was not age dependent. However, in calves receiving a high lactose intake, insulin-dependent glucose utilization was enhanced in the early phases, but was reduced in the late stages of the growth trial. Insulin-dependent glucose utilization behaved inversely with atom % excess of [¹³C₆]glucose, but changed in a manner similar to that of the rate of glucose appearance. Inhibition of endogenous glucose output, exhalation of ¹³CO₂ and amounts of oxidized glucose exhibited no group differences. More glucose was therefore stored in lactose-supplemented calves. A reduced insulin receptor number in skeletal muscle in calves fed high amounts of lactose likely contributed to low insulin-dependent glucose utilization.

	INTRODUCTION

Abstract Introduction Methods Results Discussion References

Plasma glucose and insulin levels typically decrease during and after weaning in calves (Breier et al. 1988, Hugi and Blum 1997). In contrast, veal calves, especially towards the end of fattening, often exhibit postprandial hyperglycemia, glucosuria, excessive hyperinsulinemia and marked insulin resistance, indicating insufficient homeostatic control of glucose metabolism (Doppenberg and Palmquist 1991, Hostettler-Allen et al. 1994, Hugi and Blum 1997, Hugi et al. 1997a, 1997b and 1997c, Palmquist et al. 1992, Ronge and Blum 1989). Because a normal insulin status (concentration, biological effect) is required for the stimulation of anabolic processes, it is surprising that high growth rates are possible under conditions of such dramatic insulin resistance as observed in veal calves. It may be hypothesized that high circulating insulin may mediate some of its effects through the type I insulin-like growth factor receptor (Jones and Clemmons 1995). In addition, the high amounts of circulating insulin may be responsible for the increase during fattening of plasma concentrations of insulin-like growth factor-I (Hostetter-Allen et al. 1994, Hugi et al. 1997 b), through which growth may be stimulated.

The etiology of insulin resistance in veal calves is not clear. Because breeding calves receive only small amounts of milk for a limited time until weaning, whereas veal calves are fed milk and milk replacers (MR)⁴ almost exclusively, nutrition and nutritional factors are of great importance. Veal calves are fed intensively to allow average gains of ~1.4 kg/d; there may be a constitutive inability of veal calves to handle the high amounts of absorbed nutrient components, especially glucose. Thus feeding high amounts of lactose, which is often added to low fat MR, may cause high plasma levels of glucose and insulin resistance because calves fed high lactose/low fat MR diets develop hyperglycemia, glucosuria and insulin resistance (Wijayashinge et al. 1984). However, hyperinsulinemia after MR intake is markedly greater than after oral administration of lactose or oral and parenteral administration of glucose, indicating that factors other than lactose or glucose cause hyperinsulinemia (Hostettler-Allen et al. 1994, Hugi et al. 1997b and 1997c). High fat intake may be important, although this has been questioned (Doppenberg and Palmquist 1991, Palmquist et al. 1992). We have not seen improvements in glucose tolerance by feeding veal calves a synthetic Cr(III) preparation [containing Cr(III), nicotinic acid, glycine, glutamic acid and cysteine], which functions in vitro as an insulin tolerance factor, suggesting that Cr (III) deficiency is not likely (Blum, Bruckmaier and Gautschi, unpublished observations). Because calves in different experiments were fed 20-30 mg FeSO₄/kg MR, iron excess, which causes insulin resistance, is etiologically not involved (Hostettler-Allen et al. 1993 and 1994; Hugi et al. 1997b and 1997c). High circulating amounts of insulin per se are likely important because insulin down-regulates its own receptors. Furthermore, we have found markedly higher amounts of noradrenaline and dopamine excreted in the urine per kilogram body weight (BW)^0.75 during 24 h at the end of fattening of veal calves than at the start (Hugi et al. 1997b), suggesting that an obviously enhanced sympathetic activity with increasing age may contribute to insulin resistance. On the other hand, excesses of growth hormone and cortisol, which induce insulin resistance, could be excluded as etiological factors in veal calves (Hostettler-Allen et al. 1994, Hugi et al. 1997b). Glucose-dependent insulinotropic polypeptide could be excluded also as an etiological factor of hyperinsulinemia and secondary insulin resistance because glucose-dependent insulinotropic polypeptide is not insulinogenic in ruminants (Guilloteau et al. 1995), and its postprandial rise was similar in calves fed normal or high amounts of lactose and at the start and end of fattening (Hugi et al. 1997b).

It is not clear in veal calves whether 1) insulin resistance exists in the preprandial stage, 2) preprandial insulin-dependent glucose utilization is dependent on the developmental stage, 3) preprandial insulin-dependent glucose utilization expression is modified by high lactose intake, 4) insulin receptor number and/or affinity of skeletal muscle (the organ with greatest glucose uptake during growth) are influenced by age and lactose intake and 5) glucose oxidation is affected and thus contributes to hyperglycemia. On the basis of these questions, we have tested the following hypotheses: 1) insulin resistance exists in the preprandial stage; 2) insulin-dependent glucose utilization increases preprandially with increasing age; 3) preprandial insulin-dependent glucose utilization expression is enhanced by high lactose intake; 4) insulin receptor number and/or affinity of skeletal muscle is reduced with increasing age and by high lactose intake; and 5) glucose oxidation becomes reduced with increasing age and by high lactose intake.

	MATERIALS AND METHODS

Abstract Introduction Methods Results Discussion References

Animals and husbandry.  Twenty Simmenthal × Red Holstein bull calves were housed at the Experimental Station, Posieux, Switzerland. They were allowed free movement in loose housing systems strewn with straw. During wk 1 after arrival, calves were fed whole milk and given a prophylactic antimicrobial treatment for 4 d. Then calves were weighed, ear-tagged and divided into two groups of 10 calves each. Calves of both groups had similar initial mean BW, hemoglobin and plasma iron concentrations.

Feeding.  Diets were produced by Provimi AG, Cossonay, Switzerland. The MR was composed primarily of milk-derived products (Table 1). Antimicrobial substances were not included. On the basis of chemical analyses, the MR fed to lactose-supplemented group (group L) contained much greater amounts of lactose and nitrogen-free extracts (lactose plus starch) than the MR fed to the control group (group C), but barely differed with respect to contents of crude protein, crude fat, gross energy, ash and iron (Table 1). According to calculations, there were no significant differences with respect to essential and nonessential amino acid composition and contents of linoleic acid, fat and water-soluble vitamins and minerals of the MR fed to group C and group L. The feeding plan was designed to allow an average daily gain of up to ~1.4 kg. The rate of feeding was adjusted weekly on the basis of individual BW. Calves were fed daily at 0730 and 1630 h. Milk refusals were weighed daily for each calf.

Experimental procedures.  Experimental procedures were approved by the commission overseeing animal experiments of the Canton of Freiburg, Switzerland and of the Federal Veterinary Administration, Berne, Switzerland and followed the actual Swiss law of animal protection.

Euglycemic-hyperinsulinemic glucose clamps. Experiments were performed in early and late phases of the growth trial, i.e., at ages 8 and 16 wk. In group C, calves weighed 99.4 ± 2.4 and 173.1 ± 1.8 kg; in group L, they weighed 102.4 ± 1.4 and 175.1 ± 1.1 kg. On the day before the experiment, catheters were inserted into both jugular veins. One experiment was performed per day. Calves were tethered with freedom to stand or lie. Experiments started after an overnight period of 15 h without feed. Blood samples were collected from the right jugular catheter.

At a relative time of 150 min, administration of [¹³C₆]glucose (D-glucose, 90% enrichment; no. CLM-1396-90, Cambridge Isotope Laboratories, Woburg, WA) was started. First, a bolus of 1.11 µmol of [¹³C₆]glucose/kg BW was injected over a period of 15 s, followed immediately by constant infusion of [¹³C₆]glucose at a rate of 0.1 mL/min [11.1 nmol/(kg BW·min)] for 5.5 h (i.e., from 150 min to 180 min).

Saline (9 g/L; 0.92 mL/min) was infused through a second catheter from 30 to 0 min, followed by infusion of bovine insulin in amounts of 13.8 pmol/(kg·min) from 0 to 5 min and of 6.9 pmol/(kg·min) at a rate of 0.6 mL/min from 5 to 240 min. Insulin (194 nmol/mg, lot 64 F 0349; Sigma Chemical, St. Louis, MO) was dissolved in saline (9 g/L) and bovine serum albumin (1 g/L). The calf's own plasma was added to the infusate in amounts of 4% to prevent absorption of insulin into glassware and infusion lines. Potassium chloride was infused (12.4 mg K⁺/min) to prevent hypokalemia.

Glucose extracted from sugar beets (Hausmann Laboratories AG, St. Gallen, Switzerland) was used for infusions because its [¹³C₆]glucose enrichment (1.083 atom % excess) was similar to basal values. Glucose infusion started 4 min after the start of the insulin infusion at a rate of ~5.55 mol/(kg·min) with the use of two peristaltic pumps, delivering solutions containing 1.4 or 2.8 mol glucose/L, depending on needs. Glucose infusion rates (GIR) were continually adjusted according to plasma glucose measurements. The infusion rate was increased in a step-wise fashion in the stabilization phase (0-60 min); after constant concentrations were reached between 60 to 180 min, only small changes in GIR were necessary. Clamp experiments were accepted only if >90% of the steady-state glucose values did not deviate ±10% from basal concentrations.

Blood samples were taken at 30, 20, 10 and 0 min to calculate basal plasma [¹³C₆]glucose enrichment and to obtain basal values of insulin and glucose. From 0 to 180 min, samples of 0.25 mL every 5 min and of 5 mL every 10 min were collected. Every 5 min, 0.25 mL of blood was centrifuged at 12,000 × g for 10 s, and 10 µL of plasma was then immediately analyzed for glucose as previously described (Hostettler-Allen et al. 1994). The remainder of the blood collected every 10 min was added to tubes containing sodium-EDTA, cooled, and then centrifuged at 3000 × g for 20 min within 1 h. Plasma was then frozen for later analysis of insulin and [¹³C₆]glucose enrichment.

At 150, 20, 10 min and 0 min and during the 3 h of the clamp experiment, a face mask, with two inspiratory and one expiratory unidirectional valves and connected with a flexible tube (6 cm diameter) to a pneumotachograph and from there to a gas analyzing system (Ergo-Oxyscreen; Jaeger, Würzburg, Germany), was placed on the head every 30 min to measure respiratory rate, respiratory minute volume, oxygen consumption (V_O2) and carbon dioxide production (V_CO2) as previously described (Bruckmaier and Blum 1992). During the measurement, data of respiratory rate, respiratory minute volume, V_O2, V_CO2 and calculated respiratory exchange rates (RE, i.e. V_CO2/ V_O2 ratio) were printed every 30 s. Toward the end of each respiratory measurement period, air was collected with an evacuated gas sample container (no. EX10/Z11E75, Batch 4255; Labco High Wycombe, Numelec AG, Geneva, Switzerland) to determine breath ¹³CO₂ enrichment by infrared mass spectrometry at the Institute of Physiology, University of Lausanne, Lausanne, Switzerland.

Results of the glucose clamp during the first 60 min of the clamp were excluded because this was considered the stabilization phase.

Rates of glucose turnover from 30 to 0 min and from 60 to 180 min were calculated using Steele's equation (DeBodo et al. 1963) for nonsteady-state conditions: where F is the rate of [¹³C₆]glucose infusion [µmol/(kg·min)], pV is the fraction of the total extracellular glucose pool (with p = 0.75 and V = 0.2L/kg), where p is the pool fraction and V is the distribution space d for glucose ; C₂, C₁ are the glucose concentrations in plasma (mmol/L) at times 2 and 1; E₂, E₁ are the enrichment of [¹³C₆]glucose (%); and t₂, t₁ represent the time interval.

To our knowledge, pool fraction and distribution space values for glucose have not been specifically determined in veal calves; thus we have used the values usually employed in human and rat experiments. Although the use of inaccurate values for these parameters may lead to moderate inaccuracies in the calculation of the glucose turnover rate, this did not interfere with the comparison of the two groups in this study.

Endogenous glucose production was calculated as R_a  GIR. Glucose oxidation was calculated as with ¹³CO₂ and [¹³C₆] glucose in atom % excess. The recovery in breath of ¹³CO₂ derived from [¹³CO₂]glucose oxidation was 0.56; 0.134 mL CO₂ was produced during oxidation of 1 µmol of glucose. V_CO2 is expressed in mL/min.

Nonoxidative glucose disposal during euglycemic clamps was calculated as [R_a glucose]  [glucose oxidation] and served as an indirect measure of glycogen synthesis.

Hyperglycemic glucose clamps. Clamps were performed in all 20 calves. Experiments were conducted in the early and late phases of the growth trial, i.e., at ages 7 and 15 wk. Calves in group C weighed 87.8 ± 2.1 and 162.5 ± 2.0 kg; in group L, they weighed 87.9 ± 1.8 and 165.7 ± 0.6 kg. Jugular veins were catheterized the afternoon before the procedure. Two experiments were performed per day. After an overnight period of 15 h without feed, the experiment with the first calf was started at ~0800 h and the second, ~3 h later.

Three blood samples were taken at 20, 10 and 0 min to determine basal concentrations of insulin and glucose. The goal was to raise plasma glucose concentration by 4.4 mmol/L above basal concentrations and to keep it clamped at that concentration for 2 h. To achieve the hyperglycemic concentration, a 15-min priming glucose infusion was started. The initial rate was 0.55 mmol glucose/(kg BW·min) and decreased every 5 min on the basis of reaching and maintaining the planned glucose level. From 0 to 180 min, the GIR was adjusted every 5 min to maintain the hyperglycemia. Glucose extracted from sugar beets was purchased from Hausmann Laboratories, St. Gallen, Switzerland and was infused in concentrations of 7.63 or 15.26 mmol/L as required.

Blood was sampled every 5 min for 180 min and prepared immediately after collection for glucose measurements, as described for euglycemic clamps. Plasma from every second sample was frozen for later determination of insulin. Results from the first hour of the clamp were excluded from calculations because this period was used to stabilize glucose concentrations.

Laboratory analyses.  Concentrations of plasma glucose were measured with a Beckmann Glucose Analyzer 2 (Beckmann Instruments International SA, Zürich, Switzerland), concentrations of plasma insulin were determined by RIA (Hostettler-Allen et al. 1993) and concentrations of [¹³C₆]glucose and ¹³CO₂ were analyzed at the Institute of Physiology, University of Lausanne, CH-Lausanne as described by Hostettler-Allen et al. (1993 and 1994). Mean intra- and interassay coefficients of variation for the determination of [¹³C₆]glucose were 1.4 and 6.5 %, respectively, and for the analysis of ¹³CO₂ were 1.5 and 4.5 %, respectively. Insulin receptor number and affinity in soleus muscle were measured at the Institute for Physiology, Technical University of Munich, Freising-Weihenstephan, Germany according to Boge et al. (1994).

Feed analyses were performed as described by Hugi et al. (1997b). Dry matter was determined in milk replacers after evaporation of the water by heating at 110°C for 24 h, concentration of crude protein was measured by determination of nitrogen using the Kjeldahl method and crude fat was determined after Soxleth extraction according to standard procedures. Total sugar was calculated as the difference between dry matter and crude fat, crude protein and crude ash contents. Lactose was measured enzymatically using a kit (#176303) from Boehringer AG, Mannheim, Germany, in which lactose is hydrolyzed to D-glucose and D-galactose by -galactosidase and D-galactose by -galactose dehydrogenase to galactonic acid. Gross energy was measured using adiabatic bomb calorimeter. Iron was determined by atomic absorption spectrophotometry. Analyses of milk replacer contents were performed at the Federal Research Station of Animal Production, Posieux, Switzerland.

Statistical analyses.  Values are expressed as means ± SEM. Within groups, differences between mean basal values and mean concentrations in the case of stable concentrations during clamps were calculated. These measures were compared for experiments performed at the start and end of the growth trial and between groups.

Data were analyzed by ANOVA using the General Linear Models procedure of the SAS System (Release 6.08, SAS Institute, Cary, NC). The model used was Y_ijkl = µ + calf_i + treatment_j + period_k + (treatment_j × period_k) + e_ijkl, where Y_ijkl is the measured value, µ is the general mean, treatment_j represents the feeding of milk replacer with or without supplemental lactose, period_k is the early or late period of the growth trial, (treatment_j × period_k) is the interaction between the amounts of ingested lactose and age, and e_ijkl is the residual error. Paired t test was used to evaluate differences of values within groups and Student's t test was used to evaluate the significance of differences (P < 0.05) between groups.

	RESULTS

Abstract Introduction Methods Results Discussion References

Feeding and growth performance.  Calves of groups C and L entered the trial at the age of 5 wk with a BW of 69.2 ± 1.4 and 69.0 ± 1.6 kg, respectively. During the 98 d of the trial, calves of group C and L gained 124.1± 2.2 and 127.0 ± 1.8 kg, respectively. BW increased throughout the trial (P < 0.01). The average daily gain of groups C and L was 1.27 ± 0.05 and 1.29 ± 0.03 kg, respectively. Feed refusals were negligible. MR intake (on a dry matter basis) increased during the trial in group C from 0.91 ± 0.03 to 2.48 ± 0.48 kg/(calf·d) and in group L from 0.87 ± 0.01 to 2.67 ± 0.01 kg/(calf·d). Gain:feed ratios in group C and L were 0.60 ± 0.1 and 0.61 ± 0.1 kg/kg, respectively. Mean intakes in group C and group L of crude protein [12.5 ± 0.2 and 12.8 ± 0.2 g/(kg BW^0.75·d), respectively], crude fat [12.1 ± 0.2 and 11.9 ± 0.2 g/(kg BW^0.75·d), respectively] and gross energy [1.09 ± 0.02 and 1.14 ± 0.02 MJ/(kg BW^0.75·d), respectively] were similar, but intake of lactose was lower in group C than in group L [15.4 ± 0.3 vs. 23.1 ± 0.3 g/(kg BW^0.75·d), respectively].

Euglycemic-hyperinsulinemic glucose clamps.  Insulin concentration (Fig. 1) rapidly increased during infusions up to 30 min and then remained steadily elevated (P < 0.001). Insulin concentrations between 60 and 180 min were tested with linear regression analysis against time for stability; the slopes were not different from zero. Metabolic clearance rates of insulin in groups C and L were 13.9 ± 0.6 and 15.0 ± 1.5 mL/(min·kg), respectively, at the start of the growth trial and 12.4 ± 1.5 and 12.3 ± 0.5 mL/(min·kg), respectively, at the end of the growth trial. There were no significant differences of metabolic clearance rates of insulin at the start and end of the growth trial and between groups.

View larger version (26K): [in this window] [in a new window]   Fig 1. Plasma glucose and insulin concentrations and glucose infusion rate in calves during the euglycemic-hyperinsulinemic glucose clamp at the start (A) and end (B) of the growth trial. After an overnight period without food, calves were intravenously infused with [13C6]glucose [11.1 nmol/(kg·min)], insulin [13.8 pmol/(kg·min)] and glucose (varying rate to maintain euglycemia). Calves were fed a milk replacer containing normal amounts of lactose (group C: control; 290 g/kg milk replacer) or high amounts of lactose (group L: lactose; 423 g/kg milk replacer). Data are means ± SEM, n = 10 per group. Different uppercase letters (A, B) on right end of curves indicate significant differences (P < 0.05) of mean values (from 60 to 180 min during clamps) between group L (lactose) and group C (control). Different lowercase letters (a, b) on right end of curves indicate significant differences (P < 0.05) of mean values (from 60 to 180 min during clamps) between the start and end of the growth trial within a group. An asterisk on the right end of curves indicates significant differences (P < 0.05) of mean values during clamps (from 60 to 180 min) from mean basal (preclamp) values within a group. Lack of symbol indicates P > 0.05.

Basal glucose concentrations (Fig. 1) at the start and end of the growth trial and in both groups did not differ. Concentrations from 0 to 180 min were not different from basal concentrations, between the start and end of the trial or between groups. Linear regression analysis of the glucose concentration against time during the clamp revealed that the slope was not significantly different from zero.

From 0 to 30 min, GIR (Fig. 1) increased markedly (P < 0.05) and then slightly with time. GIR from 60 to 180 min, required to maintain steady-state euglycemia during clamps, were higher (P < 0.01) in group L at the start than at the end of the growth trial, but did not differ in group C [26.5 ± 2.4 and 25.0 ± 1.4 µmol/(kg·min), respectively, at the start and end of the growth trial; in group L, GIR were 33.2 ± 2.0 and 19.6 ± 1.3 µmol/(kg·min), respectively, at the start and end of the growth trial]. GIR from 60 to 180 min were higher (P < 0.01) in group L than group C at the start of the growth trial, but were lower (P < 0.05) in group L than in group C at the end of the growth trial.

As shown in Figure 2, after a 150 min equilibrium period, i.e., prime-continuous infusion of [¹³C₆]glucose, average atom % excess of [¹³C₆]glucose in blood plasma at 30, 20 and 10 min (basal values) was higher (P < 0.05) at the end than at the start of the growth trial, but was similar in both groups. After 60 min of insulin and glucose infusions, atom % excess [¹³C₆]glucose in groups C and L had fallen (P < 0.05). During the period from 60 to 180 min, atom % excess of [¹³C₆]glucose was higher (P < 0.001) in group L at the end than at the start of the growth trial, but was not significantly different in group C. Atom % excess of [¹³C₆]glucose was lower (P < 0.01) in group L than in group C at the start, but higher (P < 0.05) in group L than in group C at the end of the growth trial.

View larger version (26K): [in this window] [in a new window]   Fig 2. Atom % excess [13C6]glucose, rate of appearance of glucose and endogenous glucose output during the euglycemic-hyperinsulinemic glucose clamp at the start (A) and end (B) of the growth trial in calves fed a milk replacer containing normal amounts of lactose (control) or high amounts of lactose (lactose). Data are means ± SEM, n = 10 per group. For further information see legend to Figure 1.

The average basal rate of appearance of glucose in blood plasma (Fig. 2) between 30 and 10 min was similar in both groups. During the infusion of insulin and glucose, starting at 0 min, rates of appearance of glucose increased steadily (P < 0.05). At 60 min, the rates of appearance of glucose at the start and end of the growth trial did not differ in group C, but were higher (P < 0.05) at the start than at the end of the growth trial in group L. From 60 to 180 min, rates of appearance of glucose were higher (P < 0.001) in group L at the start than at the end of the growth trial, but did not differ in group C. Rates of appearance of glucose from 60 to 180 min in group C were higher (P < 0.001) than in group L at the start of the growth trial, but were lower (P < 0.05) than in group L at the end of the growth trial.

Endogenous glucose outputs (Fig. 2) before the beginning of insulin infusions in groups C and L were similar. After the start (at 0 min) of the insulin infusion, endogenous glucose output within 30 min decreased (P < 0.05) to values that did not differ between groups from 60 to 180 min in experiments at the start and end of the growth trial.

Respiratory rates were lower (P < 0.01), whereas minute volume, O₂ consumption, CO₂ production and respiratory exchange ratios were higher (P < 0.01; except in O₂ consumption in group L: P > 0.05) at the end than at the start of the growth trial (Table 2). Respiratory minute volume at the start and end of the growth trial, O₂ consumption at the end of the growth trial, CO₂ production at the start and end of the growth trial and respiratory exchange ratio at the start of the growth trial were different (P < 0.05) in group L and C. 

Atom % excess of ¹³CO₂ in exhaled air between 30 and 10 min at the start of the growth trial was lower (P < 0.05) in group C than at the end of the growth trial (Fig. 3), but did not differ in group L, and there were no group differences. At 60 min, atom % excess was increased (P < 0.01). From 60 to 180 min, atom % excess in group C at the start of the growth trial was lower (P < 0.05) than at the end of the growth trial, but there were no group differences.

View larger version (26K): [in this window] [in a new window]   Fig 3. Atom % excess 13CO2 in exhaled air and oxidized glucose during the euglycemic-hyperinsulinemic glucose clamp at the start (A) and end (B) of the growth trial in calves fed a milk replacer containing normal amounts of lactose (control) or high amounts of lactose (lactose). Data are means ± SEM, n = 10 per group. For further information see legend to Figure 1.

From 30 to 10 min, oxidized glucose was higher (P < 0.01) in both groups at the start than at the end of the growth trial (Fig. 3). Oxidized glucose increased (P < 0.05) at 60 min of the infusion and was higher (P < 0.05) at 60 min of the infusion in both groups at the start than at the end of the growth trial. Oxidized glucose from 60 to 180 min was higher (P < 0.05) at the start than at the end of the growth trial in group L, but did not differ in group C at the start and the end of the growth trial. Glucose oxidation at the start and the end of the growth trial did not differ between groups.

Hyperglycemic glucose clamps.  Basal glucose concentrations did not differ at the start and end of the growth trial or between groups (Fig. 4). During glucose infusions, glucose concentration increased (P < 0.001) to stable concentrations within 45 min. The stability of the new glucose plateau was evaluated with regression analysis; the slope did not differ significantly from zero.

View larger version (28K): [in this window] [in a new window]   Fig 4. Plasma glucose and insulin concentration and glucose infusion rate in the hyperglycemic glucose clamp at the start (A) and end (B) of the growth trial in calves fed a milk replacer containing normal amounts of lactose (control) or high amounts of lactose (lactose). Glucose was infused intravenously to maintain hyperglycemia by 4.4 mmol/L above basal concentrations for 2 h. Data are means ± SEM, n = 10 per group. For further details see legend to Figure 1.

Basal insulin concentrations did not differ at the start and end of the growth period or between groups. Insulin concentrations increased continuously (P < 0.001) after the commencement of glucose infusions. Concentrations during clamps from 60 to 180 min were not different at the start and end of the growth trial or between groups.

The mean GIR from 120 to 180 min, required to maintain steady-state hyperglycemia in clamps at the start and end of the growth trial in groups C and L, were higher (P < 0.001) in group L at the start than at the end of the growth trial, but did not differ in group C. GIR from 60 to 180 min at the start of the growth trial in group L was higher (P < 0.02) than in group C, but at the end of the growth trial was lower (P < 0.05) in group L than in group C.

Insulin receptors in soleus muscle.  Insulin receptor number was lower (P < 0.05) in group L (13.5 ± 0.6 fmol/mg protein) than in group C (25.6 ± 0.7 fmol/mg protein), whereas affinity in group L (10.2 ± 5.5 × 10¹⁰ mol/mg protein) and group C (9.0 ± 6.0 × 10¹⁰ mol/mg protein) did not differ.

	DISCUSSION

Abstract Introduction Methods Results Discussion References

This study was performed under conditions as planned, i.e,., markedly different intakes of total sugar (primarily lactose), but otherwise very similar intakes of feed, crude protein, crude fat, gross energy, amino acids, vitamins, minerals (including iron) and very similar average daily gains and growth/feed ratios. Starch, which was added to the diet fed to the control group, was expected to be less well digested than lactose (Toullec and Lallès 1995). Because salivary amylin activity is not detectable in calves, in the milk-fed calf, pancreatic -amylase activity and secretion are low, and activities and secretion of -amylase, maltase and isomaltase from 1 to 4 mo of age increase less in preruminating than in ruminating (weaned) calves (Le Huerou et al. 1992a and 1992b). Metabolizable energy intake exerts a major influence on pancreatic -amylase activity and secretion, whereas increasing starch intake may even lower pancreatic -amylase activity in ruminants (Croom et al. 1992, Harmon 1992). Because gross energy intake was nearly identical in both groups, -amylase activity and secretion were probably not particularly enhanced in the control group in which lactose was replaced by starch. On this basis, small intestinal absorption of glucose was presumably greater in the group fed high amounts of lactose than in the control group. Different physiologic changes in the lactose-supplemented group compared with the control group were therefore likely due primarily to differences in lactose intake, not starch intake. This is supported by a previous study, in which the same diets were fed as in this study and in which plasma glucose concentration at the start and end of the growth trial increased postprandially significantly more in calves fed the lactose-supplemented diet than in those fed normal amounts of lactose (Hugi et al. 1997a).

Although not measured directly, there were no obvious differences in body composition at slaughter, i.e., apparent increase of muscle mass and body fat from high (relative to normal) lactose intake (not shown). Hemoglobin and plasma iron concentrations (not shown) were not different between groups, as expected, because calves were fed the same amounts of iron. This is of importance, because differences in iron intake in veal calves affect insulin-dependent glucose utilization (Hostettler-Allen et al. 1993).

Respiratory measurements during euglycemic clamps demonstrated significantly or numerically smaller respiration rates, greater minute volumes, O₂ consumption and CO₂ production at the end than at the start of the growth trial, as expected. Significant differences of respiratory traits between groups were usually small and not consistent for all variables. Values of respiratory exchange ratios indicated that carbohydrate oxidation was small, which could be expected after a period of 15 h without food. There were discrepancies regarding respiratory exchange ratio and [¹³C₆]glucose oxidation data at the start of the growth trial: calorimetry indicated essentially no oxidation of carbohydrates, whereas ¹³CO₂ production clearly showed that labeled glucose was oxidized. The most likely explanation is that it is difficult to obtain accurate values of both CO₂ and O₂ exchanges, i.e., small errors in the two determinations may yield large errors in substrate oxidation. As a comparison, values obtained with ¹³CO₂ (although also subjected to error when CO₂ production is accentuated) are much more robust.

Values of insulin in both euglycemic and hyperglycemic clamps were within the physiologic range. Results of the euglycemic-hyperinsulinemic clamps show that there were no differences in metabolic clearance rates of insulin. On that basis, the rise of insulin seen during hyperglycemic clamps mirrored insulin secretion rates. The data of hyperglycemic clamps demonstrate that insulin secretion rates during constant and identical glucose levels were not different at the start and end of the growth trial or between groups. A similar lack of differences in insulin responses (in meal tolerance tests, oral and intravenous glucose tolerance tests) between groups C and L was also found in another trial, in which calves were fed the same MR as in this experiment (Hugi et al. 1997b). However, in intravenous glucose tolerance tests, performed preprandially in the other growth trial, insulin responses were greater at the end than at the start of the growth trial, although glucose responses were similar. Reasons for these differences are not clear. It can be speculated that differences in the kinetics of insulin secretion, affecting different insulin pools in B cells, may have contributed to differences between insulin responses in hyperglycemic clamps and intravenous glucose tolerance tests.

During both euglycemic and hyperglycemic glucose clamps, glucose infusion rates required to maintain euglycemia or hyperglycemia in group L were higher than those in group C at the start, but lower than those in group C at the end of the growth trial. Glucose infusion rates in both euglycemic and hyperglycemic clamps were higher at the start than at the end of the growth trial in group L, but comparable in group C. Under the given conditions of nearly identical insulin concentrations, this indicates that the insulin-dependent glucose utilization was enhanced at the start, but reduced toward the end of the growth trial in group L compared with group C, but not changed in group C. Toward the end of another growth trial, in which the same MR was fed as in this study, evidence for markedly reduced insulin-dependent glucose utilization especially if performed postprandially has been found with other, more traditional tests, i.e., meal, oral glucose, intravenous glucose and insulin tolerance tests (Hugi et al. 1997b) and in postprandial studies on glucose kinetics using [6,6²d]glucose (Hugi et al. 1997a). In contrast to these tests, the euglycemic and hyperglycemic clamps were performed in calves after a period of 15 h without feed, i.e., preprandially. The present data therefore demonstrate that glucose metabolism towards the end of the growth trial was also disturbed in the unfed state if calves had prior exposure to high lactose intakes.

Causes for the enhanced sensitivity or responsiveness to insulin at the start of the growth trial in group L are not obvious, but such sensitivity would help to enhance glucose utilization. On the other hand, the reduced insulin-dependent glucose utilization towards the end of the growth trial was likely a consequence of the reduced number (but not affinity) of insulin receptors found in this study in the soleus muscle. This is of importance because skeletal muscles in growing animals take up most of the glucose of all organs (Weekes 1991). Postprandial hyperinsulinemia for several months may have been an important per se as a cause of down-regulation of insulin receptors at the end of the growth trial. On the other hand, because insulin concentrations during clamps were nearly identical in both experimental groups and very similar at the start and end of the growth trial, circulating insulin was not a cause of differences in insulin-dependent glucose utilization during clamps.

The application of the euglycemic glucose clamp technique, in combination with [¹³C₆]glucose, whose atom % excess was measured in plasma and as ¹³CO₂ in exhaled air, allowed for discrimination of changes in glucose metabolism between group C and group L at the start and at the end of the growth trial.

Because [¹³C₆]glucose was infused on a per kilogram BW basis, the higher basal atom % excess [¹³C₆]glucose values at the end than at the start of the growth trial, seen in both groups before and during clamps, may have been due in part to greater amounts of [¹³C₆]glucose infused on a metabolic BW basis (i.e., per kg^0.75). The continuous fall of basal atom % excess [¹³C₆]glucose values during clamps was due to dilution of the infused tracer by endogenous glucose production and by exogenously infused glucose and therefore provided an estimate of total (oxidative and nonoxidative) glucose utilization. During clamps, the atom % of [¹³C₆]glucose excess was lower in group L than in group C at the start of the growth trial, but higher at the end of the growth trial. Therefore, in group L, insulin-dependent glucose utilization, i.e., effects of insulin on glucose utilization, were greater at the start, but lower at the end of the growth trial.

Rates of glucose appearance equaled glucose utilization rates under steady-state conditions of plasma glucose and insulin as existed during euglycemic-hyperinsulinemic clamps. Findings of lower glucose appearance rates in group L at the end than at the start of the growth trial, although glucose appearance rates in group C were not different, were therefore in accordance with conclusions drawn from glucose infusion rates, indicating reduced insulin-dependent glucose utilization. Greater glucose appearance rates in group L than in group C at the start of the growth trial and smaller glucose appearance rates of group L than group C at the end of the trial reflected differences in glucose utilization rates.

Because clamps were performed after an overnight (15 h) period without feed, no exogenous glucose was added to the plasma pool from the gastrointestinal tract. Hence, the basal rate of glucose production was primarily from endogenous sources. As a result of effects of insulin during euglycemic glucose clamps, endogenous glucose output decreased very rapidly. The inhibition was not complete, a finding typical even in young ruminants and in agreement with previous studies (Hostettler-Allen et al. 1994). Importantly, the inhibition of endogenous glucose production by insulin infusions was not influenced by age and lactose intake.

The ¹³CO₂ % excess in exhaled air increased during the experiments, as expected. Reduced ¹³CO₂ % excess in exhaled air in group C at the end compared with at the start of the growth trial and the smaller values in group L than in group C at the end of the growth trial reflected [¹³C₆]glucose oxidation. Smaller glucose oxidation rates in group L at the end than at the start of the growth trial were likely the consequence of reduced insulin effects. Because less glucose was oxidized in group L at the end of the growth trial, more glucose had to be disposed of through nonoxidative pathways. Because glucose concentrations were much below the renal threshold in euglycemic clamps (Hostettler-Allen et al. 1994), renal losses in this situation were unlikely. Therefore, more glucose was probably stored as glycogen in group L than in group C. In studies on postprandial glucose kinetics in veal calves toward the end of another growth trial, we also arrived at the conclusion that glucose oxidation did not make a major contribution to plasma glucose disposal (Hugi et al. 1997a).

This study demonstrates that in contrast to the postprandial stage, during which insulin resistance develops with increasing age, there was no evidence for the development of insulin resistance during the growth trial after a period of 15 h without feed. However, evidence for development of insulin resistance could be demonstrated in calves fed high amounts of lactose for a prolonged period; this is probably causally associated with a reduced insulin receptor number in skeletal muscle.

	FOOTNOTES

Manuscript received 3 September 1997. Initial reviews completed 1 December 1997. Revision accepted 26 January 1998.

	LITERATURE CITED

Abstract Introduction Methods Results Discussion References

• Boge A., Sauerwein H., Meyer H.H.D. An enzyme immunoreceptor assay for the quantitation of insulin-like growth factor-I and insulin receptors in bovine muscle tissue. Anal. Biochem. 1994; 216:406-412[Medline]
• Breier B.H.P., Gluckman D., Bass., J. J. Plasma concentrations of insulin-like growth factor I and insulin in the infant calf: ontogeny and influence of altered nutrition. J. Endocrinol. 1988; 119:43-50[Abstract]
• Bruckmaier R. M., Blum J. W. Responses of calves to treadmill exercise during -adrenergic agonist administration. J. Anim. Sci. 1992; 70:2809-2821[Abstract]
• Croom W. J. Jr., Bull L. S., Taylor I. L. Regulation of pancreatic exocrine secretion in ruminants: a review. J. Nutr. 1992; 122:191-202
• DeBodo R., Steele R., Altszuler N., Dunn A., Bishop J. On the hormonal regulation of carbohydrate metabolism. Recent Prog. Horm. Res. 1963; 19:445-488
• Doppenberg J., Palmquist D. L. Effect of dietary fat level on feed intake, growth, plasma metabolites and hormones of calves fed dry or liquid diets. Livestock Prod. Sci. 1991; 29:151-166
• Guilloteau, P., Le Huërou-Luron., I., Toullec, R., Chayvialle, J. A. & Blum, J. W. (1995) Regulatory peptides in young ruminants. In: Ruminant Physiology: Digestion, Metabolism, Growth and Reproduction (von Engelhardt, W., Leonhard-Marek, S., Breves, G. & Giesecke, D., eds.), pp. 519-537. F. Enke Verlag, Stuttgart, Germany.
• Harmon D. L. Dietary influences on carbohydrates and small intestinal starch hydrolysis capacity in ruminants. J. Nutr. 1992; 122:203-210
• Hostettler-Allen R., Tappy L., Blum J. W. Enhanced insulin-dependent glucose utilization in iron-deficient veal calves. J. Nutr. 1993; 123:1656-1667
• Hostettler-Allen R., Tappy L., Blum J. W. Insulin resistance, hyperglycemia and glucosuria in intensively milk-fed calves. J. Anim. Sci. 1994; 72:160-173[Abstract]
• Hugi D., Blum J. W. Changes of blood metabolites and hormones in breeding calves associated with weaning. J. Vet. Med A. 1997; 44:99-108
• Hugi D. P., Bracher V. D., Tappy L., Blum J. W. Postprandial hydrogen breath excretion, plasma lactate concentration, glucose metabolism and insulin levels in veal calves. J. Anim. Physiol. Anim. Nutr. 1997a; 78:42-48
• Hugi D., Bruckmaier R. M., Blum J. W. Insulin resistance, hyperglycemia, glucosuria, and galactosuria in intensively milk-fed calves: dependency on age and effects of high lactose intake. J. Anim. Sci. 1997b; 75:469-482[Abstract/Free Full Text]
• Hugi D., Gut S. H., Blum J. W. Blood metabolites and hormonesespecially glucose and insulin in veal calves: effects of age and nutrition. J. Vet. Med. A 1997c; 44:407-416
• Jones J. I., Clemmons D. R Insulin-like growth factors and their binding proteins: biological actions. Endocr. Rev. 1995; 16:3-34[Medline]
• Le Huërou-Luron, I., Guilloteau, P., Wicker-Planquart, C., Chayvialle, J.-A., Burton, J., Mouats, A., Toullec, R. & Puigserver, A. (1992 a) Gastric and pancreatic enzyme activities and their relationship with some gut regulatory peptides during postnatal development and weaning in the calf. J. Nutr. 122: 1434-1445.
• Le Huërou-Luron, I., Guilloteau, P., Wicker-Planquart, C., Mouats, A., Chayvialle, J.-A., Bernard, C., Burton, J., Toullec, R. & Puigserver, A. (1992 b) Activity distribution of seven digestive enzymes along small intestine in calves during development and weaning. Dig. Dis. Sci. 37: 40-46.
• Palmquist D. L., Doppenberg J., Roehrig K. L., Kinsey D. J. . Glucose and insulin metabolism in ruminating and veal calves fed high and low fat diets. Domest. Anim. Endocrinol. 1992; 9:233-242[Medline]
• Ronge H., Blum J. W. Insulin-like growth factor I binding proteins in dairy cows, calves and bulls. Acta Endocrinol. (Kbh.) 1989; 120:105-111
• Toullec, R. & Lallès, J.-P. (1995) Digestion dans la caillette et l'intestin grêle. In: Nutrition des Ruminants Domestiques (Jarrige, R., Ruckebusch, Y., Demarquilly, C., Farce, M.-H. & Journet, M. K., eds.), pp. 527-582. Institut National de la Recherche Agronomique, Paris, France.
• Weekes, T.E.C. (1991). Hormonal control of glucose metabolism. In: Physiological Aspects of Digestion and Metabolism in Ruminants (Tsuda, T., Sasaki, Y. & Kawashima, R.), pp. 183-200. Academic Press, San Diego, CA.
• Wijayashinghe M. S., Smith N. E., Baldwin, R. L.. Growth, health, and blood glucose concentrations of calves fed high-glucose and high-fat milk replacers. J. Dairy Sci. 1984; 67:2949-2958

0022-3166/98 $3.00 ©1998 American Society for Nutritional Sciences